{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9989658738366081,
  "eval_steps": 500,
  "global_step": 966,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 353.0,
      "completions/max_terminated_length": 353.0,
      "completions/mean_length": 172.6666717529297,
      "completions/mean_terminated_length": 172.6666717529297,
      "completions/min_length": 122.0,
      "completions/min_terminated_length": 122.0,
      "epoch": 0.001034126163391934,
      "grad_norm": 5.071864592095285,
      "kl": 0.3359375,
      "learning_rate": 1e-06,
      "loss": 0.0134,
      "num_tokens": 78552.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.40627965331077576,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.5036101341247559,
      "step": 1
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 247.0,
      "completions/max_terminated_length": 247.0,
      "completions/mean_length": 165.5416717529297,
      "completions/mean_terminated_length": 165.5416717529297,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.002068252326783868,
      "grad_norm": 4.059461930492566,
      "kl": 0.0036773681640625,
      "learning_rate": 9.999973613218312e-07,
      "loss": 0.0001,
      "num_tokens": 161365.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.4685417115688324,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 2
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 337.0,
      "completions/max_terminated_length": 337.0,
      "completions/mean_length": 188.95834350585938,
      "completions/mean_terminated_length": 188.95834350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.0031023784901758012,
      "grad_norm": 3.468953003174482,
      "kl": 0.0027923583984375,
      "learning_rate": 9.999894453151758e-07,
      "loss": 0.0001,
      "num_tokens": 239820.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 3
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 254.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 157.0,
      "completions/mean_terminated_length": 157.0,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.004136504653567736,
      "grad_norm": 3.413335444191301,
      "kl": 0.002685546875,
      "learning_rate": 9.999762520635849e-07,
      "loss": 0.0001,
      "num_tokens": 318868.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.25634264945983887,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.371054083108902,
      "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": 408.0,
      "completions/max_terminated_length": 408.0,
      "completions/mean_length": 209.45834350585938,
      "completions/mean_terminated_length": 209.45834350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.005170630816959669,
      "grad_norm": 3.8321189551506687,
      "kl": 0.0026397705078125,
      "learning_rate": 9.99957781706309e-07,
      "loss": 0.0001,
      "num_tokens": 402655.0,
      "reward": 1.027777910232544,
      "reward_std": 0.3323635756969452,
      "rewards/reasoning_reward/mean": 1.0277777910232544,
      "rewards/reasoning_reward/std": 0.5443310141563416,
      "step": 5
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 411.0,
      "completions/max_terminated_length": 411.0,
      "completions/mean_length": 181.6666717529297,
      "completions/mean_terminated_length": 181.6666717529297,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.0062047569803516025,
      "grad_norm": 4.016570280789986,
      "kl": 0.00323486328125,
      "learning_rate": 9.999340344382978e-07,
      "loss": 0.0001,
      "num_tokens": 484391.0,
      "reward": 1.3958333730697632,
      "reward_std": 0.3663109242916107,
      "rewards/reasoning_reward/mean": 1.3958333730697632,
      "rewards/reasoning_reward/std": 0.642332136631012,
      "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": 0.0,
      "completions/max_length": 358.0,
      "completions/max_terminated_length": 358.0,
      "completions/mean_length": 213.08334350585938,
      "completions/mean_terminated_length": 213.08334350585938,
      "completions/min_length": 135.0,
      "completions/min_terminated_length": 135.0,
      "epoch": 0.007238883143743537,
      "grad_norm": 2.9810718864115326,
      "kl": 0.0038604736328125,
      "learning_rate": 9.99905010510197e-07,
      "loss": 0.0002,
      "num_tokens": 572137.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.3535533845424652,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.5089774131774902,
      "step": 7
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 283.0,
      "completions/max_terminated_length": 283.0,
      "completions/mean_length": 168.58334350585938,
      "completions/mean_terminated_length": 168.58334350585938,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.008273009307135471,
      "grad_norm": 0.033223094376599045,
      "kl": 0.0020904541015625,
      "learning_rate": 9.998707102283457e-07,
      "loss": 0.0001,
      "num_tokens": 650255.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 8
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 223.0,
      "completions/max_terminated_length": 223.0,
      "completions/mean_length": 166.4166717529297,
      "completions/mean_terminated_length": 166.4166717529297,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.009307135470527405,
      "grad_norm": 2.0403595074712557,
      "kl": 0.0030975341796875,
      "learning_rate": 9.998311339547733e-07,
      "loss": 0.0001,
      "num_tokens": 734225.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.550032913684845,
      "step": 9
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 297.0,
      "completions/max_terminated_length": 297.0,
      "completions/mean_length": 185.1666717529297,
      "completions/mean_terminated_length": 185.1666717529297,
      "completions/min_length": 74.0,
      "completions/min_terminated_length": 74.0,
      "epoch": 0.010341261633919338,
      "grad_norm": 3.389792798216633,
      "kl": 0.0031890869140625,
      "learning_rate": 9.997862821071964e-07,
      "loss": 0.0001,
      "num_tokens": 819653.0,
      "reward": 0.875,
      "reward_std": 0.40812820196151733,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.5366967916488647,
      "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": 279.0,
      "completions/max_terminated_length": 279.0,
      "completions/mean_length": 186.95834350585938,
      "completions/mean_terminated_length": 186.95834350585938,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.011375387797311272,
      "grad_norm": 2.331094314362588,
      "kl": 0.0029296875,
      "learning_rate": 9.997361551590132e-07,
      "loss": 0.0001,
      "num_tokens": 904148.0,
      "reward": 0.5,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.5,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 11
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 147.2916717529297,
      "completions/mean_terminated_length": 147.2916717529297,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.012409513960703205,
      "grad_norm": 3.413766004232087,
      "kl": 0.0036163330078125,
      "learning_rate": 9.996807536392989e-07,
      "loss": 0.0001,
      "num_tokens": 984211.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.2587745785713196,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.47729232907295227,
      "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": 346.0,
      "completions/max_terminated_length": 346.0,
      "completions/mean_length": 145.625,
      "completions/mean_terminated_length": 145.625,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.01344364012409514,
      "grad_norm": 3.9659481687377354,
      "kl": 0.0033416748046875,
      "learning_rate": 9.996200781328011e-07,
      "loss": 0.0001,
      "num_tokens": 1067738.0,
      "reward": 0.9375,
      "reward_std": 0.41282182931900024,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.5954993963241577,
      "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": 277.0,
      "completions/max_terminated_length": 277.0,
      "completions/mean_length": 172.9166717529297,
      "completions/mean_terminated_length": 172.9166717529297,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.014477766287487074,
      "grad_norm": 2.6924838670848303,
      "kl": 0.0023956298828125,
      "learning_rate": 9.99554129279932e-07,
      "loss": 0.0001,
      "num_tokens": 1146640.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "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": 328.0,
      "completions/max_terminated_length": 328.0,
      "completions/mean_length": 164.9166717529297,
      "completions/mean_terminated_length": 164.9166717529297,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.015511892450879007,
      "grad_norm": 4.1278369792222325,
      "kl": 0.0029144287109375,
      "learning_rate": 9.99482907776763e-07,
      "loss": 0.0001,
      "num_tokens": 1224398.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.42645785212516785,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "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": 223.0,
      "completions/max_terminated_length": 223.0,
      "completions/mean_length": 170.6666717529297,
      "completions/mean_terminated_length": 170.6666717529297,
      "completions/min_length": 122.0,
      "completions/min_terminated_length": 122.0,
      "epoch": 0.016546018614270942,
      "grad_norm": 0.03303182322940017,
      "kl": 0.002471923828125,
      "learning_rate": 9.994064143750165e-07,
      "loss": 0.0001,
      "num_tokens": 1303678.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "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": 349.0,
      "completions/max_terminated_length": 349.0,
      "completions/mean_length": 185.20834350585938,
      "completions/mean_terminated_length": 185.20834350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.017580144777662874,
      "grad_norm": 4.190633270648795,
      "kl": 0.003997802734375,
      "learning_rate": 9.993246498820587e-07,
      "loss": 0.0002,
      "num_tokens": 1387411.0,
      "reward": 0.875,
      "reward_std": 0.47920867800712585,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.5943574905395508,
      "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": 196.0,
      "completions/max_terminated_length": 196.0,
      "completions/mean_length": 138.75,
      "completions/mean_terminated_length": 138.75,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.01861427094105481,
      "grad_norm": 4.124744220407965,
      "kl": 0.0032958984375,
      "learning_rate": 9.992376151608897e-07,
      "loss": 0.0001,
      "num_tokens": 1465757.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.4198887050151825,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.7469745874404907,
      "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": 201.0,
      "completions/max_terminated_length": 201.0,
      "completions/mean_length": 122.29167175292969,
      "completions/mean_terminated_length": 122.29167175292969,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.01964839710444674,
      "grad_norm": 4.2126320398251105,
      "kl": 0.0028839111328125,
      "learning_rate": 9.991453111301365e-07,
      "loss": 0.0001,
      "num_tokens": 1546284.0,
      "reward": 0.3333333432674408,
      "reward_std": 0.3900056481361389,
      "rewards/reasoning_reward/mean": 0.3333333432674408,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "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": 312.0,
      "completions/max_terminated_length": 312.0,
      "completions/mean_length": 169.5,
      "completions/mean_terminated_length": 169.5,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.020682523267838676,
      "grad_norm": 3.9035749186278075,
      "kl": 0.00433349609375,
      "learning_rate": 9.990477387640415e-07,
      "loss": 0.0002,
      "num_tokens": 1629112.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.43810173869132996,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4871538281440735,
      "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": 255.0,
      "completions/max_terminated_length": 255.0,
      "completions/mean_length": 183.33334350585938,
      "completions/mean_terminated_length": 183.33334350585938,
      "completions/min_length": 127.0,
      "completions/min_terminated_length": 127.0,
      "epoch": 0.02171664943123061,
      "grad_norm": 2.7742036215600043,
      "kl": 0.0037078857421875,
      "learning_rate": 9.989448990924528e-07,
      "loss": 0.0001,
      "num_tokens": 1706600.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "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": 326.0,
      "completions/max_terminated_length": 326.0,
      "completions/mean_length": 179.125,
      "completions/mean_terminated_length": 179.125,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.022750775594622543,
      "grad_norm": 2.9495492352769728,
      "kl": 0.007080078125,
      "learning_rate": 9.988367932008138e-07,
      "loss": 0.0003,
      "num_tokens": 1791203.0,
      "reward": 0.9375,
      "reward_std": 0.30551642179489136,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.37044334411621094,
      "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": 394.0,
      "completions/max_terminated_length": 394.0,
      "completions/mean_length": 208.70834350585938,
      "completions/mean_terminated_length": 208.70834350585938,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.023784901758014478,
      "grad_norm": 3.890766517596311,
      "kl": 0.00811767578125,
      "learning_rate": 9.98723422230151e-07,
      "loss": 0.0003,
      "num_tokens": 1871036.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.40397143363952637,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.6197241544723511,
      "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": 278.0,
      "completions/max_terminated_length": 278.0,
      "completions/mean_length": 161.375,
      "completions/mean_terminated_length": 161.375,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.02481902792140641,
      "grad_norm": 3.3269487482285345,
      "kl": 0.0029144287109375,
      "learning_rate": 9.986047873770624e-07,
      "loss": 0.0001,
      "num_tokens": 1949213.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "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": 262.0,
      "completions/max_terminated_length": 262.0,
      "completions/mean_length": 146.1666717529297,
      "completions/mean_terminated_length": 146.1666717529297,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.025853154084798345,
      "grad_norm": 3.9477140851301993,
      "kl": 0.0059814453125,
      "learning_rate": 9.98480889893705e-07,
      "loss": 0.0002,
      "num_tokens": 2028025.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.5605560541152954,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.6507381796836853,
      "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": 361.0,
      "completions/max_terminated_length": 361.0,
      "completions/mean_length": 219.9166717529297,
      "completions/mean_terminated_length": 219.9166717529297,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.02688728024819028,
      "grad_norm": 4.080956404876102,
      "kl": 0.00860595703125,
      "learning_rate": 9.98351731087781e-07,
      "loss": 0.0003,
      "num_tokens": 2107687.0,
      "reward": 0.875,
      "reward_std": 0.46288391947746277,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.5757792592048645,
      "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": 293.0,
      "completions/max_terminated_length": 293.0,
      "completions/mean_length": 163.5416717529297,
      "completions/mean_terminated_length": 163.5416717529297,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.027921406411582212,
      "grad_norm": 3.5156776593492745,
      "kl": 0.0048828125,
      "learning_rate": 9.982173123225243e-07,
      "loss": 0.0002,
      "num_tokens": 2186948.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.40529346466064453,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.4995468854904175,
      "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": 265.0,
      "completions/max_terminated_length": 265.0,
      "completions/mean_length": 121.75,
      "completions/mean_terminated_length": 121.75,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.028955532574974147,
      "grad_norm": 2.437538390860398,
      "kl": 0.0029449462890625,
      "learning_rate": 9.980776350166867e-07,
      "loss": 0.0001,
      "num_tokens": 2267470.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "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": 218.0,
      "completions/max_terminated_length": 218.0,
      "completions/mean_length": 151.7916717529297,
      "completions/mean_terminated_length": 151.7916717529297,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.02998965873836608,
      "grad_norm": 3.9269223056078095,
      "kl": 0.004730224609375,
      "learning_rate": 9.979327006445216e-07,
      "loss": 0.0002,
      "num_tokens": 2344097.0,
      "reward": 0.4583333432674408,
      "reward_std": 0.5049939155578613,
      "rewards/reasoning_reward/mean": 0.4583333432674408,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "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": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 186.20834350585938,
      "completions/mean_terminated_length": 186.20834350585938,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.031023784901758014,
      "grad_norm": 3.386462451637892,
      "kl": 0.004547119140625,
      "learning_rate": 9.977825107357702e-07,
      "loss": 0.0002,
      "num_tokens": 2425350.0,
      "reward": 0.75,
      "reward_std": 0.33247750997543335,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "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": 278.0,
      "completions/max_terminated_length": 278.0,
      "completions/mean_length": 178.6666717529297,
      "completions/mean_terminated_length": 178.6666717529297,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.03205791106514995,
      "grad_norm": 3.3979916126388687,
      "kl": 0.01055908203125,
      "learning_rate": 9.976270668756433e-07,
      "loss": 0.0004,
      "num_tokens": 2510054.0,
      "reward": 1.0277777910232544,
      "reward_std": 0.3074157238006592,
      "rewards/reasoning_reward/mean": 1.0277777910232544,
      "rewards/reasoning_reward/std": 0.4627858102321625,
      "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": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 155.7916717529297,
      "completions/mean_terminated_length": 155.7916717529297,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.033092037228541885,
      "grad_norm": 2.8184150341115375,
      "kl": 0.004058837890625,
      "learning_rate": 9.974663707048065e-07,
      "loss": 0.0002,
      "num_tokens": 2588921.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.35634833574295044,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "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": 193.0,
      "completions/max_terminated_length": 193.0,
      "completions/mean_length": 154.70834350585938,
      "completions/mean_terminated_length": 154.70834350585938,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.03412616339193381,
      "grad_norm": 2.828806479597615,
      "kl": 0.005157470703125,
      "learning_rate": 9.973004239193618e-07,
      "loss": 0.0002,
      "num_tokens": 2671218.0,
      "reward": 1.1319444179534912,
      "reward_std": 0.13749298453330994,
      "rewards/reasoning_reward/mean": 1.1319444179534912,
      "rewards/reasoning_reward/std": 0.3474515974521637,
      "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": 301.0,
      "completions/max_terminated_length": 301.0,
      "completions/mean_length": 204.0,
      "completions/mean_terminated_length": 204.0,
      "completions/min_length": 172.0,
      "completions/min_terminated_length": 172.0,
      "epoch": 0.03516028955532575,
      "grad_norm": 3.0486725384758566,
      "kl": 0.005401611328125,
      "learning_rate": 9.971292282708296e-07,
      "loss": 0.0002,
      "num_tokens": 2755530.0,
      "reward": 0.8819444179534912,
      "reward_std": 0.3624235987663269,
      "rewards/reasoning_reward/mean": 0.8819444179534912,
      "rewards/reasoning_reward/std": 0.6226000189781189,
      "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": 322.0,
      "completions/max_terminated_length": 322.0,
      "completions/mean_length": 190.70834350585938,
      "completions/mean_terminated_length": 190.70834350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.03619441571871768,
      "grad_norm": 2.8675194738593675,
      "kl": 0.01416015625,
      "learning_rate": 9.969527855661307e-07,
      "loss": 0.0006,
      "num_tokens": 2845651.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.434057354927063,
      "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": 265.0,
      "completions/max_terminated_length": 265.0,
      "completions/mean_length": 167.6666717529297,
      "completions/mean_terminated_length": 167.6666717529297,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.03722854188210962,
      "grad_norm": 2.8365389717558775,
      "kl": 0.0047607421875,
      "learning_rate": 9.967710976675674e-07,
      "loss": 0.0002,
      "num_tokens": 2924235.0,
      "reward": 0.4166666865348816,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.4166666567325592,
      "rewards/reasoning_reward/std": 0.5036101937294006,
      "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": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 160.20834350585938,
      "completions/mean_terminated_length": 160.20834350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.038262668045501554,
      "grad_norm": 3.3596940088485066,
      "kl": 0.00372314453125,
      "learning_rate": 9.965841664928032e-07,
      "loss": 0.0001,
      "num_tokens": 3004232.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.42645785212516785,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "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": 0.04166666666666663,
      "completions/max_length": 1217.0,
      "completions/max_terminated_length": 398.0,
      "completions/mean_length": 256.79168701171875,
      "completions/mean_terminated_length": 215.04348754882812,
      "completions/min_length": 142.0,
      "completions/min_terminated_length": 142.0,
      "epoch": 0.03929679420889348,
      "grad_norm": 3.713051017958645,
      "kl": 0.006988525390625,
      "learning_rate": 9.963919940148428e-07,
      "loss": 0.0003,
      "num_tokens": 3095475.0,
      "reward": 1.2569445371627808,
      "reward_std": 0.36115893721580505,
      "rewards/reasoning_reward/mean": 1.2569445371627808,
      "rewards/reasoning_reward/std": 0.5223950147628784,
      "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": 318.0,
      "completions/max_terminated_length": 318.0,
      "completions/mean_length": 203.5416717529297,
      "completions/mean_terminated_length": 203.5416717529297,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.04033092037228542,
      "grad_norm": 4.058262730903875,
      "kl": 0.01019287109375,
      "learning_rate": 9.961945822620118e-07,
      "loss": 0.0004,
      "num_tokens": 3184992.0,
      "reward": 1.375,
      "reward_std": 0.6197125911712646,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.6954823136329651,
      "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": 0.04166666666666663,
      "completions/max_length": 255.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 177.83334350585938,
      "completions/mean_terminated_length": 174.478271484375,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.04136504653567735,
      "grad_norm": 3.0112482413882797,
      "kl": 0.0054931640625,
      "learning_rate": 9.959919333179344e-07,
      "loss": 0.0002,
      "num_tokens": 3266916.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "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": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 144.9166717529297,
      "completions/mean_terminated_length": 144.9166717529297,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.04239917269906929,
      "grad_norm": 3.571922005370386,
      "kl": 0.006378173828125,
      "learning_rate": 9.957840493215116e-07,
      "loss": 0.0003,
      "num_tokens": 3351314.0,
      "reward": 1.0625,
      "reward_std": 0.3972596824169159,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.5578004121780396,
      "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": 341.0,
      "completions/max_terminated_length": 341.0,
      "completions/mean_length": 174.2916717529297,
      "completions/mean_terminated_length": 174.2916717529297,
      "completions/min_length": 76.0,
      "completions/min_terminated_length": 76.0,
      "epoch": 0.04343329886246122,
      "grad_norm": 4.032955623530826,
      "kl": 0.006561279296875,
      "learning_rate": 9.955709324668997e-07,
      "loss": 0.0003,
      "num_tokens": 3429881.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.8427009582519531,
      "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": 293.0,
      "completions/max_terminated_length": 293.0,
      "completions/mean_length": 201.625,
      "completions/mean_terminated_length": 201.625,
      "completions/min_length": 119.0,
      "completions/min_terminated_length": 119.0,
      "epoch": 0.04446742502585315,
      "grad_norm": 3.509751245460736,
      "kl": 0.0074462890625,
      "learning_rate": 9.953525850034856e-07,
      "loss": 0.0003,
      "num_tokens": 3520448.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.3871031701564789,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.6672325730323792,
      "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": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 160.5,
      "completions/mean_terminated_length": 160.5,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.045501551189245086,
      "grad_norm": 3.847890323513219,
      "kl": 0.006622314453125,
      "learning_rate": 9.951290092358645e-07,
      "loss": 0.0003,
      "num_tokens": 3603924.0,
      "reward": 0.9652777910232544,
      "reward_std": 0.46292293071746826,
      "rewards/reasoning_reward/mean": 0.9652777314186096,
      "rewards/reasoning_reward/std": 0.7121769785881042,
      "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": 295.0,
      "completions/max_terminated_length": 295.0,
      "completions/mean_length": 146.45834350585938,
      "completions/mean_terminated_length": 146.45834350585938,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.04653567735263702,
      "grad_norm": 3.670364425188319,
      "kl": 0.0072021484375,
      "learning_rate": 9.949002075238139e-07,
      "loss": 0.0003,
      "num_tokens": 3684711.0,
      "reward": 0.875,
      "reward_std": 0.4082186222076416,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.39699962735176086,
      "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": 229.0,
      "completions/max_terminated_length": 229.0,
      "completions/mean_length": 155.08334350585938,
      "completions/mean_terminated_length": 155.08334350585938,
      "completions/min_length": 120.0,
      "completions/min_terminated_length": 120.0,
      "epoch": 0.047569803516028956,
      "grad_norm": 3.726045501753792,
      "kl": 0.0076904296875,
      "learning_rate": 9.9466618228227e-07,
      "loss": 0.0003,
      "num_tokens": 3762713.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "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": 394.0,
      "completions/max_terminated_length": 394.0,
      "completions/mean_length": 232.375,
      "completions/mean_terminated_length": 232.375,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.04860392967942089,
      "grad_norm": 3.592630193734655,
      "kl": 0.00811767578125,
      "learning_rate": 9.944269359813026e-07,
      "loss": 0.0003,
      "num_tokens": 3840002.0,
      "reward": 1.027777910232544,
      "reward_std": 0.40778279304504395,
      "rewards/reasoning_reward/mean": 1.0277777910232544,
      "rewards/reasoning_reward/std": 0.5972292423248291,
      "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": 264.0,
      "completions/max_terminated_length": 264.0,
      "completions/mean_length": 180.1666717529297,
      "completions/mean_terminated_length": 180.1666717529297,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.04963805584281282,
      "grad_norm": 4.29924922955248,
      "kl": 0.00848388671875,
      "learning_rate": 9.941824711460871e-07,
      "loss": 0.0003,
      "num_tokens": 3922982.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.48678088188171387,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.7019641399383545,
      "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": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 148.5,
      "completions/mean_terminated_length": 148.5,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.050672182006204755,
      "grad_norm": 2.8310868538516525,
      "kl": 0.0054931640625,
      "learning_rate": 9.9393279035688e-07,
      "loss": 0.0002,
      "num_tokens": 4001698.0,
      "reward": 0.625,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "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": 327.0,
      "completions/max_terminated_length": 327.0,
      "completions/mean_length": 204.9166717529297,
      "completions/mean_terminated_length": 204.9166717529297,
      "completions/min_length": 131.0,
      "completions/min_terminated_length": 131.0,
      "epoch": 0.05170630816959669,
      "grad_norm": 3.525954172380883,
      "kl": 0.0079345703125,
      "learning_rate": 9.936778962489902e-07,
      "loss": 0.0003,
      "num_tokens": 4081344.0,
      "reward": 0.4652777910232544,
      "reward_std": 0.31936562061309814,
      "rewards/reasoning_reward/mean": 0.4652777910232544,
      "rewards/reasoning_reward/std": 0.442268967628479,
      "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": 324.0,
      "completions/max_terminated_length": 324.0,
      "completions/mean_length": 159.5,
      "completions/mean_terminated_length": 159.5,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.052740434332988625,
      "grad_norm": 4.351659425272402,
      "kl": 0.01177978515625,
      "learning_rate": 9.934177915127515e-07,
      "loss": 0.0005,
      "num_tokens": 4169460.0,
      "reward": 1.125,
      "reward_std": 0.4082186818122864,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.8501917719841003,
      "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": 352.0,
      "completions/max_terminated_length": 352.0,
      "completions/mean_length": 186.08334350585938,
      "completions/mean_terminated_length": 186.08334350585938,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.05377456049638056,
      "grad_norm": 3.5125629537492675,
      "kl": 0.00982666015625,
      "learning_rate": 9.931524788934949e-07,
      "loss": 0.0004,
      "num_tokens": 4252750.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.5090917348861694,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.7210599780082703,
      "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": 416.0,
      "completions/max_terminated_length": 416.0,
      "completions/mean_length": 218.75,
      "completions/mean_terminated_length": 218.75,
      "completions/min_length": 137.0,
      "completions/min_terminated_length": 137.0,
      "epoch": 0.05480868665977249,
      "grad_norm": 3.193225562666632,
      "kl": 0.01190185546875,
      "learning_rate": 9.928819611915188e-07,
      "loss": 0.0005,
      "num_tokens": 4330064.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.20693820714950562,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.7058246731758118,
      "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": 368.0,
      "completions/max_terminated_length": 368.0,
      "completions/mean_length": 200.625,
      "completions/mean_terminated_length": 200.625,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.055842812823164424,
      "grad_norm": 2.9953108717900974,
      "kl": 0.01116943359375,
      "learning_rate": 9.9260624126206e-07,
      "loss": 0.0004,
      "num_tokens": 4415991.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "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": 225.0,
      "completions/max_terminated_length": 225.0,
      "completions/mean_length": 176.9166717529297,
      "completions/mean_terminated_length": 176.9166717529297,
      "completions/min_length": 132.0,
      "completions/min_terminated_length": 132.0,
      "epoch": 0.05687693898655636,
      "grad_norm": 3.5740247241093708,
      "kl": 0.01312255859375,
      "learning_rate": 9.923253220152627e-07,
      "loss": 0.0005,
      "num_tokens": 4498133.0,
      "reward": 0.6458333730697632,
      "reward_std": 0.3310800790786743,
      "rewards/reasoning_reward/mean": 0.6458333134651184,
      "rewards/reasoning_reward/std": 0.5610387921333313,
      "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": 392.0,
      "completions/max_terminated_length": 392.0,
      "completions/mean_length": 216.0416717529297,
      "completions/mean_terminated_length": 216.0416717529297,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.057911065149948295,
      "grad_norm": 4.117392785596258,
      "kl": 0.01153564453125,
      "learning_rate": 9.92039206416149e-07,
      "loss": 0.0005,
      "num_tokens": 4576190.0,
      "reward": 0.4583333432674408,
      "reward_std": 0.3574431836605072,
      "rewards/reasoning_reward/mean": 0.4583333432674408,
      "rewards/reasoning_reward/std": 0.4148510992527008,
      "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": 224.0,
      "completions/max_terminated_length": 224.0,
      "completions/mean_length": 153.83334350585938,
      "completions/mean_terminated_length": 153.83334350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.05894519131334023,
      "grad_norm": 4.837645018388818,
      "kl": 0.00958251953125,
      "learning_rate": 9.917478974845873e-07,
      "loss": 0.0004,
      "num_tokens": 4653658.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.46288391947746277,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "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": 307.0,
      "completions/max_terminated_length": 307.0,
      "completions/mean_length": 221.6666717529297,
      "completions/mean_terminated_length": 221.6666717529297,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.05997931747673216,
      "grad_norm": 3.7641592955112637,
      "kl": 0.011962890625,
      "learning_rate": 9.914513982952592e-07,
      "loss": 0.0005,
      "num_tokens": 4733554.0,
      "reward": 0.5902777910232544,
      "reward_std": 0.5594554543495178,
      "rewards/reasoning_reward/mean": 0.5902777910232544,
      "rewards/reasoning_reward/std": 0.5646151304244995,
      "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": 366.0,
      "completions/max_terminated_length": 366.0,
      "completions/mean_length": 196.375,
      "completions/mean_terminated_length": 196.375,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.06101344364012409,
      "grad_norm": 3.2142277856073393,
      "kl": 0.01385498046875,
      "learning_rate": 9.911497119776286e-07,
      "loss": 0.0006,
      "num_tokens": 4819339.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.34730279445648193,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.43353530764579773,
      "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": 316.0,
      "completions/max_terminated_length": 316.0,
      "completions/mean_length": 223.70834350585938,
      "completions/mean_terminated_length": 223.70834350585938,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.06204756980351603,
      "grad_norm": 3.6323983394895354,
      "kl": 0.01446533203125,
      "learning_rate": 9.908428417159078e-07,
      "loss": 0.0006,
      "num_tokens": 4902092.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.4685417115688324,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.7613869905471802,
      "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": 413.0,
      "completions/max_terminated_length": 413.0,
      "completions/mean_length": 210.08334350585938,
      "completions/mean_terminated_length": 210.08334350585938,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.06308169596690796,
      "grad_norm": 4.256974764936185,
      "kl": 0.01409912109375,
      "learning_rate": 9.905307907490242e-07,
      "loss": 0.0006,
      "num_tokens": 4982646.0,
      "reward": 0.7986111044883728,
      "reward_std": 0.7822409868240356,
      "rewards/reasoning_reward/mean": 0.7986111044883728,
      "rewards/reasoning_reward/std": 0.800028920173645,
      "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": 202.0,
      "completions/max_terminated_length": 202.0,
      "completions/mean_length": 125.83333587646484,
      "completions/mean_terminated_length": 125.83333587646484,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.0641158221302999,
      "grad_norm": 3.6447804056269857,
      "kl": 0.0086669921875,
      "learning_rate": 9.902135623705864e-07,
      "loss": 0.0003,
      "num_tokens": 5061682.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "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": 291.0,
      "completions/max_terminated_length": 291.0,
      "completions/mean_length": 194.0,
      "completions/mean_terminated_length": 194.0,
      "completions/min_length": 119.0,
      "completions/min_terminated_length": 119.0,
      "epoch": 0.06514994829369183,
      "grad_norm": 2.2300859017750354,
      "kl": 0.01177978515625,
      "learning_rate": 9.898911599288483e-07,
      "loss": 0.0005,
      "num_tokens": 5146090.0,
      "reward": 1.25,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.6079187393188477,
      "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": 337.0,
      "completions/max_terminated_length": 337.0,
      "completions/mean_length": 195.25,
      "completions/mean_terminated_length": 195.25,
      "completions/min_length": 127.0,
      "completions/min_terminated_length": 127.0,
      "epoch": 0.06618407445708377,
      "grad_norm": 3.6675785111838213,
      "kl": 0.0224609375,
      "learning_rate": 9.895635868266754e-07,
      "loss": 0.0009,
      "num_tokens": 5224032.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.3493061661720276,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.4340573847293854,
      "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": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 168.0,
      "completions/mean_terminated_length": 168.0,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.0672182006204757,
      "grad_norm": 3.434155813310191,
      "kl": 0.00909423828125,
      "learning_rate": 9.892308465215079e-07,
      "loss": 0.0004,
      "num_tokens": 5303200.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.34503278136253357,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "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": 370.0,
      "completions/max_terminated_length": 370.0,
      "completions/mean_length": 196.0,
      "completions/mean_terminated_length": 196.0,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.06825232678386763,
      "grad_norm": 2.512958656172137,
      "kl": 0.01190185546875,
      "learning_rate": 9.888929425253235e-07,
      "loss": 0.0005,
      "num_tokens": 5378760.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "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": 297.0,
      "completions/max_terminated_length": 297.0,
      "completions/mean_length": 164.5416717529297,
      "completions/mean_terminated_length": 164.5416717529297,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.06928645294725956,
      "grad_norm": 3.4490407712658686,
      "kl": 0.0157470703125,
      "learning_rate": 9.885498784046023e-07,
      "loss": 0.0006,
      "num_tokens": 5468669.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.31285393238067627,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.4402732849121094,
      "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": 327.0,
      "completions/max_terminated_length": 327.0,
      "completions/mean_length": 166.45834350585938,
      "completions/mean_terminated_length": 166.45834350585938,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.0703205791106515,
      "grad_norm": 3.4607132042836923,
      "kl": 0.01129150390625,
      "learning_rate": 9.882016577802873e-07,
      "loss": 0.0005,
      "num_tokens": 5547952.0,
      "reward": 0.8472222089767456,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.8472221493721008,
      "rewards/reasoning_reward/std": 0.7221757173538208,
      "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": 221.0,
      "completions/max_terminated_length": 221.0,
      "completions/mean_length": 151.7916717529297,
      "completions/mean_terminated_length": 151.7916717529297,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.07135470527404343,
      "grad_norm": 2.7650585026718284,
      "kl": 0.00799560546875,
      "learning_rate": 9.878482843277468e-07,
      "loss": 0.0003,
      "num_tokens": 5627683.0,
      "reward": 0.75,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "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": 311.0,
      "completions/max_terminated_length": 311.0,
      "completions/mean_length": 210.75,
      "completions/mean_terminated_length": 210.75,
      "completions/min_length": 134.0,
      "completions/min_terminated_length": 134.0,
      "epoch": 0.07238883143743537,
      "grad_norm": 4.449011636323865,
      "kl": 0.01300048828125,
      "learning_rate": 9.874897617767367e-07,
      "loss": 0.0005,
      "num_tokens": 5707485.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.3535533845424652,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "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": 0.04166666666666663,
      "completions/max_length": 459.0,
      "completions/max_terminated_length": 390.0,
      "completions/mean_length": 268.29168701171875,
      "completions/mean_terminated_length": 260.0,
      "completions/min_length": 135.0,
      "completions/min_terminated_length": 135.0,
      "epoch": 0.0734229576008273,
      "grad_norm": 3.5424047636885625,
      "kl": 0.014892578125,
      "learning_rate": 9.871260939113595e-07,
      "loss": 0.0006,
      "num_tokens": 5787164.0,
      "reward": 1.2708333730697632,
      "reward_std": 0.3915778398513794,
      "rewards/reasoning_reward/mean": 1.2708333730697632,
      "rewards/reasoning_reward/std": 0.48107290267944336,
      "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": 192.0,
      "completions/max_terminated_length": 192.0,
      "completions/mean_length": 118.54167175292969,
      "completions/mean_terminated_length": 118.54167175292969,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.07445708376421924,
      "grad_norm": 3.340246386770097,
      "kl": 0.00933837890625,
      "learning_rate": 9.867572845700245e-07,
      "loss": 0.0004,
      "num_tokens": 5866153.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "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": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 169.08334350585938,
      "completions/mean_terminated_length": 169.08334350585938,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.07549120992761117,
      "grad_norm": 4.82302140773541,
      "kl": 0.01165771484375,
      "learning_rate": 9.863833376454086e-07,
      "loss": 0.0005,
      "num_tokens": 5951243.0,
      "reward": 0.8402777910232544,
      "reward_std": 0.5060732960700989,
      "rewards/reasoning_reward/mean": 0.8402777314186096,
      "rewards/reasoning_reward/std": 0.5349594354629517,
      "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": 362.0,
      "completions/max_terminated_length": 362.0,
      "completions/mean_length": 183.58334350585938,
      "completions/mean_terminated_length": 183.58334350585938,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.07652533609100311,
      "grad_norm": 2.6240055600050596,
      "kl": 0.017578125,
      "learning_rate": 9.86004257084414e-07,
      "loss": 0.0007,
      "num_tokens": 6027409.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.24339044094085693,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.5247498154640198,
      "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": 303.0,
      "completions/max_terminated_length": 303.0,
      "completions/mean_length": 150.0416717529297,
      "completions/mean_terminated_length": 150.0416717529297,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.07755946225439504,
      "grad_norm": 3.592297508316741,
      "kl": 0.01373291015625,
      "learning_rate": 9.856200468881274e-07,
      "loss": 0.0006,
      "num_tokens": 6110490.0,
      "reward": 1.0,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.7223151326179504,
      "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": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 154.4166717529297,
      "completions/mean_terminated_length": 154.4166717529297,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.07859358841778696,
      "grad_norm": 9.783346222849755,
      "kl": 0.01300048828125,
      "learning_rate": 9.85230711111777e-07,
      "loss": 0.0005,
      "num_tokens": 6188476.0,
      "reward": 0.75,
      "reward_std": 0.5423438549041748,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.6079187393188477,
      "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": 370.0,
      "completions/max_terminated_length": 370.0,
      "completions/mean_length": 180.83334350585938,
      "completions/mean_terminated_length": 180.83334350585938,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.0796277145811789,
      "grad_norm": 4.348565558470908,
      "kl": 0.0186767578125,
      "learning_rate": 9.848362538646898e-07,
      "loss": 0.0007,
      "num_tokens": 6272632.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.33723291754722595,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.4871537983417511,
      "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": 327.0,
      "completions/max_terminated_length": 327.0,
      "completions/mean_length": 213.70834350585938,
      "completions/mean_terminated_length": 213.70834350585938,
      "completions/min_length": 136.0,
      "completions/min_terminated_length": 136.0,
      "epoch": 0.08066184074457083,
      "grad_norm": 3.2258066953920532,
      "kl": 0.01544189453125,
      "learning_rate": 9.844366793102487e-07,
      "loss": 0.0006,
      "num_tokens": 6361513.0,
      "reward": 1.3958333730697632,
      "reward_std": 0.13607725501060486,
      "rewards/reasoning_reward/mean": 1.3958333730697632,
      "rewards/reasoning_reward/std": 0.32900264859199524,
      "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": 275.0,
      "completions/max_terminated_length": 275.0,
      "completions/mean_length": 173.08334350585938,
      "completions/mean_terminated_length": 173.08334350585938,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.08169596690796277,
      "grad_norm": 3.863377237898613,
      "kl": 0.01416015625,
      "learning_rate": 9.840319916658487e-07,
      "loss": 0.0006,
      "num_tokens": 6446875.0,
      "reward": 0.868055522441864,
      "reward_std": 0.4124408960342407,
      "rewards/reasoning_reward/mean": 0.868055522441864,
      "rewards/reasoning_reward/std": 0.5793948769569397,
      "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": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 166.0416717529297,
      "completions/mean_terminated_length": 166.0416717529297,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.0827300930713547,
      "grad_norm": 3.9380960605825535,
      "kl": 0.0107421875,
      "learning_rate": 9.836221952028512e-07,
      "loss": 0.0004,
      "num_tokens": 6526732.0,
      "reward": 0.4166666865348816,
      "reward_std": 0.46854168176651,
      "rewards/reasoning_reward/mean": 0.4166666567325592,
      "rewards/reasoning_reward/std": 0.5036101937294006,
      "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": 318.0,
      "completions/max_terminated_length": 318.0,
      "completions/mean_length": 203.2916717529297,
      "completions/mean_terminated_length": 203.2916717529297,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.08376421923474664,
      "grad_norm": 2.968433728304001,
      "kl": 0.013427734375,
      "learning_rate": 9.832072942465403e-07,
      "loss": 0.0005,
      "num_tokens": 6613659.0,
      "reward": 1.0,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.8340576887130737,
      "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": 339.0,
      "completions/max_terminated_length": 339.0,
      "completions/mean_length": 216.5,
      "completions/mean_terminated_length": 216.5,
      "completions/min_length": 121.0,
      "completions/min_terminated_length": 121.0,
      "epoch": 0.08479834539813857,
      "grad_norm": 4.153184905003268,
      "kl": 0.01348876953125,
      "learning_rate": 9.827872931760762e-07,
      "loss": 0.0005,
      "num_tokens": 6692039.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.42645785212516785,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.7506036162376404,
      "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": 429.0,
      "completions/max_terminated_length": 429.0,
      "completions/mean_length": 270.625,
      "completions/mean_terminated_length": 270.625,
      "completions/min_length": 144.0,
      "completions/min_terminated_length": 144.0,
      "epoch": 0.08583247156153051,
      "grad_norm": 3.143774778506864,
      "kl": 0.01519775390625,
      "learning_rate": 9.823621964244499e-07,
      "loss": 0.0006,
      "num_tokens": 6776582.0,
      "reward": 0.75,
      "reward_std": 0.4629100561141968,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.48900964856147766,
      "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": 257.0,
      "completions/max_terminated_length": 257.0,
      "completions/mean_length": 163.75,
      "completions/mean_terminated_length": 163.75,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.08686659772492245,
      "grad_norm": 3.829881249796472,
      "kl": 0.01422119140625,
      "learning_rate": 9.81932008478435e-07,
      "loss": 0.0006,
      "num_tokens": 6853528.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.4563409090042114,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.4871538281440735,
      "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": 360.0,
      "completions/max_terminated_length": 360.0,
      "completions/mean_length": 198.33334350585938,
      "completions/mean_terminated_length": 198.33334350585938,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.08790072388831438,
      "grad_norm": 4.080367426621312,
      "kl": 0.01171875,
      "learning_rate": 9.814967338785423e-07,
      "loss": 0.0005,
      "num_tokens": 6934072.0,
      "reward": 0.8125,
      "reward_std": 0.5672780275344849,
      "rewards/reasoning_reward/mean": 0.8125,
      "rewards/reasoning_reward/std": 0.6280721426010132,
      "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": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 152.08334350585938,
      "completions/mean_terminated_length": 152.08334350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.0889348500517063,
      "grad_norm": 2.3347185584192713,
      "kl": 0.01202392578125,
      "learning_rate": 9.810563772189695e-07,
      "loss": 0.0005,
      "num_tokens": 7012290.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.20412415266036987,
      "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": 0.04166666666666663,
      "completions/max_length": 442.0,
      "completions/max_terminated_length": 442.0,
      "completions/mean_length": 231.75,
      "completions/mean_terminated_length": 228.3478240966797,
      "completions/min_length": 130.0,
      "completions/min_terminated_length": 130.0,
      "epoch": 0.08996897621509824,
      "grad_norm": 2.48983908009315,
      "kl": 0.01373291015625,
      "learning_rate": 9.806109431475548e-07,
      "loss": 0.0006,
      "num_tokens": 7092732.0,
      "reward": 0.875,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "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": 310.0,
      "completions/max_terminated_length": 310.0,
      "completions/mean_length": 198.5416717529297,
      "completions/mean_terminated_length": 198.5416717529297,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.09100310237849017,
      "grad_norm": 3.8019105243494167,
      "kl": 0.01519775390625,
      "learning_rate": 9.80160436365727e-07,
      "loss": 0.0006,
      "num_tokens": 7170505.0,
      "reward": 0.4791666865348816,
      "reward_std": 0.5551774501800537,
      "rewards/reasoning_reward/mean": 0.4791666567325592,
      "rewards/reasoning_reward/std": 0.5800893306732178,
      "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": 346.0,
      "completions/max_terminated_length": 346.0,
      "completions/mean_length": 198.6666717529297,
      "completions/mean_terminated_length": 198.6666717529297,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.09203722854188211,
      "grad_norm": 4.310499510741145,
      "kl": 0.0174560546875,
      "learning_rate": 9.797048616284557e-07,
      "loss": 0.0007,
      "num_tokens": 7256361.0,
      "reward": 0.7986111044883728,
      "reward_std": 0.5306869745254517,
      "rewards/reasoning_reward/mean": 0.7986111044883728,
      "rewards/reasoning_reward/std": 0.6984785795211792,
      "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": 0.04166666666666663,
      "completions/max_length": 270.0,
      "completions/max_terminated_length": 270.0,
      "completions/mean_length": 169.45834350585938,
      "completions/mean_terminated_length": 168.86956787109375,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.09307135470527404,
      "grad_norm": 3.256888913160928,
      "kl": 0.01373291015625,
      "learning_rate": 9.792442237442013e-07,
      "loss": 0.0006,
      "num_tokens": 7333220.0,
      "reward": 0.75,
      "reward_std": 0.33247750997543335,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "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": 461.0,
      "completions/max_terminated_length": 461.0,
      "completions/mean_length": 246.45834350585938,
      "completions/mean_terminated_length": 246.45834350585938,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.09410548086866598,
      "grad_norm": 4.176429595136656,
      "kl": 0.0230712890625,
      "learning_rate": 9.787785275748643e-07,
      "loss": 0.0009,
      "num_tokens": 7413559.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.6453183889389038,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.7220014929771423,
      "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": 0.04166666666666663,
      "completions/max_length": 331.0,
      "completions/max_terminated_length": 331.0,
      "completions/mean_length": 186.70834350585938,
      "completions/mean_terminated_length": 182.6521759033203,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.09513960703205791,
      "grad_norm": 4.0765925380672225,
      "kl": 0.0146484375,
      "learning_rate": 9.783077780357338e-07,
      "loss": 0.0006,
      "num_tokens": 7494080.0,
      "reward": 0.680555522441864,
      "reward_std": 0.5991425514221191,
      "rewards/reasoning_reward/mean": 0.680555522441864,
      "rewards/reasoning_reward/std": 0.6077531576156616,
      "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": 345.0,
      "completions/max_terminated_length": 345.0,
      "completions/mean_length": 210.75,
      "completions/mean_terminated_length": 210.75,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.09617373319544985,
      "grad_norm": 3.060802289470635,
      "kl": 0.016845703125,
      "learning_rate": 9.778319800954364e-07,
      "loss": 0.0007,
      "num_tokens": 7577778.0,
      "reward": 0.9236111044883728,
      "reward_std": 0.4386470317840576,
      "rewards/reasoning_reward/mean": 0.9236111044883728,
      "rewards/reasoning_reward/std": 0.533829391002655,
      "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": 262.0,
      "completions/max_terminated_length": 262.0,
      "completions/mean_length": 172.45834350585938,
      "completions/mean_terminated_length": 172.45834350585938,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.09720785935884178,
      "grad_norm": 5.03422567888706,
      "kl": 0.01513671875,
      "learning_rate": 9.773511387758821e-07,
      "loss": 0.0006,
      "num_tokens": 7656581.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.4434394836425781,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.878229022026062,
      "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": 317.0,
      "completions/max_terminated_length": 317.0,
      "completions/mean_length": 179.9166717529297,
      "completions/mean_terminated_length": 179.9166717529297,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.09824198552223372,
      "grad_norm": 3.3926662846097773,
      "kl": 0.01263427734375,
      "learning_rate": 9.768652591522133e-07,
      "loss": 0.0005,
      "num_tokens": 7735547.0,
      "reward": 0.75,
      "reward_std": 0.33247750997543335,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "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": 369.0,
      "completions/max_terminated_length": 369.0,
      "completions/mean_length": 217.5416717529297,
      "completions/mean_terminated_length": 217.5416717529297,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.09927611168562564,
      "grad_norm": 4.22083012756532,
      "kl": 0.0191650390625,
      "learning_rate": 9.763743463527496e-07,
      "loss": 0.0008,
      "num_tokens": 7819328.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.45032867789268494,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.5500329732894897,
      "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": 247.0,
      "completions/max_terminated_length": 247.0,
      "completions/mean_length": 165.375,
      "completions/mean_terminated_length": 165.375,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.10031023784901758,
      "grad_norm": 3.7501772543377743,
      "kl": 0.0177001953125,
      "learning_rate": 9.758784055589346e-07,
      "loss": 0.0007,
      "num_tokens": 7897641.0,
      "reward": 0.625,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "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": 322.0,
      "completions/max_terminated_length": 322.0,
      "completions/mean_length": 172.25,
      "completions/mean_terminated_length": 172.25,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.10134436401240951,
      "grad_norm": 4.3294563512717215,
      "kl": 0.0191650390625,
      "learning_rate": 9.753774420052807e-07,
      "loss": 0.0008,
      "num_tokens": 7974623.0,
      "reward": 0.8958333730697632,
      "reward_std": 0.4929513931274414,
      "rewards/reasoning_reward/mean": 0.8958333134651184,
      "rewards/reasoning_reward/std": 0.5311834216117859,
      "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": 353.0,
      "completions/max_terminated_length": 353.0,
      "completions/mean_length": 194.25,
      "completions/mean_terminated_length": 194.25,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.10237849017580145,
      "grad_norm": 4.1235008690948165,
      "kl": 0.0220947265625,
      "learning_rate": 9.748714609793147e-07,
      "loss": 0.0009,
      "num_tokens": 8054901.0,
      "reward": 0.4166666865348816,
      "reward_std": 0.5099153518676758,
      "rewards/reasoning_reward/mean": 0.4166666567325592,
      "rewards/reasoning_reward/std": 0.524749755859375,
      "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": 323.0,
      "completions/max_terminated_length": 323.0,
      "completions/mean_length": 205.58334350585938,
      "completions/mean_terminated_length": 205.58334350585938,
      "completions/min_length": 120.0,
      "completions/min_terminated_length": 120.0,
      "epoch": 0.10341261633919338,
      "grad_norm": 3.113383684598118,
      "kl": 0.01806640625,
      "learning_rate": 9.743604678215205e-07,
      "loss": 0.0007,
      "num_tokens": 8136203.0,
      "reward": 0.7777777910232544,
      "reward_std": 0.3232055902481079,
      "rewards/reasoning_reward/mean": 0.7777777314186096,
      "rewards/reasoning_reward/std": 0.8145219683647156,
      "step": 100
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 333.0,
      "completions/max_terminated_length": 333.0,
      "completions/mean_length": 202.125,
      "completions/mean_terminated_length": 202.125,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.10444674250258532,
      "grad_norm": 3.7183447125419007,
      "kl": 0.0211181640625,
      "learning_rate": 9.738444679252843e-07,
      "loss": 0.0008,
      "num_tokens": 8215654.0,
      "reward": 0.4791666865348816,
      "reward_std": 0.4671442210674286,
      "rewards/reasoning_reward/mean": 0.4791666567325592,
      "rewards/reasoning_reward/std": 0.5413181781768799,
      "step": 101
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 288.0,
      "completions/max_terminated_length": 288.0,
      "completions/mean_length": 176.625,
      "completions/mean_terminated_length": 176.625,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.10548086866597725,
      "grad_norm": 3.1873320574218478,
      "kl": 0.02392578125,
      "learning_rate": 9.733234667368368e-07,
      "loss": 0.001,
      "num_tokens": 8304221.0,
      "reward": 1.1388888359069824,
      "reward_std": 0.24966806173324585,
      "rewards/reasoning_reward/mean": 1.1388888359069824,
      "rewards/reasoning_reward/std": 0.5353825092315674,
      "step": 102
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 481.0,
      "completions/max_terminated_length": 481.0,
      "completions/mean_length": 253.0,
      "completions/mean_terminated_length": 253.0,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.10651499482936919,
      "grad_norm": 3.681506481032909,
      "kl": 0.016357421875,
      "learning_rate": 9.727974697551958e-07,
      "loss": 0.0007,
      "num_tokens": 8384285.0,
      "reward": 0.4236111044883728,
      "reward_std": 0.47574663162231445,
      "rewards/reasoning_reward/mean": 0.4236110746860504,
      "rewards/reasoning_reward/std": 0.49874040484428406,
      "step": 103
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 357.0,
      "completions/max_terminated_length": 357.0,
      "completions/mean_length": 239.0416717529297,
      "completions/mean_terminated_length": 239.0416717529297,
      "completions/min_length": 170.0,
      "completions/min_terminated_length": 170.0,
      "epoch": 0.10754912099276112,
      "grad_norm": 2.7098677939554507,
      "kl": 0.0281982421875,
      "learning_rate": 9.722664825321082e-07,
      "loss": 0.0011,
      "num_tokens": 8481646.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.1498909741640091,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.6004225611686707,
      "step": 104
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 371.0,
      "completions/max_terminated_length": 371.0,
      "completions/mean_length": 191.625,
      "completions/mean_terminated_length": 191.625,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.10858324715615306,
      "grad_norm": 3.7431786732254246,
      "kl": 0.0157470703125,
      "learning_rate": 9.717305106719916e-07,
      "loss": 0.0006,
      "num_tokens": 8560781.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.42645785212516785,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.5500329732894897,
      "step": 105
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 301.0,
      "completions/max_terminated_length": 301.0,
      "completions/mean_length": 186.20834350585938,
      "completions/mean_terminated_length": 186.20834350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.10961737331954498,
      "grad_norm": 3.620350204008683,
      "kl": 0.033447265625,
      "learning_rate": 9.71189559831875e-07,
      "loss": 0.0013,
      "num_tokens": 8650034.0,
      "reward": 1.6875,
      "reward_std": 0.33768826723098755,
      "rewards/reasoning_reward/mean": 1.6875,
      "rewards/reasoning_reward/std": 0.4618605971336365,
      "step": 106
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 373.0,
      "completions/max_terminated_length": 373.0,
      "completions/mean_length": 187.70834350585938,
      "completions/mean_terminated_length": 187.70834350585938,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.11065149948293691,
      "grad_norm": 2.325484346203483,
      "kl": 0.0216064453125,
      "learning_rate": 9.70643635721339e-07,
      "loss": 0.0009,
      "num_tokens": 8727995.0,
      "reward": 1.0625,
      "reward_std": 0.19795583188533783,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.3398369252681732,
      "step": 107
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 277.0,
      "completions/max_terminated_length": 277.0,
      "completions/mean_length": 206.5416717529297,
      "completions/mean_terminated_length": 206.5416717529297,
      "completions/min_length": 157.0,
      "completions/min_terminated_length": 157.0,
      "epoch": 0.11168562564632885,
      "grad_norm": 3.929289600039794,
      "kl": 0.0341796875,
      "learning_rate": 9.70092744102456e-07,
      "loss": 0.0014,
      "num_tokens": 8817472.0,
      "reward": 1.4305555820465088,
      "reward_std": 0.32358893752098083,
      "rewards/reasoning_reward/mean": 1.4305554628372192,
      "rewards/reasoning_reward/std": 0.5222504734992981,
      "step": 108
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 392.0,
      "completions/max_terminated_length": 392.0,
      "completions/mean_length": 221.375,
      "completions/mean_terminated_length": 221.375,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.11271975180972078,
      "grad_norm": 3.692465779722894,
      "kl": 0.0284423828125,
      "learning_rate": 9.695368907897286e-07,
      "loss": 0.0011,
      "num_tokens": 8910545.0,
      "reward": 1.3958333730697632,
      "reward_std": 0.31444376707077026,
      "rewards/reasoning_reward/mean": 1.3958333730697632,
      "rewards/reasoning_reward/std": 0.5243180394172668,
      "step": 109
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 309.0,
      "completions/max_terminated_length": 309.0,
      "completions/mean_length": 200.95834350585938,
      "completions/mean_terminated_length": 200.95834350585938,
      "completions/min_length": 137.0,
      "completions/min_terminated_length": 137.0,
      "epoch": 0.11375387797311272,
      "grad_norm": 3.73349102770315,
      "kl": 0.037109375,
      "learning_rate": 9.689760816500284e-07,
      "loss": 0.0015,
      "num_tokens": 9005792.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.31443244218826294,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 110
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 454.0,
      "completions/max_terminated_length": 454.0,
      "completions/mean_length": 228.95834350585938,
      "completions/mean_terminated_length": 228.95834350585938,
      "completions/min_length": 135.0,
      "completions/min_terminated_length": 135.0,
      "epoch": 0.11478800413650465,
      "grad_norm": 2.8820257320459106,
      "kl": 0.0201416015625,
      "learning_rate": 9.684103226025355e-07,
      "loss": 0.0008,
      "num_tokens": 9088655.0,
      "reward": 0.875,
      "reward_std": 0.2314550280570984,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.5565811395645142,
      "step": 111
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 355.0,
      "completions/max_terminated_length": 355.0,
      "completions/mean_length": 204.58334350585938,
      "completions/mean_terminated_length": 204.58334350585938,
      "completions/min_length": 149.0,
      "completions/min_terminated_length": 149.0,
      "epoch": 0.11582213029989659,
      "grad_norm": 1.9153457674664132,
      "kl": 0.0242919921875,
      "learning_rate": 9.678396196186738e-07,
      "loss": 0.001,
      "num_tokens": 9172229.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 112
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 331.0,
      "completions/max_terminated_length": 331.0,
      "completions/mean_length": 228.08334350585938,
      "completions/mean_terminated_length": 228.08334350585938,
      "completions/min_length": 129.0,
      "completions/min_terminated_length": 129.0,
      "epoch": 0.11685625646328852,
      "grad_norm": 4.091930192339665,
      "kl": 0.0234375,
      "learning_rate": 9.67263978722049e-07,
      "loss": 0.0009,
      "num_tokens": 9250847.0,
      "reward": 0.7361111044883728,
      "reward_std": 0.4770033359527588,
      "rewards/reasoning_reward/mean": 0.7361111044883728,
      "rewards/reasoning_reward/std": 0.581345796585083,
      "step": 113
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 287.0,
      "completions/max_terminated_length": 287.0,
      "completions/mean_length": 174.7916717529297,
      "completions/mean_terminated_length": 174.7916717529297,
      "completions/min_length": 122.0,
      "completions/min_terminated_length": 122.0,
      "epoch": 0.11789038262668046,
      "grad_norm": 3.8333319654298315,
      "kl": 0.0223388671875,
      "learning_rate": 9.666834059883856e-07,
      "loss": 0.0009,
      "num_tokens": 9336570.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.50337815284729,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.5706435441970825,
      "step": 114
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 357.0,
      "completions/max_terminated_length": 357.0,
      "completions/mean_length": 215.0,
      "completions/mean_terminated_length": 215.0,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.1189245087900724,
      "grad_norm": 4.4406021802940545,
      "kl": 0.0303955078125,
      "learning_rate": 9.66097907545462e-07,
      "loss": 0.0012,
      "num_tokens": 9416634.0,
      "reward": 0.8472222089767456,
      "reward_std": 0.5322877764701843,
      "rewards/reasoning_reward/mean": 0.8472221493721008,
      "rewards/reasoning_reward/std": 0.5644814372062683,
      "step": 115
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 209.0,
      "completions/max_terminated_length": 209.0,
      "completions/mean_length": 155.5416717529297,
      "completions/mean_terminated_length": 155.5416717529297,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.11995863495346432,
      "grad_norm": 2.161252175635768,
      "kl": 0.017822265625,
      "learning_rate": 9.655074895730462e-07,
      "loss": 0.0007,
      "num_tokens": 9501367.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.550032913684845,
      "step": 116
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 185.33334350585938,
      "completions/mean_terminated_length": 185.33334350585938,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.12099276111685625,
      "grad_norm": 4.450081255781044,
      "kl": 0.0238037109375,
      "learning_rate": 9.649121583028299e-07,
      "loss": 0.0009,
      "num_tokens": 9580399.0,
      "reward": 0.8125,
      "reward_std": 0.5099677443504333,
      "rewards/reasoning_reward/mean": 0.8125,
      "rewards/reasoning_reward/std": 0.5067479610443115,
      "step": 117
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 318.0,
      "completions/max_terminated_length": 318.0,
      "completions/mean_length": 211.25,
      "completions/mean_terminated_length": 211.25,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.12202688728024819,
      "grad_norm": 3.213840485912054,
      "kl": 0.025390625,
      "learning_rate": 9.643119200183637e-07,
      "loss": 0.001,
      "num_tokens": 9669941.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.24056154489517212,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.7144345045089722,
      "step": 118
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 260.0,
      "completions/max_terminated_length": 260.0,
      "completions/mean_length": 147.5,
      "completions/mean_terminated_length": 147.5,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.12306101344364012,
      "grad_norm": 3.122769680653565,
      "kl": 0.0184326171875,
      "learning_rate": 9.637067810549906e-07,
      "loss": 0.0007,
      "num_tokens": 9749513.0,
      "reward": 1.0555555820465088,
      "reward_std": 0.2238859385251999,
      "rewards/reasoning_reward/mean": 1.0555554628372192,
      "rewards/reasoning_reward/std": 0.31724458932876587,
      "step": 119
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 322.0,
      "completions/max_terminated_length": 322.0,
      "completions/mean_length": 178.7916717529297,
      "completions/mean_terminated_length": 178.7916717529297,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.12409513960703206,
      "grad_norm": 4.639227790783391,
      "kl": 0.02978515625,
      "learning_rate": 9.63096747799778e-07,
      "loss": 0.0012,
      "num_tokens": 9838452.0,
      "reward": 1.4583333730697632,
      "reward_std": 0.5078567266464233,
      "rewards/reasoning_reward/mean": 1.4583333730697632,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 120
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 131.83334350585938,
      "completions/mean_terminated_length": 131.83334350585938,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.125129265770424,
      "grad_norm": 2.4639138077233618,
      "kl": 0.0234375,
      "learning_rate": 9.624818266914519e-07,
      "loss": 0.0009,
      "num_tokens": 9916816.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.20412415266036987,
      "step": 121
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 389.0,
      "completions/max_terminated_length": 389.0,
      "completions/mean_length": 232.9166717529297,
      "completions/mean_terminated_length": 232.9166717529297,
      "completions/min_length": 141.0,
      "completions/min_terminated_length": 141.0,
      "epoch": 0.12616339193381593,
      "grad_norm": 3.8580416088238216,
      "kl": 0.0281982421875,
      "learning_rate": 9.618620242203278e-07,
      "loss": 0.0011,
      "num_tokens": 10002838.0,
      "reward": 1.1388888359069824,
      "reward_std": 0.3499924838542938,
      "rewards/reasoning_reward/mean": 1.1388888359069824,
      "rewards/reasoning_reward/std": 0.5870310664176941,
      "step": 122
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 282.0,
      "completions/max_terminated_length": 282.0,
      "completions/mean_length": 173.875,
      "completions/mean_terminated_length": 173.875,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.12719751809720786,
      "grad_norm": 2.83715772505631,
      "kl": 0.016845703125,
      "learning_rate": 9.612373469282428e-07,
      "loss": 0.0007,
      "num_tokens": 10081507.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3535533845424652,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.6580052971839905,
      "step": 123
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 199.0,
      "completions/max_terminated_length": 199.0,
      "completions/mean_length": 144.33334350585938,
      "completions/mean_terminated_length": 144.33334350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.1282316442605998,
      "grad_norm": 1.8523241506048544,
      "kl": 0.0299072265625,
      "learning_rate": 9.606078014084863e-07,
      "loss": 0.0012,
      "num_tokens": 10165467.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.19920477271080017,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.40824830532073975,
      "step": 124
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 153.875,
      "completions/mean_terminated_length": 153.875,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.12926577042399173,
      "grad_norm": 3.493963599295082,
      "kl": 0.03076171875,
      "learning_rate": 9.5997339430573e-07,
      "loss": 0.0012,
      "num_tokens": 10241968.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 125
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 153.33334350585938,
      "completions/mean_terminated_length": 153.33334350585938,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.13029989658738367,
      "grad_norm": 3.815467992926747,
      "kl": 0.02294921875,
      "learning_rate": 9.59334132315959e-07,
      "loss": 0.0009,
      "num_tokens": 10321768.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.4082186818122864,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 126
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 381.0,
      "completions/max_terminated_length": 381.0,
      "completions/mean_length": 184.9166717529297,
      "completions/mean_terminated_length": 184.9166717529297,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.1313340227507756,
      "grad_norm": 3.1667726238296843,
      "kl": 0.020751953125,
      "learning_rate": 9.586900221863996e-07,
      "loss": 0.0008,
      "num_tokens": 10400854.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.6100153923034668,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.6064269542694092,
      "step": 127
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 213.0,
      "completions/max_terminated_length": 213.0,
      "completions/mean_length": 164.0416717529297,
      "completions/mean_terminated_length": 164.0416717529297,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.13236814891416754,
      "grad_norm": 3.1674735991804623,
      "kl": 0.0181884765625,
      "learning_rate": 9.580410707154494e-07,
      "loss": 0.0007,
      "num_tokens": 10481695.0,
      "reward": 0.5625,
      "reward_std": 0.28302299976348877,
      "rewards/reasoning_reward/mean": 0.5625,
      "rewards/reasoning_reward/std": 0.49590715765953064,
      "step": 128
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 290.0,
      "completions/max_terminated_length": 290.0,
      "completions/mean_length": 159.0,
      "completions/mean_terminated_length": 159.0,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.13340227507755947,
      "grad_norm": 2.4728156183528083,
      "kl": 0.02294921875,
      "learning_rate": 9.573872847526048e-07,
      "loss": 0.0009,
      "num_tokens": 10561455.0,
      "reward": 1.0,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 129
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 286.0,
      "completions/max_terminated_length": 286.0,
      "completions/mean_length": 178.875,
      "completions/mean_terminated_length": 178.875,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.1344364012409514,
      "grad_norm": 4.017961732422752,
      "kl": 0.0291748046875,
      "learning_rate": 9.567286711983885e-07,
      "loss": 0.0012,
      "num_tokens": 10641076.0,
      "reward": 0.6736111044883728,
      "reward_std": 0.45231467485427856,
      "rewards/reasoning_reward/mean": 0.6736111044883728,
      "rewards/reasoning_reward/std": 0.5948230028152466,
      "step": 130
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 299.0,
      "completions/max_terminated_length": 299.0,
      "completions/mean_length": 200.9166717529297,
      "completions/mean_terminated_length": 198.9130401611328,
      "completions/min_length": 140.0,
      "completions/min_terminated_length": 140.0,
      "epoch": 0.13547052740434332,
      "grad_norm": 3.138708740315387,
      "kl": 0.0189208984375,
      "learning_rate": 9.560652370042771e-07,
      "loss": 0.0008,
      "num_tokens": 10719122.0,
      "reward": 1.0,
      "reward_std": 0.19500279426574707,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.2553769648075104,
      "step": 131
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 149.75,
      "completions/mean_terminated_length": 149.75,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.13650465356773525,
      "grad_norm": 3.840304727799911,
      "kl": 0.0203857421875,
      "learning_rate": 9.553969891726289e-07,
      "loss": 0.0008,
      "num_tokens": 10799388.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.4198887050151825,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "step": 132
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 272.0,
      "completions/max_terminated_length": 272.0,
      "completions/mean_length": 159.83334350585938,
      "completions/mean_terminated_length": 159.83334350585938,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.1375387797311272,
      "grad_norm": 1.7596178724801423,
      "kl": 0.01904296875,
      "learning_rate": 9.547239347566068e-07,
      "loss": 0.0008,
      "num_tokens": 10880672.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.20412415266036987,
      "step": 133
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 318.0,
      "completions/max_terminated_length": 318.0,
      "completions/mean_length": 188.58334350585938,
      "completions/mean_terminated_length": 188.58334350585938,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.13857290589451912,
      "grad_norm": 4.414440886337846,
      "kl": 0.0269775390625,
      "learning_rate": 9.540460808601069e-07,
      "loss": 0.0011,
      "num_tokens": 10958862.0,
      "reward": 1.625,
      "reward_std": 0.19801273941993713,
      "rewards/reasoning_reward/mean": 1.625,
      "rewards/reasoning_reward/std": 0.3686048984527588,
      "step": 134
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 188.0,
      "completions/max_terminated_length": 188.0,
      "completions/mean_length": 137.5,
      "completions/mean_terminated_length": 137.5,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.13960703205791106,
      "grad_norm": 0.18154064062767528,
      "kl": 0.0322265625,
      "learning_rate": 9.533634346376827e-07,
      "loss": 0.0013,
      "num_tokens": 11035034.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 135
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 235.0,
      "completions/max_terminated_length": 235.0,
      "completions/mean_length": 126.16667175292969,
      "completions/mean_terminated_length": 126.16667175292969,
      "completions/min_length": 71.0,
      "completions/min_terminated_length": 71.0,
      "epoch": 0.140641158221303,
      "grad_norm": 4.281841546042445,
      "kl": 0.01324462890625,
      "learning_rate": 9.526760032944687e-07,
      "loss": 0.0005,
      "num_tokens": 11114766.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.4082186818122864,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 136
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 210.0,
      "completions/max_terminated_length": 210.0,
      "completions/mean_length": 151.875,
      "completions/mean_terminated_length": 151.875,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.14167528438469493,
      "grad_norm": 3.112436935614297,
      "kl": 0.0211181640625,
      "learning_rate": 9.519837940861051e-07,
      "loss": 0.0008,
      "num_tokens": 11198851.0,
      "reward": 0.75,
      "reward_std": 0.34503278136253357,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 137
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 252.0,
      "completions/max_terminated_length": 252.0,
      "completions/mean_length": 164.375,
      "completions/mean_terminated_length": 164.375,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.14270941054808686,
      "grad_norm": 2.1893308302361865,
      "kl": 0.020751953125,
      "learning_rate": 9.512868143186614e-07,
      "loss": 0.0008,
      "num_tokens": 11274508.0,
      "reward": 0.375,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 138
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 263.0,
      "completions/max_terminated_length": 263.0,
      "completions/mean_length": 148.4166717529297,
      "completions/mean_terminated_length": 148.4166717529297,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.1437435367114788,
      "grad_norm": 1.9943583574276684,
      "kl": 0.03076171875,
      "learning_rate": 9.505850713485588e-07,
      "loss": 0.0012,
      "num_tokens": 11355774.0,
      "reward": 0.625,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 139
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 337.0,
      "completions/max_terminated_length": 337.0,
      "completions/mean_length": 177.58334350585938,
      "completions/mean_terminated_length": 177.58334350585938,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.14477766287487073,
      "grad_norm": 3.019726108756687,
      "kl": 0.0269775390625,
      "learning_rate": 9.498785725824927e-07,
      "loss": 0.0011,
      "num_tokens": 11432412.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4402732849121094,
      "step": 140
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 138.875,
      "completions/mean_terminated_length": 138.875,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.14581178903826267,
      "grad_norm": 3.4892636769097214,
      "kl": 0.0234375,
      "learning_rate": 9.491673254773544e-07,
      "loss": 0.0009,
      "num_tokens": 11508697.0,
      "reward": 0.4166666865348816,
      "reward_std": 0.40397143363952637,
      "rewards/reasoning_reward/mean": 0.4166666567325592,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "step": 141
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 261.0,
      "completions/max_terminated_length": 261.0,
      "completions/mean_length": 186.9166717529297,
      "completions/mean_terminated_length": 186.9166717529297,
      "completions/min_length": 149.0,
      "completions/min_terminated_length": 149.0,
      "epoch": 0.1468459152016546,
      "grad_norm": 2.4711223309041235,
      "kl": 0.0299072265625,
      "learning_rate": 9.484513375401531e-07,
      "loss": 0.0012,
      "num_tokens": 11589767.0,
      "reward": 0.8472222685813904,
      "reward_std": 0.3876555860042572,
      "rewards/reasoning_reward/mean": 0.8472222685813904,
      "rewards/reasoning_reward/std": 0.6444048285484314,
      "step": 142
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 137.4166717529297,
      "completions/mean_terminated_length": 137.4166717529297,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.14788004136504654,
      "grad_norm": 3.9836312904992175,
      "kl": 0.0179443359375,
      "learning_rate": 9.477306163279353e-07,
      "loss": 0.0007,
      "num_tokens": 11670673.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.43810173869132996,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.6064269542694092,
      "step": 143
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 382.0,
      "completions/max_terminated_length": 382.0,
      "completions/mean_length": 216.125,
      "completions/mean_terminated_length": 216.125,
      "completions/min_length": 141.0,
      "completions/min_terminated_length": 141.0,
      "epoch": 0.14891416752843847,
      "grad_norm": 3.084825540437868,
      "kl": 0.042724609375,
      "learning_rate": 9.470051694477066e-07,
      "loss": 0.0017,
      "num_tokens": 11757628.0,
      "reward": 0.868055522441864,
      "reward_std": 0.2675723433494568,
      "rewards/reasoning_reward/mean": 0.868055522441864,
      "rewards/reasoning_reward/std": 0.6331791877746582,
      "step": 144
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 187.0,
      "completions/max_terminated_length": 187.0,
      "completions/mean_length": 136.83334350585938,
      "completions/mean_terminated_length": 136.83334350585938,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.1499482936918304,
      "grad_norm": 3.603155546437747,
      "kl": 0.0255126953125,
      "learning_rate": 9.462750045563502e-07,
      "loss": 0.001,
      "num_tokens": 11835120.0,
      "reward": 0.7291666865348816,
      "reward_std": 0.33108004927635193,
      "rewards/reasoning_reward/mean": 0.7291666865348816,
      "rewards/reasoning_reward/std": 0.48854634165763855,
      "step": 145
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 170.5,
      "completions/mean_terminated_length": 170.5,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.15098241985522234,
      "grad_norm": 3.781720407335475,
      "kl": 0.060302734375,
      "learning_rate": 9.45540129360547e-07,
      "loss": 0.0024,
      "num_tokens": 11919284.0,
      "reward": 1.3541667461395264,
      "reward_std": 0.25392836332321167,
      "rewards/reasoning_reward/mean": 1.3541666269302368,
      "rewards/reasoning_reward/std": 0.4293363690376282,
      "step": 146
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 275.0,
      "completions/max_terminated_length": 275.0,
      "completions/mean_length": 159.2916717529297,
      "completions/mean_terminated_length": 159.2916717529297,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.15201654601861428,
      "grad_norm": 3.8293089763902644,
      "kl": 0.04296875,
      "learning_rate": 9.448005516166934e-07,
      "loss": 0.0017,
      "num_tokens": 12002627.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.45032867789268494,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.6064269542694092,
      "step": 147
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 245.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 172.5,
      "completions/mean_terminated_length": 169.3478240966797,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.15305067218200621,
      "grad_norm": 2.4437113547282188,
      "kl": 0.0250244140625,
      "learning_rate": 9.4405627913082e-07,
      "loss": 0.001,
      "num_tokens": 12082679.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 148
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 141.375,
      "completions/mean_terminated_length": 141.375,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.15408479834539815,
      "grad_norm": 3.1876345769180663,
      "kl": 0.036865234375,
      "learning_rate": 9.433073197585089e-07,
      "loss": 0.0015,
      "num_tokens": 12166192.0,
      "reward": 1.125,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.5366967916488647,
      "step": 149
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 265.0,
      "completions/max_terminated_length": 265.0,
      "completions/mean_length": 186.875,
      "completions/mean_terminated_length": 185.478271484375,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.15511892450879008,
      "grad_norm": 2.8956791515173244,
      "kl": 0.0299072265625,
      "learning_rate": 9.425536814048112e-07,
      "loss": 0.0012,
      "num_tokens": 12246877.0,
      "reward": 1.1875,
      "reward_std": 0.24052315950393677,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.4922405183315277,
      "step": 150
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 141.5,
      "completions/mean_terminated_length": 141.5,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.15615305067218202,
      "grad_norm": 3.578662637602958,
      "kl": 0.038330078125,
      "learning_rate": 9.417953720241633e-07,
      "loss": 0.0015,
      "num_tokens": 12331145.0,
      "reward": 0.875,
      "reward_std": 0.367926687002182,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.4484272003173828,
      "step": 151
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 281.0,
      "completions/max_terminated_length": 281.0,
      "completions/mean_length": 155.83334350585938,
      "completions/mean_terminated_length": 155.83334350585938,
      "completions/min_length": 114.0,
      "completions/min_terminated_length": 114.0,
      "epoch": 0.15718717683557393,
      "grad_norm": 2.4928642366111045,
      "kl": 0.040283203125,
      "learning_rate": 9.410323996203026e-07,
      "loss": 0.0016,
      "num_tokens": 12413805.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 152
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 279.0,
      "completions/max_terminated_length": 279.0,
      "completions/mean_length": 163.20834350585938,
      "completions/mean_terminated_length": 163.20834350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.15822130299896586,
      "grad_norm": 2.0203710118248086,
      "kl": 0.0213623046875,
      "learning_rate": 9.402647722461838e-07,
      "loss": 0.0009,
      "num_tokens": 12485866.0,
      "reward": 0.5,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.5,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 153
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 325.0,
      "completions/max_terminated_length": 325.0,
      "completions/mean_length": 182.1666717529297,
      "completions/mean_terminated_length": 182.1666717529297,
      "completions/min_length": 123.0,
      "completions/min_terminated_length": 123.0,
      "epoch": 0.1592554291623578,
      "grad_norm": 2.745126246673903,
      "kl": 0.02880859375,
      "learning_rate": 9.394924980038931e-07,
      "loss": 0.0012,
      "num_tokens": 12564518.0,
      "reward": 1.0347223281860352,
      "reward_std": 0.2400643527507782,
      "rewards/reasoning_reward/mean": 1.0347222089767456,
      "rewards/reasoning_reward/std": 0.3184320330619812,
      "step": 154
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 167.95834350585938,
      "completions/mean_terminated_length": 167.95834350585938,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.16028955532574973,
      "grad_norm": 3.2611636461773656,
      "kl": 0.01806640625,
      "learning_rate": 9.387155850445634e-07,
      "loss": 0.0007,
      "num_tokens": 12646309.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148510992527008,
      "step": 155
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 221.0,
      "completions/max_terminated_length": 221.0,
      "completions/mean_length": 162.9166717529297,
      "completions/mean_terminated_length": 162.9166717529297,
      "completions/min_length": 118.0,
      "completions/min_terminated_length": 118.0,
      "epoch": 0.16132368148914167,
      "grad_norm": 3.5046780041435315,
      "kl": 0.049072265625,
      "learning_rate": 9.379340415682877e-07,
      "loss": 0.002,
      "num_tokens": 12728907.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.3616904020309448,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.7755316495895386,
      "step": 156
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 214.0,
      "completions/max_terminated_length": 214.0,
      "completions/mean_length": 158.0416717529297,
      "completions/mean_terminated_length": 158.0416717529297,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.1623578076525336,
      "grad_norm": 3.0934531885922616,
      "kl": 0.049072265625,
      "learning_rate": 9.371478758240327e-07,
      "loss": 0.002,
      "num_tokens": 12819820.0,
      "reward": 1.2708333730697632,
      "reward_std": 0.21322892606258392,
      "rewards/reasoning_reward/mean": 1.2708333730697632,
      "rewards/reasoning_reward/std": 0.6075461506843567,
      "step": 157
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 149.75,
      "completions/mean_terminated_length": 149.75,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.16339193381592554,
      "grad_norm": 0.2825124812867749,
      "kl": 0.03076171875,
      "learning_rate": 9.363570961095522e-07,
      "loss": 0.0012,
      "num_tokens": 12902054.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 158
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 403.0,
      "completions/max_terminated_length": 403.0,
      "completions/mean_length": 188.375,
      "completions/mean_terminated_length": 188.375,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.16442605997931747,
      "grad_norm": 3.4620590744937343,
      "kl": 0.029296875,
      "learning_rate": 9.355617107712988e-07,
      "loss": 0.0012,
      "num_tokens": 12981463.0,
      "reward": 0.875,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.4484272003173828,
      "step": 159
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 405.0,
      "completions/max_terminated_length": 405.0,
      "completions/mean_length": 186.20834350585938,
      "completions/mean_terminated_length": 185.0,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.1654601861427094,
      "grad_norm": 3.67117533558172,
      "kl": 0.056396484375,
      "learning_rate": 9.347617282043361e-07,
      "loss": 0.0022,
      "num_tokens": 13061260.0,
      "reward": 1.0625,
      "reward_std": 0.294627845287323,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.39870715141296387,
      "step": 160
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 157.25,
      "completions/mean_terminated_length": 157.25,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.16649431230610134,
      "grad_norm": 3.8432953762387436,
      "kl": 0.033935546875,
      "learning_rate": 9.339571568522502e-07,
      "loss": 0.0014,
      "num_tokens": 13139362.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.3766257166862488,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.642332136631012,
      "step": 161
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 161.625,
      "completions/mean_terminated_length": 161.625,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.16752843846949328,
      "grad_norm": 2.827018325830448,
      "kl": 0.0224609375,
      "learning_rate": 9.331480052070606e-07,
      "loss": 0.0009,
      "num_tokens": 13221897.0,
      "reward": 1.0625,
      "reward_std": 0.24185511469841003,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.3398369252681732,
      "step": 162
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 151.375,
      "completions/mean_terminated_length": 151.375,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.16856256463288521,
      "grad_norm": 3.849615825816937,
      "kl": 0.03857421875,
      "learning_rate": 9.323342818091307e-07,
      "loss": 0.0015,
      "num_tokens": 13309802.0,
      "reward": 0.8472222089767456,
      "reward_std": 0.38511237502098083,
      "rewards/reasoning_reward/mean": 0.8472221493721008,
      "rewards/reasoning_reward/std": 0.8074482083320618,
      "step": 163
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 326.0,
      "completions/max_terminated_length": 326.0,
      "completions/mean_length": 214.6666717529297,
      "completions/mean_terminated_length": 214.6666717529297,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.16959669079627715,
      "grad_norm": 4.0101406070036925,
      "kl": 0.02783203125,
      "learning_rate": 9.315159952470765e-07,
      "loss": 0.0011,
      "num_tokens": 13391154.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.31285393238067627,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.5299029350280762,
      "step": 164
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 345.0,
      "completions/max_terminated_length": 345.0,
      "completions/mean_length": 175.1666717529297,
      "completions/mean_terminated_length": 175.1666717529297,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.17063081695966908,
      "grad_norm": 3.5535731972038187,
      "kl": 0.0289306640625,
      "learning_rate": 9.306931541576783e-07,
      "loss": 0.0012,
      "num_tokens": 13476590.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.4883233904838562,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.5706435441970825,
      "step": 165
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 398.0,
      "completions/max_terminated_length": 398.0,
      "completions/mean_length": 200.0,
      "completions/mean_terminated_length": 200.0,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.17166494312306102,
      "grad_norm": 2.8543563729941903,
      "kl": 0.041259765625,
      "learning_rate": 9.29865767225787e-07,
      "loss": 0.0017,
      "num_tokens": 13560134.0,
      "reward": 0.9305555820465088,
      "reward_std": 0.25030583143234253,
      "rewards/reasoning_reward/mean": 0.9305555820465088,
      "rewards/reasoning_reward/std": 0.7659995555877686,
      "step": 166
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 314.0,
      "completions/max_terminated_length": 314.0,
      "completions/mean_length": 191.4166717529297,
      "completions/mean_terminated_length": 191.4166717529297,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.17269906928645296,
      "grad_norm": 4.219399972626256,
      "kl": 0.0240478515625,
      "learning_rate": 9.29033843184234e-07,
      "loss": 0.001,
      "num_tokens": 13637928.0,
      "reward": 0.6041666865348816,
      "reward_std": 0.5158624053001404,
      "rewards/reasoning_reward/mean": 0.6041666865348816,
      "rewards/reasoning_reward/std": 0.551266610622406,
      "step": 167
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 137.1666717529297,
      "completions/mean_terminated_length": 137.1666717529297,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.1737331954498449,
      "grad_norm": 4.058689025510401,
      "kl": 0.038330078125,
      "learning_rate": 9.281973908137385e-07,
      "loss": 0.0015,
      "num_tokens": 13723524.0,
      "reward": 0.875,
      "reward_std": 0.5222300291061401,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.5366967916488647,
      "step": 168
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 222.0,
      "completions/max_terminated_length": 222.0,
      "completions/mean_length": 148.0,
      "completions/mean_terminated_length": 148.0,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.17476732161323683,
      "grad_norm": 2.82936642039495,
      "kl": 0.0220947265625,
      "learning_rate": 9.273564189428149e-07,
      "loss": 0.0009,
      "num_tokens": 13809044.0,
      "reward": 0.625,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 169
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 149.875,
      "completions/mean_terminated_length": 149.875,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.17580144777662876,
      "grad_norm": 3.7043384803322907,
      "kl": 0.0245361328125,
      "learning_rate": 9.265109364476798e-07,
      "loss": 0.001,
      "num_tokens": 13892353.0,
      "reward": 0.75,
      "reward_std": 0.41387641429901123,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 170
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 407.0,
      "completions/max_terminated_length": 407.0,
      "completions/mean_length": 220.75,
      "completions/mean_terminated_length": 216.69566345214844,
      "completions/min_length": 152.0,
      "completions/min_terminated_length": 152.0,
      "epoch": 0.1768355739400207,
      "grad_norm": 3.0294206597876343,
      "kl": 0.032470703125,
      "learning_rate": 9.256609522521578e-07,
      "loss": 0.0013,
      "num_tokens": 13976627.0,
      "reward": 0.75,
      "reward_std": 0.3142080307006836,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.8135328888893127,
      "step": 171
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 295.0,
      "completions/max_terminated_length": 295.0,
      "completions/mean_length": 165.625,
      "completions/mean_terminated_length": 165.625,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.1778697001034126,
      "grad_norm": 3.1395182912560826,
      "kl": 0.03466796875,
      "learning_rate": 9.248064753275881e-07,
      "loss": 0.0014,
      "num_tokens": 14057114.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 172
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 195.0,
      "completions/max_terminated_length": 195.0,
      "completions/mean_length": 139.2916717529297,
      "completions/mean_terminated_length": 139.2916717529297,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.17890382626680454,
      "grad_norm": 2.2716748016429715,
      "kl": 0.0263671875,
      "learning_rate": 9.239475146927289e-07,
      "loss": 0.0011,
      "num_tokens": 14135657.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 173
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 339.0,
      "completions/max_terminated_length": 339.0,
      "completions/mean_length": 224.5,
      "completions/mean_terminated_length": 224.5,
      "completions/min_length": 137.0,
      "completions/min_terminated_length": 137.0,
      "epoch": 0.17993795243019647,
      "grad_norm": 3.706550911084408,
      "kl": 0.0322265625,
      "learning_rate": 9.23084079413663e-07,
      "loss": 0.0013,
      "num_tokens": 14214229.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.5276275873184204,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.5472813844680786,
      "step": 174
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 295.0,
      "completions/max_terminated_length": 295.0,
      "completions/mean_length": 160.45834350585938,
      "completions/mean_terminated_length": 160.45834350585938,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.1809720785935884,
      "grad_norm": 3.8563106320766187,
      "kl": 0.0223388671875,
      "learning_rate": 9.222161786037017e-07,
      "loss": 0.0009,
      "num_tokens": 14292528.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.36456555128097534,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.550032913684845,
      "step": 175
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 257.0,
      "completions/max_terminated_length": 257.0,
      "completions/mean_length": 177.875,
      "completions/mean_terminated_length": 177.875,
      "completions/min_length": 123.0,
      "completions/min_terminated_length": 123.0,
      "epoch": 0.18200620475698034,
      "grad_norm": 3.522749129891572,
      "kl": 0.0235595703125,
      "learning_rate": 9.213438214232887e-07,
      "loss": 0.0009,
      "num_tokens": 14369853.0,
      "reward": 0.513888955116272,
      "reward_std": 0.4159068465232849,
      "rewards/reasoning_reward/mean": 0.5138888955116272,
      "rewards/reasoning_reward/std": 0.44482171535491943,
      "step": 176
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.08333333333333337,
      "completions/max_length": 269.0,
      "completions/max_terminated_length": 269.0,
      "completions/mean_length": 193.25,
      "completions/mean_terminated_length": 189.18182373046875,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.18304033092037228,
      "grad_norm": 4.470234549353068,
      "kl": 0.0299072265625,
      "learning_rate": 9.204670170799034e-07,
      "loss": 0.0012,
      "num_tokens": 14446075.0,
      "reward": 0.7222222089767456,
      "reward_std": 0.4285862445831299,
      "rewards/reasoning_reward/mean": 0.7222221493721008,
      "rewards/reasoning_reward/std": 0.7234289646148682,
      "step": 177
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 367.0,
      "completions/max_terminated_length": 367.0,
      "completions/mean_length": 218.83334350585938,
      "completions/mean_terminated_length": 218.83334350585938,
      "completions/min_length": 126.0,
      "completions/min_terminated_length": 126.0,
      "epoch": 0.18407445708376421,
      "grad_norm": 3.680761822106045,
      "kl": 0.0284423828125,
      "learning_rate": 9.195857748279636e-07,
      "loss": 0.0011,
      "num_tokens": 14524703.0,
      "reward": 0.1666666716337204,
      "reward_std": 0.3616904020309448,
      "rewards/reasoning_reward/mean": 0.1666666716337204,
      "rewards/reasoning_reward/std": 0.3806935250759125,
      "step": 178
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 144.33334350585938,
      "completions/mean_terminated_length": 144.33334350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.18510858324715615,
      "grad_norm": 2.0002566843886043,
      "kl": 0.0277099609375,
      "learning_rate": 9.187001039687283e-07,
      "loss": 0.0011,
      "num_tokens": 14609903.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 179
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 193.0,
      "completions/max_terminated_length": 193.0,
      "completions/mean_length": 125.45833587646484,
      "completions/mean_terminated_length": 125.45833587646484,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.18614270941054809,
      "grad_norm": 3.220863102101657,
      "kl": 0.0291748046875,
      "learning_rate": 9.178100138501987e-07,
      "loss": 0.0012,
      "num_tokens": 14691482.0,
      "reward": 1.0,
      "reward_std": 0.24339044094085693,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.3611575663089752,
      "step": 180
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 269.0,
      "completions/max_terminated_length": 269.0,
      "completions/mean_length": 169.375,
      "completions/mean_terminated_length": 169.375,
      "completions/min_length": 65.0,
      "completions/min_terminated_length": 65.0,
      "epoch": 0.18717683557394002,
      "grad_norm": 1.9372953316829284,
      "kl": 0.0308837890625,
      "learning_rate": 9.169155138670202e-07,
      "loss": 0.0012,
      "num_tokens": 14773179.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 181
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 335.0,
      "completions/max_terminated_length": 335.0,
      "completions/mean_length": 198.33334350585938,
      "completions/mean_terminated_length": 198.33334350585938,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.18821096173733196,
      "grad_norm": 4.409408065166967,
      "kl": 0.032958984375,
      "learning_rate": 9.160166134603833e-07,
      "loss": 0.0013,
      "num_tokens": 14857035.0,
      "reward": 1.0625,
      "reward_std": 0.41124165058135986,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.5379611253738403,
      "step": 182
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 266.0,
      "completions/max_terminated_length": 266.0,
      "completions/mean_length": 168.20834350585938,
      "completions/mean_terminated_length": 168.20834350585938,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.1892450879007239,
      "grad_norm": 2.7356052191198317,
      "kl": 0.032470703125,
      "learning_rate": 9.151133221179236e-07,
      "loss": 0.0013,
      "num_tokens": 14934480.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.4148510992527008,
      "step": 183
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 339.0,
      "completions/max_terminated_length": 339.0,
      "completions/mean_length": 230.5416717529297,
      "completions/mean_terminated_length": 230.5416717529297,
      "completions/min_length": 134.0,
      "completions/min_terminated_length": 134.0,
      "epoch": 0.19027921406411583,
      "grad_norm": 4.0459885315893445,
      "kl": 0.0576171875,
      "learning_rate": 9.142056493736214e-07,
      "loss": 0.0023,
      "num_tokens": 15024701.0,
      "reward": 1.0,
      "reward_std": 0.5209805369377136,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.7661308646202087,
      "step": 184
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 284.0,
      "completions/max_terminated_length": 268.0,
      "completions/mean_length": 188.33334350585938,
      "completions/mean_terminated_length": 184.17391967773438,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.19131334022750776,
      "grad_norm": 3.393759487686452,
      "kl": 0.042724609375,
      "learning_rate": 9.13293604807702e-07,
      "loss": 0.0017,
      "num_tokens": 15119653.0,
      "reward": 1.5833333730697632,
      "reward_std": 0.34930619597435,
      "rewards/reasoning_reward/mean": 1.5833333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 185
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 337.0,
      "completions/max_terminated_length": 337.0,
      "completions/mean_length": 185.83334350585938,
      "completions/mean_terminated_length": 185.83334350585938,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.1923474663908997,
      "grad_norm": 1.8836374889884402,
      "kl": 0.025390625,
      "learning_rate": 9.123771980465336e-07,
      "loss": 0.001,
      "num_tokens": 15200161.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 186
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 372.0,
      "completions/max_terminated_length": 372.0,
      "completions/mean_length": 211.5,
      "completions/mean_terminated_length": 211.5,
      "completions/min_length": 127.0,
      "completions/min_terminated_length": 127.0,
      "epoch": 0.19338159255429163,
      "grad_norm": 3.7167415446965446,
      "kl": 0.03173828125,
      "learning_rate": 9.114564387625261e-07,
      "loss": 0.0013,
      "num_tokens": 15278277.0,
      "reward": 0.4652777910232544,
      "reward_std": 0.6122889518737793,
      "rewards/reasoning_reward/mean": 0.4652777910232544,
      "rewards/reasoning_reward/std": 0.6756266355514526,
      "step": 187
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 268.0,
      "completions/max_terminated_length": 268.0,
      "completions/mean_length": 201.375,
      "completions/mean_terminated_length": 201.375,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 0.19441571871768357,
      "grad_norm": 3.074846245023018,
      "kl": 0.02587890625,
      "learning_rate": 9.105313366740295e-07,
      "loss": 0.001,
      "num_tokens": 15357622.0,
      "reward": 1.1527777910232544,
      "reward_std": 0.2801976501941681,
      "rewards/reasoning_reward/mean": 1.1527777910232544,
      "rewards/reasoning_reward/std": 0.35412225127220154,
      "step": 188
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 264.0,
      "completions/max_terminated_length": 264.0,
      "completions/mean_length": 161.625,
      "completions/mean_terminated_length": 161.625,
      "completions/min_length": 61.0,
      "completions/min_terminated_length": 61.0,
      "epoch": 0.1954498448810755,
      "grad_norm": 0.15074496627824105,
      "kl": 0.033935546875,
      "learning_rate": 9.096019015452303e-07,
      "loss": 0.0014,
      "num_tokens": 15442109.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 189
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 180.0,
      "completions/max_terminated_length": 180.0,
      "completions/mean_length": 113.41667175292969,
      "completions/mean_terminated_length": 113.41667175292969,
      "completions/min_length": 64.0,
      "completions/min_terminated_length": 64.0,
      "epoch": 0.19648397104446744,
      "grad_norm": 3.1874055691182623,
      "kl": 0.0322265625,
      "learning_rate": 9.086681431860492e-07,
      "loss": 0.0013,
      "num_tokens": 15525791.0,
      "reward": 1.0,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.7801894545555115,
      "step": 190
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 274.0,
      "completions/max_terminated_length": 274.0,
      "completions/mean_length": 158.5,
      "completions/mean_terminated_length": 158.5,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.19751809720785937,
      "grad_norm": 3.24323335305027,
      "kl": 0.0301513671875,
      "learning_rate": 9.077300714520377e-07,
      "loss": 0.0012,
      "num_tokens": 15609755.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.2965203523635864,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.5706435441970825,
      "step": 191
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 215.375,
      "completions/mean_terminated_length": 215.375,
      "completions/min_length": 133.0,
      "completions/min_terminated_length": 133.0,
      "epoch": 0.19855222337125128,
      "grad_norm": 3.5456355924138805,
      "kl": 0.0306396484375,
      "learning_rate": 9.067876962442732e-07,
      "loss": 0.0012,
      "num_tokens": 15696732.0,
      "reward": 1.0,
      "reward_std": 0.44887280464172363,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.7939992547035217,
      "step": 192
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 219.0,
      "completions/max_terminated_length": 219.0,
      "completions/mean_length": 150.5416717529297,
      "completions/mean_terminated_length": 150.5416717529297,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.19958634953464321,
      "grad_norm": 2.245179535057254,
      "kl": 0.02734375,
      "learning_rate": 9.058410275092553e-07,
      "loss": 0.0011,
      "num_tokens": 15777353.0,
      "reward": 0.5833333730697632,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.5833333134651184,
      "rewards/reasoning_reward/std": 0.5036101341247559,
      "step": 193
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 325.0,
      "completions/max_terminated_length": 325.0,
      "completions/mean_length": 180.625,
      "completions/mean_terminated_length": 180.625,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.20062047569803515,
      "grad_norm": 2.2025459230007964,
      "kl": 0.033447265625,
      "learning_rate": 9.048900752388004e-07,
      "loss": 0.0013,
      "num_tokens": 15862848.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.29546841979026794,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.5036101341247559,
      "step": 194
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 282.0,
      "completions/max_terminated_length": 282.0,
      "completions/mean_length": 165.875,
      "completions/mean_terminated_length": 165.875,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.20165460186142709,
      "grad_norm": 3.077504641125675,
      "kl": 0.04296875,
      "learning_rate": 9.039348494699366e-07,
      "loss": 0.0017,
      "num_tokens": 15945461.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.3535533845424652,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.5089774131774902,
      "step": 195
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 361.0,
      "completions/max_terminated_length": 361.0,
      "completions/mean_length": 176.45834350585938,
      "completions/mean_terminated_length": 176.45834350585938,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.20268872802481902,
      "grad_norm": 1.72449286161736,
      "kl": 0.0302734375,
      "learning_rate": 9.029753602847974e-07,
      "loss": 0.0012,
      "num_tokens": 16026360.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.10206207633018494,
      "step": 196
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 391.0,
      "completions/max_terminated_length": 391.0,
      "completions/mean_length": 199.20834350585938,
      "completions/mean_terminated_length": 199.20834350585938,
      "completions/min_length": 120.0,
      "completions/min_terminated_length": 120.0,
      "epoch": 0.20372285418821096,
      "grad_norm": 3.9491196867976286,
      "kl": 0.055908203125,
      "learning_rate": 9.020116178105153e-07,
      "loss": 0.0022,
      "num_tokens": 16115637.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.5151108503341675,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.7515081167221069,
      "step": 197
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 337.0,
      "completions/max_terminated_length": 337.0,
      "completions/mean_length": 176.0,
      "completions/mean_terminated_length": 176.0,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.2047569803516029,
      "grad_norm": 2.8175351580004078,
      "kl": 0.0289306640625,
      "learning_rate": 9.010436322191155e-07,
      "loss": 0.0012,
      "num_tokens": 16199165.0,
      "reward": 0.8472222089767456,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 0.8472221493721008,
      "rewards/reasoning_reward/std": 0.7221757769584656,
      "step": 198
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 355.0,
      "completions/max_terminated_length": 355.0,
      "completions/mean_length": 204.58334350585938,
      "completions/mean_terminated_length": 204.58334350585938,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.20579110651499483,
      "grad_norm": 4.291506948811305,
      "kl": 0.039794921875,
      "learning_rate": 9.000714137274077e-07,
      "loss": 0.0016,
      "num_tokens": 16277363.0,
      "reward": 0.75,
      "reward_std": 0.3900056481361389,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.6079187393188477,
      "step": 199
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 321.0,
      "completions/max_terminated_length": 321.0,
      "completions/mean_length": 200.9166717529297,
      "completions/mean_terminated_length": 200.9166717529297,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.20682523267838676,
      "grad_norm": 3.795582364631847,
      "kl": 0.033447265625,
      "learning_rate": 8.990949725968786e-07,
      "loss": 0.0013,
      "num_tokens": 16356617.0,
      "reward": 1.1388888359069824,
      "reward_std": 0.42590853571891785,
      "rewards/reasoning_reward/mean": 1.1388888359069824,
      "rewards/reasoning_reward/std": 0.5992480516433716,
      "step": 200
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 165.7916717529297,
      "completions/mean_terminated_length": 165.7916717529297,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.2078593588417787,
      "grad_norm": 4.0373942611533495,
      "kl": 0.0400390625,
      "learning_rate": 8.981143191335839e-07,
      "loss": 0.0016,
      "num_tokens": 16441756.0,
      "reward": 0.930555522441864,
      "reward_std": 0.42342984676361084,
      "rewards/reasoning_reward/mean": 0.930555522441864,
      "rewards/reasoning_reward/std": 0.6843958497047424,
      "step": 201
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 175.75,
      "completions/mean_terminated_length": 175.75,
      "completions/min_length": 71.0,
      "completions/min_terminated_length": 71.0,
      "epoch": 0.20889348500517063,
      "grad_norm": 4.339046231792284,
      "kl": 0.042236328125,
      "learning_rate": 8.971294636880391e-07,
      "loss": 0.0017,
      "num_tokens": 16530366.0,
      "reward": 1.4375,
      "reward_std": 0.4006626605987549,
      "rewards/reasoning_reward/mean": 1.4375,
      "rewards/reasoning_reward/std": 0.6753286719322205,
      "step": 202
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 287.0,
      "completions/max_terminated_length": 287.0,
      "completions/mean_length": 155.375,
      "completions/mean_terminated_length": 155.375,
      "completions/min_length": 62.0,
      "completions/min_terminated_length": 62.0,
      "epoch": 0.20992761116856257,
      "grad_norm": 2.8570580431999235,
      "kl": 0.08154296875,
      "learning_rate": 8.961404166551103e-07,
      "loss": 0.0033,
      "num_tokens": 16615319.0,
      "reward": 1.375,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 203
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 150.0,
      "completions/mean_terminated_length": 150.0,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.2109617373319545,
      "grad_norm": 2.980300192320368,
      "kl": 0.02880859375,
      "learning_rate": 8.951471884739051e-07,
      "loss": 0.0012,
      "num_tokens": 16696703.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.22466278076171875,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.36052998900413513,
      "step": 204
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 323.0,
      "completions/max_terminated_length": 323.0,
      "completions/mean_length": 191.58334350585938,
      "completions/mean_terminated_length": 191.58334350585938,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.21199586349534644,
      "grad_norm": 4.159029061904166,
      "kl": 0.0439453125,
      "learning_rate": 8.941497896276613e-07,
      "loss": 0.0018,
      "num_tokens": 16778829.0,
      "reward": 0.8194445371627808,
      "reward_std": 0.6607243418693542,
      "rewards/reasoning_reward/mean": 0.819444477558136,
      "rewards/reasoning_reward/std": 0.7892953753471375,
      "step": 205
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 307.0,
      "completions/max_terminated_length": 307.0,
      "completions/mean_length": 219.9166717529297,
      "completions/mean_terminated_length": 219.9166717529297,
      "completions/min_length": 140.0,
      "completions/min_terminated_length": 140.0,
      "epoch": 0.21302998965873837,
      "grad_norm": 3.6827623626462063,
      "kl": 0.06298828125,
      "learning_rate": 8.931482306436373e-07,
      "loss": 0.0025,
      "num_tokens": 16863339.0,
      "reward": 1.3958333730697632,
      "reward_std": 0.3719491958618164,
      "rewards/reasoning_reward/mean": 1.3958333730697632,
      "rewards/reasoning_reward/std": 0.6015529036521912,
      "step": 206
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 290.0,
      "completions/max_terminated_length": 290.0,
      "completions/mean_length": 166.625,
      "completions/mean_terminated_length": 166.625,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.2140641158221303,
      "grad_norm": 4.5771027363801,
      "kl": 0.07666015625,
      "learning_rate": 8.921425220930001e-07,
      "loss": 0.0031,
      "num_tokens": 16942642.0,
      "reward": 0.7569444179534912,
      "reward_std": 0.5365828275680542,
      "rewards/reasoning_reward/mean": 0.7569444179534912,
      "rewards/reasoning_reward/std": 0.5223950147628784,
      "step": 207
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 294.0,
      "completions/max_terminated_length": 294.0,
      "completions/mean_length": 148.1666717529297,
      "completions/mean_terminated_length": 148.1666717529297,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.21509824198552224,
      "grad_norm": 4.6067336602799935,
      "kl": 0.072265625,
      "learning_rate": 8.91132674590715e-07,
      "loss": 0.0029,
      "num_tokens": 17028814.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.42645785212516785,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.46430566906929016,
      "step": 208
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 253.0,
      "completions/max_terminated_length": 253.0,
      "completions/mean_length": 161.0416717529297,
      "completions/mean_terminated_length": 161.0416717529297,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.21613236814891418,
      "grad_norm": 3.888741740906291,
      "kl": 0.07861328125,
      "learning_rate": 8.901186987954319e-07,
      "loss": 0.0031,
      "num_tokens": 17112839.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.34503278136253357,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.5835920572280884,
      "step": 209
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 378.0,
      "completions/max_terminated_length": 378.0,
      "completions/mean_length": 190.7916717529297,
      "completions/mean_terminated_length": 190.7916717529297,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.2171664943123061,
      "grad_norm": 3.5678811440994815,
      "kl": 0.041259765625,
      "learning_rate": 8.891006054093739e-07,
      "loss": 0.0016,
      "num_tokens": 17193130.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.24800793826580048,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.6902530789375305,
      "step": 210
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 290.0,
      "completions/max_terminated_length": 290.0,
      "completions/mean_length": 209.75,
      "completions/mean_terminated_length": 209.75,
      "completions/min_length": 128.0,
      "completions/min_terminated_length": 128.0,
      "epoch": 0.21820062047569805,
      "grad_norm": 3.3232953547887,
      "kl": 0.046142578125,
      "learning_rate": 8.880784051782243e-07,
      "loss": 0.0018,
      "num_tokens": 17271196.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.1767766922712326,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.2321528047323227,
      "step": 211
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 334.0,
      "completions/max_terminated_length": 334.0,
      "completions/mean_length": 200.7916717529297,
      "completions/mean_terminated_length": 200.7916717529297,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.21923474663908996,
      "grad_norm": 3.17766425227838,
      "kl": 0.033935546875,
      "learning_rate": 8.870521088910129e-07,
      "loss": 0.0014,
      "num_tokens": 17348999.0,
      "reward": 0.5208333730697632,
      "reward_std": 0.30699092149734497,
      "rewards/reasoning_reward/mean": 0.5208333134651184,
      "rewards/reasoning_reward/std": 0.49954691529273987,
      "step": 212
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 334.0,
      "completions/max_terminated_length": 334.0,
      "completions/mean_length": 192.70834350585938,
      "completions/mean_terminated_length": 192.70834350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.2202688728024819,
      "grad_norm": 3.8551099627848244,
      "kl": 0.04296875,
      "learning_rate": 8.860217273800021e-07,
      "loss": 0.0017,
      "num_tokens": 17428608.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.5464199781417847,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.5985338091850281,
      "step": 213
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 154.33334350585938,
      "completions/mean_terminated_length": 154.33334350585938,
      "completions/min_length": 66.0,
      "completions/min_terminated_length": 66.0,
      "epoch": 0.22130299896587383,
      "grad_norm": 3.824252202511437,
      "kl": 0.07568359375,
      "learning_rate": 8.849872715205725e-07,
      "loss": 0.003,
      "num_tokens": 17507808.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.434057354927063,
      "step": 214
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 332.0,
      "completions/max_terminated_length": 332.0,
      "completions/mean_length": 193.70834350585938,
      "completions/mean_terminated_length": 193.70834350585938,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.22233712512926576,
      "grad_norm": 2.7022340614121916,
      "kl": 0.033447265625,
      "learning_rate": 8.839487522311086e-07,
      "loss": 0.0013,
      "num_tokens": 17598361.0,
      "reward": 1.5,
      "reward_std": 0.24966806173324585,
      "rewards/reasoning_reward/mean": 1.5,
      "rewards/reasoning_reward/std": 0.7071067690849304,
      "step": 215
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 332.0,
      "completions/max_terminated_length": 332.0,
      "completions/mean_length": 186.70834350585938,
      "completions/mean_terminated_length": 186.70834350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.2233712512926577,
      "grad_norm": 2.845013705079817,
      "kl": 0.043701171875,
      "learning_rate": 8.829061804728834e-07,
      "loss": 0.0017,
      "num_tokens": 17687378.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.39000558853149414,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.4815434515476227,
      "step": 216
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 255.0,
      "completions/max_terminated_length": 255.0,
      "completions/mean_length": 142.2916717529297,
      "completions/mean_terminated_length": 142.2916717529297,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.22440537745604963,
      "grad_norm": 3.704720807397464,
      "kl": 0.029296875,
      "learning_rate": 8.818595672499418e-07,
      "loss": 0.0012,
      "num_tokens": 17771825.0,
      "reward": 0.763888955116272,
      "reward_std": 0.32886335253715515,
      "rewards/reasoning_reward/mean": 0.7638888955116272,
      "rewards/reasoning_reward/std": 0.5152661204338074,
      "step": 217
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 360.0,
      "completions/max_terminated_length": 360.0,
      "completions/mean_length": 204.58334350585938,
      "completions/mean_terminated_length": 204.58334350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.22543950361944157,
      "grad_norm": 4.235645113018916,
      "kl": 0.048828125,
      "learning_rate": 8.808089236089857e-07,
      "loss": 0.002,
      "num_tokens": 17849271.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.6813797950744629,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.7067864537239075,
      "step": 218
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 166.6666717529297,
      "completions/mean_terminated_length": 166.6666717529297,
      "completions/min_length": 52.0,
      "completions/min_terminated_length": 52.0,
      "epoch": 0.2264736297828335,
      "grad_norm": 3.7267555026025962,
      "kl": 0.055908203125,
      "learning_rate": 8.797542606392572e-07,
      "loss": 0.0022,
      "num_tokens": 17934255.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.4042079448699951,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.6507381200790405,
      "step": 219
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 235.0,
      "completions/max_terminated_length": 235.0,
      "completions/mean_length": 165.7916717529297,
      "completions/mean_terminated_length": 165.7916717529297,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.22750775594622544,
      "grad_norm": 3.259723649327136,
      "kl": 0.036865234375,
      "learning_rate": 8.786955894724206e-07,
      "loss": 0.0015,
      "num_tokens": 18011010.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 220
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 286.0,
      "completions/max_terminated_length": 286.0,
      "completions/mean_length": 167.08334350585938,
      "completions/mean_terminated_length": 167.08334350585938,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.22854188210961737,
      "grad_norm": 3.2853923018575095,
      "kl": 0.040771484375,
      "learning_rate": 8.776329212824461e-07,
      "loss": 0.0016,
      "num_tokens": 18089092.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.30551642179489136,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.7067864537239075,
      "step": 221
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 360.0,
      "completions/max_terminated_length": 360.0,
      "completions/mean_length": 176.6666717529297,
      "completions/mean_terminated_length": 176.6666717529297,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.2295760082730093,
      "grad_norm": 3.3736718186548527,
      "kl": 0.03173828125,
      "learning_rate": 8.765662672854908e-07,
      "loss": 0.0013,
      "num_tokens": 18173180.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.2314550280570984,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.9197904467582703,
      "step": 222
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 234.0,
      "completions/max_terminated_length": 234.0,
      "completions/mean_length": 180.625,
      "completions/mean_terminated_length": 180.625,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.23061013443640124,
      "grad_norm": 4.387150850841933,
      "kl": 0.041748046875,
      "learning_rate": 8.754956387397814e-07,
      "loss": 0.0017,
      "num_tokens": 18252827.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.5432543754577637,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.5311833620071411,
      "step": 223
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 346.0,
      "completions/max_terminated_length": 346.0,
      "completions/mean_length": 179.6666717529297,
      "completions/mean_terminated_length": 179.6666717529297,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.23164426059979318,
      "grad_norm": 3.7927764070085943,
      "kl": 0.035888671875,
      "learning_rate": 8.744210469454945e-07,
      "loss": 0.0014,
      "num_tokens": 18334267.0,
      "reward": 0.8819444179534912,
      "reward_std": 0.3051118850708008,
      "rewards/reasoning_reward/mean": 0.8819444179534912,
      "rewards/reasoning_reward/std": 0.5258513689041138,
      "step": 224
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 261.0,
      "completions/max_terminated_length": 261.0,
      "completions/mean_length": 178.08334350585938,
      "completions/mean_terminated_length": 178.08334350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.2326783867631851,
      "grad_norm": 2.0858422243920622,
      "kl": 0.035400390625,
      "learning_rate": 8.73342503244638e-07,
      "loss": 0.0014,
      "num_tokens": 18417093.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 225
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 150.5416717529297,
      "completions/mean_terminated_length": 150.5416717529297,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.23371251292657705,
      "grad_norm": 1.7365069787778553,
      "kl": 0.0274658203125,
      "learning_rate": 8.722600190209303e-07,
      "loss": 0.0011,
      "num_tokens": 18494866.0,
      "reward": 0.625,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 226
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 306.0,
      "completions/max_terminated_length": 306.0,
      "completions/mean_length": 172.58334350585938,
      "completions/mean_terminated_length": 172.58334350585938,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.23474663908996898,
      "grad_norm": 4.708061391148886,
      "kl": 0.034912109375,
      "learning_rate": 8.711736056996817e-07,
      "loss": 0.0014,
      "num_tokens": 18573464.0,
      "reward": 0.8125,
      "reward_std": 0.6011933088302612,
      "rewards/reasoning_reward/mean": 0.8125,
      "rewards/reasoning_reward/std": 0.7775728702545166,
      "step": 227
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 181.7916717529297,
      "completions/mean_terminated_length": 181.7916717529297,
      "completions/min_length": 137.0,
      "completions/min_terminated_length": 137.0,
      "epoch": 0.23578076525336092,
      "grad_norm": 3.0293873866544394,
      "kl": 0.037353515625,
      "learning_rate": 8.700832747476725e-07,
      "loss": 0.0015,
      "num_tokens": 18653739.0,
      "reward": 0.5277777910232544,
      "reward_std": 0.24982166290283203,
      "rewards/reasoning_reward/mean": 0.5277777314186096,
      "rewards/reasoning_reward/std": 0.7350221872329712,
      "step": 228
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 274.0,
      "completions/max_terminated_length": 274.0,
      "completions/mean_length": 161.75,
      "completions/mean_terminated_length": 161.75,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.23681489141675285,
      "grad_norm": 2.7748190749466977,
      "kl": 0.03173828125,
      "learning_rate": 8.689890376730327e-07,
      "loss": 0.0013,
      "num_tokens": 18733685.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 229
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 265.0,
      "completions/max_terminated_length": 265.0,
      "completions/mean_length": 187.25,
      "completions/mean_terminated_length": 187.25,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.2378490175801448,
      "grad_norm": 2.810129730379906,
      "kl": 0.036865234375,
      "learning_rate": 8.678909060251201e-07,
      "loss": 0.0015,
      "num_tokens": 18816811.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.352710485458374,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.5413181781768799,
      "step": 230
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 190.0,
      "completions/max_terminated_length": 190.0,
      "completions/mean_length": 140.125,
      "completions/mean_terminated_length": 140.125,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.23888314374353672,
      "grad_norm": 2.4272615508776347,
      "kl": 0.038818359375,
      "learning_rate": 8.667888913943988e-07,
      "loss": 0.0016,
      "num_tokens": 18893230.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.20412415266036987,
      "step": 231
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 280.0,
      "completions/max_terminated_length": 280.0,
      "completions/mean_length": 159.4166717529297,
      "completions/mean_terminated_length": 159.4166717529297,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.23991726990692863,
      "grad_norm": 4.0221042286356985,
      "kl": 0.0390625,
      "learning_rate": 8.656830054123168e-07,
      "loss": 0.0016,
      "num_tokens": 18983048.0,
      "reward": 1.125,
      "reward_std": 0.4563485085964203,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.8501917719841003,
      "step": 232
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 416.0,
      "completions/max_terminated_length": 416.0,
      "completions/mean_length": 206.45834350585938,
      "completions/mean_terminated_length": 206.45834350585938,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.24095139607032057,
      "grad_norm": 4.213144645418847,
      "kl": 0.04052734375,
      "learning_rate": 8.645732597511825e-07,
      "loss": 0.0016,
      "num_tokens": 19060963.0,
      "reward": 0.875,
      "reward_std": 0.60628741979599,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.710939347743988,
      "step": 233
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 299.0,
      "completions/max_terminated_length": 299.0,
      "completions/mean_length": 173.33334350585938,
      "completions/mean_terminated_length": 173.33334350585938,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.2419855222337125,
      "grad_norm": 3.05995072143482,
      "kl": 0.048583984375,
      "learning_rate": 8.634596661240428e-07,
      "loss": 0.0019,
      "num_tokens": 19137963.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.33247750997543335,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.5036101341247559,
      "step": 234
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 388.0,
      "completions/max_terminated_length": 388.0,
      "completions/mean_length": 217.0416717529297,
      "completions/mean_terminated_length": 217.0416717529297,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.24301964839710444,
      "grad_norm": 3.020436177796217,
      "kl": 0.0296630859375,
      "learning_rate": 8.623422362845582e-07,
      "loss": 0.0012,
      "num_tokens": 19215988.0,
      "reward": 0.625,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 235
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 277.0,
      "completions/max_terminated_length": 277.0,
      "completions/mean_length": 180.125,
      "completions/mean_terminated_length": 180.125,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.24405377456049637,
      "grad_norm": 2.7937754256291765,
      "kl": 0.0296630859375,
      "learning_rate": 8.612209820268798e-07,
      "loss": 0.0012,
      "num_tokens": 19299191.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.29546841979026794,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.5036101341247559,
      "step": 236
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 362.0,
      "completions/max_terminated_length": 362.0,
      "completions/mean_length": 217.875,
      "completions/mean_terminated_length": 217.875,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.2450879007238883,
      "grad_norm": 3.360745099795255,
      "kl": 0.052490234375,
      "learning_rate": 8.600959151855241e-07,
      "loss": 0.0021,
      "num_tokens": 19380468.0,
      "reward": 0.875,
      "reward_std": 0.2985045611858368,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.5160468220710754,
      "step": 237
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 367.0,
      "completions/max_terminated_length": 367.0,
      "completions/mean_length": 167.1666717529297,
      "completions/mean_terminated_length": 167.1666717529297,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.24612202688728024,
      "grad_norm": 4.215940666816239,
      "kl": 0.032470703125,
      "learning_rate": 8.589670476352484e-07,
      "loss": 0.0013,
      "num_tokens": 19465384.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.39000558853149414,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.6197241544723511,
      "step": 238
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 415.0,
      "completions/max_terminated_length": 415.0,
      "completions/mean_length": 226.20834350585938,
      "completions/mean_terminated_length": 226.20834350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.24715615305067218,
      "grad_norm": 1.8917075772209162,
      "kl": 0.041015625,
      "learning_rate": 8.578343912909252e-07,
      "loss": 0.0016,
      "num_tokens": 19545613.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.08266931027173996,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.14947572350502014,
      "step": 239
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 292.0,
      "completions/max_terminated_length": 292.0,
      "completions/mean_length": 181.1666717529297,
      "completions/mean_terminated_length": 181.1666717529297,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.2481902792140641,
      "grad_norm": 3.3661865153565125,
      "kl": 0.037353515625,
      "learning_rate": 8.566979581074168e-07,
      "loss": 0.0015,
      "num_tokens": 19631889.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.24339044094085693,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.5247498154640198,
      "step": 240
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 230.625,
      "completions/mean_terminated_length": 230.625,
      "completions/min_length": 132.0,
      "completions/min_terminated_length": 132.0,
      "epoch": 0.24922440537745605,
      "grad_norm": 28.79446249181958,
      "kl": 0.67578125,
      "learning_rate": 8.555577600794488e-07,
      "loss": 0.0271,
      "num_tokens": 19711856.0,
      "reward": 1.0069445371627808,
      "reward_std": 0.6009811162948608,
      "rewards/reasoning_reward/mean": 1.0069445371627808,
      "rewards/reasoning_reward/std": 0.7010675072669983,
      "step": 241
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 378.0,
      "completions/max_terminated_length": 378.0,
      "completions/mean_length": 175.75,
      "completions/mean_terminated_length": 175.75,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.250258531540848,
      "grad_norm": 2.003121090259284,
      "kl": 0.037841796875,
      "learning_rate": 8.54413809241484e-07,
      "loss": 0.0015,
      "num_tokens": 19791402.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 242
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 196.0,
      "completions/max_terminated_length": 196.0,
      "completions/mean_length": 143.375,
      "completions/mean_terminated_length": 143.375,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.2512926577042399,
      "grad_norm": 0.15042965975061226,
      "kl": 0.032958984375,
      "learning_rate": 8.53266117667595e-07,
      "loss": 0.0013,
      "num_tokens": 19871107.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 243
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 252.0,
      "completions/max_terminated_length": 252.0,
      "completions/mean_length": 157.25,
      "completions/mean_terminated_length": 157.25,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.25232678386763185,
      "grad_norm": 0.13370684582648995,
      "kl": 0.042236328125,
      "learning_rate": 8.521146974713363e-07,
      "loss": 0.0017,
      "num_tokens": 19950505.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 244
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 144.375,
      "completions/mean_terminated_length": 142.6521759033203,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.2533609100310238,
      "grad_norm": 2.465382515779942,
      "kl": 0.0281982421875,
      "learning_rate": 8.50959560805617e-07,
      "loss": 0.0011,
      "num_tokens": 20031090.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 245
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 252.0,
      "completions/max_terminated_length": 252.0,
      "completions/mean_length": 161.33334350585938,
      "completions/mean_terminated_length": 161.33334350585938,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.2543950361944157,
      "grad_norm": 4.076548379222048,
      "kl": 0.038330078125,
      "learning_rate": 8.498007198625732e-07,
      "loss": 0.0015,
      "num_tokens": 20112338.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.22466278076171875,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.40322521328926086,
      "step": 246
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 219.0,
      "completions/max_terminated_length": 219.0,
      "completions/mean_length": 163.9166717529297,
      "completions/mean_terminated_length": 163.9166717529297,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.25542916235780766,
      "grad_norm": 3.9449748133141647,
      "kl": 0.051513671875,
      "learning_rate": 8.486381868734378e-07,
      "loss": 0.0021,
      "num_tokens": 20196792.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.4564814567565918,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.6833288669586182,
      "step": 247
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 247.0,
      "completions/max_terminated_length": 247.0,
      "completions/mean_length": 127.91667175292969,
      "completions/mean_terminated_length": 127.91667175292969,
      "completions/min_length": 66.0,
      "completions/min_terminated_length": 66.0,
      "epoch": 0.2564632885211996,
      "grad_norm": 4.846992572972704,
      "kl": 0.040283203125,
      "learning_rate": 8.474719741084126e-07,
      "loss": 0.0016,
      "num_tokens": 20279862.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.3900056481361389,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "step": 248
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 310.0,
      "completions/max_terminated_length": 310.0,
      "completions/mean_length": 180.70834350585938,
      "completions/mean_terminated_length": 180.70834350585938,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.25749741468459153,
      "grad_norm": 5.015013743019392,
      "kl": 0.06787109375,
      "learning_rate": 8.463020938765384e-07,
      "loss": 0.0027,
      "num_tokens": 20363271.0,
      "reward": 1.125,
      "reward_std": 0.5078567266464233,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.8501917719841003,
      "step": 249
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 255.0,
      "completions/max_terminated_length": 255.0,
      "completions/mean_length": 158.0416717529297,
      "completions/mean_terminated_length": 158.0416717529297,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.25853154084798347,
      "grad_norm": 3.3147927169947553,
      "kl": 0.06982421875,
      "learning_rate": 8.451285585255646e-07,
      "loss": 0.0028,
      "num_tokens": 20442512.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 250
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 133.83334350585938,
      "completions/mean_terminated_length": 133.83334350585938,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.2595656670113754,
      "grad_norm": 1.9961818932049205,
      "kl": 0.0390625,
      "learning_rate": 8.439513804418196e-07,
      "loss": 0.0016,
      "num_tokens": 20523788.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148510992527008,
      "step": 251
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 218.0,
      "completions/max_terminated_length": 218.0,
      "completions/mean_length": 148.6666717529297,
      "completions/mean_terminated_length": 148.6666717529297,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.26059979317476734,
      "grad_norm": 3.886022313764354,
      "kl": 0.0380859375,
      "learning_rate": 8.4277057205008e-07,
      "loss": 0.0015,
      "num_tokens": 20606012.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 252
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 258.0,
      "completions/max_terminated_length": 258.0,
      "completions/mean_length": 166.375,
      "completions/mean_terminated_length": 166.375,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.26163391933815927,
      "grad_norm": 3.611166967490902,
      "kl": 0.07421875,
      "learning_rate": 8.415861458134392e-07,
      "loss": 0.003,
      "num_tokens": 20694661.0,
      "reward": 1.625,
      "reward_std": 0.3933655619621277,
      "rewards/reasoning_reward/mean": 1.625,
      "rewards/reasoning_reward/std": 0.5366967916488647,
      "step": 253
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 347.0,
      "completions/max_terminated_length": 347.0,
      "completions/mean_length": 188.875,
      "completions/mean_terminated_length": 188.875,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.2626680455015512,
      "grad_norm": 3.9865927693728582,
      "kl": 0.076171875,
      "learning_rate": 8.403981142331758e-07,
      "loss": 0.003,
      "num_tokens": 20774098.0,
      "reward": 0.875,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 254
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 209.0,
      "completions/max_terminated_length": 209.0,
      "completions/mean_length": 133.125,
      "completions/mean_terminated_length": 133.125,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.26370217166494314,
      "grad_norm": 2.6781574727908297,
      "kl": 0.0556640625,
      "learning_rate": 8.392064898486215e-07,
      "loss": 0.0022,
      "num_tokens": 20853589.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 255
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 256.0,
      "completions/max_terminated_length": 256.0,
      "completions/mean_length": 149.70834350585938,
      "completions/mean_terminated_length": 149.70834350585938,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.2647362978283351,
      "grad_norm": 5.965736009719672,
      "kl": 0.07080078125,
      "learning_rate": 8.380112852370296e-07,
      "loss": 0.0028,
      "num_tokens": 20936110.0,
      "reward": 1.1805555820465088,
      "reward_std": 0.36751919984817505,
      "rewards/reasoning_reward/mean": 1.1805554628372192,
      "rewards/reasoning_reward/std": 0.6117146611213684,
      "step": 256
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 153.5416717529297,
      "completions/mean_terminated_length": 153.5416717529297,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.265770423991727,
      "grad_norm": 2.4421141476780917,
      "kl": 0.038818359375,
      "learning_rate": 8.368125130134414e-07,
      "loss": 0.0015,
      "num_tokens": 21024379.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.18766528367996216,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.31204676628112793,
      "step": 257
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 243.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 160.5416717529297,
      "completions/mean_terminated_length": 160.5416717529297,
      "completions/min_length": 119.0,
      "completions/min_terminated_length": 119.0,
      "epoch": 0.26680455015511895,
      "grad_norm": 3.149878852517023,
      "kl": 0.0556640625,
      "learning_rate": 8.356101858305528e-07,
      "loss": 0.0022,
      "num_tokens": 21110432.0,
      "reward": 1.1875,
      "reward_std": 0.23709973692893982,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.7042186260223389,
      "step": 258
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 271.0,
      "completions/max_terminated_length": 271.0,
      "completions/mean_length": 181.375,
      "completions/mean_terminated_length": 181.375,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.2678386763185109,
      "grad_norm": 4.173006267735779,
      "kl": 0.07080078125,
      "learning_rate": 8.344043163785823e-07,
      "loss": 0.0028,
      "num_tokens": 21193913.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.5727972984313965,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.8329709768295288,
      "step": 259
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 264.0,
      "completions/max_terminated_length": 264.0,
      "completions/mean_length": 161.7916717529297,
      "completions/mean_terminated_length": 161.7916717529297,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.2688728024819028,
      "grad_norm": 3.930560040382512,
      "kl": 0.07470703125,
      "learning_rate": 8.331949173851354e-07,
      "loss": 0.003,
      "num_tokens": 21270796.0,
      "reward": 0.75,
      "reward_std": 0.33247750997543335,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 260
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 242.0,
      "completions/max_terminated_length": 242.0,
      "completions/mean_length": 137.5,
      "completions/mean_terminated_length": 137.5,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.26990692864529475,
      "grad_norm": 4.470134236685902,
      "kl": 0.06298828125,
      "learning_rate": 8.319820016150706e-07,
      "loss": 0.0025,
      "num_tokens": 21353616.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.39814266562461853,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.4402732849121094,
      "step": 261
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 352.0,
      "completions/max_terminated_length": 352.0,
      "completions/mean_length": 184.08334350585938,
      "completions/mean_terminated_length": 184.08334350585938,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.27094105480868663,
      "grad_norm": 2.50371678079303,
      "kl": 0.06298828125,
      "learning_rate": 8.307655818703657e-07,
      "loss": 0.0025,
      "num_tokens": 21432634.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.1259881556034088,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.24077169597148895,
      "step": 262
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 296.0,
      "completions/max_terminated_length": 296.0,
      "completions/mean_length": 188.20834350585938,
      "completions/mean_terminated_length": 188.20834350585938,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.27197518097207857,
      "grad_norm": 2.890189899345422,
      "kl": 0.0284423828125,
      "learning_rate": 8.295456709899816e-07,
      "loss": 0.0011,
      "num_tokens": 21519879.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.33247750997543335,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.746974527835846,
      "step": 263
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 253.0,
      "completions/max_terminated_length": 253.0,
      "completions/mean_length": 153.33334350585938,
      "completions/mean_terminated_length": 153.33334350585938,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.2730093071354705,
      "grad_norm": 4.254336132840291,
      "kl": 0.0634765625,
      "learning_rate": 8.283222818497269e-07,
      "loss": 0.0025,
      "num_tokens": 21599551.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 264
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 240.0,
      "completions/max_terminated_length": 240.0,
      "completions/mean_length": 142.125,
      "completions/mean_terminated_length": 147.0,
      "completions/min_length": 30.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.27404343329886244,
      "grad_norm": 4.319565951453572,
      "kl": 0.060302734375,
      "learning_rate": 8.270954273621228e-07,
      "loss": 0.0024,
      "num_tokens": 21682506.0,
      "reward": 0.75,
      "reward_std": 0.45846566557884216,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.7071067690849304,
      "step": 265
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 163.6666717529297,
      "completions/mean_terminated_length": 163.6666717529297,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.2750775594622544,
      "grad_norm": 3.4635605581504056,
      "kl": 0.05859375,
      "learning_rate": 8.258651204762657e-07,
      "loss": 0.0023,
      "num_tokens": 21771346.0,
      "reward": 1.5416667461395264,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.5416666269302368,
      "rewards/reasoning_reward/std": 0.6580052971839905,
      "step": 266
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 242.0,
      "completions/max_terminated_length": 242.0,
      "completions/mean_length": 157.08334350585938,
      "completions/mean_terminated_length": 157.08334350585938,
      "completions/min_length": 64.0,
      "completions/min_terminated_length": 64.0,
      "epoch": 0.2761116856256463,
      "grad_norm": 4.0415573834840535,
      "kl": 0.052490234375,
      "learning_rate": 8.24631374177691e-07,
      "loss": 0.0021,
      "num_tokens": 21850356.0,
      "reward": 0.8472222089767456,
      "reward_std": 0.4722983241081238,
      "rewards/reasoning_reward/mean": 0.8472221493721008,
      "rewards/reasoning_reward/std": 0.7612547874450684,
      "step": 267
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 235.0,
      "completions/max_terminated_length": 235.0,
      "completions/mean_length": 120.41667175292969,
      "completions/mean_terminated_length": 120.41667175292969,
      "completions/min_length": 54.0,
      "completions/min_terminated_length": 54.0,
      "epoch": 0.27714581178903824,
      "grad_norm": 3.246047157021264,
      "kl": 0.058349609375,
      "learning_rate": 8.233942014882369e-07,
      "loss": 0.0023,
      "num_tokens": 21930830.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "step": 268
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 362.0,
      "completions/max_terminated_length": 362.0,
      "completions/mean_length": 160.625,
      "completions/mean_terminated_length": 160.625,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.2781799379524302,
      "grad_norm": 4.562799738250913,
      "kl": 0.052490234375,
      "learning_rate": 8.221536154659054e-07,
      "loss": 0.0021,
      "num_tokens": 22015157.0,
      "reward": 1.125,
      "reward_std": 0.5863928198814392,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.7408866882324219,
      "step": 269
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 553.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 163.33334350585938,
      "completions/mean_terminated_length": 146.3913116455078,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.2792140641158221,
      "grad_norm": 4.181549595465215,
      "kl": 0.07568359375,
      "learning_rate": 8.209096292047257e-07,
      "loss": 0.003,
      "num_tokens": 22101757.0,
      "reward": 0.75,
      "reward_std": 0.4446708858013153,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.675663948059082,
      "step": 270
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 254.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 165.08334350585938,
      "completions/mean_terminated_length": 165.08334350585938,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.28024819027921405,
      "grad_norm": 7.933565453714079,
      "kl": 0.2265625,
      "learning_rate": 8.196622558346152e-07,
      "loss": 0.0091,
      "num_tokens": 22180183.0,
      "reward": 0.875,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 271
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 189.0,
      "completions/max_terminated_length": 189.0,
      "completions/mean_length": 129.70834350585938,
      "completions/mean_terminated_length": 132.13043212890625,
      "completions/min_length": 67.0,
      "completions/min_terminated_length": 67.0,
      "epoch": 0.281282316442606,
      "grad_norm": 3.1643095072157363,
      "kl": 0.07421875,
      "learning_rate": 8.184115085212413e-07,
      "loss": 0.003,
      "num_tokens": 22256120.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 272
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 146.33334350585938,
      "completions/mean_terminated_length": 146.33334350585938,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.2823164426059979,
      "grad_norm": 3.819168262230962,
      "kl": 0.062255859375,
      "learning_rate": 8.171574004658828e-07,
      "loss": 0.0025,
      "num_tokens": 22344216.0,
      "reward": 1.375,
      "reward_std": 0.31285393238067627,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 273
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 223.0,
      "completions/max_terminated_length": 223.0,
      "completions/mean_length": 161.08334350585938,
      "completions/mean_terminated_length": 161.08334350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.28335056876938985,
      "grad_norm": 3.2604091114453984,
      "kl": 0.058837890625,
      "learning_rate": 8.158999449052898e-07,
      "loss": 0.0024,
      "num_tokens": 22419034.0,
      "reward": 1.0625,
      "reward_std": 0.13607725501060486,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.1689159870147705,
      "step": 274
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 254.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 148.45834350585938,
      "completions/mean_terminated_length": 148.45834350585938,
      "completions/min_length": 73.0,
      "completions/min_terminated_length": 73.0,
      "epoch": 0.2843846949327818,
      "grad_norm": 5.802068488952265,
      "kl": 0.1806640625,
      "learning_rate": 8.146391551115442e-07,
      "loss": 0.0072,
      "num_tokens": 22497061.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.23144195973873138,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.4164854884147644,
      "step": 275
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 184.0,
      "completions/max_terminated_length": 184.0,
      "completions/mean_length": 135.4166717529297,
      "completions/mean_terminated_length": 135.4166717529297,
      "completions/min_length": 57.0,
      "completions/min_terminated_length": 57.0,
      "epoch": 0.2854188210961737,
      "grad_norm": 2.5403712266985683,
      "kl": 0.0615234375,
      "learning_rate": 8.133750443919205e-07,
      "loss": 0.0025,
      "num_tokens": 22575655.0,
      "reward": 1.375,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 276
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 449.0,
      "completions/max_terminated_length": 449.0,
      "completions/mean_length": 164.0416717529297,
      "completions/mean_terminated_length": 164.0416717529297,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.28645294725956566,
      "grad_norm": 2.660324010030384,
      "kl": 0.056884765625,
      "learning_rate": 8.121076260887436e-07,
      "loss": 0.0023,
      "num_tokens": 22653648.0,
      "reward": 0.5,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.5,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 277
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 155.70834350585938,
      "completions/mean_terminated_length": 155.70834350585938,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.2874870734229576,
      "grad_norm": 3.979951013030403,
      "kl": 0.052978515625,
      "learning_rate": 8.108369135792498e-07,
      "loss": 0.0021,
      "num_tokens": 22739169.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.39324939250946045,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.7755315899848938,
      "step": 278
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 163.125,
      "completions/mean_terminated_length": 163.125,
      "completions/min_length": 59.0,
      "completions/min_terminated_length": 59.0,
      "epoch": 0.28852119958634953,
      "grad_norm": 2.664386460601889,
      "kl": 0.07568359375,
      "learning_rate": 8.095629202754447e-07,
      "loss": 0.003,
      "num_tokens": 22816052.0,
      "reward": 0.6458333730697632,
      "reward_std": 0.13908717036247253,
      "rewards/reasoning_reward/mean": 0.6458333134651184,
      "rewards/reasoning_reward/std": 0.5208514332771301,
      "step": 279
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 349.0,
      "completions/max_terminated_length": 349.0,
      "completions/mean_length": 154.875,
      "completions/mean_terminated_length": 154.875,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.28955532574974147,
      "grad_norm": 3.1543556872497907,
      "kl": 0.060791015625,
      "learning_rate": 8.082856596239613e-07,
      "loss": 0.0024,
      "num_tokens": 22893649.0,
      "reward": 0.5,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.5,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 280
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 247.0,
      "completions/max_terminated_length": 247.0,
      "completions/mean_length": 147.75,
      "completions/mean_terminated_length": 147.75,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.2905894519131334,
      "grad_norm": 3.3080012627332933,
      "kl": 0.07275390625,
      "learning_rate": 8.070051451059188e-07,
      "loss": 0.0029,
      "num_tokens": 22971419.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.1849137246608734,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.5706435441970825,
      "step": 281
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 143.83334350585938,
      "completions/mean_terminated_length": 143.83334350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.29162357807652534,
      "grad_norm": 1.9572479604131818,
      "kl": 0.0556640625,
      "learning_rate": 8.057213902367801e-07,
      "loss": 0.0022,
      "num_tokens": 23048719.0,
      "reward": 0.75,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 282
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 379.0,
      "completions/max_terminated_length": 379.0,
      "completions/mean_length": 176.20834350585938,
      "completions/mean_terminated_length": 176.20834350585938,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.29265770423991727,
      "grad_norm": 2.3492148432005755,
      "kl": 0.0556640625,
      "learning_rate": 8.044344085662092e-07,
      "loss": 0.0022,
      "num_tokens": 23127580.0,
      "reward": 0.125,
      "reward_std": 0.24800793826580048,
      "rewards/reasoning_reward/mean": 0.125,
      "rewards/reasoning_reward/std": 0.4484272003173828,
      "step": 283
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 276.0,
      "completions/max_terminated_length": 276.0,
      "completions/mean_length": 178.75,
      "completions/mean_terminated_length": 178.75,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.2936918304033092,
      "grad_norm": 3.9640676648988027,
      "kl": 0.076171875,
      "learning_rate": 8.031442136779271e-07,
      "loss": 0.0031,
      "num_tokens": 23210798.0,
      "reward": 1.0625,
      "reward_std": 0.3596132695674896,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.7798991799354553,
      "step": 284
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 296.0,
      "completions/max_terminated_length": 296.0,
      "completions/mean_length": 184.0416717529297,
      "completions/mean_terminated_length": 184.0416717529297,
      "completions/min_length": 114.0,
      "completions/min_terminated_length": 114.0,
      "epoch": 0.29472595656670114,
      "grad_norm": 4.300616429230677,
      "kl": 0.057861328125,
      "learning_rate": 8.018508191895712e-07,
      "loss": 0.0023,
      "num_tokens": 23294247.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.40489405393600464,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.6507381200790405,
      "step": 285
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 324.0,
      "completions/max_terminated_length": 324.0,
      "completions/mean_length": 190.375,
      "completions/mean_terminated_length": 190.375,
      "completions/min_length": 128.0,
      "completions/min_terminated_length": 128.0,
      "epoch": 0.2957600827300931,
      "grad_norm": 2.685078105274284,
      "kl": 0.06884765625,
      "learning_rate": 8.005542387525479e-07,
      "loss": 0.0028,
      "num_tokens": 23383472.0,
      "reward": 1.3125,
      "reward_std": 0.23709973692893982,
      "rewards/reasoning_reward/mean": 1.3125,
      "rewards/reasoning_reward/std": 0.38483479619026184,
      "step": 286
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 161.375,
      "completions/mean_terminated_length": 161.375,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.296794208893485,
      "grad_norm": 3.908207128618161,
      "kl": 0.056640625,
      "learning_rate": 7.992544860518915e-07,
      "loss": 0.0023,
      "num_tokens": 23464673.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.3900056481361389,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 287
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 260.0,
      "completions/max_terminated_length": 260.0,
      "completions/mean_length": 174.95834350585938,
      "completions/mean_terminated_length": 174.95834350585938,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.29782833505687695,
      "grad_norm": 3.2451345960277793,
      "kl": 0.051513671875,
      "learning_rate": 7.979515748061181e-07,
      "loss": 0.0021,
      "num_tokens": 23549896.0,
      "reward": 1.0625,
      "reward_std": 0.3194752335548401,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.47348156571388245,
      "step": 288
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 296.0,
      "completions/max_terminated_length": 296.0,
      "completions/mean_length": 174.75,
      "completions/mean_terminated_length": 174.75,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.2988624612202689,
      "grad_norm": 4.7409903359224295,
      "kl": 0.060791015625,
      "learning_rate": 7.966455187670819e-07,
      "loss": 0.0024,
      "num_tokens": 23627058.0,
      "reward": 0.4791666865348816,
      "reward_std": 0.4373263716697693,
      "rewards/reasoning_reward/mean": 0.4791666567325592,
      "rewards/reasoning_reward/std": 0.5985338091850281,
      "step": 289
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 269.0,
      "completions/max_terminated_length": 269.0,
      "completions/mean_length": 152.9166717529297,
      "completions/mean_terminated_length": 152.9166717529297,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.2998965873836608,
      "grad_norm": 1.7959395053181861,
      "kl": 0.038818359375,
      "learning_rate": 7.953363317198287e-07,
      "loss": 0.0016,
      "num_tokens": 23705888.0,
      "reward": 1.1597223281860352,
      "reward_std": 0.12751007080078125,
      "rewards/reasoning_reward/mean": 1.1597222089767456,
      "rewards/reasoning_reward/std": 0.3126911520957947,
      "step": 290
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 406.0,
      "completions/max_terminated_length": 406.0,
      "completions/mean_length": 231.45834350585938,
      "completions/mean_terminated_length": 231.45834350585938,
      "completions/min_length": 142.0,
      "completions/min_terminated_length": 142.0,
      "epoch": 0.30093071354705275,
      "grad_norm": 2.7320267048254983,
      "kl": 0.046630859375,
      "learning_rate": 7.940240274824519e-07,
      "loss": 0.0019,
      "num_tokens": 23786971.0,
      "reward": 0.8125,
      "reward_std": 0.2836732268333435,
      "rewards/reasoning_reward/mean": 0.8125,
      "rewards/reasoning_reward/std": 0.44589513540267944,
      "step": 291
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 260.0,
      "completions/max_terminated_length": 260.0,
      "completions/mean_length": 148.83334350585938,
      "completions/mean_terminated_length": 148.83334350585938,
      "completions/min_length": 74.0,
      "completions/min_terminated_length": 74.0,
      "epoch": 0.3019648397104447,
      "grad_norm": 3.1680818855775654,
      "kl": 0.0556640625,
      "learning_rate": 7.927086199059457e-07,
      "loss": 0.0022,
      "num_tokens": 23866479.0,
      "reward": 1.1875,
      "reward_std": 0.24185511469841003,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.4618605971336365,
      "step": 292
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 248.0,
      "completions/max_terminated_length": 248.0,
      "completions/mean_length": 156.625,
      "completions/mean_terminated_length": 156.625,
      "completions/min_length": 73.0,
      "completions/min_terminated_length": 73.0,
      "epoch": 0.3029989658738366,
      "grad_norm": 1.8750379438190574,
      "kl": 0.045166015625,
      "learning_rate": 7.913901228740589e-07,
      "loss": 0.0018,
      "num_tokens": 23952070.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.550032913684845,
      "step": 293
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 218.0,
      "completions/max_terminated_length": 218.0,
      "completions/mean_length": 138.58334350585938,
      "completions/mean_terminated_length": 138.58334350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.30403309203722856,
      "grad_norm": 3.2244709897237893,
      "kl": 0.0419921875,
      "learning_rate": 7.90068550303149e-07,
      "loss": 0.0017,
      "num_tokens": 24034804.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 294
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 143.20834350585938,
      "completions/mean_terminated_length": 143.20834350585938,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.3050672182006205,
      "grad_norm": 3.2322747610220564,
      "kl": 0.06494140625,
      "learning_rate": 7.887439161420346e-07,
      "loss": 0.0026,
      "num_tokens": 24115961.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.21026216447353363,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.6743220090866089,
      "step": 295
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 151.0,
      "completions/max_terminated_length": 151.0,
      "completions/mean_length": 102.5,
      "completions/mean_terminated_length": 102.5,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.30610134436401243,
      "grad_norm": 4.065413966797424,
      "kl": 0.050048828125,
      "learning_rate": 7.874162343718489e-07,
      "loss": 0.002,
      "num_tokens": 24197397.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 296
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 223.0,
      "completions/max_terminated_length": 223.0,
      "completions/mean_length": 156.58334350585938,
      "completions/mean_terminated_length": 156.58334350585938,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.30713547052740436,
      "grad_norm": 35.175482523866044,
      "kl": 0.20703125,
      "learning_rate": 7.860855190058913e-07,
      "loss": 0.0083,
      "num_tokens": 24281619.0,
      "reward": 1.0625,
      "reward_std": 0.4130779206752777,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.6806725859642029,
      "step": 297
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 211.0,
      "completions/max_terminated_length": 211.0,
      "completions/mean_length": 143.125,
      "completions/mean_terminated_length": 143.125,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.3081695966907963,
      "grad_norm": 3.422532515079038,
      "kl": 0.0654296875,
      "learning_rate": 7.847517840894803e-07,
      "loss": 0.0026,
      "num_tokens": 24361918.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.23709973692893982,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.40322521328926086,
      "step": 298
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 332.0,
      "completions/max_terminated_length": 332.0,
      "completions/mean_length": 171.08334350585938,
      "completions/mean_terminated_length": 171.08334350585938,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.30920372285418823,
      "grad_norm": 4.1270544852144475,
      "kl": 0.0498046875,
      "learning_rate": 7.834150436998046e-07,
      "loss": 0.002,
      "num_tokens": 24443768.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.4082317352294922,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.7937139868736267,
      "step": 299
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 316.0,
      "completions/max_terminated_length": 316.0,
      "completions/mean_length": 138.45834350585938,
      "completions/mean_terminated_length": 138.45834350585938,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.31023784901758017,
      "grad_norm": 3.611002627039542,
      "kl": 0.0537109375,
      "learning_rate": 7.820753119457751e-07,
      "loss": 0.0021,
      "num_tokens": 24521315.0,
      "reward": 0.6458333730697632,
      "reward_std": 0.1767766922712326,
      "rewards/reasoning_reward/mean": 0.6458333134651184,
      "rewards/reasoning_reward/std": 0.47729232907295227,
      "step": 300
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 149.20834350585938,
      "completions/mean_terminated_length": 149.20834350585938,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.3112719751809721,
      "grad_norm": 3.152472722978425,
      "kl": 0.068359375,
      "learning_rate": 7.807326029678753e-07,
      "loss": 0.0027,
      "num_tokens": 24603760.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.36585909128189087,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.8164966702461243,
      "step": 301
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 359.0,
      "completions/max_terminated_length": 359.0,
      "completions/mean_length": 205.4166717529297,
      "completions/mean_terminated_length": 205.4166717529297,
      "completions/min_length": 138.0,
      "completions/min_terminated_length": 138.0,
      "epoch": 0.31230610134436404,
      "grad_norm": 2.786475190664922,
      "kl": 0.051025390625,
      "learning_rate": 7.793869309380128e-07,
      "loss": 0.002,
      "num_tokens": 24683322.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.42052432894706726,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.6538625359535217,
      "step": 302
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 347.0,
      "completions/max_terminated_length": 347.0,
      "completions/mean_length": 180.625,
      "completions/mean_terminated_length": 180.625,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.3133402275077559,
      "grad_norm": 4.57864897319015,
      "kl": 0.09521484375,
      "learning_rate": 7.780383100593692e-07,
      "loss": 0.0038,
      "num_tokens": 24766529.0,
      "reward": 1.4583333730697632,
      "reward_std": 0.35799640417099,
      "rewards/reasoning_reward/mean": 1.4583333730697632,
      "rewards/reasoning_reward/std": 0.4402732849121094,
      "step": 303
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 261.0,
      "completions/max_terminated_length": 261.0,
      "completions/mean_length": 163.95834350585938,
      "completions/mean_terminated_length": 163.95834350585938,
      "completions/min_length": 119.0,
      "completions/min_terminated_length": 119.0,
      "epoch": 0.31437435367114785,
      "grad_norm": 3.623601369487776,
      "kl": 0.053955078125,
      "learning_rate": 7.766867545662506e-07,
      "loss": 0.0022,
      "num_tokens": 24850544.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.31726133823394775,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.5800893306732178,
      "step": 304
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 314.0,
      "completions/max_terminated_length": 314.0,
      "completions/mean_length": 138.9166717529297,
      "completions/mean_terminated_length": 138.9166717529297,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.3154084798345398,
      "grad_norm": 3.8825218639045667,
      "kl": 0.0732421875,
      "learning_rate": 7.753322787239365e-07,
      "loss": 0.0029,
      "num_tokens": 24931478.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.31285393238067627,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.32693126797676086,
      "step": 305
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 252.0,
      "completions/max_terminated_length": 252.0,
      "completions/mean_length": 164.4166717529297,
      "completions/mean_terminated_length": 164.4166717529297,
      "completions/min_length": 62.0,
      "completions/min_terminated_length": 62.0,
      "epoch": 0.3164426059979317,
      "grad_norm": 2.1908650833519956,
      "kl": 0.048828125,
      "learning_rate": 7.739748968285305e-07,
      "loss": 0.0019,
      "num_tokens": 25008608.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.318511039018631,
      "step": 306
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 194.0,
      "completions/max_terminated_length": 194.0,
      "completions/mean_length": 145.4166717529297,
      "completions/mean_terminated_length": 145.4166717529297,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.31747673216132366,
      "grad_norm": 4.314449957580246,
      "kl": 0.07275390625,
      "learning_rate": 7.726146232068083e-07,
      "loss": 0.0029,
      "num_tokens": 25090962.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.5232069492340088,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.7172814607620239,
      "step": 307
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 289.0,
      "completions/max_terminated_length": 289.0,
      "completions/mean_length": 162.0416717529297,
      "completions/mean_terminated_length": 162.0416717529297,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.3185108583247156,
      "grad_norm": 3.710576665295931,
      "kl": 0.05859375,
      "learning_rate": 7.712514722160673e-07,
      "loss": 0.0023,
      "num_tokens": 25179115.0,
      "reward": 1.5416667461395264,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.5416666269302368,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 308
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 272.0,
      "completions/max_terminated_length": 272.0,
      "completions/mean_length": 160.83334350585938,
      "completions/mean_terminated_length": 160.83334350585938,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.31954498448810753,
      "grad_norm": 4.256835002888248,
      "kl": 0.09521484375,
      "learning_rate": 7.698854582439744e-07,
      "loss": 0.0038,
      "num_tokens": 25267727.0,
      "reward": 1.0555555820465088,
      "reward_std": 0.6559478044509888,
      "rewards/reasoning_reward/mean": 1.0555554628372192,
      "rewards/reasoning_reward/std": 0.7656052112579346,
      "step": 309
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 305.0,
      "completions/max_terminated_length": 305.0,
      "completions/mean_length": 188.5416717529297,
      "completions/mean_terminated_length": 188.5416717529297,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.32057911065149947,
      "grad_norm": 1.8170985697472053,
      "kl": 0.052001953125,
      "learning_rate": 7.685165957084147e-07,
      "loss": 0.0021,
      "num_tokens": 25358292.0,
      "reward": 0.625,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.5366967916488647,
      "step": 310
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 222.0,
      "completions/max_terminated_length": 222.0,
      "completions/mean_length": 155.1666717529297,
      "completions/mean_terminated_length": 155.1666717529297,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.3216132368148914,
      "grad_norm": 3.812637593183448,
      "kl": 0.08935546875,
      "learning_rate": 7.671448990573391e-07,
      "loss": 0.0036,
      "num_tokens": 25446560.0,
      "reward": 1.2708333730697632,
      "reward_std": 0.5512506365776062,
      "rewards/reasoning_reward/mean": 1.2708333730697632,
      "rewards/reasoning_reward/std": 0.6753286719322205,
      "step": 311
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 351.0,
      "completions/max_terminated_length": 351.0,
      "completions/mean_length": 170.20834350585938,
      "completions/mean_terminated_length": 170.20834350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.32264736297828334,
      "grad_norm": 2.218366125481505,
      "kl": 0.06689453125,
      "learning_rate": 7.657703827686115e-07,
      "loss": 0.0027,
      "num_tokens": 25523845.0,
      "reward": 0.6875,
      "reward_std": 0.25877460837364197,
      "rewards/reasoning_reward/mean": 0.6875,
      "rewards/reasoning_reward/std": 0.6562823057174683,
      "step": 312
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 275.0,
      "completions/max_terminated_length": 275.0,
      "completions/mean_length": 179.6666717529297,
      "completions/mean_terminated_length": 179.6666717529297,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.32368148914167527,
      "grad_norm": 3.4872202871188547,
      "kl": 0.06298828125,
      "learning_rate": 7.643930613498561e-07,
      "loss": 0.0025,
      "num_tokens": 25606917.0,
      "reward": 1.1736111640930176,
      "reward_std": 0.3726871907711029,
      "rewards/reasoning_reward/mean": 1.173611044883728,
      "rewards/reasoning_reward/std": 0.464901328086853,
      "step": 313
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 166.33334350585938,
      "completions/mean_terminated_length": 166.33334350585938,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.3247156153050672,
      "grad_norm": 3.037599272229024,
      "kl": 0.06982421875,
      "learning_rate": 7.630129493383052e-07,
      "loss": 0.0028,
      "num_tokens": 25685437.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.23709973692893982,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.6833289265632629,
      "step": 314
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 245.0,
      "completions/max_terminated_length": 245.0,
      "completions/mean_length": 146.95834350585938,
      "completions/mean_terminated_length": 146.95834350585938,
      "completions/min_length": 74.0,
      "completions/min_terminated_length": 74.0,
      "epoch": 0.32574974146845914,
      "grad_norm": 4.078205524302534,
      "kl": 0.047119140625,
      "learning_rate": 7.616300613006442e-07,
      "loss": 0.0019,
      "num_tokens": 25765932.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.48112308979034424,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 315
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 272.0,
      "completions/max_terminated_length": 272.0,
      "completions/mean_length": 133.7916717529297,
      "completions/mean_terminated_length": 133.7916717529297,
      "completions/min_length": 63.0,
      "completions/min_terminated_length": 63.0,
      "epoch": 0.3267838676318511,
      "grad_norm": 0.2821083532299412,
      "kl": 0.06787109375,
      "learning_rate": 7.602444118328592e-07,
      "loss": 0.0027,
      "num_tokens": 25840439.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 316
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 134.875,
      "completions/mean_terminated_length": 134.875,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.327817993795243,
      "grad_norm": 3.5599707379709162,
      "kl": 0.091796875,
      "learning_rate": 7.588560155600823e-07,
      "loss": 0.0037,
      "num_tokens": 25921308.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.43601590394973755,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.7587054967880249,
      "step": 317
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 304.0,
      "completions/max_terminated_length": 304.0,
      "completions/mean_length": 153.6666717529297,
      "completions/mean_terminated_length": 153.6666717529297,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.32885211995863495,
      "grad_norm": 3.1856291943344077,
      "kl": 0.07666015625,
      "learning_rate": 7.574648871364368e-07,
      "loss": 0.0031,
      "num_tokens": 26000556.0,
      "reward": 1.2013888359069824,
      "reward_std": 0.1516503393650055,
      "rewards/reasoning_reward/mean": 1.2013888359069824,
      "rewards/reasoning_reward/std": 0.3068428933620453,
      "step": 318
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 228.0,
      "completions/max_terminated_length": 228.0,
      "completions/mean_length": 152.1666717529297,
      "completions/mean_terminated_length": 152.1666717529297,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.3298862461220269,
      "grad_norm": 3.283679581462454,
      "kl": 0.07080078125,
      "learning_rate": 7.560710412448838e-07,
      "loss": 0.0028,
      "num_tokens": 26080200.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.32520395517349243,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.416485458612442,
      "step": 319
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 260.0,
      "completions/max_terminated_length": 260.0,
      "completions/mean_length": 164.875,
      "completions/mean_terminated_length": 164.875,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.3309203722854188,
      "grad_norm": 4.160648948343019,
      "kl": 0.054443359375,
      "learning_rate": 7.546744925970664e-07,
      "loss": 0.0022,
      "num_tokens": 26157397.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.3205421268939972,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.35864076018333435,
      "step": 320
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 339.0,
      "completions/max_terminated_length": 339.0,
      "completions/mean_length": 214.25,
      "completions/mean_terminated_length": 209.0869598388672,
      "completions/min_length": 123.0,
      "completions/min_terminated_length": 123.0,
      "epoch": 0.33195449844881075,
      "grad_norm": 3.5547184484445915,
      "kl": 0.07373046875,
      "learning_rate": 7.532752559331539e-07,
      "loss": 0.003,
      "num_tokens": 26235427.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 321
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 307.0,
      "completions/max_terminated_length": 307.0,
      "completions/mean_length": 175.25,
      "completions/mean_terminated_length": 175.25,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.3329886246122027,
      "grad_norm": 4.066280261564092,
      "kl": 0.0791015625,
      "learning_rate": 7.518733460216875e-07,
      "loss": 0.0032,
      "num_tokens": 26319289.0,
      "reward": 1.125,
      "reward_std": 0.499225378036499,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.6298723220825195,
      "step": 322
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 256.0,
      "completions/max_terminated_length": 256.0,
      "completions/mean_length": 158.33334350585938,
      "completions/mean_terminated_length": 158.33334350585938,
      "completions/min_length": 64.0,
      "completions/min_terminated_length": 64.0,
      "epoch": 0.3340227507755946,
      "grad_norm": 2.980136617187653,
      "kl": 0.051025390625,
      "learning_rate": 7.504687776594234e-07,
      "loss": 0.002,
      "num_tokens": 26395665.0,
      "reward": 0.9375,
      "reward_std": 0.3255884051322937,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.39870715141296387,
      "step": 323
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 219.0,
      "completions/max_terminated_length": 219.0,
      "completions/mean_length": 136.83334350585938,
      "completions/mean_terminated_length": 136.83334350585938,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.33505687693898656,
      "grad_norm": 3.0212039876952943,
      "kl": 0.0576171875,
      "learning_rate": 7.490615656711771e-07,
      "loss": 0.0023,
      "num_tokens": 26472917.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 324
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 257.0,
      "completions/max_terminated_length": 257.0,
      "completions/mean_length": 172.125,
      "completions/mean_terminated_length": 172.125,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.3360910031023785,
      "grad_norm": 3.015813746260023,
      "kl": 0.07568359375,
      "learning_rate": 7.47651724909667e-07,
      "loss": 0.003,
      "num_tokens": 26557080.0,
      "reward": 0.6041666865348816,
      "reward_std": 0.3116035461425781,
      "rewards/reasoning_reward/mean": 0.6041666865348816,
      "rewards/reasoning_reward/std": 0.7220015525817871,
      "step": 325
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 617.0,
      "completions/max_terminated_length": 617.0,
      "completions/mean_length": 188.75,
      "completions/mean_terminated_length": 188.75,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.33712512926577043,
      "grad_norm": 4.08859100969837,
      "kl": 0.06494140625,
      "learning_rate": 7.46239270255357e-07,
      "loss": 0.0026,
      "num_tokens": 26641714.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.39000558853149414,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 326
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 342.0,
      "completions/max_terminated_length": 342.0,
      "completions/mean_length": 181.6666717529297,
      "completions/mean_terminated_length": 181.6666717529297,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.33815925542916236,
      "grad_norm": 4.0866348173332545,
      "kl": 0.07080078125,
      "learning_rate": 7.448242166163003e-07,
      "loss": 0.0028,
      "num_tokens": 26724962.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.32801350951194763,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.48900964856147766,
      "step": 327
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 248.0,
      "completions/max_terminated_length": 248.0,
      "completions/mean_length": 152.625,
      "completions/mean_terminated_length": 152.625,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.3391933815925543,
      "grad_norm": 2.1510475663169353,
      "kl": 0.057373046875,
      "learning_rate": 7.434065789279815e-07,
      "loss": 0.0023,
      "num_tokens": 26803649.0,
      "reward": 0.875,
      "reward_std": 0.14773420989513397,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.7260674238204956,
      "step": 328
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 159.9166717529297,
      "completions/mean_terminated_length": 159.9166717529297,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.34022750775594623,
      "grad_norm": 3.4038270866830955,
      "kl": 0.07275390625,
      "learning_rate": 7.41986372153159e-07,
      "loss": 0.0029,
      "num_tokens": 26895591.0,
      "reward": 1.375,
      "reward_std": 0.273722380399704,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.4484272003173828,
      "step": 329
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 224.0,
      "completions/max_terminated_length": 224.0,
      "completions/mean_length": 157.625,
      "completions/mean_terminated_length": 157.625,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.34126163391933817,
      "grad_norm": 3.2332472627018083,
      "kl": 0.060791015625,
      "learning_rate": 7.405636112817071e-07,
      "loss": 0.0024,
      "num_tokens": 26974222.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.35864076018333435,
      "step": 330
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 252.0,
      "completions/max_terminated_length": 252.0,
      "completions/mean_length": 164.45834350585938,
      "completions/mean_terminated_length": 164.45834350585938,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.3422957600827301,
      "grad_norm": 3.3744277472463686,
      "kl": 0.07275390625,
      "learning_rate": 7.391383113304583e-07,
      "loss": 0.0029,
      "num_tokens": 27058377.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.3508317470550537,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.6833289265632629,
      "step": 331
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 189.0,
      "completions/max_terminated_length": 189.0,
      "completions/mean_length": 134.7916717529297,
      "completions/mean_terminated_length": 134.7916717529297,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.34332988624612204,
      "grad_norm": 3.766622375322823,
      "kl": 0.08056640625,
      "learning_rate": 7.377104873430438e-07,
      "loss": 0.0032,
      "num_tokens": 27138204.0,
      "reward": 0.75,
      "reward_std": 0.34503278136253357,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 332
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 177.0,
      "completions/max_terminated_length": 177.0,
      "completions/mean_length": 134.4166717529297,
      "completions/mean_terminated_length": 134.4166717529297,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.344364012409514,
      "grad_norm": 3.8370132926239346,
      "kl": 0.0859375,
      "learning_rate": 7.362801543897357e-07,
      "loss": 0.0034,
      "num_tokens": 27219846.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.23144195973873138,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.7220014929771423,
      "step": 333
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 259.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 144.625,
      "completions/mean_terminated_length": 139.6521759033203,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.3453981385729059,
      "grad_norm": 2.1634706010744544,
      "kl": 0.058349609375,
      "learning_rate": 7.348473275672873e-07,
      "loss": 0.0023,
      "num_tokens": 27300293.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.18722420930862427,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.31565436720848083,
      "step": 334
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 261.0,
      "completions/max_terminated_length": 261.0,
      "completions/mean_length": 166.625,
      "completions/mean_terminated_length": 166.625,
      "completions/min_length": 69.0,
      "completions/min_terminated_length": 69.0,
      "epoch": 0.34643226473629785,
      "grad_norm": 3.494854866170849,
      "kl": 0.0751953125,
      "learning_rate": 7.334120219987741e-07,
      "loss": 0.003,
      "num_tokens": 27379068.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 335
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 243.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 162.75,
      "completions/mean_terminated_length": 162.75,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.3474663908996898,
      "grad_norm": 0.35319764403800336,
      "kl": 0.0810546875,
      "learning_rate": 7.319742528334339e-07,
      "loss": 0.0032,
      "num_tokens": 27455982.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 336
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 167.0,
      "completions/max_terminated_length": 167.0,
      "completions/mean_length": 126.70833587646484,
      "completions/mean_terminated_length": 126.70833587646484,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.3485005170630817,
      "grad_norm": 4.842479635487001,
      "kl": 0.053955078125,
      "learning_rate": 7.305340352465071e-07,
      "loss": 0.0022,
      "num_tokens": 27535311.0,
      "reward": 0.75,
      "reward_std": 0.41387641429901123,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 337
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 228.0,
      "completions/max_terminated_length": 228.0,
      "completions/mean_length": 144.6666717529297,
      "completions/mean_terminated_length": 144.6666717529297,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.34953464322647365,
      "grad_norm": 3.960259845440186,
      "kl": 0.0625,
      "learning_rate": 7.290913844390765e-07,
      "loss": 0.0025,
      "num_tokens": 27613047.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 338
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 187.0,
      "completions/max_terminated_length": 187.0,
      "completions/mean_length": 127.20833587646484,
      "completions/mean_terminated_length": 127.20833587646484,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.3505687693898656,
      "grad_norm": 2.53370412691966,
      "kl": 0.052001953125,
      "learning_rate": 7.276463156379069e-07,
      "loss": 0.0021,
      "num_tokens": 27691236.0,
      "reward": 0.625,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 339
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 169.0,
      "completions/mean_terminated_length": 167.56521606445312,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.3516028955532575,
      "grad_norm": 13.18852870542549,
      "kl": 0.390625,
      "learning_rate": 7.261988440952844e-07,
      "loss": 0.0157,
      "num_tokens": 27775724.0,
      "reward": 1.4305557012557983,
      "reward_std": 0.40538716316223145,
      "rewards/reasoning_reward/mean": 1.4305557012557983,
      "rewards/reasoning_reward/std": 0.648142397403717,
      "step": 340
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 285.0,
      "completions/max_terminated_length": 285.0,
      "completions/mean_length": 186.875,
      "completions/mean_terminated_length": 186.875,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.35263702171664946,
      "grad_norm": 3.7251808322917257,
      "kl": 0.08154296875,
      "learning_rate": 7.247489850888551e-07,
      "loss": 0.0033,
      "num_tokens": 27857497.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.3808860182762146,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.40824830532073975,
      "step": 341
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 286.0,
      "completions/max_terminated_length": 286.0,
      "completions/mean_length": 146.125,
      "completions/mean_terminated_length": 146.125,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.3536711478800414,
      "grad_norm": 4.446040086211323,
      "kl": 0.06884765625,
      "learning_rate": 7.232967539214643e-07,
      "loss": 0.0028,
      "num_tokens": 27938476.0,
      "reward": 1.5416667461395264,
      "reward_std": 0.2314550280570984,
      "rewards/reasoning_reward/mean": 1.5416666269302368,
      "rewards/reasoning_reward/std": 0.4871537983417511,
      "step": 342
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 165.0,
      "completions/max_terminated_length": 165.0,
      "completions/mean_length": 122.0,
      "completions/mean_terminated_length": 122.0,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.35470527404343327,
      "grad_norm": 4.159860302213038,
      "kl": 0.04296875,
      "learning_rate": 7.218421659209948e-07,
      "loss": 0.0017,
      "num_tokens": 28018204.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.46854168176651,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 343
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 210.0,
      "completions/max_terminated_length": 210.0,
      "completions/mean_length": 135.9166717529297,
      "completions/mean_terminated_length": 137.04348754882812,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.3557394002068252,
      "grad_norm": 3.9456854124043956,
      "kl": 0.087890625,
      "learning_rate": 7.203852364402048e-07,
      "loss": 0.0035,
      "num_tokens": 28103106.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 344
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 273.0,
      "completions/max_terminated_length": 273.0,
      "completions/mean_length": 159.70834350585938,
      "completions/mean_terminated_length": 159.70834350585938,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.35677352637021714,
      "grad_norm": 3.8039878218613024,
      "kl": 0.07275390625,
      "learning_rate": 7.189259808565664e-07,
      "loss": 0.0029,
      "num_tokens": 28184387.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 345
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 242.0,
      "completions/max_terminated_length": 242.0,
      "completions/mean_length": 154.95834350585938,
      "completions/mean_terminated_length": 154.95834350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.3578076525336091,
      "grad_norm": 3.8927448717227846,
      "kl": 0.07373046875,
      "learning_rate": 7.174644145721031e-07,
      "loss": 0.003,
      "num_tokens": 28262346.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.39000558853149414,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 346
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 174.33334350585938,
      "completions/mean_terminated_length": 174.33334350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.358841778697001,
      "grad_norm": 3.9724913747791044,
      "kl": 0.07373046875,
      "learning_rate": 7.16000553013227e-07,
      "loss": 0.003,
      "num_tokens": 28338434.0,
      "reward": 0.6041666865348816,
      "reward_std": 0.4981047809123993,
      "rewards/reasoning_reward/mean": 0.6041666865348816,
      "rewards/reasoning_reward/std": 0.5311833620071411,
      "step": 347
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 243.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 154.9166717529297,
      "completions/mean_terminated_length": 154.9166717529297,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.35987590486039295,
      "grad_norm": 2.0828938523114524,
      "kl": 0.07763671875,
      "learning_rate": 7.145344116305762e-07,
      "loss": 0.0031,
      "num_tokens": 28415024.0,
      "reward": 1.125,
      "reward_std": 0.10603483021259308,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.25180506706237793,
      "step": 348
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 159.2916717529297,
      "completions/mean_terminated_length": 159.2916717529297,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.3609100310237849,
      "grad_norm": 4.458540222973645,
      "kl": 0.11328125,
      "learning_rate": 7.13066005898852e-07,
      "loss": 0.0045,
      "num_tokens": 28497783.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.6395318508148193,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.6412736177444458,
      "step": 349
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 222.0,
      "completions/max_terminated_length": 222.0,
      "completions/mean_length": 161.875,
      "completions/mean_terminated_length": 161.875,
      "completions/min_length": 114.0,
      "completions/min_terminated_length": 114.0,
      "epoch": 0.3619441571871768,
      "grad_norm": 3.7766746412265,
      "kl": 0.125,
      "learning_rate": 7.115953513166549e-07,
      "loss": 0.005,
      "num_tokens": 28574772.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 350
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 202.0,
      "completions/max_terminated_length": 202.0,
      "completions/mean_length": 114.16667175292969,
      "completions/mean_terminated_length": 114.16667175292969,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.36297828335056875,
      "grad_norm": 3.461708376714296,
      "kl": 0.0732421875,
      "learning_rate": 7.101224634063212e-07,
      "loss": 0.0029,
      "num_tokens": 28659984.0,
      "reward": 0.993055522441864,
      "reward_std": 0.18709687888622284,
      "rewards/reasoning_reward/mean": 0.993055522441864,
      "rewards/reasoning_reward/std": 0.7557815313339233,
      "step": 351
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 127.95833587646484,
      "completions/mean_terminated_length": 127.95833587646484,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.3640124095139607,
      "grad_norm": 2.183902034581425,
      "kl": 0.06396484375,
      "learning_rate": 7.086473577137598e-07,
      "loss": 0.0026,
      "num_tokens": 28750087.0,
      "reward": 1.5,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 1.5,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 352
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 208.0,
      "completions/max_terminated_length": 208.0,
      "completions/mean_length": 137.7916717529297,
      "completions/mean_terminated_length": 137.7916717529297,
      "completions/min_length": 70.0,
      "completions/min_terminated_length": 70.0,
      "epoch": 0.3650465356773526,
      "grad_norm": 3.440182240952645,
      "kl": 0.07421875,
      "learning_rate": 7.071700498082873e-07,
      "loss": 0.003,
      "num_tokens": 28835346.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.4082186222076416,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 353
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 215.0,
      "completions/max_terminated_length": 215.0,
      "completions/mean_length": 159.70834350585938,
      "completions/mean_terminated_length": 159.70834350585938,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.36608066184074456,
      "grad_norm": 3.7647272928819335,
      "kl": 0.083984375,
      "learning_rate": 7.056905552824644e-07,
      "loss": 0.0033,
      "num_tokens": 28919371.0,
      "reward": 1.0625,
      "reward_std": 0.3944129943847656,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.5954993963241577,
      "step": 354
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 150.4166717529297,
      "completions/mean_terminated_length": 150.4166717529297,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.3671147880041365,
      "grad_norm": 3.3768970946788697,
      "kl": 0.059814453125,
      "learning_rate": 7.042088897519307e-07,
      "loss": 0.0024,
      "num_tokens": 28997701.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3535533845424652,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4871538281440735,
      "step": 355
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 243.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 163.95834350585938,
      "completions/mean_terminated_length": 163.95834350585938,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.36814891416752843,
      "grad_norm": 4.085471865989243,
      "kl": 0.10986328125,
      "learning_rate": 7.027250688552399e-07,
      "loss": 0.0044,
      "num_tokens": 29086372.0,
      "reward": 1.5416667461395264,
      "reward_std": 0.49276697635650635,
      "rewards/reasoning_reward/mean": 1.5416666269302368,
      "rewards/reasoning_reward/std": 0.4871537983417511,
      "step": 356
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 310.0,
      "completions/max_terminated_length": 310.0,
      "completions/mean_length": 160.75,
      "completions/mean_terminated_length": 160.75,
      "completions/min_length": 69.0,
      "completions/min_terminated_length": 69.0,
      "epoch": 0.36918304033092036,
      "grad_norm": 3.1239460885673687,
      "kl": 0.055908203125,
      "learning_rate": 7.012391082536955e-07,
      "loss": 0.0022,
      "num_tokens": 29163254.0,
      "reward": 1.1388888359069824,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 1.1388888359069824,
      "rewards/reasoning_reward/std": 0.6052640676498413,
      "step": 357
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 251.0,
      "completions/max_terminated_length": 251.0,
      "completions/mean_length": 181.9166717529297,
      "completions/mean_terminated_length": 181.9166717529297,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.3702171664943123,
      "grad_norm": 3.5139169095587626,
      "kl": 0.091796875,
      "learning_rate": 6.997510236311846e-07,
      "loss": 0.0037,
      "num_tokens": 29245260.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.49076026678085327,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.5694518089294434,
      "step": 358
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 174.0,
      "completions/max_terminated_length": 174.0,
      "completions/mean_length": 127.41667175292969,
      "completions/mean_terminated_length": 127.41667175292969,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.37125129265770423,
      "grad_norm": 3.785316738310359,
      "kl": 0.06396484375,
      "learning_rate": 6.982608306940128e-07,
      "loss": 0.0025,
      "num_tokens": 29322550.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148510992527008,
      "step": 359
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 281.0,
      "completions/max_terminated_length": 281.0,
      "completions/mean_length": 177.58334350585938,
      "completions/mean_terminated_length": 177.58334350585938,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.37228541882109617,
      "grad_norm": 2.241778037699298,
      "kl": 0.09375,
      "learning_rate": 6.967685451707383e-07,
      "loss": 0.0037,
      "num_tokens": 29399980.0,
      "reward": 0.3333333432674408,
      "reward_std": 0.2182178944349289,
      "rewards/reasoning_reward/mean": 0.3333333432674408,
      "rewards/reasoning_reward/std": 0.601929247379303,
      "step": 360
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 178.95834350585938,
      "completions/mean_terminated_length": 178.95834350585938,
      "completions/min_length": 129.0,
      "completions/min_terminated_length": 129.0,
      "epoch": 0.3733195449844881,
      "grad_norm": 0.25717311965945855,
      "kl": 0.07666015625,
      "learning_rate": 6.952741828120062e-07,
      "loss": 0.0031,
      "num_tokens": 29484387.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.637022078037262,
      "step": 361
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 280.0,
      "completions/max_terminated_length": 280.0,
      "completions/mean_length": 159.875,
      "completions/mean_terminated_length": 159.875,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.37435367114788004,
      "grad_norm": 4.722848836354921,
      "kl": 0.09765625,
      "learning_rate": 6.937777593903817e-07,
      "loss": 0.0039,
      "num_tokens": 29568552.0,
      "reward": 1.1875,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.28788962960243225,
      "step": 362
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 222.0,
      "completions/max_terminated_length": 222.0,
      "completions/mean_length": 152.70834350585938,
      "completions/mean_terminated_length": 152.70834350585938,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.375387797311272,
      "grad_norm": 3.6743007364748133,
      "kl": 0.062255859375,
      "learning_rate": 6.922792907001842e-07,
      "loss": 0.0025,
      "num_tokens": 29646721.0,
      "reward": 0.4166666865348816,
      "reward_std": 0.33247750997543335,
      "rewards/reasoning_reward/mean": 0.4166666567325592,
      "rewards/reasoning_reward/std": 0.5036101937294006,
      "step": 363
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 282.0,
      "completions/max_terminated_length": 282.0,
      "completions/mean_length": 181.08334350585938,
      "completions/mean_terminated_length": 181.08334350585938,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.3764219234746639,
      "grad_norm": 3.384160520560316,
      "kl": 0.095703125,
      "learning_rate": 6.9077879255732e-07,
      "loss": 0.0038,
      "num_tokens": 29731443.0,
      "reward": 1.2222222089767456,
      "reward_std": 0.4460780620574951,
      "rewards/reasoning_reward/mean": 1.2222222089767456,
      "rewards/reasoning_reward/std": 0.5787431597709656,
      "step": 364
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 150.75,
      "completions/mean_terminated_length": 150.75,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.37745604963805585,
      "grad_norm": 3.8887377118282904,
      "kl": 0.087890625,
      "learning_rate": 6.892762807991159e-07,
      "loss": 0.0035,
      "num_tokens": 29814029.0,
      "reward": 1.0069444179534912,
      "reward_std": 0.44658660888671875,
      "rewards/reasoning_reward/mean": 1.0069444179534912,
      "rewards/reasoning_reward/std": 0.7573778629302979,
      "step": 365
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 184.0,
      "completions/max_terminated_length": 184.0,
      "completions/mean_length": 120.66667175292969,
      "completions/mean_terminated_length": 120.66667175292969,
      "completions/min_length": 59.0,
      "completions/min_terminated_length": 59.0,
      "epoch": 0.3784901758014478,
      "grad_norm": 2.581428574067341,
      "kl": 0.08203125,
      "learning_rate": 6.87771771284152e-07,
      "loss": 0.0033,
      "num_tokens": 29897901.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.40824830532073975,
      "step": 366
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 223.0,
      "completions/max_terminated_length": 223.0,
      "completions/mean_length": 165.1666717529297,
      "completions/mean_terminated_length": 165.1666717529297,
      "completions/min_length": 122.0,
      "completions/min_terminated_length": 122.0,
      "epoch": 0.3795243019648397,
      "grad_norm": 2.9198024039411745,
      "kl": 0.0634765625,
      "learning_rate": 6.862652798920938e-07,
      "loss": 0.0025,
      "num_tokens": 29976281.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.27053868770599365,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.34512653946876526,
      "step": 367
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 331.0,
      "completions/max_terminated_length": 331.0,
      "completions/mean_length": 193.375,
      "completions/mean_terminated_length": 193.375,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.38055842812823165,
      "grad_norm": 2.9072434267769616,
      "kl": 0.09326171875,
      "learning_rate": 6.84756822523525e-07,
      "loss": 0.0037,
      "num_tokens": 30054058.0,
      "reward": 0.8125,
      "reward_std": 0.3433460593223572,
      "rewards/reasoning_reward/mean": 0.8125,
      "rewards/reasoning_reward/std": 0.4848240315914154,
      "step": 368
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 144.58334350585938,
      "completions/mean_terminated_length": 144.58334350585938,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.3815925542916236,
      "grad_norm": 3.5358384891912107,
      "kl": 0.08984375,
      "learning_rate": 6.832464150997798e-07,
      "loss": 0.0036,
      "num_tokens": 30137656.0,
      "reward": 1.2986111640930176,
      "reward_std": 0.20151451230049133,
      "rewards/reasoning_reward/mean": 1.298611044883728,
      "rewards/reasoning_reward/std": 0.41985490918159485,
      "step": 369
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 149.75,
      "completions/mean_terminated_length": 149.75,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.3826266804550155,
      "grad_norm": 0.19189388153538847,
      "kl": 0.06298828125,
      "learning_rate": 6.817340735627745e-07,
      "loss": 0.0025,
      "num_tokens": 30223114.0,
      "reward": 0.3333333432674408,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.3333333432674408,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 370
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 234.0,
      "completions/max_terminated_length": 234.0,
      "completions/mean_length": 138.375,
      "completions/mean_terminated_length": 138.375,
      "completions/min_length": 70.0,
      "completions/min_terminated_length": 70.0,
      "epoch": 0.38366080661840746,
      "grad_norm": 2.5621136837530125,
      "kl": 0.056640625,
      "learning_rate": 6.802198138748397e-07,
      "loss": 0.0023,
      "num_tokens": 30305003.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 371
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 243.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 168.875,
      "completions/mean_terminated_length": 168.875,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.3846949327817994,
      "grad_norm": 3.730162920596462,
      "kl": 0.095703125,
      "learning_rate": 6.78703652018551e-07,
      "loss": 0.0038,
      "num_tokens": 30387952.0,
      "reward": 1.2777777910232544,
      "reward_std": 0.4832340478897095,
      "rewards/reasoning_reward/mean": 1.2777777910232544,
      "rewards/reasoning_reward/std": 0.6287527680397034,
      "step": 372
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 228.0,
      "completions/max_terminated_length": 228.0,
      "completions/mean_length": 143.70834350585938,
      "completions/mean_terminated_length": 143.70834350585938,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.3857290589451913,
      "grad_norm": 2.227976386200522,
      "kl": 0.055908203125,
      "learning_rate": 6.771856039965615e-07,
      "loss": 0.0022,
      "num_tokens": 30474265.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.8064504265785217,
      "step": 373
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 215.0,
      "completions/max_terminated_length": 215.0,
      "completions/mean_length": 154.625,
      "completions/mean_terminated_length": 154.625,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.38676318510858326,
      "grad_norm": 3.974774217751018,
      "kl": 0.12451171875,
      "learning_rate": 6.756656858314318e-07,
      "loss": 0.005,
      "num_tokens": 30559688.0,
      "reward": 1.4513888359069824,
      "reward_std": 0.39070504903793335,
      "rewards/reasoning_reward/mean": 1.4513888359069824,
      "rewards/reasoning_reward/std": 0.5613973736763,
      "step": 374
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 238.0,
      "completions/max_terminated_length": 238.0,
      "completions/mean_length": 175.58334350585938,
      "completions/mean_terminated_length": 175.58334350585938,
      "completions/min_length": 132.0,
      "completions/min_terminated_length": 132.0,
      "epoch": 0.3877973112719752,
      "grad_norm": 2.8084365803963354,
      "kl": 0.08203125,
      "learning_rate": 6.741439135654612e-07,
      "loss": 0.0033,
      "num_tokens": 30639974.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.7019641399383545,
      "step": 375
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 228.0,
      "completions/max_terminated_length": 228.0,
      "completions/mean_length": 143.45834350585938,
      "completions/mean_terminated_length": 143.45834350585938,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.38883143743536713,
      "grad_norm": 3.1747783499460858,
      "kl": 0.07568359375,
      "learning_rate": 6.726203032605189e-07,
      "loss": 0.003,
      "num_tokens": 30721217.0,
      "reward": 0.75,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 376
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 215.0,
      "completions/max_terminated_length": 215.0,
      "completions/mean_length": 154.20834350585938,
      "completions/mean_terminated_length": 154.20834350585938,
      "completions/min_length": 70.0,
      "completions/min_terminated_length": 70.0,
      "epoch": 0.38986556359875907,
      "grad_norm": 2.466045662278419,
      "kl": 0.1279296875,
      "learning_rate": 6.710948709978741e-07,
      "loss": 0.0051,
      "num_tokens": 30804974.0,
      "reward": 0.8958333730697632,
      "reward_std": 0.15268757939338684,
      "rewards/reasoning_reward/mean": 0.8958333134651184,
      "rewards/reasoning_reward/std": 0.29411497712135315,
      "step": 377
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 249.0,
      "completions/max_terminated_length": 249.0,
      "completions/mean_length": 149.58334350585938,
      "completions/mean_terminated_length": 149.58334350585938,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.390899689762151,
      "grad_norm": 4.338062118799447,
      "kl": 0.06640625,
      "learning_rate": 6.695676328780256e-07,
      "loss": 0.0027,
      "num_tokens": 30886508.0,
      "reward": 0.9861111044883728,
      "reward_std": 0.46136391162872314,
      "rewards/reasoning_reward/mean": 0.9861111044883728,
      "rewards/reasoning_reward/std": 0.6406455039978027,
      "step": 378
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 158.45834350585938,
      "completions/mean_terminated_length": 158.45834350585938,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.39193381592554294,
      "grad_norm": 3.275227479605371,
      "kl": 0.10693359375,
      "learning_rate": 6.680386050205332e-07,
      "loss": 0.0043,
      "num_tokens": 30969583.0,
      "reward": 1.3958333730697632,
      "reward_std": 0.24185511469841003,
      "rewards/reasoning_reward/mean": 1.3958333730697632,
      "rewards/reasoning_reward/std": 0.4657664895057678,
      "step": 379
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 271.0,
      "completions/max_terminated_length": 271.0,
      "completions/mean_length": 160.6666717529297,
      "completions/mean_terminated_length": 160.6666717529297,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.3929679420889349,
      "grad_norm": 2.8767415849776694,
      "kl": 0.11279296875,
      "learning_rate": 6.665078035638465e-07,
      "loss": 0.0045,
      "num_tokens": 31046431.0,
      "reward": 0.9861111044883728,
      "reward_std": 0.2921907305717468,
      "rewards/reasoning_reward/mean": 0.9861111044883728,
      "rewards/reasoning_reward/std": 0.3867262005805969,
      "step": 380
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 264.0,
      "completions/max_terminated_length": 264.0,
      "completions/mean_length": 143.08334350585938,
      "completions/mean_terminated_length": 143.08334350585938,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.3940020682523268,
      "grad_norm": 2.3645686686795626,
      "kl": 0.09375,
      "learning_rate": 6.649752446651352e-07,
      "loss": 0.0037,
      "num_tokens": 31125817.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 381
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 157.33334350585938,
      "completions/mean_terminated_length": 157.33334350585938,
      "completions/min_length": 121.0,
      "completions/min_terminated_length": 121.0,
      "epoch": 0.39503619441571874,
      "grad_norm": 3.7353394391876265,
      "kl": 0.056884765625,
      "learning_rate": 6.634409445001181e-07,
      "loss": 0.0023,
      "num_tokens": 31206177.0,
      "reward": 0.75,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 382
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 251.0,
      "completions/max_terminated_length": 251.0,
      "completions/mean_length": 173.08334350585938,
      "completions/mean_terminated_length": 173.08334350585938,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.3960703205791106,
      "grad_norm": 3.022792894461881,
      "kl": 0.08740234375,
      "learning_rate": 6.619049192628924e-07,
      "loss": 0.0035,
      "num_tokens": 31287907.0,
      "reward": 0.375,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 383
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 182.0,
      "completions/max_terminated_length": 182.0,
      "completions/mean_length": 133.70834350585938,
      "completions/mean_terminated_length": 133.70834350585938,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.39710444674250256,
      "grad_norm": 2.7266764014944096,
      "kl": 0.0771484375,
      "learning_rate": 6.603671851657634e-07,
      "loss": 0.0031,
      "num_tokens": 31365276.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.12400396913290024,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.25448867678642273,
      "step": 384
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 273.0,
      "completions/max_terminated_length": 273.0,
      "completions/mean_length": 115.70833587646484,
      "completions/mean_terminated_length": 115.70833587646484,
      "completions/min_length": 53.0,
      "completions/min_terminated_length": 53.0,
      "epoch": 0.3981385729058945,
      "grad_norm": 3.4941613374593965,
      "kl": 0.09228515625,
      "learning_rate": 6.588277584390725e-07,
      "loss": 0.0037,
      "num_tokens": 31442029.0,
      "reward": 0.75,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 385
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 214.0,
      "completions/max_terminated_length": 214.0,
      "completions/mean_length": 135.83334350585938,
      "completions/mean_terminated_length": 135.83334350585938,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.39917269906928643,
      "grad_norm": 3.558030153703449,
      "kl": 0.09619140625,
      "learning_rate": 6.572866553310265e-07,
      "loss": 0.0039,
      "num_tokens": 31525481.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 386
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 174.0,
      "completions/max_terminated_length": 174.0,
      "completions/mean_length": 132.0,
      "completions/mean_terminated_length": 132.0,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.40020682523267836,
      "grad_norm": 3.432909868088115,
      "kl": 0.052734375,
      "learning_rate": 6.557438921075258e-07,
      "loss": 0.0021,
      "num_tokens": 31603905.0,
      "reward": 0.5,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.5,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 387
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 209.0,
      "completions/max_terminated_length": 209.0,
      "completions/mean_length": 152.625,
      "completions/mean_terminated_length": 152.625,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.4012409513960703,
      "grad_norm": 3.199320861926171,
      "kl": 0.08251953125,
      "learning_rate": 6.541994850519933e-07,
      "loss": 0.0033,
      "num_tokens": 31684208.0,
      "reward": 0.875,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 388
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 245.0,
      "completions/max_terminated_length": 245.0,
      "completions/mean_length": 184.45834350585938,
      "completions/mean_terminated_length": 184.45834350585938,
      "completions/min_length": 120.0,
      "completions/min_terminated_length": 120.0,
      "epoch": 0.40227507755946224,
      "grad_norm": 2.917820029257383,
      "kl": 0.08935546875,
      "learning_rate": 6.526534504652013e-07,
      "loss": 0.0036,
      "num_tokens": 31768707.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.6370220184326172,
      "step": 389
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 257.0,
      "completions/max_terminated_length": 257.0,
      "completions/mean_length": 147.6666717529297,
      "completions/mean_terminated_length": 147.6666717529297,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.40330920372285417,
      "grad_norm": 2.3537166439553805,
      "kl": 0.0615234375,
      "learning_rate": 6.511058046651011e-07,
      "loss": 0.0025,
      "num_tokens": 31847507.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 390
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 240.0,
      "completions/max_terminated_length": 240.0,
      "completions/mean_length": 183.95834350585938,
      "completions/mean_terminated_length": 183.95834350585938,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.4043433298862461,
      "grad_norm": 2.9414248666748626,
      "kl": 0.11572265625,
      "learning_rate": 6.49556563986649e-07,
      "loss": 0.0046,
      "num_tokens": 31932354.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.2832478880882263,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 391
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 158.33334350585938,
      "completions/mean_terminated_length": 158.33334350585938,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.40537745604963804,
      "grad_norm": 4.303254887652743,
      "kl": 0.07666015625,
      "learning_rate": 6.480057447816355e-07,
      "loss": 0.0031,
      "num_tokens": 32017602.0,
      "reward": 0.7291666865348816,
      "reward_std": 0.48977774381637573,
      "rewards/reasoning_reward/mean": 0.7291666865348816,
      "rewards/reasoning_reward/std": 0.5311833620071411,
      "step": 392
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 325.0,
      "completions/max_terminated_length": 325.0,
      "completions/mean_length": 190.20834350585938,
      "completions/mean_terminated_length": 190.20834350585938,
      "completions/min_length": 140.0,
      "completions/min_terminated_length": 140.0,
      "epoch": 0.40641158221303,
      "grad_norm": 3.84726477694061,
      "kl": 0.099609375,
      "learning_rate": 6.464533634185117e-07,
      "loss": 0.004,
      "num_tokens": 32101527.0,
      "reward": 0.7847222685813904,
      "reward_std": 0.3665553033351898,
      "rewards/reasoning_reward/mean": 0.7847222685813904,
      "rewards/reasoning_reward/std": 0.5394557118415833,
      "step": 393
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 138.125,
      "completions/mean_terminated_length": 138.125,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.4074457083764219,
      "grad_norm": 3.5955314731898214,
      "kl": 0.087890625,
      "learning_rate": 6.448994362822167e-07,
      "loss": 0.0035,
      "num_tokens": 32181610.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3917974829673767,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 394
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 253.0,
      "completions/max_terminated_length": 253.0,
      "completions/mean_length": 175.25,
      "completions/mean_terminated_length": 175.25,
      "completions/min_length": 123.0,
      "completions/min_terminated_length": 123.0,
      "epoch": 0.40847983453981385,
      "grad_norm": 2.047840627858506,
      "kl": 0.087890625,
      "learning_rate": 6.433439797740049e-07,
      "loss": 0.0035,
      "num_tokens": 32258896.0,
      "reward": 0.75,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 395
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 155.4166717529297,
      "completions/mean_terminated_length": 155.4166717529297,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.4095139607032058,
      "grad_norm": 3.1714239189466853,
      "kl": 0.11279296875,
      "learning_rate": 6.417870103112731e-07,
      "loss": 0.0045,
      "num_tokens": 32341714.0,
      "reward": 1.5069444179534912,
      "reward_std": 0.2113002985715866,
      "rewards/reasoning_reward/mean": 1.5069442987442017,
      "rewards/reasoning_reward/std": 0.44363224506378174,
      "step": 396
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 360.0,
      "completions/max_terminated_length": 360.0,
      "completions/mean_length": 162.5416717529297,
      "completions/mean_terminated_length": 165.60870361328125,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.4105480868665977,
      "grad_norm": 3.3579511466794707,
      "kl": 0.08154296875,
      "learning_rate": 6.402285443273865e-07,
      "loss": 0.0033,
      "num_tokens": 32426735.0,
      "reward": 1.25,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.5316095352172852,
      "step": 397
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 148.25,
      "completions/mean_terminated_length": 148.25,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.41158221302998965,
      "grad_norm": 3.718010993661941,
      "kl": 0.0771484375,
      "learning_rate": 6.386685982715056e-07,
      "loss": 0.0031,
      "num_tokens": 32511453.0,
      "reward": 1.25,
      "reward_std": 0.3120119273662567,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.5897678136825562,
      "step": 398
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 129.125,
      "completions/mean_terminated_length": 129.52174377441406,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.4126163391933816,
      "grad_norm": 2.040886107836913,
      "kl": 0.0830078125,
      "learning_rate": 6.371071886084132e-07,
      "loss": 0.0033,
      "num_tokens": 32597216.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.20412413775920868,
      "step": 399
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 173.0,
      "completions/mean_terminated_length": 173.0,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.4136504653567735,
      "grad_norm": 3.8617028180789013,
      "kl": 0.09326171875,
      "learning_rate": 6.355443318183394e-07,
      "loss": 0.0037,
      "num_tokens": 32682504.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.21322892606258392,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.32900264859199524,
      "step": 400
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 216.0,
      "completions/max_terminated_length": 216.0,
      "completions/mean_length": 155.20834350585938,
      "completions/mean_terminated_length": 155.20834350585938,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.41468459152016546,
      "grad_norm": 3.8146003140109523,
      "kl": 0.10205078125,
      "learning_rate": 6.339800443967884e-07,
      "loss": 0.0041,
      "num_tokens": 32763605.0,
      "reward": 0.9375,
      "reward_std": 0.43075722455978394,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.5379611253738403,
      "step": 401
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 192.0,
      "completions/max_terminated_length": 192.0,
      "completions/mean_length": 123.0,
      "completions/mean_terminated_length": 123.0,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.4157187176835574,
      "grad_norm": 2.025558543566053,
      "kl": 0.07666015625,
      "learning_rate": 6.324143428543647e-07,
      "loss": 0.0031,
      "num_tokens": 32840709.0,
      "reward": 1.0625,
      "reward_std": 0.12400396913290024,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.2242136001586914,
      "step": 402
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 211.0,
      "completions/max_terminated_length": 211.0,
      "completions/mean_length": 156.4166717529297,
      "completions/mean_terminated_length": 156.4166717529297,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.41675284384694933,
      "grad_norm": 3.726281910392908,
      "kl": 0.12158203125,
      "learning_rate": 6.308472437165982e-07,
      "loss": 0.0049,
      "num_tokens": 32934543.0,
      "reward": 1.6805555820465088,
      "reward_std": 0.28358834981918335,
      "rewards/reasoning_reward/mean": 1.6805554628372192,
      "rewards/reasoning_reward/std": 0.3143764138221741,
      "step": 403
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 260.0,
      "completions/max_terminated_length": 260.0,
      "completions/mean_length": 146.6666717529297,
      "completions/mean_terminated_length": 146.6666717529297,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.41778697001034126,
      "grad_norm": 4.354273396992502,
      "kl": 0.078125,
      "learning_rate": 6.292787635237699e-07,
      "loss": 0.0031,
      "num_tokens": 33015679.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.4993361234664917,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.6370220184326172,
      "step": 404
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 173.58334350585938,
      "completions/mean_terminated_length": 173.58334350585938,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.4188210961737332,
      "grad_norm": 3.983196444253434,
      "kl": 0.11767578125,
      "learning_rate": 6.277089188307378e-07,
      "loss": 0.0047,
      "num_tokens": 33110301.0,
      "reward": 1.375,
      "reward_std": 0.483431339263916,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.6796738505363464,
      "step": 405
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 199.0,
      "completions/max_terminated_length": 199.0,
      "completions/mean_length": 138.7916717529297,
      "completions/mean_terminated_length": 138.7916717529297,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.41985522233712513,
      "grad_norm": 3.0815905327353788,
      "kl": 0.07763671875,
      "learning_rate": 6.261377262067615e-07,
      "loss": 0.0031,
      "num_tokens": 33186976.0,
      "reward": 0.625,
      "reward_std": 0.2553258538246155,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.5757792592048645,
      "step": 406
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 254.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 142.625,
      "completions/mean_terminated_length": 142.625,
      "completions/min_length": 73.0,
      "completions/min_terminated_length": 73.0,
      "epoch": 0.42088934850051707,
      "grad_norm": 3.2668731146325487,
      "kl": 0.111328125,
      "learning_rate": 6.245652022353276e-07,
      "loss": 0.0045,
      "num_tokens": 33269199.0,
      "reward": 0.7291666865348816,
      "reward_std": 0.38959476351737976,
      "rewards/reasoning_reward/mean": 0.7291666865348816,
      "rewards/reasoning_reward/std": 0.5311833620071411,
      "step": 407
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 263.0,
      "completions/max_terminated_length": 263.0,
      "completions/mean_length": 140.875,
      "completions/mean_terminated_length": 136.95652770996094,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.421923474663909,
      "grad_norm": 2.7293326482277154,
      "kl": 0.138671875,
      "learning_rate": 6.229913635139748e-07,
      "loss": 0.0056,
      "num_tokens": 33357652.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.9168313145637512,
      "step": 408
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 174.0,
      "completions/max_terminated_length": 174.0,
      "completions/mean_length": 131.5,
      "completions/mean_terminated_length": 131.5,
      "completions/min_length": 61.0,
      "completions/min_terminated_length": 61.0,
      "epoch": 0.42295760082730094,
      "grad_norm": 3.6612477267420185,
      "kl": 0.08984375,
      "learning_rate": 6.214162266541187e-07,
      "loss": 0.0036,
      "num_tokens": 33437576.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.3907342851161957,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 409
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 199.0,
      "completions/max_terminated_length": 199.0,
      "completions/mean_length": 120.5,
      "completions/mean_terminated_length": 120.5,
      "completions/min_length": 66.0,
      "completions/min_terminated_length": 66.0,
      "epoch": 0.4239917269906929,
      "grad_norm": 0.3608860547487124,
      "kl": 0.06689453125,
      "learning_rate": 6.198398082808763e-07,
      "loss": 0.0027,
      "num_tokens": 33518076.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 410
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 305.0,
      "completions/max_terminated_length": 305.0,
      "completions/mean_length": 159.625,
      "completions/mean_terminated_length": 159.625,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.4250258531540848,
      "grad_norm": 1.8961111171035863,
      "kl": 0.06787109375,
      "learning_rate": 6.182621250328905e-07,
      "loss": 0.0027,
      "num_tokens": 33603971.0,
      "reward": 1.25,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.6079187393188477,
      "step": 411
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 156.0,
      "completions/max_terminated_length": 156.0,
      "completions/mean_length": 121.16667175292969,
      "completions/mean_terminated_length": 121.16667175292969,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.42605997931747674,
      "grad_norm": 3.9688045476328773,
      "kl": 0.08642578125,
      "learning_rate": 6.166831935621546e-07,
      "loss": 0.0035,
      "num_tokens": 33684095.0,
      "reward": 1.0,
      "reward_std": 0.34503278136253357,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 412
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 258.0,
      "completions/max_terminated_length": 258.0,
      "completions/mean_length": 166.70834350585938,
      "completions/mean_terminated_length": 166.70834350585938,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.4270941054808687,
      "grad_norm": 3.51307920136012,
      "kl": 0.072265625,
      "learning_rate": 6.151030305338367e-07,
      "loss": 0.0029,
      "num_tokens": 33763608.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.355445921421051,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.5450701117515564,
      "step": 413
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 214.0,
      "completions/max_terminated_length": 214.0,
      "completions/mean_length": 142.7916717529297,
      "completions/mean_terminated_length": 142.7916717529297,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.4281282316442606,
      "grad_norm": 2.4788553923664116,
      "kl": 0.08642578125,
      "learning_rate": 6.135216526261036e-07,
      "loss": 0.0035,
      "num_tokens": 33848787.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.8139966726303101,
      "step": 414
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 213.0,
      "completions/max_terminated_length": 213.0,
      "completions/mean_length": 150.125,
      "completions/mean_terminated_length": 150.125,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.42916235780765255,
      "grad_norm": 2.8834248799500073,
      "kl": 0.10009765625,
      "learning_rate": 6.119390765299447e-07,
      "loss": 0.004,
      "num_tokens": 33931398.0,
      "reward": 0.5902777910232544,
      "reward_std": 0.19089612364768982,
      "rewards/reasoning_reward/mean": 0.5902777910232544,
      "rewards/reasoning_reward/std": 0.4661984443664551,
      "step": 415
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 157.1666717529297,
      "completions/mean_terminated_length": 157.1666717529297,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.4301964839710445,
      "grad_norm": 3.9268847081933598,
      "kl": 0.0859375,
      "learning_rate": 6.103553189489959e-07,
      "loss": 0.0034,
      "num_tokens": 34013978.0,
      "reward": 1.7430557012557983,
      "reward_std": 0.36321234703063965,
      "rewards/reasoning_reward/mean": 1.7430557012557983,
      "rewards/reasoning_reward/std": 0.3961748480796814,
      "step": 416
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 294.0,
      "completions/max_terminated_length": 294.0,
      "completions/mean_length": 179.83334350585938,
      "completions/mean_terminated_length": 179.83334350585938,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.4312306101344364,
      "grad_norm": 2.896040987170424,
      "kl": 0.09619140625,
      "learning_rate": 6.087703965993636e-07,
      "loss": 0.0038,
      "num_tokens": 34098558.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.26726123690605164,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.458415687084198,
      "step": 417
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 262.0,
      "completions/max_terminated_length": 262.0,
      "completions/mean_length": 148.125,
      "completions/mean_terminated_length": 148.125,
      "completions/min_length": 67.0,
      "completions/min_terminated_length": 67.0,
      "epoch": 0.43226473629782836,
      "grad_norm": 3.0726708113736834,
      "kl": 0.08642578125,
      "learning_rate": 6.071843262094476e-07,
      "loss": 0.0035,
      "num_tokens": 34176129.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.4130779504776001,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.5311833620071411,
      "step": 418
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 212.0,
      "completions/max_terminated_length": 212.0,
      "completions/mean_length": 158.2916717529297,
      "completions/mean_terminated_length": 158.2916717529297,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.4332988624612203,
      "grad_norm": 3.651158491177238,
      "kl": 0.115234375,
      "learning_rate": 6.055971245197652e-07,
      "loss": 0.0046,
      "num_tokens": 34258992.0,
      "reward": 1.1319445371627808,
      "reward_std": 0.40136003494262695,
      "rewards/reasoning_reward/mean": 1.1319445371627808,
      "rewards/reasoning_reward/std": 0.6407633423805237,
      "step": 419
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 369.0,
      "completions/max_terminated_length": 369.0,
      "completions/mean_length": 165.75,
      "completions/mean_terminated_length": 165.75,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.4343329886246122,
      "grad_norm": 3.011267910041691,
      "kl": 0.1337890625,
      "learning_rate": 6.040088082827744e-07,
      "loss": 0.0053,
      "num_tokens": 34342034.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.43203312158584595,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.9609683156013489,
      "step": 420
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 223.0,
      "completions/max_terminated_length": 223.0,
      "completions/mean_length": 142.125,
      "completions/mean_terminated_length": 142.125,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.43536711478800416,
      "grad_norm": 3.934648597506697,
      "kl": 0.048828125,
      "learning_rate": 6.024193942626961e-07,
      "loss": 0.002,
      "num_tokens": 34423189.0,
      "reward": 0.4583333432674408,
      "reward_std": 0.4082186818122864,
      "rewards/reasoning_reward/mean": 0.4583333432674408,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 421
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 262.0,
      "completions/max_terminated_length": 262.0,
      "completions/mean_length": 146.0,
      "completions/mean_terminated_length": 146.0,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.4364012409513961,
      "grad_norm": 2.895700821484859,
      "kl": 0.103515625,
      "learning_rate": 6.008288992353396e-07,
      "loss": 0.0041,
      "num_tokens": 34500773.0,
      "reward": 0.625,
      "reward_std": 0.19416078925132751,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.5565811395645142,
      "step": 422
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 154.5416717529297,
      "completions/mean_terminated_length": 154.5416717529297,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.43743536711478803,
      "grad_norm": 3.7713820783469303,
      "kl": 0.07177734375,
      "learning_rate": 5.99237339987922e-07,
      "loss": 0.0029,
      "num_tokens": 34577722.0,
      "reward": 0.5625,
      "reward_std": 0.5429885983467102,
      "rewards/reasoning_reward/mean": 0.5625,
      "rewards/reasoning_reward/std": 0.5379611253738403,
      "step": 423
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 148.70834350585938,
      "completions/mean_terminated_length": 148.70834350585938,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.4384694932781799,
      "grad_norm": 3.373544653686179,
      "kl": 0.050537109375,
      "learning_rate": 5.976447333188944e-07,
      "loss": 0.002,
      "num_tokens": 34652923.0,
      "reward": 0.5,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 0.5,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 424
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 309.0,
      "completions/max_terminated_length": 309.0,
      "completions/mean_length": 147.125,
      "completions/mean_terminated_length": 147.125,
      "completions/min_length": 76.0,
      "completions/min_terminated_length": 76.0,
      "epoch": 0.43950361944157185,
      "grad_norm": 4.7102217326975415,
      "kl": 0.09033203125,
      "learning_rate": 5.960510960377626e-07,
      "loss": 0.0036,
      "num_tokens": 34736582.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.354950875043869,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.3753018081188202,
      "step": 425
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 213.0,
      "completions/max_terminated_length": 213.0,
      "completions/mean_length": 153.95834350585938,
      "completions/mean_terminated_length": 153.95834350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.4405377456049638,
      "grad_norm": 4.16978047474189,
      "kl": 0.10693359375,
      "learning_rate": 5.944564449649099e-07,
      "loss": 0.0043,
      "num_tokens": 34813797.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.5391935706138611,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.6743220090866089,
      "step": 426
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 290.0,
      "completions/max_terminated_length": 290.0,
      "completions/mean_length": 128.875,
      "completions/mean_terminated_length": 128.875,
      "completions/min_length": 76.0,
      "completions/min_terminated_length": 76.0,
      "epoch": 0.4415718717683557,
      "grad_norm": 4.140299837667284,
      "kl": 0.10791015625,
      "learning_rate": 5.928607969314201e-07,
      "loss": 0.0043,
      "num_tokens": 34894010.0,
      "reward": 0.75,
      "reward_std": 0.47364258766174316,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.6079187393188477,
      "step": 427
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 178.0,
      "completions/max_terminated_length": 178.0,
      "completions/mean_length": 124.45833587646484,
      "completions/mean_terminated_length": 124.45833587646484,
      "completions/min_length": 51.0,
      "completions/min_terminated_length": 51.0,
      "epoch": 0.44260599793174765,
      "grad_norm": 3.8213113040181987,
      "kl": 0.064453125,
      "learning_rate": 5.912641687789002e-07,
      "loss": 0.0026,
      "num_tokens": 34968525.0,
      "reward": 0.625,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 428
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 249.0,
      "completions/max_terminated_length": 249.0,
      "completions/mean_length": 136.4166717529297,
      "completions/mean_terminated_length": 136.4166717529297,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.4436401240951396,
      "grad_norm": 2.733283075630004,
      "kl": 0.058837890625,
      "learning_rate": 5.896665773593012e-07,
      "loss": 0.0023,
      "num_tokens": 35046975.0,
      "reward": 1.0625,
      "reward_std": 0.08625819534063339,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.1689159870147705,
      "step": 429
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 151.58334350585938,
      "completions/mean_terminated_length": 151.58334350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.4446742502585315,
      "grad_norm": 2.848539827633773,
      "kl": 0.07177734375,
      "learning_rate": 5.880680395347418e-07,
      "loss": 0.0029,
      "num_tokens": 35132493.0,
      "reward": 1.2638888359069824,
      "reward_std": 0.26332971453666687,
      "rewards/reasoning_reward/mean": 1.2638888359069824,
      "rewards/reasoning_reward/std": 0.8899827003479004,
      "step": 430
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 214.0,
      "completions/max_terminated_length": 214.0,
      "completions/mean_length": 133.0,
      "completions/mean_terminated_length": 133.0,
      "completions/min_length": 73.0,
      "completions/min_terminated_length": 73.0,
      "epoch": 0.44570837642192346,
      "grad_norm": 2.221455031748688,
      "kl": 0.057861328125,
      "learning_rate": 5.864685721773293e-07,
      "loss": 0.0023,
      "num_tokens": 35212989.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.550032913684845,
      "step": 431
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 249.0,
      "completions/max_terminated_length": 249.0,
      "completions/mean_length": 157.0,
      "completions/mean_terminated_length": 157.0,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.4467425025853154,
      "grad_norm": 3.710718263121241,
      "kl": 0.07958984375,
      "learning_rate": 5.848681921689819e-07,
      "loss": 0.0032,
      "num_tokens": 35298725.0,
      "reward": 1.1319445371627808,
      "reward_std": 0.3816637396812439,
      "rewards/reasoning_reward/mean": 1.1319444179534912,
      "rewards/reasoning_reward/std": 0.6255030035972595,
      "step": 432
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 214.0,
      "completions/max_terminated_length": 214.0,
      "completions/mean_length": 169.08334350585938,
      "completions/mean_terminated_length": 169.08334350585938,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.44777662874870733,
      "grad_norm": 3.6503587546386016,
      "kl": 0.0810546875,
      "learning_rate": 5.832669164012513e-07,
      "loss": 0.0032,
      "num_tokens": 35381551.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.4745539426803589,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.5036101341247559,
      "step": 433
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 211.0,
      "completions/max_terminated_length": 211.0,
      "completions/mean_length": 173.0416717529297,
      "completions/mean_terminated_length": 173.0416717529297,
      "completions/min_length": 131.0,
      "completions/min_terminated_length": 131.0,
      "epoch": 0.44881075491209926,
      "grad_norm": 3.4278772902345302,
      "kl": 0.07421875,
      "learning_rate": 5.816647617751424e-07,
      "loss": 0.003,
      "num_tokens": 35463120.0,
      "reward": 1.1875,
      "reward_std": 0.3177132308483124,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.5067479610443115,
      "step": 434
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 181.0,
      "completions/max_terminated_length": 181.0,
      "completions/mean_length": 141.0,
      "completions/mean_terminated_length": 141.0,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.4498448810754912,
      "grad_norm": 2.8560665845210957,
      "kl": 0.07666015625,
      "learning_rate": 5.800617452009375e-07,
      "loss": 0.0031,
      "num_tokens": 35548048.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.550032913684845,
      "step": 435
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 161.1666717529297,
      "completions/mean_terminated_length": 161.1666717529297,
      "completions/min_length": 119.0,
      "completions/min_terminated_length": 119.0,
      "epoch": 0.45087900723888313,
      "grad_norm": 4.509744320701168,
      "kl": 0.119140625,
      "learning_rate": 5.784578835980157e-07,
      "loss": 0.0048,
      "num_tokens": 35642260.0,
      "reward": 1.4791667461395264,
      "reward_std": 0.41317981481552124,
      "rewards/reasoning_reward/mean": 1.4791666269302368,
      "rewards/reasoning_reward/std": 0.4293363690376282,
      "step": 436
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 134.7916717529297,
      "completions/mean_terminated_length": 134.7916717529297,
      "completions/min_length": 61.0,
      "completions/min_terminated_length": 61.0,
      "epoch": 0.45191313340227507,
      "grad_norm": 2.5466178790890446,
      "kl": 0.07080078125,
      "learning_rate": 5.768531938946756e-07,
      "loss": 0.0028,
      "num_tokens": 35719927.0,
      "reward": 0.6458333730697632,
      "reward_std": 0.13908717036247253,
      "rewards/reasoning_reward/mean": 0.6458333134651184,
      "rewards/reasoning_reward/std": 0.5208514928817749,
      "step": 437
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 148.4166717529297,
      "completions/mean_terminated_length": 148.4166717529297,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.452947259565667,
      "grad_norm": 4.2489173473223145,
      "kl": 0.107421875,
      "learning_rate": 5.752476930279557e-07,
      "loss": 0.0043,
      "num_tokens": 35804233.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.4671573042869568,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.5694518089294434,
      "step": 438
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 230.0,
      "completions/max_terminated_length": 230.0,
      "completions/mean_length": 147.125,
      "completions/mean_terminated_length": 147.125,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.45398138572905894,
      "grad_norm": 4.008875539956656,
      "kl": 0.08837890625,
      "learning_rate": 5.736413979434566e-07,
      "loss": 0.0035,
      "num_tokens": 35882052.0,
      "reward": 1.0277777910232544,
      "reward_std": 0.329608291387558,
      "rewards/reasoning_reward/mean": 1.0277777910232544,
      "rewards/reasoning_reward/std": 0.3859447240829468,
      "step": 439
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 266.0,
      "completions/max_terminated_length": 266.0,
      "completions/mean_length": 155.25,
      "completions/mean_terminated_length": 155.25,
      "completions/min_length": 49.0,
      "completions/min_terminated_length": 49.0,
      "epoch": 0.4550155118924509,
      "grad_norm": 3.550772237674307,
      "kl": 0.091796875,
      "learning_rate": 5.720343255951611e-07,
      "loss": 0.0037,
      "num_tokens": 35961122.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 440
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 178.1666717529297,
      "completions/mean_terminated_length": 178.1666717529297,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.4560496380558428,
      "grad_norm": 4.361928899053349,
      "kl": 0.11083984375,
      "learning_rate": 5.704264929452562e-07,
      "loss": 0.0044,
      "num_tokens": 36050486.0,
      "reward": 1.1597222089767456,
      "reward_std": 0.41859424114227295,
      "rewards/reasoning_reward/mean": 1.1597222089767456,
      "rewards/reasoning_reward/std": 0.7573778033256531,
      "step": 441
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 253.0,
      "completions/max_terminated_length": 253.0,
      "completions/mean_length": 158.70834350585938,
      "completions/mean_terminated_length": 158.70834350585938,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.45708376421923474,
      "grad_norm": 4.446970005435705,
      "kl": 0.109375,
      "learning_rate": 5.688179169639537e-07,
      "loss": 0.0044,
      "num_tokens": 36130319.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 442
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 165.5,
      "completions/mean_terminated_length": 165.5,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.4581178903826267,
      "grad_norm": 4.394083775689496,
      "kl": 0.138671875,
      "learning_rate": 5.672086146293108e-07,
      "loss": 0.0056,
      "num_tokens": 36219107.0,
      "reward": 1.3819444179534912,
      "reward_std": 0.48597466945648193,
      "rewards/reasoning_reward/mean": 1.3819442987442017,
      "rewards/reasoning_reward/std": 0.7428876161575317,
      "step": 443
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 154.7916717529297,
      "completions/mean_terminated_length": 154.7916717529297,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.4591520165460186,
      "grad_norm": 3.2132018317716478,
      "kl": 0.1513671875,
      "learning_rate": 5.65598602927051e-07,
      "loss": 0.006,
      "num_tokens": 36301518.0,
      "reward": 0.9375,
      "reward_std": 0.26507532596588135,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.6309499144554138,
      "step": 444
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 279.0,
      "completions/max_terminated_length": 279.0,
      "completions/mean_length": 171.95834350585938,
      "completions/mean_terminated_length": 171.95834350585938,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.46018614270941055,
      "grad_norm": 4.320892935715968,
      "kl": 0.10791015625,
      "learning_rate": 5.639878988503858e-07,
      "loss": 0.0043,
      "num_tokens": 36377789.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.436039537191391,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.4871537983417511,
      "step": 445
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 146.0416717529297,
      "completions/mean_terminated_length": 146.0416717529297,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.4612202688728025,
      "grad_norm": 3.1035680233792466,
      "kl": 0.09765625,
      "learning_rate": 5.623765193998333e-07,
      "loss": 0.0039,
      "num_tokens": 36461686.0,
      "reward": 1.125,
      "reward_std": 0.2553258538246155,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.6634888052940369,
      "step": 446
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 216.0,
      "completions/max_terminated_length": 216.0,
      "completions/mean_length": 133.45834350585938,
      "completions/mean_terminated_length": 133.45834350585938,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.4622543950361944,
      "grad_norm": 3.4348649831091187,
      "kl": 0.09130859375,
      "learning_rate": 5.607644815830412e-07,
      "loss": 0.0037,
      "num_tokens": 36539977.0,
      "reward": 0.930555522441864,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.930555522441864,
      "rewards/reasoning_reward/std": 0.405030757188797,
      "step": 447
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 160.75,
      "completions/mean_terminated_length": 160.75,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.46328852119958636,
      "grad_norm": 3.0932846774731515,
      "kl": 0.050048828125,
      "learning_rate": 5.591518024146049e-07,
      "loss": 0.002,
      "num_tokens": 36621971.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.243839293718338,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.6064269542694092,
      "step": 448
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 265.0,
      "completions/max_terminated_length": 265.0,
      "completions/mean_length": 161.5,
      "completions/mean_terminated_length": 161.5,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.4643226473629783,
      "grad_norm": 1.5405981712007193,
      "kl": 0.058349609375,
      "learning_rate": 5.5753849891589e-07,
      "loss": 0.0023,
      "num_tokens": 36698823.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 449
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 139.0,
      "completions/mean_terminated_length": 139.0,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.4653567735263702,
      "grad_norm": 3.2989222982872035,
      "kl": 0.080078125,
      "learning_rate": 5.559245881148513e-07,
      "loss": 0.0032,
      "num_tokens": 36779479.0,
      "reward": 1.125,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.4484272003173828,
      "step": 450
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 249.0,
      "completions/max_terminated_length": 249.0,
      "completions/mean_length": 170.2916717529297,
      "completions/mean_terminated_length": 170.2916717529297,
      "completions/min_length": 118.0,
      "completions/min_terminated_length": 118.0,
      "epoch": 0.46639089968976216,
      "grad_norm": 3.9459517839159957,
      "kl": 0.06591796875,
      "learning_rate": 5.543100870458537e-07,
      "loss": 0.0026,
      "num_tokens": 36855542.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.3492930829524994,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.4657664895057678,
      "step": 451
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 213.0,
      "completions/max_terminated_length": 213.0,
      "completions/mean_length": 151.0416717529297,
      "completions/mean_terminated_length": 151.0416717529297,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.4674250258531541,
      "grad_norm": 3.080169913436108,
      "kl": 0.09619140625,
      "learning_rate": 5.526950127494918e-07,
      "loss": 0.0039,
      "num_tokens": 36939599.0,
      "reward": 1.3263888359069824,
      "reward_std": 0.259171724319458,
      "rewards/reasoning_reward/mean": 1.3263888359069824,
      "rewards/reasoning_reward/std": 0.5549060106277466,
      "step": 452
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 238.0,
      "completions/max_terminated_length": 238.0,
      "completions/mean_length": 144.95834350585938,
      "completions/mean_terminated_length": 144.95834350585938,
      "completions/min_length": 54.0,
      "completions/min_terminated_length": 54.0,
      "epoch": 0.46845915201654603,
      "grad_norm": 3.478946037062411,
      "kl": 0.09619140625,
      "learning_rate": 5.510793822724111e-07,
      "loss": 0.0039,
      "num_tokens": 37024790.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.2616034746170044,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.35864076018333435,
      "step": 453
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 159.4166717529297,
      "completions/mean_terminated_length": 159.4166717529297,
      "completions/min_length": 120.0,
      "completions/min_terminated_length": 120.0,
      "epoch": 0.46949327817993797,
      "grad_norm": 4.69807004009168,
      "kl": 0.07421875,
      "learning_rate": 5.494632126671268e-07,
      "loss": 0.003,
      "num_tokens": 37105936.0,
      "reward": 0.75,
      "reward_std": 0.46631526947021484,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.5897678136825562,
      "step": 454
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 145.125,
      "completions/mean_terminated_length": 145.125,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.4705274043433299,
      "grad_norm": 3.5314961615269183,
      "kl": 0.08642578125,
      "learning_rate": 5.478465209918449e-07,
      "loss": 0.0035,
      "num_tokens": 37190787.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.37612998485565186,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.7658352255821228,
      "step": 455
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 247.0,
      "completions/max_terminated_length": 247.0,
      "completions/mean_length": 169.2916717529297,
      "completions/mean_terminated_length": 169.2916717529297,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.47156153050672184,
      "grad_norm": 3.157082111223005,
      "kl": 0.06298828125,
      "learning_rate": 5.462293243102815e-07,
      "loss": 0.0025,
      "num_tokens": 37270178.0,
      "reward": 1.0625,
      "reward_std": 0.2644323706626892,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.5954993963241577,
      "step": 456
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 144.70834350585938,
      "completions/mean_terminated_length": 144.70834350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.4725956566701138,
      "grad_norm": 2.2530317338339554,
      "kl": 0.09326171875,
      "learning_rate": 5.44611639691483e-07,
      "loss": 0.0037,
      "num_tokens": 37356739.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "step": 457
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 194.0,
      "completions/max_terminated_length": 194.0,
      "completions/mean_length": 129.58334350585938,
      "completions/mean_terminated_length": 129.58334350585938,
      "completions/min_length": 67.0,
      "completions/min_terminated_length": 67.0,
      "epoch": 0.4736297828335057,
      "grad_norm": 3.530213579949582,
      "kl": 0.06884765625,
      "learning_rate": 5.429934842096453e-07,
      "loss": 0.0028,
      "num_tokens": 37444233.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 458
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 453.0,
      "completions/max_terminated_length": 230.0,
      "completions/mean_length": 159.45834350585938,
      "completions/mean_terminated_length": 146.69564819335938,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.47466390899689764,
      "grad_norm": 10.937724197833212,
      "kl": 0.07373046875,
      "learning_rate": 5.41374874943935e-07,
      "loss": 0.0029,
      "num_tokens": 37530724.0,
      "reward": 0.8958333730697632,
      "reward_std": 0.15268757939338684,
      "rewards/reasoning_reward/mean": 0.8958333134651184,
      "rewards/reasoning_reward/std": 0.7515081763267517,
      "step": 459
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 126.41667175292969,
      "completions/mean_terminated_length": 126.41667175292969,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.4756980351602896,
      "grad_norm": 0.23424191596610372,
      "kl": 0.07568359375,
      "learning_rate": 5.397558289783079e-07,
      "loss": 0.003,
      "num_tokens": 37614622.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 460
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 144.2916717529297,
      "completions/mean_terminated_length": 144.2916717529297,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.4767321613236815,
      "grad_norm": 3.5303048844343343,
      "kl": 0.068359375,
      "learning_rate": 5.381363634013285e-07,
      "loss": 0.0027,
      "num_tokens": 37697165.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 461
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 187.0,
      "completions/max_terminated_length": 187.0,
      "completions/mean_length": 127.625,
      "completions/mean_terminated_length": 127.625,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.47776628748707345,
      "grad_norm": 2.4796497555216117,
      "kl": 0.042236328125,
      "learning_rate": 5.365164953059911e-07,
      "loss": 0.0017,
      "num_tokens": 37776012.0,
      "reward": 0.75,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 462
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 226.0,
      "completions/max_terminated_length": 226.0,
      "completions/mean_length": 145.0,
      "completions/mean_terminated_length": 145.0,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.4788004136504654,
      "grad_norm": 3.9203231363884554,
      "kl": 0.07177734375,
      "learning_rate": 5.348962417895378e-07,
      "loss": 0.0029,
      "num_tokens": 37852260.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.3027648627758026,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.7798991799354553,
      "step": 463
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 189.0,
      "completions/max_terminated_length": 189.0,
      "completions/mean_length": 128.33334350585938,
      "completions/mean_terminated_length": 128.33334350585938,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.47983453981385726,
      "grad_norm": 2.3870572364816045,
      "kl": 0.0498046875,
      "learning_rate": 5.332756199532791e-07,
      "loss": 0.002,
      "num_tokens": 37932140.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 464
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 252.0,
      "completions/max_terminated_length": 252.0,
      "completions/mean_length": 159.45834350585938,
      "completions/mean_terminated_length": 159.45834350585938,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.4808686659772492,
      "grad_norm": 3.4551439032369498,
      "kl": 0.0703125,
      "learning_rate": 5.316546469024127e-07,
      "loss": 0.0028,
      "num_tokens": 38008975.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.3493061661720276,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.5835920572280884,
      "step": 465
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 254.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 151.58334350585938,
      "completions/mean_terminated_length": 151.58334350585938,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.48190279214064113,
      "grad_norm": 2.2638186168786993,
      "kl": 0.050048828125,
      "learning_rate": 5.300333397458436e-07,
      "loss": 0.002,
      "num_tokens": 38086597.0,
      "reward": 0.7291666865348816,
      "reward_std": 0.12400396913290024,
      "rewards/reasoning_reward/mean": 0.7291666865348816,
      "rewards/reasoning_reward/std": 0.4418136179447174,
      "step": 466
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 201.0,
      "completions/max_terminated_length": 201.0,
      "completions/mean_length": 135.6666717529297,
      "completions/mean_terminated_length": 135.6666717529297,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.48293691830403307,
      "grad_norm": 3.4589069312702914,
      "kl": 0.0751953125,
      "learning_rate": 5.284117155960025e-07,
      "loss": 0.003,
      "num_tokens": 38165749.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 467
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 219.0,
      "completions/max_terminated_length": 219.0,
      "completions/mean_length": 145.83334350585938,
      "completions/mean_terminated_length": 145.83334350585938,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.483971044467425,
      "grad_norm": 3.777619829419609,
      "kl": 0.056640625,
      "learning_rate": 5.267897915686668e-07,
      "loss": 0.0023,
      "num_tokens": 38247225.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.28029152750968933,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.458415687084198,
      "step": 468
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 202.0,
      "completions/max_terminated_length": 202.0,
      "completions/mean_length": 158.6666717529297,
      "completions/mean_terminated_length": 158.6666717529297,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.48500517063081694,
      "grad_norm": 3.5113858847079924,
      "kl": 0.08740234375,
      "learning_rate": 5.251675847827784e-07,
      "loss": 0.0035,
      "num_tokens": 38331265.0,
      "reward": 0.9375,
      "reward_std": 0.2725489139556885,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.8884831070899963,
      "step": 469
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 367.0,
      "completions/max_terminated_length": 367.0,
      "completions/mean_length": 166.875,
      "completions/mean_terminated_length": 166.875,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.4860392967942089,
      "grad_norm": 3.0917385670170185,
      "kl": 0.07177734375,
      "learning_rate": 5.235451123602641e-07,
      "loss": 0.0029,
      "num_tokens": 38408302.0,
      "reward": 0.9375,
      "reward_std": 0.29339051246643066,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.7705517411231995,
      "step": 470
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 127.5,
      "completions/mean_terminated_length": 127.5,
      "completions/min_length": 56.0,
      "completions/min_terminated_length": 56.0,
      "epoch": 0.4870734229576008,
      "grad_norm": 0.3175955227430425,
      "kl": 0.06494140625,
      "learning_rate": 5.219223914258538e-07,
      "loss": 0.0026,
      "num_tokens": 38491450.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 471
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 283.0,
      "completions/max_terminated_length": 283.0,
      "completions/mean_length": 150.125,
      "completions/mean_terminated_length": 150.125,
      "completions/min_length": 56.0,
      "completions/min_terminated_length": 56.0,
      "epoch": 0.48810754912099275,
      "grad_norm": 2.4136738947782126,
      "kl": 0.06884765625,
      "learning_rate": 5.202994391069008e-07,
      "loss": 0.0028,
      "num_tokens": 38570469.0,
      "reward": 0.6111111640930176,
      "reward_std": 0.23193079233169556,
      "rewards/reasoning_reward/mean": 0.6111111044883728,
      "rewards/reasoning_reward/std": 0.589084804058075,
      "step": 472
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 247.0,
      "completions/max_terminated_length": 247.0,
      "completions/mean_length": 166.5416717529297,
      "completions/mean_terminated_length": 166.5416717529297,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.4891416752843847,
      "grad_norm": 3.1274290142900023,
      "kl": 0.08984375,
      "learning_rate": 5.186762725332008e-07,
      "loss": 0.0036,
      "num_tokens": 38655282.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.2717381715774536,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.8272421956062317,
      "step": 473
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 296.0,
      "completions/max_terminated_length": 296.0,
      "completions/mean_length": 175.125,
      "completions/mean_terminated_length": 175.125,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.4901758014477766,
      "grad_norm": 4.7856676348279015,
      "kl": 0.07958984375,
      "learning_rate": 5.170529088368103e-07,
      "loss": 0.0032,
      "num_tokens": 38745685.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.4939148426055908,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 474
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 165.9166717529297,
      "completions/mean_terminated_length": 165.9166717529297,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.49120992761116855,
      "grad_norm": 3.991471510599265,
      "kl": 0.07568359375,
      "learning_rate": 5.154293651518666e-07,
      "loss": 0.003,
      "num_tokens": 38822715.0,
      "reward": 0.7777777910232544,
      "reward_std": 0.44400954246520996,
      "rewards/reasoning_reward/mean": 0.7777777314186096,
      "rewards/reasoning_reward/std": 0.6344891786575317,
      "step": 475
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 195.0,
      "completions/max_terminated_length": 195.0,
      "completions/mean_length": 146.375,
      "completions/mean_terminated_length": 146.375,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.4922440537745605,
      "grad_norm": 3.6198040700819236,
      "kl": 0.091796875,
      "learning_rate": 5.138056586144071e-07,
      "loss": 0.0037,
      "num_tokens": 38905196.0,
      "reward": 1.0625,
      "reward_std": 0.33768826723098755,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.5954993963241577,
      "step": 476
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 159.58334350585938,
      "completions/mean_terminated_length": 159.58334350585938,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.4932781799379524,
      "grad_norm": 3.4082780232755794,
      "kl": 0.0556640625,
      "learning_rate": 5.121818063621877e-07,
      "loss": 0.0022,
      "num_tokens": 38995162.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.36204060912132263,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.5706435441970825,
      "step": 477
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 214.0,
      "completions/max_terminated_length": 214.0,
      "completions/mean_length": 135.375,
      "completions/mean_terminated_length": 135.375,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.49431230610134436,
      "grad_norm": 3.0224789729725345,
      "kl": 0.07373046875,
      "learning_rate": 5.105578255345021e-07,
      "loss": 0.003,
      "num_tokens": 39077227.0,
      "reward": 0.8125,
      "reward_std": 0.22466278076171875,
      "rewards/reasoning_reward/mean": 0.8125,
      "rewards/reasoning_reward/std": 0.7042186260223389,
      "step": 478
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 228.0,
      "completions/max_terminated_length": 228.0,
      "completions/mean_length": 160.0416717529297,
      "completions/mean_terminated_length": 160.0416717529297,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.4953464322647363,
      "grad_norm": 3.053679665790895,
      "kl": 0.07861328125,
      "learning_rate": 5.089337332720016e-07,
      "loss": 0.0031,
      "num_tokens": 39165884.0,
      "reward": 1.3125,
      "reward_std": 0.2041093111038208,
      "rewards/reasoning_reward/mean": 1.3125,
      "rewards/reasoning_reward/std": 0.8945790529251099,
      "step": 479
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 225.0,
      "completions/max_terminated_length": 225.0,
      "completions/mean_length": 162.08334350585938,
      "completions/mean_terminated_length": 162.08334350585938,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.4963805584281282,
      "grad_norm": 3.1381029630629733,
      "kl": 0.06640625,
      "learning_rate": 5.073095467165134e-07,
      "loss": 0.0027,
      "num_tokens": 39248414.0,
      "reward": 1.0555555820465088,
      "reward_std": 0.22960862517356873,
      "rewards/reasoning_reward/mean": 1.0555554628372192,
      "rewards/reasoning_reward/std": 0.5766525864601135,
      "step": 480
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 202.0,
      "completions/max_terminated_length": 202.0,
      "completions/mean_length": 148.75,
      "completions/mean_terminated_length": 148.75,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.49741468459152016,
      "grad_norm": 2.6932656248524913,
      "kl": 0.09765625,
      "learning_rate": 5.056852830108598e-07,
      "loss": 0.0039,
      "num_tokens": 39330824.0,
      "reward": 1.5,
      "reward_std": 0.28029152750968933,
      "rewards/reasoning_reward/mean": 1.5,
      "rewards/reasoning_reward/std": 0.48900964856147766,
      "step": 481
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 240.0,
      "completions/max_terminated_length": 240.0,
      "completions/mean_length": 133.25,
      "completions/mean_terminated_length": 133.25,
      "completions/min_length": 71.0,
      "completions/min_terminated_length": 71.0,
      "epoch": 0.4984488107549121,
      "grad_norm": 4.963868420618398,
      "kl": 0.08935546875,
      "learning_rate": 5.040609592986775e-07,
      "loss": 0.0036,
      "num_tokens": 39413486.0,
      "reward": 0.5208333730697632,
      "reward_std": 0.36753225326538086,
      "rewards/reasoning_reward/mean": 0.5208333134651184,
      "rewards/reasoning_reward/std": 0.6507381796836853,
      "step": 482
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 235.0,
      "completions/max_terminated_length": 235.0,
      "completions/mean_length": 153.25,
      "completions/mean_terminated_length": 153.25,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.49948293691830403,
      "grad_norm": 0.19760209135973245,
      "kl": 0.049560546875,
      "learning_rate": 5.024365927242367e-07,
      "loss": 0.002,
      "num_tokens": 39493220.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 483
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 145.2916717529297,
      "completions/mean_terminated_length": 145.2916717529297,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.500517063081696,
      "grad_norm": 4.265668637720173,
      "kl": 0.09423828125,
      "learning_rate": 5.008122004322597e-07,
      "loss": 0.0038,
      "num_tokens": 39575411.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.45082372426986694,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.47729235887527466,
      "step": 484
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 235.0,
      "completions/max_terminated_length": 235.0,
      "completions/mean_length": 155.875,
      "completions/mean_terminated_length": 155.875,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.5015511892450879,
      "grad_norm": 3.577115117621814,
      "kl": 0.0791015625,
      "learning_rate": 4.991877995677404e-07,
      "loss": 0.0032,
      "num_tokens": 39659448.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.367926687002182,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.721060037612915,
      "step": 485
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 182.0,
      "completions/max_terminated_length": 182.0,
      "completions/mean_length": 131.4166717529297,
      "completions/mean_terminated_length": 131.4166717529297,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.5025853154084798,
      "grad_norm": 3.025540625801558,
      "kl": 0.07470703125,
      "learning_rate": 4.975634072757634e-07,
      "loss": 0.003,
      "num_tokens": 39742018.0,
      "reward": 1.4166667461395264,
      "reward_std": 0.19500279426574707,
      "rewards/reasoning_reward/mean": 1.4166666269302368,
      "rewards/reasoning_reward/std": 0.524749755859375,
      "step": 486
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 167.08334350585938,
      "completions/mean_terminated_length": 167.08334350585938,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.5036194415718718,
      "grad_norm": 3.4418671262934426,
      "kl": 0.09375,
      "learning_rate": 4.959390407013226e-07,
      "loss": 0.0038,
      "num_tokens": 39819180.0,
      "reward": 1.0,
      "reward_std": 0.42724665999412537,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 487
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 274.0,
      "completions/max_terminated_length": 274.0,
      "completions/mean_length": 174.375,
      "completions/mean_terminated_length": 174.375,
      "completions/min_length": 133.0,
      "completions/min_terminated_length": 133.0,
      "epoch": 0.5046535677352637,
      "grad_norm": 3.319472739975236,
      "kl": 0.06884765625,
      "learning_rate": 4.943147169891402e-07,
      "loss": 0.0028,
      "num_tokens": 39896373.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.35864076018333435,
      "step": 488
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 249.0,
      "completions/max_terminated_length": 249.0,
      "completions/mean_length": 162.9166717529297,
      "completions/mean_terminated_length": 162.9166717529297,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.5056876938986556,
      "grad_norm": 3.3304049158014823,
      "kl": 0.054931640625,
      "learning_rate": 4.926904532834866e-07,
      "loss": 0.0022,
      "num_tokens": 39984259.0,
      "reward": 0.375,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 489
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 213.0,
      "completions/max_terminated_length": 213.0,
      "completions/mean_length": 151.5416717529297,
      "completions/mean_terminated_length": 151.5416717529297,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.5067218200620476,
      "grad_norm": 0.18358736944603146,
      "kl": 0.0576171875,
      "learning_rate": 4.910662667279983e-07,
      "loss": 0.0023,
      "num_tokens": 40060984.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 490
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 277.0,
      "completions/max_terminated_length": 277.0,
      "completions/mean_length": 158.0,
      "completions/mean_terminated_length": 158.0,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.5077559462254395,
      "grad_norm": 3.9194876650899495,
      "kl": 0.061767578125,
      "learning_rate": 4.894421744654979e-07,
      "loss": 0.0025,
      "num_tokens": 40147480.0,
      "reward": 0.8125,
      "reward_std": 0.4737785756587982,
      "rewards/reasoning_reward/mean": 0.8125,
      "rewards/reasoning_reward/std": 0.6726408004760742,
      "step": 491
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 283.0,
      "completions/max_terminated_length": 283.0,
      "completions/mean_length": 144.75,
      "completions/mean_terminated_length": 144.75,
      "completions/min_length": 66.0,
      "completions/min_terminated_length": 66.0,
      "epoch": 0.5087900723888314,
      "grad_norm": 3.096539621007116,
      "kl": 0.07666015625,
      "learning_rate": 4.878181936378124e-07,
      "loss": 0.0031,
      "num_tokens": 40224658.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.30366337299346924,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.45792141556739807,
      "step": 492
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 247.0,
      "completions/max_terminated_length": 247.0,
      "completions/mean_length": 170.95834350585938,
      "completions/mean_terminated_length": 170.95834350585938,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.5098241985522234,
      "grad_norm": 3.575630215884605,
      "kl": 0.06396484375,
      "learning_rate": 4.861943413855928e-07,
      "loss": 0.0026,
      "num_tokens": 40301689.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.3493061661720276,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.5450701117515564,
      "step": 493
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 248.0,
      "completions/max_terminated_length": 248.0,
      "completions/mean_length": 143.70834350585938,
      "completions/mean_terminated_length": 143.70834350585938,
      "completions/min_length": 65.0,
      "completions/min_terminated_length": 65.0,
      "epoch": 0.5108583247156153,
      "grad_norm": 3.7344523341009213,
      "kl": 0.044677734375,
      "learning_rate": 4.845706348481333e-07,
      "loss": 0.0018,
      "num_tokens": 40381394.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 494
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 265.0,
      "completions/max_terminated_length": 265.0,
      "completions/mean_length": 160.2916717529297,
      "completions/mean_terminated_length": 160.2916717529297,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.5118924508790073,
      "grad_norm": 1.866445752533792,
      "kl": 0.11328125,
      "learning_rate": 4.829470911631898e-07,
      "loss": 0.0045,
      "num_tokens": 40458361.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 495
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 224.0,
      "completions/max_terminated_length": 224.0,
      "completions/mean_length": 135.33334350585938,
      "completions/mean_terminated_length": 135.33334350585938,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.5129265770423992,
      "grad_norm": 4.4003611615962095,
      "kl": 0.080078125,
      "learning_rate": 4.813237274667993e-07,
      "loss": 0.0032,
      "num_tokens": 40548409.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.48112308979034424,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.8329709768295288,
      "step": 496
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 204.0,
      "completions/max_terminated_length": 204.0,
      "completions/mean_length": 152.7916717529297,
      "completions/mean_terminated_length": 152.7916717529297,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.5139607032057911,
      "grad_norm": 3.3720583603504815,
      "kl": 0.046142578125,
      "learning_rate": 4.797005608930991e-07,
      "loss": 0.0018,
      "num_tokens": 40628820.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 497
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 187.0,
      "completions/max_terminated_length": 187.0,
      "completions/mean_length": 142.58334350585938,
      "completions/mean_terminated_length": 142.58334350585938,
      "completions/min_length": 114.0,
      "completions/min_terminated_length": 114.0,
      "epoch": 0.5149948293691831,
      "grad_norm": 0.15329245374819234,
      "kl": 0.045166015625,
      "learning_rate": 4.780776085741462e-07,
      "loss": 0.0018,
      "num_tokens": 40704098.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.24077169597148895,
      "step": 498
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 249.0,
      "completions/max_terminated_length": 249.0,
      "completions/mean_length": 165.33334350585938,
      "completions/mean_terminated_length": 165.33334350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.516028955532575,
      "grad_norm": 3.4000622681936936,
      "kl": 0.08349609375,
      "learning_rate": 4.76454887639736e-07,
      "loss": 0.0033,
      "num_tokens": 40787242.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.41387641429901123,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.5835920572280884,
      "step": 499
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 152.20834350585938,
      "completions/mean_terminated_length": 152.20834350585938,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.5170630816959669,
      "grad_norm": 0.1571488063532437,
      "kl": 0.048583984375,
      "learning_rate": 4.7483241521722154e-07,
      "loss": 0.0019,
      "num_tokens": 40865215.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 500
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 261.0,
      "completions/max_terminated_length": 261.0,
      "completions/mean_length": 138.70834350585938,
      "completions/mean_terminated_length": 138.70834350585938,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.5180972078593589,
      "grad_norm": 2.487583254626514,
      "kl": 0.06689453125,
      "learning_rate": 4.7321020843133326e-07,
      "loss": 0.0027,
      "num_tokens": 40950344.0,
      "reward": 1.3402776718139648,
      "reward_std": 0.09820928424596786,
      "rewards/reasoning_reward/mean": 1.3402776718139648,
      "rewards/reasoning_reward/std": 0.4622962176799774,
      "step": 501
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 209.0,
      "completions/max_terminated_length": 209.0,
      "completions/mean_length": 145.0416717529297,
      "completions/mean_terminated_length": 145.0416717529297,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.5191313340227508,
      "grad_norm": 4.3586596392892565,
      "kl": 0.0693359375,
      "learning_rate": 4.7158828440399747e-07,
      "loss": 0.0028,
      "num_tokens": 41034001.0,
      "reward": 1.2361111640930176,
      "reward_std": 0.382678747177124,
      "rewards/reasoning_reward/mean": 1.236111044883728,
      "rewards/reasoning_reward/std": 0.7691463828086853,
      "step": 502
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 224.0,
      "completions/max_terminated_length": 224.0,
      "completions/mean_length": 150.58334350585938,
      "completions/mean_terminated_length": 150.58334350585938,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.5201654601861427,
      "grad_norm": 3.884759725833334,
      "kl": 0.059326171875,
      "learning_rate": 4.699666602541565e-07,
      "loss": 0.0024,
      "num_tokens": 41110591.0,
      "reward": 0.75,
      "reward_std": 0.44819486141204834,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.7372097969055176,
      "step": 503
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 253.0,
      "completions/max_terminated_length": 253.0,
      "completions/mean_length": 176.5,
      "completions/mean_terminated_length": 176.5,
      "completions/min_length": 119.0,
      "completions/min_terminated_length": 119.0,
      "epoch": 0.5211995863495347,
      "grad_norm": 3.6687504446873556,
      "kl": 0.06884765625,
      "learning_rate": 4.683453530975872e-07,
      "loss": 0.0028,
      "num_tokens": 41188139.0,
      "reward": 0.75,
      "reward_std": 0.45045679807662964,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.5897678136825562,
      "step": 504
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 192.0,
      "completions/max_terminated_length": 192.0,
      "completions/mean_length": 145.375,
      "completions/mean_terminated_length": 145.375,
      "completions/min_length": 74.0,
      "completions/min_terminated_length": 74.0,
      "epoch": 0.5222337125129266,
      "grad_norm": 4.270390001851114,
      "kl": 0.056396484375,
      "learning_rate": 4.6672438004672074e-07,
      "loss": 0.0023,
      "num_tokens": 41265292.0,
      "reward": 0.8958333730697632,
      "reward_std": 0.24185511469841003,
      "rewards/reasoning_reward/mean": 0.8958333134651184,
      "rewards/reasoning_reward/std": 0.29411497712135315,
      "step": 505
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 284.0,
      "completions/max_terminated_length": 284.0,
      "completions/mean_length": 160.83334350585938,
      "completions/mean_terminated_length": 160.83334350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.5232678386763185,
      "grad_norm": 3.860866061268796,
      "kl": 0.07177734375,
      "learning_rate": 4.6510375821046204e-07,
      "loss": 0.0029,
      "num_tokens": 41349392.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.47301241755485535,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.6915640830993652,
      "step": 506
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 132.4166717529297,
      "completions/mean_terminated_length": 132.4166717529297,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.5243019648397105,
      "grad_norm": 3.5645240907272,
      "kl": 0.050048828125,
      "learning_rate": 4.6348350469400885e-07,
      "loss": 0.002,
      "num_tokens": 41428746.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 507
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 162.58334350585938,
      "completions/mean_terminated_length": 162.58334350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.5253360910031024,
      "grad_norm": 4.091927437075089,
      "kl": 0.06396484375,
      "learning_rate": 4.618636365986714e-07,
      "loss": 0.0026,
      "num_tokens": 41509568.0,
      "reward": 0.7291666865348816,
      "reward_std": 0.4690367579460144,
      "rewards/reasoning_reward/mean": 0.7291666865348816,
      "rewards/reasoning_reward/std": 0.5706435441970825,
      "step": 508
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 245.0,
      "completions/max_terminated_length": 245.0,
      "completions/mean_length": 151.70834350585938,
      "completions/mean_terminated_length": 151.70834350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.5263702171664943,
      "grad_norm": 3.4807276011577883,
      "kl": 0.07177734375,
      "learning_rate": 4.602441710216922e-07,
      "loss": 0.0029,
      "num_tokens": 41588865.0,
      "reward": 1.25,
      "reward_std": 0.3314744830131531,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.675663948059082,
      "step": 509
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 238.0,
      "completions/max_terminated_length": 238.0,
      "completions/mean_length": 123.95833587646484,
      "completions/mean_terminated_length": 123.95833587646484,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.5274043433298863,
      "grad_norm": 3.5624810608659483,
      "kl": 0.0751953125,
      "learning_rate": 4.586251250560648e-07,
      "loss": 0.003,
      "num_tokens": 41672816.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.3493061661720276,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.434057354927063,
      "step": 510
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 194.0,
      "completions/max_terminated_length": 194.0,
      "completions/mean_length": 129.125,
      "completions/mean_terminated_length": 129.125,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.5284384694932782,
      "grad_norm": 2.1052711972616454,
      "kl": 0.07275390625,
      "learning_rate": 4.5700651579035453e-07,
      "loss": 0.0029,
      "num_tokens": 41754563.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.8164966106414795,
      "step": 511
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 147.58334350585938,
      "completions/mean_terminated_length": 147.58334350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.5294725956566702,
      "grad_norm": 3.6467154627194254,
      "kl": 0.06982421875,
      "learning_rate": 4.55388360308517e-07,
      "loss": 0.0028,
      "num_tokens": 41836945.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.4629100561141968,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.7755315899848938,
      "step": 512
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 197.0,
      "completions/max_terminated_length": 197.0,
      "completions/mean_length": 142.20834350585938,
      "completions/mean_terminated_length": 142.20834350585938,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.5305067218200621,
      "grad_norm": 2.9628853082893154,
      "kl": 0.07861328125,
      "learning_rate": 4.5377067568971837e-07,
      "loss": 0.0031,
      "num_tokens": 41925614.0,
      "reward": 1.1875,
      "reward_std": 0.183963343501091,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.28788962960243225,
      "step": 513
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 166.5416717529297,
      "completions/mean_terminated_length": 166.5416717529297,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.531540847983454,
      "grad_norm": 1.7751423102606667,
      "kl": 0.040771484375,
      "learning_rate": 4.521534790081549e-07,
      "loss": 0.0016,
      "num_tokens": 42009907.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.10206207633018494,
      "step": 514
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 214.0,
      "completions/max_terminated_length": 214.0,
      "completions/mean_length": 157.70834350585938,
      "completions/mean_terminated_length": 157.70834350585938,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.532574974146846,
      "grad_norm": 2.6567778397813617,
      "kl": 0.0771484375,
      "learning_rate": 4.505367873328731e-07,
      "loss": 0.0031,
      "num_tokens": 42100620.0,
      "reward": 1.5208333730697632,
      "reward_std": 0.21309106051921844,
      "rewards/reasoning_reward/mean": 1.5208333730697632,
      "rewards/reasoning_reward/std": 0.453948050737381,
      "step": 515
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 138.9166717529297,
      "completions/mean_terminated_length": 138.9166717529297,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.5336091003102379,
      "grad_norm": 3.1909559071085325,
      "kl": 0.0458984375,
      "learning_rate": 4.489206177275889e-07,
      "loss": 0.0018,
      "num_tokens": 42185474.0,
      "reward": 1.125,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.4484272003173828,
      "step": 516
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 262.0,
      "completions/max_terminated_length": 262.0,
      "completions/mean_length": 151.33334350585938,
      "completions/mean_terminated_length": 151.33334350585938,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.5346432264736298,
      "grad_norm": 2.715317039850287,
      "kl": 0.06982421875,
      "learning_rate": 4.473049872505081e-07,
      "loss": 0.0028,
      "num_tokens": 42269130.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.20693820714950562,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.721060037612915,
      "step": 517
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 208.0,
      "completions/max_terminated_length": 208.0,
      "completions/mean_length": 133.7916717529297,
      "completions/mean_terminated_length": 133.7916717529297,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.5356773526370218,
      "grad_norm": 4.433835698458295,
      "kl": 0.07763671875,
      "learning_rate": 4.4568991295414637e-07,
      "loss": 0.0031,
      "num_tokens": 42358405.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.3900056481361389,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "step": 518
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 230.0,
      "completions/max_terminated_length": 230.0,
      "completions/mean_length": 126.04167175292969,
      "completions/mean_terminated_length": 126.04167175292969,
      "completions/min_length": 69.0,
      "completions/min_terminated_length": 69.0,
      "epoch": 0.5367114788004137,
      "grad_norm": 2.490946137525958,
      "kl": 0.07861328125,
      "learning_rate": 4.440754118851486e-07,
      "loss": 0.0032,
      "num_tokens": 42437086.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 519
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 147.125,
      "completions/mean_terminated_length": 147.125,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.5377456049638056,
      "grad_norm": 4.116295710152532,
      "kl": 0.0703125,
      "learning_rate": 4.424615010841099e-07,
      "loss": 0.0028,
      "num_tokens": 42521801.0,
      "reward": 1.0138888359069824,
      "reward_std": 0.4999142587184906,
      "rewards/reasoning_reward/mean": 1.0138888359069824,
      "rewards/reasoning_reward/std": 0.547097384929657,
      "step": 520
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 263.0,
      "completions/max_terminated_length": 263.0,
      "completions/mean_length": 169.375,
      "completions/mean_terminated_length": 169.375,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.5387797311271976,
      "grad_norm": 3.031621149484474,
      "kl": 0.10595703125,
      "learning_rate": 4.4084819758539506e-07,
      "loss": 0.0043,
      "num_tokens": 42604818.0,
      "reward": 1.2986111640930176,
      "reward_std": 0.1608150601387024,
      "rewards/reasoning_reward/mean": 1.298611044883728,
      "rewards/reasoning_reward/std": 0.39311453700065613,
      "step": 521
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 294.0,
      "completions/max_terminated_length": 294.0,
      "completions/mean_length": 187.875,
      "completions/mean_terminated_length": 187.875,
      "completions/min_length": 127.0,
      "completions/min_terminated_length": 127.0,
      "epoch": 0.5398138572905895,
      "grad_norm": 1.8896709692670526,
      "kl": 0.0458984375,
      "learning_rate": 4.3923551841695885e-07,
      "loss": 0.0018,
      "num_tokens": 42684495.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 522
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 201.0,
      "completions/max_terminated_length": 201.0,
      "completions/mean_length": 141.25,
      "completions/mean_terminated_length": 141.25,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.5408479834539814,
      "grad_norm": 2.5348710252262485,
      "kl": 0.048583984375,
      "learning_rate": 4.376234806001665e-07,
      "loss": 0.0019,
      "num_tokens": 42762557.0,
      "reward": 1.4375,
      "reward_std": 0.08625819534063339,
      "rewards/reasoning_reward/mean": 1.4375,
      "rewards/reasoning_reward/std": 0.37044334411621094,
      "step": 523
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 152.0,
      "completions/mean_terminated_length": 152.0,
      "completions/min_length": 57.0,
      "completions/min_terminated_length": 57.0,
      "epoch": 0.5418821096173733,
      "grad_norm": 4.4258462156938965,
      "kl": 0.06298828125,
      "learning_rate": 4.360121011496142e-07,
      "loss": 0.0025,
      "num_tokens": 42841621.0,
      "reward": 0.6458333730697632,
      "reward_std": 0.41873571276664734,
      "rewards/reasoning_reward/mean": 0.6458333134651184,
      "rewards/reasoning_reward/std": 0.580089271068573,
      "step": 524
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 279.0,
      "completions/max_terminated_length": 279.0,
      "completions/mean_length": 165.2916717529297,
      "completions/mean_terminated_length": 165.2916717529297,
      "completions/min_length": 125.0,
      "completions/min_terminated_length": 125.0,
      "epoch": 0.5429162357807652,
      "grad_norm": 2.973908534715797,
      "kl": 0.056640625,
      "learning_rate": 4.344013970729489e-07,
      "loss": 0.0023,
      "num_tokens": 42926292.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.1767766922712326,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.7868369221687317,
      "step": 525
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 228.0,
      "completions/max_terminated_length": 228.0,
      "completions/mean_length": 155.625,
      "completions/mean_terminated_length": 155.625,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.5439503619441571,
      "grad_norm": 4.478416886750731,
      "kl": 0.0732421875,
      "learning_rate": 4.327913853706893e-07,
      "loss": 0.0029,
      "num_tokens": 43011243.0,
      "reward": 1.375,
      "reward_std": 0.39814266562461853,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.5366967916488647,
      "step": 526
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 251.0,
      "completions/max_terminated_length": 251.0,
      "completions/mean_length": 133.4166717529297,
      "completions/mean_terminated_length": 133.4166717529297,
      "completions/min_length": 64.0,
      "completions/min_terminated_length": 64.0,
      "epoch": 0.5449844881075491,
      "grad_norm": 0.24843813220604533,
      "kl": 0.07666015625,
      "learning_rate": 4.3118208303604635e-07,
      "loss": 0.0031,
      "num_tokens": 43096325.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.8340576887130737,
      "step": 527
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 254.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 148.875,
      "completions/mean_terminated_length": 148.875,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.546018614270941,
      "grad_norm": 3.677049891807094,
      "kl": 0.09423828125,
      "learning_rate": 4.295735070547438e-07,
      "loss": 0.0038,
      "num_tokens": 43184346.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.3438849151134491,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.8905196785926819,
      "step": 528
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 222.0,
      "completions/max_terminated_length": 222.0,
      "completions/mean_length": 159.4166717529297,
      "completions/mean_terminated_length": 159.4166717529297,
      "completions/min_length": 120.0,
      "completions/min_terminated_length": 120.0,
      "epoch": 0.5470527404343329,
      "grad_norm": 3.3666791301046612,
      "kl": 0.059326171875,
      "learning_rate": 4.2796567440483904e-07,
      "loss": 0.0024,
      "num_tokens": 43270244.0,
      "reward": 0.8125,
      "reward_std": 0.3438849151134491,
      "rewards/reasoning_reward/mean": 0.8125,
      "rewards/reasoning_reward/std": 0.5277618765830994,
      "step": 529
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 162.08334350585938,
      "completions/mean_terminated_length": 162.08334350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.5480868665977249,
      "grad_norm": 4.015099024696647,
      "kl": 0.07568359375,
      "learning_rate": 4.263586020565436e-07,
      "loss": 0.003,
      "num_tokens": 43349494.0,
      "reward": 0.6458333730697632,
      "reward_std": 0.3310800790786743,
      "rewards/reasoning_reward/mean": 0.6458333134651184,
      "rewards/reasoning_reward/std": 0.5610387921333313,
      "step": 530
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 226.0,
      "completions/max_terminated_length": 226.0,
      "completions/mean_length": 142.125,
      "completions/mean_terminated_length": 142.125,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.5491209927611168,
      "grad_norm": 1.6909730449484282,
      "kl": 0.07861328125,
      "learning_rate": 4.2475230697204446e-07,
      "loss": 0.0032,
      "num_tokens": 43439009.0,
      "reward": 1.3125,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 1.3125,
      "rewards/reasoning_reward/std": 0.4618605971336365,
      "step": 531
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 156.375,
      "completions/mean_terminated_length": 156.375,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.5501551189245087,
      "grad_norm": 3.64263703695767,
      "kl": 0.08056640625,
      "learning_rate": 4.2314680610532445e-07,
      "loss": 0.0032,
      "num_tokens": 43528666.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.23894576728343964,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 532
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 156.9166717529297,
      "completions/mean_terminated_length": 156.9166717529297,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.5511892450879007,
      "grad_norm": 3.424907168401309,
      "kl": 0.05859375,
      "learning_rate": 4.2154211640198426e-07,
      "loss": 0.0023,
      "num_tokens": 43611552.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 533
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 154.2916717529297,
      "completions/mean_terminated_length": 154.2916717529297,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.5522233712512926,
      "grad_norm": 3.695112863190559,
      "kl": 0.1298828125,
      "learning_rate": 4.199382547990625e-07,
      "loss": 0.0052,
      "num_tokens": 43690327.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.38613972067832947,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "step": 534
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 133.75,
      "completions/mean_terminated_length": 133.75,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.5532574974146846,
      "grad_norm": 2.0790329349480148,
      "kl": 0.047119140625,
      "learning_rate": 4.1833523822485766e-07,
      "loss": 0.0019,
      "num_tokens": 43774529.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 535
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 170.58334350585938,
      "completions/mean_terminated_length": 170.58334350585938,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.5542916235780765,
      "grad_norm": 2.323046338652214,
      "kl": 0.0830078125,
      "learning_rate": 4.167330835987489e-07,
      "loss": 0.0033,
      "num_tokens": 43852103.0,
      "reward": 0.7291666865348816,
      "reward_std": 0.19795583188533783,
      "rewards/reasoning_reward/mean": 0.7291666865348816,
      "rewards/reasoning_reward/std": 0.6251811385154724,
      "step": 536
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 248.0,
      "completions/max_terminated_length": 248.0,
      "completions/mean_length": 170.5,
      "completions/mean_terminated_length": 170.5,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.5553257497414684,
      "grad_norm": 4.195661413844518,
      "kl": 0.08544921875,
      "learning_rate": 4.1513180783101807e-07,
      "loss": 0.0034,
      "num_tokens": 43936347.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.4177326560020447,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.580089271068573,
      "step": 537
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 202.0,
      "completions/max_terminated_length": 202.0,
      "completions/mean_length": 149.9166717529297,
      "completions/mean_terminated_length": 149.9166717529297,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.5563598759048604,
      "grad_norm": 4.132578751644871,
      "kl": 0.07080078125,
      "learning_rate": 4.135314278226708e-07,
      "loss": 0.0028,
      "num_tokens": 44017537.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.41387641429901123,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.8030737638473511,
      "step": 538
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 132.5416717529297,
      "completions/mean_terminated_length": 132.5416717529297,
      "completions/min_length": 57.0,
      "completions/min_terminated_length": 57.0,
      "epoch": 0.5573940020682523,
      "grad_norm": 6.445791655514141,
      "kl": 0.0498046875,
      "learning_rate": 4.119319604652583e-07,
      "loss": 0.002,
      "num_tokens": 44097166.0,
      "reward": 0.625,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 539
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 230.0,
      "completions/max_terminated_length": 230.0,
      "completions/mean_length": 145.83334350585938,
      "completions/mean_terminated_length": 145.83334350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.5584281282316442,
      "grad_norm": 3.455599464978524,
      "kl": 0.09423828125,
      "learning_rate": 4.1033342264069887e-07,
      "loss": 0.0038,
      "num_tokens": 44176954.0,
      "reward": 1.2222222089767456,
      "reward_std": 0.24338775873184204,
      "rewards/reasoning_reward/mean": 1.2222222089767456,
      "rewards/reasoning_reward/std": 0.47565528750419617,
      "step": 540
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 151.375,
      "completions/mean_terminated_length": 151.375,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.5594622543950362,
      "grad_norm": 31.684587209903366,
      "kl": 0.333984375,
      "learning_rate": 4.0873583122109986e-07,
      "loss": 0.0134,
      "num_tokens": 44265163.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 541
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 189.0,
      "completions/max_terminated_length": 189.0,
      "completions/mean_length": 135.0,
      "completions/mean_terminated_length": 135.0,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.5604963805584281,
      "grad_norm": 2.58155802773976,
      "kl": 0.061279296875,
      "learning_rate": 4.071392030685799e-07,
      "loss": 0.0025,
      "num_tokens": 44344491.0,
      "reward": 0.8958333730697632,
      "reward_std": 0.19795583188533783,
      "rewards/reasoning_reward/mean": 0.8958333134651184,
      "rewards/reasoning_reward/std": 0.3605300188064575,
      "step": 542
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 309.0,
      "completions/max_terminated_length": 309.0,
      "completions/mean_length": 177.08334350585938,
      "completions/mean_terminated_length": 177.60870361328125,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.56153050672182,
      "grad_norm": 3.9169957578271273,
      "kl": 0.057373046875,
      "learning_rate": 4.055435550350903e-07,
      "loss": 0.0023,
      "num_tokens": 44430293.0,
      "reward": 1.2986111640930176,
      "reward_std": 0.34180140495300293,
      "rewards/reasoning_reward/mean": 1.298611044883728,
      "rewards/reasoning_reward/std": 0.46878182888031006,
      "step": 543
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 125.625,
      "completions/mean_terminated_length": 125.625,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.562564632885212,
      "grad_norm": 3.5379641241095476,
      "kl": 0.049560546875,
      "learning_rate": 4.039489039622376e-07,
      "loss": 0.002,
      "num_tokens": 44511572.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 544
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 225.0,
      "completions/max_terminated_length": 225.0,
      "completions/mean_length": 136.7916717529297,
      "completions/mean_terminated_length": 136.7916717529297,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.5635987590486039,
      "grad_norm": 2.7760862641762913,
      "kl": 0.0576171875,
      "learning_rate": 4.023552666811056e-07,
      "loss": 0.0023,
      "num_tokens": 44591183.0,
      "reward": 0.6875,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 0.6875,
      "rewards/reasoning_reward/std": 0.5067479610443115,
      "step": 545
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 156.70834350585938,
      "completions/mean_terminated_length": 156.70834350585938,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.5646328852119958,
      "grad_norm": 3.308306062539149,
      "kl": 0.0439453125,
      "learning_rate": 4.0076266001207796e-07,
      "loss": 0.0018,
      "num_tokens": 44672504.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.3255884051322937,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.40322521328926086,
      "step": 546
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 182.0,
      "completions/max_terminated_length": 182.0,
      "completions/mean_length": 109.91667175292969,
      "completions/mean_terminated_length": 109.91667175292969,
      "completions/min_length": 69.0,
      "completions/min_terminated_length": 69.0,
      "epoch": 0.5656670113753878,
      "grad_norm": 2.3476228247469417,
      "kl": 0.0478515625,
      "learning_rate": 3.9917110076466054e-07,
      "loss": 0.0019,
      "num_tokens": 44748886.0,
      "reward": 0.375,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 547
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 210.0,
      "completions/max_terminated_length": 210.0,
      "completions/mean_length": 139.375,
      "completions/mean_terminated_length": 139.375,
      "completions/min_length": 71.0,
      "completions/min_terminated_length": 71.0,
      "epoch": 0.5667011375387797,
      "grad_norm": 1.9607809752536416,
      "kl": 0.04296875,
      "learning_rate": 3.9758060573730376e-07,
      "loss": 0.0017,
      "num_tokens": 44826023.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.2314550280570984,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4871538281440735,
      "step": 548
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 225.0,
      "completions/max_terminated_length": 225.0,
      "completions/mean_length": 136.20834350585938,
      "completions/mean_terminated_length": 136.20834350585938,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.5677352637021716,
      "grad_norm": 4.07950906711803,
      "kl": 0.057861328125,
      "learning_rate": 3.9599119171722575e-07,
      "loss": 0.0023,
      "num_tokens": 44906724.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.416355699300766,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.5450701713562012,
      "step": 549
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 295.0,
      "completions/max_terminated_length": 295.0,
      "completions/mean_length": 157.875,
      "completions/mean_terminated_length": 157.875,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.5687693898655636,
      "grad_norm": 2.389180002667207,
      "kl": 0.0634765625,
      "learning_rate": 3.9440287548023484e-07,
      "loss": 0.0025,
      "num_tokens": 44984433.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.07715167850255966,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4402732849121094,
      "step": 550
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 143.75,
      "completions/mean_terminated_length": 143.75,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.5698035160289555,
      "grad_norm": 3.4273893764803365,
      "kl": 0.083984375,
      "learning_rate": 3.928156737905525e-07,
      "loss": 0.0034,
      "num_tokens": 45067011.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.34503278136253357,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 551
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 146.6666717529297,
      "completions/mean_terminated_length": 146.6666717529297,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.5708376421923474,
      "grad_norm": 3.5104041268905806,
      "kl": 0.0830078125,
      "learning_rate": 3.912296034006365e-07,
      "loss": 0.0033,
      "num_tokens": 45143987.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.24999213218688965,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.3120467960834503,
      "step": 552
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 149.95834350585938,
      "completions/mean_terminated_length": 149.95834350585938,
      "completions/min_length": 64.0,
      "completions/min_terminated_length": 64.0,
      "epoch": 0.5718717683557394,
      "grad_norm": 0.29409247983573755,
      "kl": 0.059814453125,
      "learning_rate": 3.896446810510041e-07,
      "loss": 0.0024,
      "num_tokens": 45229122.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 553
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 254.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 177.75,
      "completions/mean_terminated_length": 177.75,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.5729058945191313,
      "grad_norm": 3.9211078693164234,
      "kl": 0.07275390625,
      "learning_rate": 3.880609234700554e-07,
      "loss": 0.0029,
      "num_tokens": 45306684.0,
      "reward": 0.25,
      "reward_std": 0.43459486961364746,
      "rewards/reasoning_reward/mean": 0.25,
      "rewards/reasoning_reward/std": 0.41702884435653687,
      "step": 554
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 297.0,
      "completions/max_terminated_length": 297.0,
      "completions/mean_length": 144.0,
      "completions/mean_terminated_length": 144.0,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.5739400206825233,
      "grad_norm": 12.819817865471965,
      "kl": 0.080078125,
      "learning_rate": 3.8647834737389637e-07,
      "loss": 0.0032,
      "num_tokens": 45389860.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 555
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 165.0,
      "completions/max_terminated_length": 165.0,
      "completions/mean_length": 111.45833587646484,
      "completions/mean_terminated_length": 111.45833587646484,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.5749741468459152,
      "grad_norm": 2.5262171083580074,
      "kl": 0.048583984375,
      "learning_rate": 3.8489696946616334e-07,
      "loss": 0.0019,
      "num_tokens": 45469223.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 556
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 249.0,
      "completions/max_terminated_length": 249.0,
      "completions/mean_length": 158.625,
      "completions/mean_terminated_length": 158.625,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.5760082730093071,
      "grad_norm": 2.651554531736017,
      "kl": 0.0595703125,
      "learning_rate": 3.833168064378455e-07,
      "loss": 0.0024,
      "num_tokens": 45546182.0,
      "reward": 0.375,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 557
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 185.0,
      "completions/max_terminated_length": 185.0,
      "completions/mean_length": 119.25,
      "completions/mean_terminated_length": 119.25,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.5770423991726991,
      "grad_norm": 3.386660184100593,
      "kl": 0.0625,
      "learning_rate": 3.817378749671095e-07,
      "loss": 0.0025,
      "num_tokens": 45630044.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.34503278136253357,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.5835920572280884,
      "step": 558
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 261.0,
      "completions/max_terminated_length": 261.0,
      "completions/mean_length": 171.08334350585938,
      "completions/mean_terminated_length": 171.08334350585938,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.578076525336091,
      "grad_norm": 3.757186136491232,
      "kl": 0.0712890625,
      "learning_rate": 3.801601917191237e-07,
      "loss": 0.0029,
      "num_tokens": 45714430.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.34930619597435,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.6197241544723511,
      "step": 559
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 264.0,
      "completions/max_terminated_length": 264.0,
      "completions/mean_length": 133.375,
      "completions/mean_terminated_length": 133.375,
      "completions/min_length": 71.0,
      "completions/min_terminated_length": 71.0,
      "epoch": 0.5791106514994829,
      "grad_norm": 4.509762581052894,
      "kl": 0.060302734375,
      "learning_rate": 3.7858377334588127e-07,
      "loss": 0.0024,
      "num_tokens": 45800863.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.30217814445495605,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.6590369343757629,
      "step": 560
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 302.0,
      "completions/max_terminated_length": 302.0,
      "completions/mean_length": 169.9166717529297,
      "completions/mean_terminated_length": 169.9166717529297,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.5801447776628749,
      "grad_norm": 4.1190032329728385,
      "kl": 0.07421875,
      "learning_rate": 3.7700863648602516e-07,
      "loss": 0.003,
      "num_tokens": 45882437.0,
      "reward": 1.25,
      "reward_std": 0.48957228660583496,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.7518094182014465,
      "step": 561
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 226.0,
      "completions/max_terminated_length": 226.0,
      "completions/mean_length": 145.33334350585938,
      "completions/mean_terminated_length": 145.33334350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.5811789038262668,
      "grad_norm": 3.870893194153702,
      "kl": 0.09912109375,
      "learning_rate": 3.7543479776467244e-07,
      "loss": 0.004,
      "num_tokens": 45962437.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.3900056481361389,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.868114709854126,
      "step": 562
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 185.625,
      "completions/mean_terminated_length": 185.625,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.5822130299896587,
      "grad_norm": 3.3202411092809654,
      "kl": 0.09814453125,
      "learning_rate": 3.7386227379323855e-07,
      "loss": 0.0039,
      "num_tokens": 46040004.0,
      "reward": 0.1875,
      "reward_std": 0.33108004927635193,
      "rewards/reasoning_reward/mean": 0.1875,
      "rewards/reasoning_reward/std": 0.4121128022670746,
      "step": 563
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 194.0,
      "completions/max_terminated_length": 194.0,
      "completions/mean_length": 139.0,
      "completions/mean_terminated_length": 139.0,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.5832471561530507,
      "grad_norm": 4.0347158309096764,
      "kl": 0.0654296875,
      "learning_rate": 3.7229108116926223e-07,
      "loss": 0.0026,
      "num_tokens": 46116284.0,
      "reward": 0.9375,
      "reward_std": 0.35970625281333923,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.7845311760902405,
      "step": 564
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 312.0,
      "completions/max_terminated_length": 312.0,
      "completions/mean_length": 166.75,
      "completions/mean_terminated_length": 166.75,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.5842812823164426,
      "grad_norm": 4.136783811313354,
      "kl": 0.091796875,
      "learning_rate": 3.707212364762301e-07,
      "loss": 0.0037,
      "num_tokens": 46194998.0,
      "reward": 0.6875,
      "reward_std": 0.5491421222686768,
      "rewards/reasoning_reward/mean": 0.6875,
      "rewards/reasoning_reward/std": 0.7634660601615906,
      "step": 565
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 148.2916717529297,
      "completions/mean_terminated_length": 148.2916717529297,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.5853154084798345,
      "grad_norm": 3.7478667802868526,
      "kl": 0.07421875,
      "learning_rate": 3.691527562834018e-07,
      "loss": 0.003,
      "num_tokens": 46278789.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.6902530789375305,
      "step": 566
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 323.0,
      "completions/max_terminated_length": 323.0,
      "completions/mean_length": 167.75,
      "completions/mean_terminated_length": 167.75,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.5863495346432265,
      "grad_norm": 2.8655968955809286,
      "kl": 0.08056640625,
      "learning_rate": 3.6758565714563534e-07,
      "loss": 0.0032,
      "num_tokens": 46355567.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.2985045611858368,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.8198179602622986,
      "step": 567
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 274.0,
      "completions/max_terminated_length": 274.0,
      "completions/mean_length": 160.58334350585938,
      "completions/mean_terminated_length": 160.58334350585938,
      "completions/min_length": 73.0,
      "completions/min_terminated_length": 73.0,
      "epoch": 0.5873836608066184,
      "grad_norm": 4.048409611227816,
      "kl": 0.06640625,
      "learning_rate": 3.6601995560321164e-07,
      "loss": 0.0027,
      "num_tokens": 46435733.0,
      "reward": 0.7291666865348816,
      "reward_std": 0.47280198335647583,
      "rewards/reasoning_reward/mean": 0.7291666865348816,
      "rewards/reasoning_reward/std": 0.48854634165763855,
      "step": 568
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 159.9166717529297,
      "completions/mean_terminated_length": 159.9166717529297,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.5884177869700103,
      "grad_norm": 4.048728445965589,
      "kl": 0.078125,
      "learning_rate": 3.6445566818166075e-07,
      "loss": 0.0031,
      "num_tokens": 46518451.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.3610706031322479,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 569
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 212.0,
      "completions/max_terminated_length": 212.0,
      "completions/mean_length": 131.6666717529297,
      "completions/mean_terminated_length": 131.6666717529297,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.5894519131334023,
      "grad_norm": 3.9194927568414175,
      "kl": 0.06298828125,
      "learning_rate": 3.6289281139158685e-07,
      "loss": 0.0025,
      "num_tokens": 46601875.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.42645785212516785,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.8064504861831665,
      "step": 570
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 279.0,
      "completions/max_terminated_length": 279.0,
      "completions/mean_length": 170.6666717529297,
      "completions/mean_terminated_length": 170.6666717529297,
      "completions/min_length": 120.0,
      "completions/min_terminated_length": 120.0,
      "epoch": 0.5904860392967942,
      "grad_norm": 3.375166233888406,
      "kl": 0.08935546875,
      "learning_rate": 3.613314017284943e-07,
      "loss": 0.0036,
      "num_tokens": 46677899.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.23144195973873138,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.7442411184310913,
      "step": 571
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 188.0,
      "completions/max_terminated_length": 188.0,
      "completions/mean_length": 139.95834350585938,
      "completions/mean_terminated_length": 139.95834350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.5915201654601862,
      "grad_norm": 2.187391650107713,
      "kl": 0.0478515625,
      "learning_rate": 3.5977145567261355e-07,
      "loss": 0.0019,
      "num_tokens": 46755594.0,
      "reward": 0.5,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.5,
      "rewards/reasoning_reward/std": 0.5107539296150208,
      "step": 572
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 144.9166717529297,
      "completions/mean_terminated_length": 144.9166717529297,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.5925542916235781,
      "grad_norm": 3.5297891429966186,
      "kl": 0.0615234375,
      "learning_rate": 3.5821298968872696e-07,
      "loss": 0.0025,
      "num_tokens": 46835112.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.243839293718338,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.2917960286140442,
      "step": 573
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 240.0,
      "completions/max_terminated_length": 240.0,
      "completions/mean_length": 145.0416717529297,
      "completions/mean_terminated_length": 145.0416717529297,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.59358841778697,
      "grad_norm": 3.362879841575184,
      "kl": 0.0732421875,
      "learning_rate": 3.566560202259951e-07,
      "loss": 0.0029,
      "num_tokens": 46911793.0,
      "reward": 0.9375,
      "reward_std": 0.3688412308692932,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.49590715765953064,
      "step": 574
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 116.29167175292969,
      "completions/mean_terminated_length": 116.29167175292969,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.594622543950362,
      "grad_norm": 2.429845277216557,
      "kl": 0.0625,
      "learning_rate": 3.5510056371778337e-07,
      "loss": 0.0025,
      "num_tokens": 46990464.0,
      "reward": 1.125,
      "reward_std": 0.14773420989513397,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.30395936965942383,
      "step": 575
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 261.0,
      "completions/max_terminated_length": 261.0,
      "completions/mean_length": 162.83334350585938,
      "completions/mean_terminated_length": 162.83334350585938,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.5956566701137539,
      "grad_norm": 2.922280213132429,
      "kl": 0.0654296875,
      "learning_rate": 3.5354663658148834e-07,
      "loss": 0.0026,
      "num_tokens": 47079588.0,
      "reward": 1.1875,
      "reward_std": 0.2041093111038208,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.8945790529251099,
      "step": 576
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 240.0,
      "completions/max_terminated_length": 240.0,
      "completions/mean_length": 143.20834350585938,
      "completions/mean_terminated_length": 143.21739196777344,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.5966907962771458,
      "grad_norm": 4.143240683081162,
      "kl": 0.041259765625,
      "learning_rate": 3.5199425521836445e-07,
      "loss": 0.0017,
      "num_tokens": 47157417.0,
      "reward": 0.5833333730697632,
      "reward_std": 0.3900056481361389,
      "rewards/reasoning_reward/mean": 0.5833333134651184,
      "rewards/reasoning_reward/std": 0.5036101341247559,
      "step": 577
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 201.0,
      "completions/max_terminated_length": 201.0,
      "completions/mean_length": 140.4166717529297,
      "completions/mean_terminated_length": 140.4166717529297,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.5977249224405378,
      "grad_norm": 1.9659319807373157,
      "kl": 0.0634765625,
      "learning_rate": 3.50443436013351e-07,
      "loss": 0.0025,
      "num_tokens": 47235395.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.20412415266036987,
      "step": 578
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 204.0,
      "completions/max_terminated_length": 204.0,
      "completions/mean_length": 123.33333587646484,
      "completions/mean_terminated_length": 123.33333587646484,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.5987590486039297,
      "grad_norm": 2.2199998500371265,
      "kl": 0.06396484375,
      "learning_rate": 3.4889419533489895e-07,
      "loss": 0.0026,
      "num_tokens": 47321179.0,
      "reward": 1.2152777910232544,
      "reward_std": 0.07534459978342056,
      "rewards/reasoning_reward/mean": 1.2152777910232544,
      "rewards/reasoning_reward/std": 0.33506491780281067,
      "step": 579
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 308.0,
      "completions/max_terminated_length": 308.0,
      "completions/mean_length": 179.20834350585938,
      "completions/mean_terminated_length": 179.20834350585938,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.5997931747673216,
      "grad_norm": 4.027929778275415,
      "kl": 0.056640625,
      "learning_rate": 3.4734654953479863e-07,
      "loss": 0.0023,
      "num_tokens": 47401472.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.48678088188171387,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 580
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 212.0,
      "completions/max_terminated_length": 212.0,
      "completions/mean_length": 145.2916717529297,
      "completions/mean_terminated_length": 145.2916717529297,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.6008273009307136,
      "grad_norm": 4.203672290672796,
      "kl": 0.08349609375,
      "learning_rate": 3.458005149480068e-07,
      "loss": 0.0033,
      "num_tokens": 47483511.0,
      "reward": 0.9861111044883728,
      "reward_std": 0.3927455544471741,
      "rewards/reasoning_reward/mean": 0.9861111044883728,
      "rewards/reasoning_reward/std": 0.6683251857757568,
      "step": 581
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 136.0,
      "completions/mean_terminated_length": 136.0,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.6018614270941055,
      "grad_norm": 0.1514238170358154,
      "kl": 0.042236328125,
      "learning_rate": 3.4425610789247415e-07,
      "loss": 0.0017,
      "num_tokens": 47564375.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 582
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 224.0,
      "completions/max_terminated_length": 224.0,
      "completions/mean_length": 143.0,
      "completions/mean_terminated_length": 143.0,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.6028955532574974,
      "grad_norm": 3.123715432165474,
      "kl": 0.06689453125,
      "learning_rate": 3.4271334466897353e-07,
      "loss": 0.0027,
      "num_tokens": 47646583.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.23709973692893982,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.40322521328926086,
      "step": 583
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 290.0,
      "completions/max_terminated_length": 290.0,
      "completions/mean_length": 156.70834350585938,
      "completions/mean_terminated_length": 156.70834350585938,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.6039296794208894,
      "grad_norm": 3.5821953064727694,
      "kl": 0.0439453125,
      "learning_rate": 3.411722415609275e-07,
      "loss": 0.0018,
      "num_tokens": 47731776.0,
      "reward": 1.125,
      "reward_std": 0.2314550280570984,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.47204458713531494,
      "step": 584
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 150.08334350585938,
      "completions/mean_terminated_length": 150.08334350585938,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.6049638055842813,
      "grad_norm": 2.9553822886142007,
      "kl": 0.046630859375,
      "learning_rate": 3.396328148342366e-07,
      "loss": 0.0019,
      "num_tokens": 47811218.0,
      "reward": 0.4791666865348816,
      "reward_std": 0.31580695509910583,
      "rewards/reasoning_reward/mean": 0.4791666567325592,
      "rewards/reasoning_reward/std": 0.5413181781768799,
      "step": 585
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 136.4166717529297,
      "completions/mean_terminated_length": 136.4166717529297,
      "completions/min_length": 74.0,
      "completions/min_terminated_length": 74.0,
      "epoch": 0.6059979317476732,
      "grad_norm": 3.452553119930219,
      "kl": 0.059326171875,
      "learning_rate": 3.3809508073710754e-07,
      "loss": 0.0024,
      "num_tokens": 47899740.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.40824830532073975,
      "step": 586
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 282.0,
      "completions/max_terminated_length": 282.0,
      "completions/mean_length": 180.4166717529297,
      "completions/mean_terminated_length": 180.4166717529297,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.6070320579110652,
      "grad_norm": 2.0970296040353067,
      "kl": 0.052001953125,
      "learning_rate": 3.365590554998819e-07,
      "loss": 0.0021,
      "num_tokens": 47978398.0,
      "reward": 1.25,
      "reward_std": 0.1259881556034088,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.41702884435653687,
      "step": 587
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 243.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 151.625,
      "completions/mean_terminated_length": 147.7391357421875,
      "completions/min_length": 58.0,
      "completions/min_terminated_length": 58.0,
      "epoch": 0.6080661840744571,
      "grad_norm": 4.189419981041684,
      "kl": 0.0546875,
      "learning_rate": 3.350247553348647e-07,
      "loss": 0.0022,
      "num_tokens": 48060621.0,
      "reward": 0.875,
      "reward_std": 0.4082186818122864,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.4484272003173828,
      "step": 588
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 336.0,
      "completions/max_terminated_length": 336.0,
      "completions/mean_length": 158.25,
      "completions/mean_terminated_length": 158.25,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.609100310237849,
      "grad_norm": 3.5703666580841418,
      "kl": 0.054443359375,
      "learning_rate": 3.3349219643615344e-07,
      "loss": 0.0022,
      "num_tokens": 48139971.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.27392348647117615,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.35098204016685486,
      "step": 589
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 251.0,
      "completions/max_terminated_length": 251.0,
      "completions/mean_length": 182.5,
      "completions/mean_terminated_length": 182.5,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.610134436401241,
      "grad_norm": 3.272644263091865,
      "kl": 0.060791015625,
      "learning_rate": 3.319613949794668e-07,
      "loss": 0.0024,
      "num_tokens": 48223079.0,
      "reward": 0.6111111640930176,
      "reward_std": 0.31463193893432617,
      "rewards/reasoning_reward/mean": 0.6111111044883728,
      "rewards/reasoning_reward/std": 0.48070666193962097,
      "step": 590
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 295.0,
      "completions/max_terminated_length": 295.0,
      "completions/mean_length": 178.2916717529297,
      "completions/mean_terminated_length": 178.2916717529297,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.6111685625646329,
      "grad_norm": 3.8491229946452425,
      "kl": 0.06884765625,
      "learning_rate": 3.304323671219744e-07,
      "loss": 0.0028,
      "num_tokens": 48302822.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.24602244794368744,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.7708455324172974,
      "step": 591
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 168.0,
      "completions/max_terminated_length": 168.0,
      "completions/mean_length": 108.08333587646484,
      "completions/mean_terminated_length": 108.08333587646484,
      "completions/min_length": 55.0,
      "completions/min_terminated_length": 55.0,
      "epoch": 0.6122026887280249,
      "grad_norm": 3.2825922300107773,
      "kl": 0.052978515625,
      "learning_rate": 3.2890512900212585e-07,
      "loss": 0.0021,
      "num_tokens": 48380512.0,
      "reward": 0.5833333730697632,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.5833333134651184,
      "rewards/reasoning_reward/std": 0.5036101341247559,
      "step": 592
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 152.375,
      "completions/mean_terminated_length": 152.375,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.6132368148914168,
      "grad_norm": 2.5916120299999803,
      "kl": 0.053955078125,
      "learning_rate": 3.273796967394809e-07,
      "loss": 0.0022,
      "num_tokens": 48465905.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.6580052971839905,
      "step": 593
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 161.0,
      "completions/mean_terminated_length": 161.0,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.6142709410548087,
      "grad_norm": 4.020983723037926,
      "kl": 0.050048828125,
      "learning_rate": 3.2585608643453867e-07,
      "loss": 0.002,
      "num_tokens": 48543945.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.1767766922712326,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.8139966726303101,
      "step": 594
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 133.6666717529297,
      "completions/mean_terminated_length": 133.6666717529297,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.6153050672182007,
      "grad_norm": 2.05299852274566,
      "kl": 0.04052734375,
      "learning_rate": 3.2433431416856816e-07,
      "loss": 0.0016,
      "num_tokens": 48622017.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.20412415266036987,
      "step": 595
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 265.0,
      "completions/max_terminated_length": 265.0,
      "completions/mean_length": 171.20834350585938,
      "completions/mean_terminated_length": 171.20834350585938,
      "completions/min_length": 114.0,
      "completions/min_terminated_length": 114.0,
      "epoch": 0.6163391933815926,
      "grad_norm": 3.1066601718856584,
      "kl": 0.087890625,
      "learning_rate": 3.2281439600343835e-07,
      "loss": 0.0035,
      "num_tokens": 48704886.0,
      "reward": 0.5833333730697632,
      "reward_std": 0.21161314845085144,
      "rewards/reasoning_reward/mean": 0.5833333134651184,
      "rewards/reasoning_reward/std": 0.8427009582519531,
      "step": 596
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 262.0,
      "completions/max_terminated_length": 262.0,
      "completions/mean_length": 165.2916717529297,
      "completions/mean_terminated_length": 165.2916717529297,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.6173733195449845,
      "grad_norm": 4.305444073516952,
      "kl": 0.07763671875,
      "learning_rate": 3.2129634798144885e-07,
      "loss": 0.0031,
      "num_tokens": 48787885.0,
      "reward": 1.5625,
      "reward_std": 0.37917613983154297,
      "rewards/reasoning_reward/mean": 1.5625,
      "rewards/reasoning_reward/std": 0.517361581325531,
      "step": 597
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 215.0,
      "completions/max_terminated_length": 215.0,
      "completions/mean_length": 157.1666717529297,
      "completions/mean_terminated_length": 157.1666717529297,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.6184074457083765,
      "grad_norm": 3.1348740582949395,
      "kl": 0.07470703125,
      "learning_rate": 3.1978018612516024e-07,
      "loss": 0.003,
      "num_tokens": 48870625.0,
      "reward": 0.7291666865348816,
      "reward_std": 0.12400396913290024,
      "rewards/reasoning_reward/mean": 0.7291666865348816,
      "rewards/reasoning_reward/std": 0.9438492059707642,
      "step": 598
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 184.0,
      "completions/max_terminated_length": 184.0,
      "completions/mean_length": 137.45834350585938,
      "completions/mean_terminated_length": 137.45834350585938,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.6194415718717684,
      "grad_norm": 3.57892132145044,
      "kl": 0.0751953125,
      "learning_rate": 3.182659264372254e-07,
      "loss": 0.003,
      "num_tokens": 48953788.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.637022078037262,
      "step": 599
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 230.0,
      "completions/max_terminated_length": 230.0,
      "completions/mean_length": 142.2916717529297,
      "completions/mean_terminated_length": 142.2916717529297,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.6204756980351603,
      "grad_norm": 4.301884451387064,
      "kl": 0.0673828125,
      "learning_rate": 3.1675358490022006e-07,
      "loss": 0.0027,
      "num_tokens": 49032971.0,
      "reward": 0.375,
      "reward_std": 0.48112308979034424,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 600
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 155.4166717529297,
      "completions/mean_terminated_length": 155.4166717529297,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.6215098241985523,
      "grad_norm": 2.256674358684405,
      "kl": 0.056884765625,
      "learning_rate": 3.1524317747647487e-07,
      "loss": 0.0023,
      "num_tokens": 49112829.0,
      "reward": 1.1805555820465088,
      "reward_std": 0.12858611345291138,
      "rewards/reasoning_reward/mean": 1.1805554628372192,
      "rewards/reasoning_reward/std": 0.3366382122039795,
      "step": 601
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 154.95834350585938,
      "completions/mean_terminated_length": 154.95834350585938,
      "completions/min_length": 76.0,
      "completions/min_terminated_length": 76.0,
      "epoch": 0.6225439503619442,
      "grad_norm": 2.155514941763339,
      "kl": 0.0771484375,
      "learning_rate": 3.1373472010790613e-07,
      "loss": 0.0031,
      "num_tokens": 49189396.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.267261266708374,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.5036101937294006,
      "step": 602
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 266.0,
      "completions/max_terminated_length": 266.0,
      "completions/mean_length": 158.375,
      "completions/mean_terminated_length": 158.375,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.6235780765253361,
      "grad_norm": 3.1569287636174064,
      "kl": 0.0810546875,
      "learning_rate": 3.122282287158479e-07,
      "loss": 0.0032,
      "num_tokens": 49267197.0,
      "reward": 0.5625,
      "reward_std": 0.2848889231681824,
      "rewards/reasoning_reward/mean": 0.5625,
      "rewards/reasoning_reward/std": 0.517361581325531,
      "step": 603
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 149.125,
      "completions/mean_terminated_length": 149.125,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.6246122026887281,
      "grad_norm": 3.5123089746645637,
      "kl": 0.058349609375,
      "learning_rate": 3.1072371920088393e-07,
      "loss": 0.0023,
      "num_tokens": 49352248.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.243839293718338,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.4871537983417511,
      "step": 604
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 225.0,
      "completions/max_terminated_length": 225.0,
      "completions/mean_length": 164.95834350585938,
      "completions/mean_terminated_length": 164.95834350585938,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.6256463288521199,
      "grad_norm": 2.617225700730397,
      "kl": 0.080078125,
      "learning_rate": 3.092212074426799e-07,
      "loss": 0.0032,
      "num_tokens": 49430807.0,
      "reward": 1.0625,
      "reward_std": 0.08625819534063339,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.1689159870147705,
      "step": 605
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 159.70834350585938,
      "completions/mean_terminated_length": 159.70834350585938,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.6266804550155118,
      "grad_norm": 3.775214071699908,
      "kl": 0.07568359375,
      "learning_rate": 3.0772070929981587e-07,
      "loss": 0.003,
      "num_tokens": 49520960.0,
      "reward": 1.3125,
      "reward_std": 0.36753225326538086,
      "rewards/reasoning_reward/mean": 1.3125,
      "rewards/reasoning_reward/std": 0.6222766637802124,
      "step": 606
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 218.0,
      "completions/max_terminated_length": 218.0,
      "completions/mean_length": 166.20834350585938,
      "completions/mean_terminated_length": 166.20834350585938,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.6277145811789038,
      "grad_norm": 3.7546362665727453,
      "kl": 0.0615234375,
      "learning_rate": 3.062222406096183e-07,
      "loss": 0.0025,
      "num_tokens": 49602421.0,
      "reward": 0.6041666865348816,
      "reward_std": 0.4294546842575073,
      "rewards/reasoning_reward/mean": 0.6041666865348816,
      "rewards/reasoning_reward/std": 0.5893837213516235,
      "step": 607
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 226.0,
      "completions/max_terminated_length": 226.0,
      "completions/mean_length": 110.54167175292969,
      "completions/mean_terminated_length": 110.54167175292969,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.6287487073422957,
      "grad_norm": 0.1828600047194931,
      "kl": 0.039794921875,
      "learning_rate": 3.047258171879939e-07,
      "loss": 0.0016,
      "num_tokens": 49682514.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 608
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 224.0,
      "completions/max_terminated_length": 224.0,
      "completions/mean_length": 170.1666717529297,
      "completions/mean_terminated_length": 170.1666717529297,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.6297828335056876,
      "grad_norm": 1.781481839446534,
      "kl": 0.033447265625,
      "learning_rate": 3.032314548292618e-07,
      "loss": 0.0013,
      "num_tokens": 49762454.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 609
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 145.1666717529297,
      "completions/mean_terminated_length": 145.1666717529297,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.6308169596690796,
      "grad_norm": 3.893664449459787,
      "kl": 0.057861328125,
      "learning_rate": 3.0173916930598743e-07,
      "loss": 0.0023,
      "num_tokens": 49848010.0,
      "reward": 1.0763888359069824,
      "reward_std": 0.3868226408958435,
      "rewards/reasoning_reward/mean": 1.0763888359069824,
      "rewards/reasoning_reward/std": 0.6078773736953735,
      "step": 610
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 269.0,
      "completions/max_terminated_length": 269.0,
      "completions/mean_length": 154.375,
      "completions/mean_terminated_length": 154.375,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.6318510858324715,
      "grad_norm": 2.91304378338177,
      "kl": 0.07275390625,
      "learning_rate": 3.0024897636881556e-07,
      "loss": 0.0029,
      "num_tokens": 49926275.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.24966806173324585,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.458415687084198,
      "step": 611
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 248.0,
      "completions/max_terminated_length": 248.0,
      "completions/mean_length": 160.33334350585938,
      "completions/mean_terminated_length": 160.33334350585938,
      "completions/min_length": 123.0,
      "completions/min_terminated_length": 123.0,
      "epoch": 0.6328852119958635,
      "grad_norm": 4.645327728337031,
      "kl": 0.08935546875,
      "learning_rate": 2.9876089174630465e-07,
      "loss": 0.0036,
      "num_tokens": 50008955.0,
      "reward": 1.0625,
      "reward_std": 0.3766257166862488,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.5954993963241577,
      "step": 612
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 224.0,
      "completions/max_terminated_length": 224.0,
      "completions/mean_length": 114.58333587646484,
      "completions/mean_terminated_length": 114.58333587646484,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.6339193381592554,
      "grad_norm": 2.3500928948101976,
      "kl": 0.043701171875,
      "learning_rate": 2.972749311447602e-07,
      "loss": 0.0017,
      "num_tokens": 50086817.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 613
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 216.0,
      "completions/max_terminated_length": 216.0,
      "completions/mean_length": 144.33334350585938,
      "completions/mean_terminated_length": 144.33334350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.6349534643226473,
      "grad_norm": 4.815665886134638,
      "kl": 0.06689453125,
      "learning_rate": 2.957911102480694e-07,
      "loss": 0.0027,
      "num_tokens": 50165777.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.5201427936553955,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.5512666702270508,
      "step": 614
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 256.0,
      "completions/max_terminated_length": 256.0,
      "completions/mean_length": 150.7916717529297,
      "completions/mean_terminated_length": 150.7916717529297,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.6359875904860393,
      "grad_norm": 0.5275728635268186,
      "kl": 0.04052734375,
      "learning_rate": 2.943094447175356e-07,
      "loss": 0.0016,
      "num_tokens": 50248660.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 615
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 253.0,
      "completions/max_terminated_length": 253.0,
      "completions/mean_length": 169.6666717529297,
      "completions/mean_terminated_length": 169.6666717529297,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.6370217166494312,
      "grad_norm": 2.0323032948060398,
      "kl": 0.058837890625,
      "learning_rate": 2.9282995019171276e-07,
      "loss": 0.0024,
      "num_tokens": 50334548.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.8064504265785217,
      "step": 616
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 154.20834350585938,
      "completions/mean_terminated_length": 154.20834350585938,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.6380558428128231,
      "grad_norm": 3.175018020932587,
      "kl": 0.068359375,
      "learning_rate": 2.9135264228624036e-07,
      "loss": 0.0027,
      "num_tokens": 50413561.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.31925714015960693,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.417752206325531,
      "step": 617
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 153.6666717529297,
      "completions/mean_terminated_length": 153.6666717529297,
      "completions/min_length": 74.0,
      "completions/min_terminated_length": 74.0,
      "epoch": 0.6390899689762151,
      "grad_norm": 4.35022573120695,
      "kl": 0.07275390625,
      "learning_rate": 2.8987753659367884e-07,
      "loss": 0.0029,
      "num_tokens": 50488753.0,
      "reward": 0.9375,
      "reward_std": 0.24185511469841003,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.6645119190216064,
      "step": 618
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 210.0,
      "completions/max_terminated_length": 210.0,
      "completions/mean_length": 137.25,
      "completions/mean_terminated_length": 137.25,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.640124095139607,
      "grad_norm": 3.9637606696357435,
      "kl": 0.064453125,
      "learning_rate": 2.884046486833453e-07,
      "loss": 0.0026,
      "num_tokens": 50569367.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.705824613571167,
      "step": 619
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 313.0,
      "completions/max_terminated_length": 313.0,
      "completions/mean_length": 181.25,
      "completions/mean_terminated_length": 181.25,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.6411582213029989,
      "grad_norm": 2.8397850153757553,
      "kl": 0.060546875,
      "learning_rate": 2.8693399410114793e-07,
      "loss": 0.0024,
      "num_tokens": 50649053.0,
      "reward": 1.0694444179534912,
      "reward_std": 0.2472916543483734,
      "rewards/reasoning_reward/mean": 1.0694444179534912,
      "rewards/reasoning_reward/std": 0.3506951928138733,
      "step": 620
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 286.0,
      "completions/max_terminated_length": 286.0,
      "completions/mean_length": 176.33334350585938,
      "completions/mean_terminated_length": 176.33334350585938,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.6421923474663909,
      "grad_norm": 3.382273478222385,
      "kl": 0.07080078125,
      "learning_rate": 2.854655883694238e-07,
      "loss": 0.0028,
      "num_tokens": 50733821.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.28302299976348877,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.416485458612442,
      "step": 621
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 216.0,
      "completions/max_terminated_length": 216.0,
      "completions/mean_length": 172.2916717529297,
      "completions/mean_terminated_length": 172.2916717529297,
      "completions/min_length": 139.0,
      "completions/min_terminated_length": 139.0,
      "epoch": 0.6432264736297828,
      "grad_norm": 2.8244945361941944,
      "kl": 0.06982421875,
      "learning_rate": 2.83999446986773e-07,
      "loss": 0.0028,
      "num_tokens": 50815580.0,
      "reward": 0.8958333730697632,
      "reward_std": 0.2041093111038208,
      "rewards/reasoning_reward/mean": 0.8958333134651184,
      "rewards/reasoning_reward/std": 0.7937139868736267,
      "step": 622
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 211.0,
      "completions/max_terminated_length": 211.0,
      "completions/mean_length": 159.1666717529297,
      "completions/mean_terminated_length": 159.1666717529297,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.6442605997931747,
      "grad_norm": 3.773616346150942,
      "kl": 0.08056640625,
      "learning_rate": 2.82535585427897e-07,
      "loss": 0.0032,
      "num_tokens": 50897280.0,
      "reward": 0.8958333730697632,
      "reward_std": 0.33768826723098755,
      "rewards/reasoning_reward/mean": 0.8958333134651184,
      "rewards/reasoning_reward/std": 0.4418136179447174,
      "step": 623
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 238.0,
      "completions/max_terminated_length": 238.0,
      "completions/mean_length": 159.5,
      "completions/mean_terminated_length": 159.5,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.6452947259565667,
      "grad_norm": 3.3186602952698903,
      "kl": 0.07470703125,
      "learning_rate": 2.8107401914343363e-07,
      "loss": 0.003,
      "num_tokens": 50980988.0,
      "reward": 1.2708333730697632,
      "reward_std": 0.33086174726486206,
      "rewards/reasoning_reward/mean": 1.2708333730697632,
      "rewards/reasoning_reward/std": 0.4418136179447174,
      "step": 624
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 148.4166717529297,
      "completions/mean_terminated_length": 148.4166717529297,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.6463288521199586,
      "grad_norm": 4.081124219760009,
      "kl": 0.06689453125,
      "learning_rate": 2.796147635597954e-07,
      "loss": 0.0027,
      "num_tokens": 51065942.0,
      "reward": 1.4583333730697632,
      "reward_std": 0.36124157905578613,
      "rewards/reasoning_reward/mean": 1.4583333730697632,
      "rewards/reasoning_reward/std": 0.550032913684845,
      "step": 625
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 155.5416717529297,
      "completions/mean_terminated_length": 155.5416717529297,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.6473629782833505,
      "grad_norm": 4.171951829209572,
      "kl": 0.07373046875,
      "learning_rate": 2.781578340790053e-07,
      "loss": 0.003,
      "num_tokens": 51148627.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.39000558853149414,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.6863049864768982,
      "step": 626
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 345.0,
      "completions/max_terminated_length": 345.0,
      "completions/mean_length": 175.4166717529297,
      "completions/mean_terminated_length": 175.4166717529297,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.6483971044467425,
      "grad_norm": 3.816589069723601,
      "kl": 0.07763671875,
      "learning_rate": 2.767032460785356e-07,
      "loss": 0.0031,
      "num_tokens": 51238597.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.42099714279174805,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.6164266467094421,
      "step": 627
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 153.625,
      "completions/mean_terminated_length": 153.625,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.6494312306101344,
      "grad_norm": 1.955866878147104,
      "kl": 0.05078125,
      "learning_rate": 2.752510149111449e-07,
      "loss": 0.002,
      "num_tokens": 51316820.0,
      "reward": 1.1875,
      "reward_std": 0.13908717036247253,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.355469673871994,
      "step": 628
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 195.0,
      "completions/max_terminated_length": 195.0,
      "completions/mean_length": 150.58334350585938,
      "completions/mean_terminated_length": 150.58334350585938,
      "completions/min_length": 118.0,
      "completions/min_terminated_length": 118.0,
      "epoch": 0.6504653567735263,
      "grad_norm": 3.0788457748288915,
      "kl": 0.06396484375,
      "learning_rate": 2.738011559047155e-07,
      "loss": 0.0026,
      "num_tokens": 51402106.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.1767766922712326,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.5985338687896729,
      "step": 629
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 255.0,
      "completions/max_terminated_length": 255.0,
      "completions/mean_length": 162.20834350585938,
      "completions/mean_terminated_length": 162.20834350585938,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.6514994829369183,
      "grad_norm": 3.631670896990646,
      "kl": 0.064453125,
      "learning_rate": 2.723536843620931e-07,
      "loss": 0.0026,
      "num_tokens": 51486543.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.4769924581050873,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.6833289265632629,
      "step": 630
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 168.5,
      "completions/mean_terminated_length": 168.5,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.6525336091003102,
      "grad_norm": 3.121087345839574,
      "kl": 0.07275390625,
      "learning_rate": 2.7090861556092347e-07,
      "loss": 0.0029,
      "num_tokens": 51570619.0,
      "reward": 0.9236111044883728,
      "reward_std": 0.18046043813228607,
      "rewards/reasoning_reward/mean": 0.9236111044883728,
      "rewards/reasoning_reward/std": 0.7239504456520081,
      "step": 631
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 287.0,
      "completions/max_terminated_length": 287.0,
      "completions/mean_length": 173.33334350585938,
      "completions/mean_terminated_length": 173.33334350585938,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.6535677352637022,
      "grad_norm": 4.0132486156921,
      "kl": 0.07275390625,
      "learning_rate": 2.6946596475349305e-07,
      "loss": 0.0029,
      "num_tokens": 51655891.0,
      "reward": 1.125,
      "reward_std": 0.40037286281585693,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.5299029350280762,
      "step": 632
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 193.45834350585938,
      "completions/mean_terminated_length": 193.45834350585938,
      "completions/min_length": 132.0,
      "completions/min_terminated_length": 132.0,
      "epoch": 0.6546018614270941,
      "grad_norm": 3.686435293590089,
      "kl": 0.0712890625,
      "learning_rate": 2.680257471665661e-07,
      "loss": 0.0029,
      "num_tokens": 51745486.0,
      "reward": 1.25,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.4662523865699768,
      "step": 633
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 154.08334350585938,
      "completions/mean_terminated_length": 154.08334350585938,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.655635987590486,
      "grad_norm": 3.142110321096429,
      "kl": 0.058837890625,
      "learning_rate": 2.66587978001226e-07,
      "loss": 0.0024,
      "num_tokens": 51829440.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 634
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 215.0,
      "completions/max_terminated_length": 215.0,
      "completions/mean_length": 147.5,
      "completions/mean_terminated_length": 147.5,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.656670113753878,
      "grad_norm": 4.105530331463831,
      "kl": 0.09521484375,
      "learning_rate": 2.651526724327127e-07,
      "loss": 0.0038,
      "num_tokens": 51906116.0,
      "reward": 1.0,
      "reward_std": 0.4981178641319275,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.5316095352172852,
      "step": 635
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 210.0,
      "completions/max_terminated_length": 210.0,
      "completions/mean_length": 130.1666717529297,
      "completions/mean_terminated_length": 130.1666717529297,
      "completions/min_length": 70.0,
      "completions/min_terminated_length": 70.0,
      "epoch": 0.6577042399172699,
      "grad_norm": 3.203427593680448,
      "kl": 0.0537109375,
      "learning_rate": 2.6371984561026416e-07,
      "loss": 0.0021,
      "num_tokens": 51985168.0,
      "reward": 0.9861111044883728,
      "reward_std": 0.29100531339645386,
      "rewards/reasoning_reward/mean": 0.9861111044883728,
      "rewards/reasoning_reward/std": 0.35751646757125854,
      "step": 636
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 155.5,
      "completions/mean_terminated_length": 155.5,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.6587383660806618,
      "grad_norm": 4.913600177708958,
      "kl": 0.06396484375,
      "learning_rate": 2.622895126569562e-07,
      "loss": 0.0026,
      "num_tokens": 52063404.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.4082186818122864,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 637
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 289.0,
      "completions/max_terminated_length": 289.0,
      "completions/mean_length": 174.1666717529297,
      "completions/mean_terminated_length": 174.1666717529297,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.6597724922440538,
      "grad_norm": 2.4339757197483873,
      "kl": 0.058837890625,
      "learning_rate": 2.6086168866954175e-07,
      "loss": 0.0024,
      "num_tokens": 52148776.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.08908708393573761,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 638
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 187.0,
      "completions/max_terminated_length": 187.0,
      "completions/mean_length": 128.25,
      "completions/mean_terminated_length": 128.25,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.6608066184074457,
      "grad_norm": 2.3656571936353594,
      "kl": 0.072265625,
      "learning_rate": 2.5943638871829296e-07,
      "loss": 0.0029,
      "num_tokens": 52232894.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.8064504265785217,
      "step": 639
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 258.0,
      "completions/max_terminated_length": 258.0,
      "completions/mean_length": 154.0416717529297,
      "completions/mean_terminated_length": 154.0416717529297,
      "completions/min_length": 55.0,
      "completions/min_terminated_length": 55.0,
      "epoch": 0.6618407445708376,
      "grad_norm": 2.370538964682846,
      "kl": 0.060302734375,
      "learning_rate": 2.5801362784684104e-07,
      "loss": 0.0024,
      "num_tokens": 52311279.0,
      "reward": 1.0,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.39009472727775574,
      "step": 640
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 187.0,
      "completions/max_terminated_length": 187.0,
      "completions/mean_length": 136.70834350585938,
      "completions/mean_terminated_length": 136.70834350585938,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.6628748707342296,
      "grad_norm": 4.307337721384269,
      "kl": 0.0693359375,
      "learning_rate": 2.5659342107201857e-07,
      "loss": 0.0028,
      "num_tokens": 52395696.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.4240165948867798,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.642332136631012,
      "step": 641
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 201.0,
      "completions/max_terminated_length": 201.0,
      "completions/mean_length": 126.04167175292969,
      "completions/mean_terminated_length": 126.04167175292969,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.6639089968976215,
      "grad_norm": 3.4267578238060703,
      "kl": 0.047119140625,
      "learning_rate": 2.551757833836996e-07,
      "loss": 0.0019,
      "num_tokens": 52477881.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 642
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 247.0,
      "completions/max_terminated_length": 247.0,
      "completions/mean_length": 150.125,
      "completions/mean_terminated_length": 150.125,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.6649431230610134,
      "grad_norm": 3.0652745827085965,
      "kl": 0.06640625,
      "learning_rate": 2.537607297446428e-07,
      "loss": 0.0027,
      "num_tokens": 52567396.0,
      "reward": 1.4583333730697632,
      "reward_std": 0.1451837718486786,
      "rewards/reasoning_reward/mean": 1.4583333730697632,
      "rewards/reasoning_reward/std": 0.3877657949924469,
      "step": 643
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 276.0,
      "completions/max_terminated_length": 276.0,
      "completions/mean_length": 133.25,
      "completions/mean_terminated_length": 133.25,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.6659772492244054,
      "grad_norm": 4.114200525159867,
      "kl": 0.0546875,
      "learning_rate": 2.5234827509033294e-07,
      "loss": 0.0022,
      "num_tokens": 52643658.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 644
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 262.0,
      "completions/max_terminated_length": 262.0,
      "completions/mean_length": 162.95834350585938,
      "completions/mean_terminated_length": 162.95834350585938,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.6670113753877973,
      "grad_norm": 4.283824327880422,
      "kl": 0.061279296875,
      "learning_rate": 2.509384343288227e-07,
      "loss": 0.0024,
      "num_tokens": 52728177.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.35970625281333923,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.8720186948776245,
      "step": 645
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 158.0,
      "completions/max_terminated_length": 158.0,
      "completions/mean_length": 112.29167175292969,
      "completions/mean_terminated_length": 112.29167175292969,
      "completions/min_length": 76.0,
      "completions/min_terminated_length": 76.0,
      "epoch": 0.6680455015511892,
      "grad_norm": 2.5631822617842257,
      "kl": 0.0517578125,
      "learning_rate": 2.495312223405766e-07,
      "loss": 0.0021,
      "num_tokens": 52805880.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 646
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 288.0,
      "completions/max_terminated_length": 288.0,
      "completions/mean_length": 159.5,
      "completions/mean_terminated_length": 159.5,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.6690796277145812,
      "grad_norm": 3.1846394660646826,
      "kl": 0.0693359375,
      "learning_rate": 2.4812665397831243e-07,
      "loss": 0.0028,
      "num_tokens": 52888628.0,
      "reward": 1.1597222089767456,
      "reward_std": 0.2325192391872406,
      "rewards/reasoning_reward/mean": 1.1597222089767456,
      "rewards/reasoning_reward/std": 0.5118857026100159,
      "step": 647
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 264.0,
      "completions/max_terminated_length": 264.0,
      "completions/mean_length": 157.95834350585938,
      "completions/mean_terminated_length": 157.95834350585938,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.6701137538779731,
      "grad_norm": 2.9464142757577507,
      "kl": 0.0771484375,
      "learning_rate": 2.467247440668462e-07,
      "loss": 0.0031,
      "num_tokens": 52966915.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.28029152750968933,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4584156572818756,
      "step": 648
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 161.625,
      "completions/mean_terminated_length": 161.625,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.671147880041365,
      "grad_norm": 3.492845045813914,
      "kl": 0.06689453125,
      "learning_rate": 2.453255074029336e-07,
      "loss": 0.0027,
      "num_tokens": 53052506.0,
      "reward": 1.0625,
      "reward_std": 0.2041093111038208,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.3061862289905548,
      "step": 649
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 165.58334350585938,
      "completions/mean_terminated_length": 165.58334350585938,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.672182006204757,
      "grad_norm": 3.471526846228643,
      "kl": 0.06494140625,
      "learning_rate": 2.4392895875511613e-07,
      "loss": 0.0026,
      "num_tokens": 53128112.0,
      "reward": 0.25,
      "reward_std": 0.30416232347488403,
      "rewards/reasoning_reward/mean": 0.25,
      "rewards/reasoning_reward/std": 0.41702884435653687,
      "step": 650
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 256.0,
      "completions/max_terminated_length": 256.0,
      "completions/mean_length": 172.58334350585938,
      "completions/mean_terminated_length": 172.58334350585938,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.6732161323681489,
      "grad_norm": 3.4426169831993034,
      "kl": 0.06787109375,
      "learning_rate": 2.425351128635632e-07,
      "loss": 0.0027,
      "num_tokens": 53211558.0,
      "reward": 1.5,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 1.5,
      "rewards/reasoning_reward/std": 0.6593804359436035,
      "step": 651
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 221.0,
      "completions/max_terminated_length": 221.0,
      "completions/mean_length": 151.5,
      "completions/mean_terminated_length": 151.5,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.6742502585315409,
      "grad_norm": 2.6977128922942453,
      "kl": 0.06787109375,
      "learning_rate": 2.411439844399177e-07,
      "loss": 0.0027,
      "num_tokens": 53290618.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.18766528367996216,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.3753018081188202,
      "step": 652
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 210.0,
      "completions/max_terminated_length": 210.0,
      "completions/mean_length": 135.70834350585938,
      "completions/mean_terminated_length": 135.70834350585938,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.6752843846949328,
      "grad_norm": 4.273684646889255,
      "kl": 0.06396484375,
      "learning_rate": 2.3975558816714073e-07,
      "loss": 0.0026,
      "num_tokens": 53366675.0,
      "reward": 1.0,
      "reward_std": 0.28029152750968933,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.39009472727775574,
      "step": 653
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 276.0,
      "completions/max_terminated_length": 276.0,
      "completions/mean_length": 175.58334350585938,
      "completions/mean_terminated_length": 178.56521606445312,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 121.0,
      "epoch": 0.6763185108583247,
      "grad_norm": 3.922489743480712,
      "kl": 0.08544921875,
      "learning_rate": 2.383699386993557e-07,
      "loss": 0.0034,
      "num_tokens": 53451353.0,
      "reward": 1.1111111640930176,
      "reward_std": 0.4623008966445923,
      "rewards/reasoning_reward/mean": 1.111111044883728,
      "rewards/reasoning_reward/std": 0.647676408290863,
      "step": 654
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 177.2916717529297,
      "completions/mean_terminated_length": 177.2916717529297,
      "completions/min_length": 130.0,
      "completions/min_terminated_length": 130.0,
      "epoch": 0.6773526370217167,
      "grad_norm": 3.720452216088832,
      "kl": 0.08984375,
      "learning_rate": 2.3698705066169483e-07,
      "loss": 0.0036,
      "num_tokens": 53539944.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.6266384720802307,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.6538625359535217,
      "step": 655
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 251.0,
      "completions/max_terminated_length": 251.0,
      "completions/mean_length": 160.70834350585938,
      "completions/mean_terminated_length": 160.70834350585938,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.6783867631851086,
      "grad_norm": 13.755429616291899,
      "kl": 0.310546875,
      "learning_rate": 2.356069386501438e-07,
      "loss": 0.0125,
      "num_tokens": 53618241.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.21322892606258392,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.5893837213516235,
      "step": 656
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 193.0,
      "completions/max_terminated_length": 193.0,
      "completions/mean_length": 147.58334350585938,
      "completions/mean_terminated_length": 147.58334350585938,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.6794208893485005,
      "grad_norm": 3.308998342120999,
      "kl": 0.0732421875,
      "learning_rate": 2.342296172313886e-07,
      "loss": 0.0029,
      "num_tokens": 53706335.0,
      "reward": 1.4722222089767456,
      "reward_std": 0.2105759084224701,
      "rewards/reasoning_reward/mean": 1.4722222089767456,
      "rewards/reasoning_reward/std": 0.4441424012184143,
      "step": 657
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 160.6666717529297,
      "completions/mean_terminated_length": 160.6666717529297,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.6804550155118925,
      "grad_norm": 2.0128009083781153,
      "kl": 0.07666015625,
      "learning_rate": 2.3285510094266087e-07,
      "loss": 0.0031,
      "num_tokens": 53784703.0,
      "reward": 1.1875,
      "reward_std": 0.10681166499853134,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.3234494626522064,
      "step": 658
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 144.875,
      "completions/mean_terminated_length": 144.875,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.6814891416752844,
      "grad_norm": 2.9811727952229545,
      "kl": 0.06591796875,
      "learning_rate": 2.3148340429158526e-07,
      "loss": 0.0027,
      "num_tokens": 53870556.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.35864076018333435,
      "step": 659
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 271.0,
      "completions/max_terminated_length": 271.0,
      "completions/mean_length": 154.2916717529297,
      "completions/mean_terminated_length": 154.2916717529297,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.6825232678386763,
      "grad_norm": 3.6867423359764477,
      "kl": 0.040283203125,
      "learning_rate": 2.3011454175602558e-07,
      "loss": 0.0016,
      "num_tokens": 53949691.0,
      "reward": 0.875,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 660
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 221.0,
      "completions/max_terminated_length": 215.0,
      "completions/mean_length": 156.875,
      "completions/mean_terminated_length": 154.0869598388672,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.6835573940020683,
      "grad_norm": 3.4323592527418394,
      "kl": 0.111328125,
      "learning_rate": 2.2874852778393266e-07,
      "loss": 0.0045,
      "num_tokens": 54037808.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.2985045611858368,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 661
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 143.4166717529297,
      "completions/mean_terminated_length": 139.95652770996094,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.6845915201654602,
      "grad_norm": 3.3556977958017895,
      "kl": 0.07958984375,
      "learning_rate": 2.273853767931918e-07,
      "loss": 0.0032,
      "num_tokens": 54116794.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.5500329732894897,
      "step": 662
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 211.0,
      "completions/max_terminated_length": 211.0,
      "completions/mean_length": 160.25,
      "completions/mean_terminated_length": 160.25,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.6856256463288521,
      "grad_norm": 2.337364555066446,
      "kl": 0.049072265625,
      "learning_rate": 2.2602510317146956e-07,
      "loss": 0.002,
      "num_tokens": 54201880.0,
      "reward": 1.125,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.7408866882324219,
      "step": 663
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 154.83334350585938,
      "completions/mean_terminated_length": 154.83334350585938,
      "completions/min_length": 56.0,
      "completions/min_terminated_length": 56.0,
      "epoch": 0.6866597724922441,
      "grad_norm": 3.4959986502490086,
      "kl": 0.09521484375,
      "learning_rate": 2.246677212760636e-07,
      "loss": 0.0038,
      "num_tokens": 54290276.0,
      "reward": 1.4930555820465088,
      "reward_std": 0.35425281524658203,
      "rewards/reasoning_reward/mean": 1.4930554628372192,
      "rewards/reasoning_reward/std": 0.7010673880577087,
      "step": 664
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 261.0,
      "completions/max_terminated_length": 261.0,
      "completions/mean_length": 184.875,
      "completions/mean_terminated_length": 184.875,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.687693898655636,
      "grad_norm": 3.6836605606951633,
      "kl": 0.07763671875,
      "learning_rate": 2.233132454337494e-07,
      "loss": 0.0031,
      "num_tokens": 54368369.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.41387641429901123,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.524749755859375,
      "step": 665
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 141.70834350585938,
      "completions/mean_terminated_length": 141.70834350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.688728024819028,
      "grad_norm": 3.8448609470179624,
      "kl": 0.07275390625,
      "learning_rate": 2.2196168994063075e-07,
      "loss": 0.0029,
      "num_tokens": 54446450.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.23709973692893982,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.40322521328926086,
      "step": 666
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 181.58334350585938,
      "completions/mean_terminated_length": 181.58334350585938,
      "completions/min_length": 134.0,
      "completions/min_terminated_length": 134.0,
      "epoch": 0.6897621509824199,
      "grad_norm": 3.2290937591662097,
      "kl": 0.08544921875,
      "learning_rate": 2.2061306906198707e-07,
      "loss": 0.0034,
      "num_tokens": 54530368.0,
      "reward": 1.3541667461395264,
      "reward_std": 0.27830731868743896,
      "rewards/reasoning_reward/mean": 1.3541666269302368,
      "rewards/reasoning_reward/std": 0.453948050737381,
      "step": 667
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 347.0,
      "completions/max_terminated_length": 347.0,
      "completions/mean_length": 184.33334350585938,
      "completions/mean_terminated_length": 184.33334350585938,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.6907962771458118,
      "grad_norm": 3.9502785850753477,
      "kl": 0.08154296875,
      "learning_rate": 2.1926739703212472e-07,
      "loss": 0.0033,
      "num_tokens": 54613976.0,
      "reward": 0.9375,
      "reward_std": 0.34602540731430054,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.6806725859642029,
      "step": 668
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 333.0,
      "completions/max_terminated_length": 333.0,
      "completions/mean_length": 171.9166717529297,
      "completions/mean_terminated_length": 171.9166717529297,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.6918304033092038,
      "grad_norm": 2.466267818183153,
      "kl": 0.056884765625,
      "learning_rate": 2.1792468805422487e-07,
      "loss": 0.0023,
      "num_tokens": 54695686.0,
      "reward": 1.0625,
      "reward_std": 0.1767766922712326,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.3061862289905548,
      "step": 669
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 201.0,
      "completions/max_terminated_length": 201.0,
      "completions/mean_length": 149.1666717529297,
      "completions/mean_terminated_length": 149.1666717529297,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.6928645294725957,
      "grad_norm": 3.9637637220123643,
      "kl": 0.08447265625,
      "learning_rate": 2.1658495630019518e-07,
      "loss": 0.0034,
      "num_tokens": 54782322.0,
      "reward": 1.3680555820465088,
      "reward_std": 0.36238986253738403,
      "rewards/reasoning_reward/mean": 1.3680554628372192,
      "rewards/reasoning_reward/std": 0.5059528946876526,
      "step": 670
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 313.0,
      "completions/max_terminated_length": 313.0,
      "completions/mean_length": 176.25,
      "completions/mean_terminated_length": 176.25,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.6938986556359876,
      "grad_norm": 3.8385551187680393,
      "kl": 0.0771484375,
      "learning_rate": 2.152482159105196e-07,
      "loss": 0.0031,
      "num_tokens": 54865144.0,
      "reward": 0.930555522441864,
      "reward_std": 0.4847185015678406,
      "rewards/reasoning_reward/mean": 0.930555522441864,
      "rewards/reasoning_reward/std": 0.8411469459533691,
      "step": 671
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 248.0,
      "completions/max_terminated_length": 248.0,
      "completions/mean_length": 173.33334350585938,
      "completions/mean_terminated_length": 173.33334350585938,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.6949327817993796,
      "grad_norm": 3.164139243922661,
      "kl": 0.0791015625,
      "learning_rate": 2.1391448099410853e-07,
      "loss": 0.0032,
      "num_tokens": 54948368.0,
      "reward": 1.1597222089767456,
      "reward_std": 0.33324792981147766,
      "rewards/reasoning_reward/mean": 1.1597222089767456,
      "rewards/reasoning_reward/std": 0.5118856430053711,
      "step": 672
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 147.875,
      "completions/mean_terminated_length": 147.875,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.6959669079627715,
      "grad_norm": 3.4067201589070613,
      "kl": 0.054443359375,
      "learning_rate": 2.1258376562815112e-07,
      "loss": 0.0022,
      "num_tokens": 55029237.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.32025060057640076,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "step": 673
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 162.58334350585938,
      "completions/mean_terminated_length": 162.58334350585938,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.6970010341261634,
      "grad_norm": 2.5831577803082353,
      "kl": 0.07275390625,
      "learning_rate": 2.112560838579653e-07,
      "loss": 0.0029,
      "num_tokens": 55120027.0,
      "reward": 1.5763888359069824,
      "reward_std": 0.1376926600933075,
      "rewards/reasoning_reward/mean": 1.5763888359069824,
      "rewards/reasoning_reward/std": 0.45038676261901855,
      "step": 674
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 339.0,
      "completions/max_terminated_length": 339.0,
      "completions/mean_length": 172.4166717529297,
      "completions/mean_terminated_length": 172.4166717529297,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.6980351602895554,
      "grad_norm": 3.9629234668552953,
      "kl": 0.06640625,
      "learning_rate": 2.0993144969685106e-07,
      "loss": 0.0027,
      "num_tokens": 55195797.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.39000558853149414,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.6863049864768982,
      "step": 675
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 248.0,
      "completions/max_terminated_length": 248.0,
      "completions/mean_length": 166.375,
      "completions/mean_terminated_length": 166.375,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.6990692864529473,
      "grad_norm": 2.8177613643424504,
      "kl": 0.09228515625,
      "learning_rate": 2.0860987712594103e-07,
      "loss": 0.0037,
      "num_tokens": 55284382.0,
      "reward": 1.2708333730697632,
      "reward_std": 0.20249390602111816,
      "rewards/reasoning_reward/mean": 1.2708333730697632,
      "rewards/reasoning_reward/std": 0.8509904742240906,
      "step": 676
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 210.0,
      "completions/max_terminated_length": 210.0,
      "completions/mean_length": 151.70834350585938,
      "completions/mean_terminated_length": 151.70834350585938,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.7001034126163392,
      "grad_norm": 0.17490723954585877,
      "kl": 0.0546875,
      "learning_rate": 2.0729138009405417e-07,
      "loss": 0.0022,
      "num_tokens": 55363807.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 677
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 263.0,
      "completions/max_terminated_length": 263.0,
      "completions/mean_length": 151.70834350585938,
      "completions/mean_terminated_length": 151.70834350585938,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.7011375387797312,
      "grad_norm": 2.1768886150741418,
      "kl": 0.07861328125,
      "learning_rate": 2.05975972517548e-07,
      "loss": 0.0031,
      "num_tokens": 55452096.0,
      "reward": 1.6458333730697632,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 1.6458333730697632,
      "rewards/reasoning_reward/std": 0.47729232907295227,
      "step": 678
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 200.0,
      "completions/max_terminated_length": 200.0,
      "completions/mean_length": 125.66667175292969,
      "completions/mean_terminated_length": 125.66667175292969,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.7021716649431231,
      "grad_norm": 3.6068735599469863,
      "kl": 0.044189453125,
      "learning_rate": 2.0466366828017113e-07,
      "loss": 0.0018,
      "num_tokens": 55538384.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.20693820714950562,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 679
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 156.375,
      "completions/mean_terminated_length": 156.375,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.703205791106515,
      "grad_norm": 3.5043504652823727,
      "kl": 0.08251953125,
      "learning_rate": 2.033544812329181e-07,
      "loss": 0.0033,
      "num_tokens": 55626745.0,
      "reward": 1.3541667461395264,
      "reward_std": 0.32196044921875,
      "rewards/reasoning_reward/mean": 1.3541666269302368,
      "rewards/reasoning_reward/std": 0.5800893306732178,
      "step": 680
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 143.875,
      "completions/mean_terminated_length": 143.875,
      "completions/min_length": 70.0,
      "completions/min_terminated_length": 70.0,
      "epoch": 0.704239917269907,
      "grad_norm": 2.9250034907846185,
      "kl": 0.076171875,
      "learning_rate": 2.020484251938817e-07,
      "loss": 0.0031,
      "num_tokens": 55712742.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.8998792171478271,
      "step": 681
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 139.5416717529297,
      "completions/mean_terminated_length": 139.5416717529297,
      "completions/min_length": 66.0,
      "completions/min_terminated_length": 66.0,
      "epoch": 0.7052740434332989,
      "grad_norm": 2.899108027116337,
      "kl": 0.0595703125,
      "learning_rate": 2.007455139481085e-07,
      "loss": 0.0024,
      "num_tokens": 55792819.0,
      "reward": 0.75,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 682
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 279.0,
      "completions/max_terminated_length": 279.0,
      "completions/mean_length": 157.70834350585938,
      "completions/mean_terminated_length": 157.70834350585938,
      "completions/min_length": 70.0,
      "completions/min_terminated_length": 70.0,
      "epoch": 0.7063081695966908,
      "grad_norm": 0.19921601907261913,
      "kl": 0.055908203125,
      "learning_rate": 1.9944576124745205e-07,
      "loss": 0.0022,
      "num_tokens": 55869540.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.24077169597148895,
      "step": 683
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 331.0,
      "completions/max_terminated_length": 331.0,
      "completions/mean_length": 183.33334350585938,
      "completions/mean_terminated_length": 183.33334350585938,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.7073422957600828,
      "grad_norm": 3.8943822086746436,
      "kl": 0.0791015625,
      "learning_rate": 1.9814918081042887e-07,
      "loss": 0.0032,
      "num_tokens": 55948236.0,
      "reward": 1.0625,
      "reward_std": 0.3584126830101013,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.42509588599205017,
      "step": 684
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 257.0,
      "completions/max_terminated_length": 257.0,
      "completions/mean_length": 161.4166717529297,
      "completions/mean_terminated_length": 161.4166717529297,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.7083764219234746,
      "grad_norm": 2.3459301471771914,
      "kl": 0.06591796875,
      "learning_rate": 1.9685578632207268e-07,
      "loss": 0.0026,
      "num_tokens": 56033366.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.9315329194068909,
      "step": 685
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 234.0,
      "completions/max_terminated_length": 234.0,
      "completions/mean_length": 142.20834350585938,
      "completions/mean_terminated_length": 142.20834350585938,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.7094105480868665,
      "grad_norm": 3.786217018380023,
      "kl": 0.060302734375,
      "learning_rate": 1.9556559143379097e-07,
      "loss": 0.0024,
      "num_tokens": 56110987.0,
      "reward": 0.875,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 686
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 290.0,
      "completions/max_terminated_length": 290.0,
      "completions/mean_length": 174.875,
      "completions/mean_terminated_length": 174.875,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.7104446742502585,
      "grad_norm": 3.277700175814856,
      "kl": 0.0634765625,
      "learning_rate": 1.9427860976321996e-07,
      "loss": 0.0025,
      "num_tokens": 56191280.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 687
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 139.70834350585938,
      "completions/mean_terminated_length": 139.70834350585938,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.7114788004136504,
      "grad_norm": 2.2482590984095934,
      "kl": 0.0830078125,
      "learning_rate": 1.9299485489408125e-07,
      "loss": 0.0033,
      "num_tokens": 56275553.0,
      "reward": 0.9722222089767456,
      "reward_std": 0.0514344647526741,
      "rewards/reasoning_reward/mean": 0.9722221493721008,
      "rewards/reasoning_reward/std": 0.8040757775306702,
      "step": 688
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 278.0,
      "completions/max_terminated_length": 278.0,
      "completions/mean_length": 168.4166717529297,
      "completions/mean_terminated_length": 168.4166717529297,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.7125129265770423,
      "grad_norm": 1.952148524225847,
      "kl": 0.06982421875,
      "learning_rate": 1.9171434037603883e-07,
      "loss": 0.0028,
      "num_tokens": 56354091.0,
      "reward": 1.2013888359069824,
      "reward_std": 0.058925554156303406,
      "rewards/reasoning_reward/mean": 1.2013888359069824,
      "rewards/reasoning_reward/std": 0.3068428933620453,
      "step": 689
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 151.1666717529297,
      "completions/mean_terminated_length": 151.1666717529297,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.7135470527404343,
      "grad_norm": 3.0579537299955377,
      "kl": 0.06591796875,
      "learning_rate": 1.9043707972455537e-07,
      "loss": 0.0026,
      "num_tokens": 56437463.0,
      "reward": 1.0,
      "reward_std": 0.31100785732269287,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.6370220184326172,
      "step": 690
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 319.0,
      "completions/max_terminated_length": 319.0,
      "completions/mean_length": 154.875,
      "completions/mean_terminated_length": 154.875,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.7145811789038262,
      "grad_norm": 2.1711645432342626,
      "kl": 0.12158203125,
      "learning_rate": 1.8916308642075007e-07,
      "loss": 0.0049,
      "num_tokens": 56525892.0,
      "reward": 1.8125,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 1.8125,
      "rewards/reasoning_reward/std": 0.28788962960243225,
      "step": 691
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 266.0,
      "completions/max_terminated_length": 266.0,
      "completions/mean_length": 162.75,
      "completions/mean_terminated_length": 162.75,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.7156153050672182,
      "grad_norm": 2.18124506294311,
      "kl": 0.10888671875,
      "learning_rate": 1.8789237391125644e-07,
      "loss": 0.0044,
      "num_tokens": 56620214.0,
      "reward": 1.375,
      "reward_std": 0.21362332999706268,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.5757792592048645,
      "step": 692
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 141.5416717529297,
      "completions/mean_terminated_length": 141.5416717529297,
      "completions/min_length": 76.0,
      "completions/min_terminated_length": 76.0,
      "epoch": 0.7166494312306101,
      "grad_norm": 2.1892743478043566,
      "kl": 0.06689453125,
      "learning_rate": 1.8662495560807957e-07,
      "loss": 0.0027,
      "num_tokens": 56702419.0,
      "reward": 1.1875,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.28788962960243225,
      "step": 693
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 300.0,
      "completions/max_terminated_length": 300.0,
      "completions/mean_length": 165.2916717529297,
      "completions/mean_terminated_length": 165.2916717529297,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.717683557394002,
      "grad_norm": 4.212537614597707,
      "kl": 0.09326171875,
      "learning_rate": 1.8536084488845583e-07,
      "loss": 0.0037,
      "num_tokens": 56779346.0,
      "reward": 0.5625,
      "reward_std": 0.37478944659233093,
      "rewards/reasoning_reward/mean": 0.5625,
      "rewards/reasoning_reward/std": 0.7270025014877319,
      "step": 694
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 304.0,
      "completions/max_terminated_length": 304.0,
      "completions/mean_length": 177.7916717529297,
      "completions/mean_terminated_length": 177.7916717529297,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.718717683557394,
      "grad_norm": 2.0102770916091472,
      "kl": 0.056396484375,
      "learning_rate": 1.8410005509471028e-07,
      "loss": 0.0023,
      "num_tokens": 56860157.0,
      "reward": 1.0625,
      "reward_std": 0.12400396913290024,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.2242136001586914,
      "step": 695
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 228.0,
      "completions/max_terminated_length": 228.0,
      "completions/mean_length": 130.375,
      "completions/mean_terminated_length": 130.375,
      "completions/min_length": 69.0,
      "completions/min_terminated_length": 69.0,
      "epoch": 0.7197518097207859,
      "grad_norm": 2.329478391630926,
      "kl": 0.0439453125,
      "learning_rate": 1.828425995341173e-07,
      "loss": 0.0018,
      "num_tokens": 56939902.0,
      "reward": 0.375,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 696
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 151.875,
      "completions/mean_terminated_length": 151.875,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.7207859358841778,
      "grad_norm": 3.1877025746249577,
      "kl": 0.055908203125,
      "learning_rate": 1.815884914787587e-07,
      "loss": 0.0022,
      "num_tokens": 57024539.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.24056154489517212,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.3753018081188202,
      "step": 697
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 263.0,
      "completions/max_terminated_length": 263.0,
      "completions/mean_length": 159.7916717529297,
      "completions/mean_terminated_length": 159.0,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.7218200620475698,
      "grad_norm": 3.3274373464580114,
      "kl": 0.055908203125,
      "learning_rate": 1.80337744165385e-07,
      "loss": 0.0022,
      "num_tokens": 57108438.0,
      "reward": 0.6041666865348816,
      "reward_std": 0.3803371787071228,
      "rewards/reasoning_reward/mean": 0.6041666865348816,
      "rewards/reasoning_reward/std": 0.642332136631012,
      "step": 698
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 212.0,
      "completions/max_terminated_length": 212.0,
      "completions/mean_length": 144.2916717529297,
      "completions/mean_terminated_length": 144.2916717529297,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.7228541882109617,
      "grad_norm": 0.20006012207095064,
      "kl": 0.07568359375,
      "learning_rate": 1.7909037079527433e-07,
      "loss": 0.003,
      "num_tokens": 57192853.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.9630867838859558,
      "step": 699
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 146.83334350585938,
      "completions/mean_terminated_length": 146.83334350585938,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.7238883143743536,
      "grad_norm": 3.3580145045189655,
      "kl": 0.07666015625,
      "learning_rate": 1.7784638453409451e-07,
      "loss": 0.0031,
      "num_tokens": 57271785.0,
      "reward": 1.0555555820465088,
      "reward_std": 0.3299504518508911,
      "rewards/reasoning_reward/mean": 1.0555555820465088,
      "rewards/reasoning_reward/std": 0.4189550578594208,
      "step": 700
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 224.0,
      "completions/max_terminated_length": 224.0,
      "completions/mean_length": 140.95834350585938,
      "completions/mean_terminated_length": 140.95834350585938,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.7249224405377456,
      "grad_norm": 0.22405126131157563,
      "kl": 0.059814453125,
      "learning_rate": 1.7660579851176317e-07,
      "loss": 0.0024,
      "num_tokens": 57347536.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 701
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 315.0,
      "completions/max_terminated_length": 245.0,
      "completions/mean_length": 180.5,
      "completions/mean_terminated_length": 174.6521759033203,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.7259565667011375,
      "grad_norm": 2.9389018302570014,
      "kl": 0.055419921875,
      "learning_rate": 1.7536862582230893e-07,
      "loss": 0.0022,
      "num_tokens": 57430948.0,
      "reward": 1.5833333730697632,
      "reward_std": 0.27392348647117615,
      "rewards/reasoning_reward/mean": 1.5833333730697632,
      "rewards/reasoning_reward/std": 0.4584156572818756,
      "step": 702
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 263.0,
      "completions/max_terminated_length": 263.0,
      "completions/mean_length": 178.0416717529297,
      "completions/mean_terminated_length": 174.60870361328125,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.7269906928645294,
      "grad_norm": 2.282134417830866,
      "kl": 0.0576171875,
      "learning_rate": 1.7413487952373455e-07,
      "loss": 0.0023,
      "num_tokens": 57508453.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.40824830532073975,
      "step": 703
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 305.0,
      "completions/max_terminated_length": 305.0,
      "completions/mean_length": 184.58334350585938,
      "completions/mean_terminated_length": 181.69564819335938,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.7280248190279214,
      "grad_norm": 4.1836131721858765,
      "kl": 0.08935546875,
      "learning_rate": 1.7290457263787728e-07,
      "loss": 0.0036,
      "num_tokens": 57593995.0,
      "reward": 1.3125,
      "reward_std": 0.3492930829524994,
      "rewards/reasoning_reward/mean": 1.3125,
      "rewards/reasoning_reward/std": 0.7775728702545166,
      "step": 704
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 166.70834350585938,
      "completions/mean_terminated_length": 166.70834350585938,
      "completions/min_length": 122.0,
      "completions/min_terminated_length": 122.0,
      "epoch": 0.7290589451913133,
      "grad_norm": 3.165264879865031,
      "kl": 0.058837890625,
      "learning_rate": 1.7167771815027317e-07,
      "loss": 0.0024,
      "num_tokens": 57672700.0,
      "reward": 0.6041666865348816,
      "reward_std": 0.30217814445495605,
      "rewards/reasoning_reward/mean": 0.6041666865348816,
      "rewards/reasoning_reward/std": 0.6590369343757629,
      "step": 705
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 225.0,
      "completions/max_terminated_length": 225.0,
      "completions/mean_length": 156.625,
      "completions/mean_terminated_length": 156.625,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.7300930713547052,
      "grad_norm": 2.672806554030225,
      "kl": 0.054931640625,
      "learning_rate": 1.7045432901001844e-07,
      "loss": 0.0022,
      "num_tokens": 57751843.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 706
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 197.0,
      "completions/max_terminated_length": 197.0,
      "completions/mean_length": 136.70834350585938,
      "completions/mean_terminated_length": 136.70834350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.7311271975180972,
      "grad_norm": 3.035864140075777,
      "kl": 0.07080078125,
      "learning_rate": 1.6923441812963434e-07,
      "loss": 0.0028,
      "num_tokens": 57839420.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 707
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 162.95834350585938,
      "completions/mean_terminated_length": 162.95834350585938,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.7321613236814891,
      "grad_norm": 2.6352705999114425,
      "kl": 0.07861328125,
      "learning_rate": 1.6801799838492942e-07,
      "loss": 0.0032,
      "num_tokens": 57917803.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 708
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 341.0,
      "completions/max_terminated_length": 341.0,
      "completions/mean_length": 177.25,
      "completions/mean_terminated_length": 177.25,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.733195449844881,
      "grad_norm": 2.839232223213205,
      "kl": 0.07666015625,
      "learning_rate": 1.6680508261486465e-07,
      "loss": 0.0031,
      "num_tokens": 58001793.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.33034375309944153,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.7649476528167725,
      "step": 709
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 268.0,
      "completions/max_terminated_length": 268.0,
      "completions/mean_length": 166.20834350585938,
      "completions/mean_terminated_length": 166.20834350585938,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.734229576008273,
      "grad_norm": 3.7820274298450687,
      "kl": 0.07177734375,
      "learning_rate": 1.6559568362141769e-07,
      "loss": 0.0029,
      "num_tokens": 58086726.0,
      "reward": 0.8125,
      "reward_std": 0.4476938843727112,
      "rewards/reasoning_reward/mean": 0.8125,
      "rewards/reasoning_reward/std": 0.6562823057174683,
      "step": 710
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 284.0,
      "completions/max_terminated_length": 284.0,
      "completions/mean_length": 154.5416717529297,
      "completions/mean_terminated_length": 154.5416717529297,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.7352637021716649,
      "grad_norm": 3.1124009121892198,
      "kl": 0.09228515625,
      "learning_rate": 1.6438981416944708e-07,
      "loss": 0.0037,
      "num_tokens": 58173611.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.8164966106414795,
      "step": 711
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 276.0,
      "completions/max_terminated_length": 276.0,
      "completions/mean_length": 196.0416717529297,
      "completions/mean_terminated_length": 196.0416717529297,
      "completions/min_length": 128.0,
      "completions/min_terminated_length": 128.0,
      "epoch": 0.7362978283350569,
      "grad_norm": 3.1443322611380684,
      "kl": 0.08544921875,
      "learning_rate": 1.631874869865587e-07,
      "loss": 0.0034,
      "num_tokens": 58251924.0,
      "reward": 1.3958333730697632,
      "reward_std": 0.33768826723098755,
      "rewards/reasoning_reward/mean": 1.3958333730697632,
      "rewards/reasoning_reward/std": 0.48854637145996094,
      "step": 712
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 200.0,
      "completions/max_terminated_length": 200.0,
      "completions/mean_length": 124.125,
      "completions/mean_terminated_length": 124.125,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.7373319544984488,
      "grad_norm": 2.3752422112829867,
      "kl": 0.054443359375,
      "learning_rate": 1.6198871476297033e-07,
      "loss": 0.0022,
      "num_tokens": 58332575.0,
      "reward": 0.8958333730697632,
      "reward_std": 0.19795583188533783,
      "rewards/reasoning_reward/mean": 0.8958333134651184,
      "rewards/reasoning_reward/std": 0.3605300188064575,
      "step": 713
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 288.0,
      "completions/max_terminated_length": 288.0,
      "completions/mean_length": 150.4166717529297,
      "completions/mean_terminated_length": 150.4166717529297,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.7383660806618407,
      "grad_norm": 2.0212470615420246,
      "kl": 0.051025390625,
      "learning_rate": 1.607935101513785e-07,
      "loss": 0.002,
      "num_tokens": 58418201.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 714
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 242.0,
      "completions/max_terminated_length": 242.0,
      "completions/mean_length": 171.58334350585938,
      "completions/mean_terminated_length": 171.58334350585938,
      "completions/min_length": 122.0,
      "completions/min_terminated_length": 122.0,
      "epoch": 0.7394002068252327,
      "grad_norm": 3.251685083157416,
      "kl": 0.061767578125,
      "learning_rate": 1.596018857668242e-07,
      "loss": 0.0025,
      "num_tokens": 58495583.0,
      "reward": 0.6875,
      "reward_std": 0.3709374666213989,
      "rewards/reasoning_reward/mean": 0.6875,
      "rewards/reasoning_reward/std": 0.7194880843162537,
      "step": 715
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 413.0,
      "completions/max_terminated_length": 413.0,
      "completions/mean_length": 212.08334350585938,
      "completions/mean_terminated_length": 212.08334350585938,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.7404343329886246,
      "grad_norm": 3.242133442856897,
      "kl": 0.07177734375,
      "learning_rate": 1.5841385418656068e-07,
      "loss": 0.0029,
      "num_tokens": 58581425.0,
      "reward": 1.4583333730697632,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 1.4583333730697632,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 716
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 311.0,
      "completions/max_terminated_length": 311.0,
      "completions/mean_length": 168.33334350585938,
      "completions/mean_terminated_length": 168.33334350585938,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.7414684591520165,
      "grad_norm": 3.0656410513536554,
      "kl": 0.07373046875,
      "learning_rate": 1.5722942794991995e-07,
      "loss": 0.003,
      "num_tokens": 58665161.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.39486488699913025,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.6251811385154724,
      "step": 717
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 308.0,
      "completions/max_terminated_length": 308.0,
      "completions/mean_length": 144.2916717529297,
      "completions/mean_terminated_length": 144.2916717529297,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.7425025853154085,
      "grad_norm": 3.1287480808791184,
      "kl": 0.07177734375,
      "learning_rate": 1.5604861955818038e-07,
      "loss": 0.0029,
      "num_tokens": 58752048.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.7172814607620239,
      "step": 718
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 321.0,
      "completions/max_terminated_length": 321.0,
      "completions/mean_length": 161.9166717529297,
      "completions/mean_terminated_length": 161.9166717529297,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.7435367114788004,
      "grad_norm": 4.0449765747453545,
      "kl": 0.08740234375,
      "learning_rate": 1.548714414744356e-07,
      "loss": 0.0035,
      "num_tokens": 58834862.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.39814266562461853,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.7359800934791565,
      "step": 719
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 258.0,
      "completions/max_terminated_length": 258.0,
      "completions/mean_length": 172.08334350585938,
      "completions/mean_terminated_length": 172.08334350585938,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.7445708376421923,
      "grad_norm": 3.011337708200572,
      "kl": 0.06298828125,
      "learning_rate": 1.5369790612346168e-07,
      "loss": 0.0025,
      "num_tokens": 58914176.0,
      "reward": 1.0694444179534912,
      "reward_std": 0.2472916543483734,
      "rewards/reasoning_reward/mean": 1.0694444179534912,
      "rewards/reasoning_reward/std": 0.3506951928138733,
      "step": 720
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 170.5,
      "completions/mean_terminated_length": 170.5,
      "completions/min_length": 121.0,
      "completions/min_terminated_length": 121.0,
      "epoch": 0.7456049638055843,
      "grad_norm": 4.6480620380414965,
      "kl": 0.0625,
      "learning_rate": 1.5252802589158737e-07,
      "loss": 0.0025,
      "num_tokens": 58996844.0,
      "reward": 1.0694445371627808,
      "reward_std": 0.30431777238845825,
      "rewards/reasoning_reward/mean": 1.0694445371627808,
      "rewards/reasoning_reward/std": 0.7659994959831238,
      "step": 721
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 148.83334350585938,
      "completions/mean_terminated_length": 148.83334350585938,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.7466390899689762,
      "grad_norm": 3.1996652428167045,
      "kl": 0.056884765625,
      "learning_rate": 1.513618131265621e-07,
      "loss": 0.0023,
      "num_tokens": 59076408.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.24056154489517212,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.6672325730323792,
      "step": 722
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 231.0,
      "completions/max_terminated_length": 231.0,
      "completions/mean_length": 172.625,
      "completions/mean_terminated_length": 172.625,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.7476732161323681,
      "grad_norm": 3.280704854907078,
      "kl": 0.06982421875,
      "learning_rate": 1.5019928013742682e-07,
      "loss": 0.0028,
      "num_tokens": 59163431.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.9315329790115356,
      "step": 723
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 200.0,
      "completions/max_terminated_length": 200.0,
      "completions/mean_length": 134.70834350585938,
      "completions/mean_terminated_length": 134.70834350585938,
      "completions/min_length": 73.0,
      "completions/min_terminated_length": 73.0,
      "epoch": 0.7487073422957601,
      "grad_norm": 2.185541011879563,
      "kl": 0.06640625,
      "learning_rate": 1.490404391943829e-07,
      "loss": 0.0026,
      "num_tokens": 59246096.0,
      "reward": 1.25,
      "reward_std": 0.07273930311203003,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 724
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 304.0,
      "completions/max_terminated_length": 304.0,
      "completions/mean_length": 165.0416717529297,
      "completions/mean_terminated_length": 165.0416717529297,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.749741468459152,
      "grad_norm": 2.51252824919629,
      "kl": 0.09521484375,
      "learning_rate": 1.4788530252866372e-07,
      "loss": 0.0038,
      "num_tokens": 59325185.0,
      "reward": 0.625,
      "reward_std": 0.21362332999706268,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.5757792592048645,
      "step": 725
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 347.0,
      "completions/max_terminated_length": 347.0,
      "completions/mean_length": 162.6666717529297,
      "completions/mean_terminated_length": 162.6666717529297,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.750775594622544,
      "grad_norm": 2.8053048835881818,
      "kl": 0.060546875,
      "learning_rate": 1.4673388233240502e-07,
      "loss": 0.0024,
      "num_tokens": 59416513.0,
      "reward": 1.4375,
      "reward_std": 0.12400396913290024,
      "rewards/reasoning_reward/mean": 1.4375,
      "rewards/reasoning_reward/std": 0.47348156571388245,
      "step": 726
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 351.0,
      "completions/max_terminated_length": 351.0,
      "completions/mean_length": 181.45834350585938,
      "completions/mean_terminated_length": 178.3913116455078,
      "completions/min_length": 121.0,
      "completions/min_terminated_length": 121.0,
      "epoch": 0.7518097207859359,
      "grad_norm": 4.163295102377276,
      "kl": 0.060546875,
      "learning_rate": 1.455861907585158e-07,
      "loss": 0.0024,
      "num_tokens": 59494236.0,
      "reward": 0.9375,
      "reward_std": 0.5304333567619324,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.613480806350708,
      "step": 727
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 169.83334350585938,
      "completions/mean_terminated_length": 169.83334350585938,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.7528438469493278,
      "grad_norm": 3.420329539869673,
      "kl": 0.08251953125,
      "learning_rate": 1.4444223992055116e-07,
      "loss": 0.0033,
      "num_tokens": 59578752.0,
      "reward": 1.0,
      "reward_std": 0.36585909128189087,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.4662523865699768,
      "step": 728
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 212.0,
      "completions/max_terminated_length": 212.0,
      "completions/mean_length": 147.9166717529297,
      "completions/mean_terminated_length": 147.9166717529297,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.7538779731127198,
      "grad_norm": 4.210005348002778,
      "kl": 0.06298828125,
      "learning_rate": 1.4330204189258327e-07,
      "loss": 0.0025,
      "num_tokens": 59655494.0,
      "reward": 0.6180555820465088,
      "reward_std": 0.47668325901031494,
      "rewards/reasoning_reward/mean": 0.618055522441864,
      "rewards/reasoning_reward/std": 0.5948230028152466,
      "step": 729
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 151.7916717529297,
      "completions/mean_terminated_length": 151.7916717529297,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.7549120992761117,
      "grad_norm": 3.359678282929136,
      "kl": 0.0771484375,
      "learning_rate": 1.4216560870907496e-07,
      "loss": 0.0031,
      "num_tokens": 59733049.0,
      "reward": 1.1111111640930176,
      "reward_std": 0.3074157238006592,
      "rewards/reasoning_reward/mean": 1.1111111640930176,
      "rewards/reasoning_reward/std": 0.3796345293521881,
      "step": 730
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 286.0,
      "completions/max_terminated_length": 286.0,
      "completions/mean_length": 169.6666717529297,
      "completions/mean_terminated_length": 169.43478393554688,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.7559462254395036,
      "grad_norm": 4.109241986433982,
      "kl": 0.072265625,
      "learning_rate": 1.4103295236475166e-07,
      "loss": 0.0029,
      "num_tokens": 59811105.0,
      "reward": 1.125,
      "reward_std": 0.42645785212516785,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.6123724579811096,
      "step": 731
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 251.0,
      "completions/max_terminated_length": 251.0,
      "completions/mean_length": 164.0416717529297,
      "completions/mean_terminated_length": 164.0416717529297,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.7569803516028956,
      "grad_norm": 3.609914878076435,
      "kl": 0.07958984375,
      "learning_rate": 1.3990408481447596e-07,
      "loss": 0.0032,
      "num_tokens": 59895338.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.4563485085964203,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.832970917224884,
      "step": 732
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 350.0,
      "completions/max_terminated_length": 350.0,
      "completions/mean_length": 170.9166717529297,
      "completions/mean_terminated_length": 170.9166717529297,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.7580144777662875,
      "grad_norm": 3.4849706899707638,
      "kl": 0.051513671875,
      "learning_rate": 1.387790179731202e-07,
      "loss": 0.0021,
      "num_tokens": 59974200.0,
      "reward": 1.0625,
      "reward_std": 0.29339051246643066,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.4499396085739136,
      "step": 733
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 243.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 151.625,
      "completions/mean_terminated_length": 151.625,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.7590486039296794,
      "grad_norm": 4.775586282531071,
      "kl": 0.06298828125,
      "learning_rate": 1.3765776371544173e-07,
      "loss": 0.0025,
      "num_tokens": 60050775.0,
      "reward": 0.8472222685813904,
      "reward_std": 0.3458244502544403,
      "rewards/reasoning_reward/mean": 0.8472222685813904,
      "rewards/reasoning_reward/std": 0.38359054923057556,
      "step": 734
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 266.0,
      "completions/max_terminated_length": 266.0,
      "completions/mean_length": 176.20834350585938,
      "completions/mean_terminated_length": 176.20834350585938,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.7600827300930714,
      "grad_norm": 3.511611967896255,
      "kl": 0.083984375,
      "learning_rate": 1.3654033387595732e-07,
      "loss": 0.0034,
      "num_tokens": 60134140.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.30078065395355225,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.4148510992527008,
      "step": 735
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 125.5,
      "completions/mean_terminated_length": 125.5,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.7611168562564633,
      "grad_norm": 4.1099857019944395,
      "kl": 0.10205078125,
      "learning_rate": 1.3542674024881746e-07,
      "loss": 0.0041,
      "num_tokens": 60211512.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.3877657949924469,
      "step": 736
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 172.0,
      "completions/max_terminated_length": 172.0,
      "completions/mean_length": 123.95833587646484,
      "completions/mean_terminated_length": 123.95833587646484,
      "completions/min_length": 66.0,
      "completions/min_terminated_length": 66.0,
      "epoch": 0.7621509824198552,
      "grad_norm": 2.5865983458888597,
      "kl": 0.056396484375,
      "learning_rate": 1.3431699458768332e-07,
      "loss": 0.0023,
      "num_tokens": 60289567.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.20412415266036987,
      "step": 737
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 148.33334350585938,
      "completions/mean_terminated_length": 148.33334350585938,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.7631851085832472,
      "grad_norm": 4.027481526615197,
      "kl": 0.07470703125,
      "learning_rate": 1.332111086056011e-07,
      "loss": 0.003,
      "num_tokens": 60370247.0,
      "reward": 1.3125,
      "reward_std": 0.39615845680236816,
      "rewards/reasoning_reward/mean": 1.3125,
      "rewards/reasoning_reward/std": 0.6045569777488708,
      "step": 738
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 175.0,
      "completions/max_terminated_length": 175.0,
      "completions/mean_length": 134.70834350585938,
      "completions/mean_terminated_length": 134.70834350585938,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.7642192347466391,
      "grad_norm": 3.3303426892145067,
      "kl": 0.0419921875,
      "learning_rate": 1.3210909397487995e-07,
      "loss": 0.0017,
      "num_tokens": 60450768.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 739
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 212.0,
      "completions/max_terminated_length": 212.0,
      "completions/mean_length": 137.2916717529297,
      "completions/mean_terminated_length": 137.2916717529297,
      "completions/min_length": 73.0,
      "completions/min_terminated_length": 73.0,
      "epoch": 0.765253360910031,
      "grad_norm": 4.295523478132368,
      "kl": 0.087890625,
      "learning_rate": 1.3101096232696735e-07,
      "loss": 0.0035,
      "num_tokens": 60540887.0,
      "reward": 0.993055522441864,
      "reward_std": 0.25392836332321167,
      "rewards/reasoning_reward/mean": 0.993055522441864,
      "rewards/reasoning_reward/std": 0.6245368719100952,
      "step": 740
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 159.7916717529297,
      "completions/mean_terminated_length": 159.7916717529297,
      "completions/min_length": 114.0,
      "completions/min_terminated_length": 114.0,
      "epoch": 0.766287487073423,
      "grad_norm": 3.7021465221363217,
      "kl": 0.08740234375,
      "learning_rate": 1.2991672525232756e-07,
      "loss": 0.0035,
      "num_tokens": 60629106.0,
      "reward": 1.5416667461395264,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.5416666269302368,
      "rewards/reasoning_reward/std": 0.6580053567886353,
      "step": 741
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 311.0,
      "completions/max_terminated_length": 311.0,
      "completions/mean_length": 172.6666717529297,
      "completions/mean_terminated_length": 172.6666717529297,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.7673216132368149,
      "grad_norm": 3.9994888009560037,
      "kl": 0.07763671875,
      "learning_rate": 1.2882639430031833e-07,
      "loss": 0.0031,
      "num_tokens": 60718066.0,
      "reward": 1.5625,
      "reward_std": 0.25392839312553406,
      "rewards/reasoning_reward/mean": 1.5625,
      "rewards/reasoning_reward/std": 0.5578004121780396,
      "step": 742
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 260.0,
      "completions/max_terminated_length": 260.0,
      "completions/mean_length": 165.375,
      "completions/mean_terminated_length": 165.375,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.7683557394002068,
      "grad_norm": 0.3657715562784098,
      "kl": 0.0693359375,
      "learning_rate": 1.2773998097906962e-07,
      "loss": 0.0028,
      "num_tokens": 60800979.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.8340576887130737,
      "step": 743
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 159.125,
      "completions/mean_terminated_length": 159.125,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.7693898655635988,
      "grad_norm": 2.1996165323485397,
      "kl": 0.052978515625,
      "learning_rate": 1.2665749675536209e-07,
      "loss": 0.0021,
      "num_tokens": 60879894.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 744
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 266.0,
      "completions/max_terminated_length": 266.0,
      "completions/mean_length": 159.125,
      "completions/mean_terminated_length": 159.125,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.7704239917269907,
      "grad_norm": 3.291864531573179,
      "kl": 0.056640625,
      "learning_rate": 1.2557895305450533e-07,
      "loss": 0.0023,
      "num_tokens": 60962449.0,
      "reward": 1.125,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.6123724579811096,
      "step": 745
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 197.0,
      "completions/max_terminated_length": 197.0,
      "completions/mean_length": 159.875,
      "completions/mean_terminated_length": 159.875,
      "completions/min_length": 121.0,
      "completions/min_terminated_length": 121.0,
      "epoch": 0.7714581178903827,
      "grad_norm": 4.09716328217157,
      "kl": 0.06396484375,
      "learning_rate": 1.2450436126021863e-07,
      "loss": 0.0026,
      "num_tokens": 61045190.0,
      "reward": 1.2569444179534912,
      "reward_std": 0.40088510513305664,
      "rewards/reasoning_reward/mean": 1.2569444179534912,
      "rewards/reasoning_reward/std": 0.6738367080688477,
      "step": 746
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 370.0,
      "completions/max_terminated_length": 370.0,
      "completions/mean_length": 156.33334350585938,
      "completions/mean_terminated_length": 156.33334350585938,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.7724922440537746,
      "grad_norm": 3.237616260351,
      "kl": 0.05859375,
      "learning_rate": 1.234337327145092e-07,
      "loss": 0.0023,
      "num_tokens": 61130822.0,
      "reward": 1.2777777910232544,
      "reward_std": 0.15713486075401306,
      "rewards/reasoning_reward/mean": 1.2777777910232544,
      "rewards/reasoning_reward/std": 0.5353825092315674,
      "step": 747
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 182.0,
      "completions/max_terminated_length": 182.0,
      "completions/mean_length": 128.625,
      "completions/mean_terminated_length": 128.625,
      "completions/min_length": 61.0,
      "completions/min_terminated_length": 61.0,
      "epoch": 0.7735263702171665,
      "grad_norm": 2.731762557253584,
      "kl": 0.05859375,
      "learning_rate": 1.2236707871755403e-07,
      "loss": 0.0023,
      "num_tokens": 61209893.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.16517187654972076,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.2750164568424225,
      "step": 748
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 257.0,
      "completions/max_terminated_length": 257.0,
      "completions/mean_length": 183.75,
      "completions/mean_terminated_length": 183.75,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.7745604963805585,
      "grad_norm": 3.5907800559210017,
      "kl": 0.07275390625,
      "learning_rate": 1.2130441052757939e-07,
      "loss": 0.0029,
      "num_tokens": 61289695.0,
      "reward": 0.75,
      "reward_std": 0.3247893452644348,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.48900964856147766,
      "step": 749
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 149.375,
      "completions/mean_terminated_length": 149.375,
      "completions/min_length": 71.0,
      "completions/min_terminated_length": 71.0,
      "epoch": 0.7755946225439504,
      "grad_norm": 2.491098383777402,
      "kl": 0.095703125,
      "learning_rate": 1.2024573936074274e-07,
      "loss": 0.0038,
      "num_tokens": 61373152.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.16517187654972076,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.7144344449043274,
      "step": 750
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 307.0,
      "completions/max_terminated_length": 307.0,
      "completions/mean_length": 188.75,
      "completions/mean_terminated_length": 188.75,
      "completions/min_length": 138.0,
      "completions/min_terminated_length": 138.0,
      "epoch": 0.7766287487073423,
      "grad_norm": 2.7802792084915904,
      "kl": 0.062255859375,
      "learning_rate": 1.1919107639101423e-07,
      "loss": 0.0025,
      "num_tokens": 61452002.0,
      "reward": 0.75,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.5897678136825562,
      "step": 751
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 191.0,
      "completions/max_terminated_length": 191.0,
      "completions/mean_length": 150.2916717529297,
      "completions/mean_terminated_length": 150.2916717529297,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.7776628748707343,
      "grad_norm": 0.23470415565994152,
      "kl": 0.07177734375,
      "learning_rate": 1.181404327500582e-07,
      "loss": 0.0029,
      "num_tokens": 61528641.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 752
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 122.625,
      "completions/mean_terminated_length": 122.625,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.7786970010341262,
      "grad_norm": 3.327920369451495,
      "kl": 0.07470703125,
      "learning_rate": 1.1709381952711667e-07,
      "loss": 0.003,
      "num_tokens": 61609248.0,
      "reward": 1.0,
      "reward_std": 0.19500279426574707,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.2553769648075104,
      "step": 753
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 277.0,
      "completions/max_terminated_length": 277.0,
      "completions/mean_length": 144.08334350585938,
      "completions/mean_terminated_length": 144.08334350585938,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.7797311271975181,
      "grad_norm": 3.335800938480737,
      "kl": 0.057373046875,
      "learning_rate": 1.1605124776889125e-07,
      "loss": 0.0023,
      "num_tokens": 61685530.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.637022078037262,
      "step": 754
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 248.0,
      "completions/max_terminated_length": 248.0,
      "completions/mean_length": 175.08334350585938,
      "completions/mean_terminated_length": 175.08334350585938,
      "completions/min_length": 114.0,
      "completions/min_terminated_length": 114.0,
      "epoch": 0.7807652533609101,
      "grad_norm": 3.294221512666946,
      "kl": 0.06884765625,
      "learning_rate": 1.150127284794275e-07,
      "loss": 0.0027,
      "num_tokens": 61768532.0,
      "reward": 1.3888888359069824,
      "reward_std": 0.11878277361392975,
      "rewards/reasoning_reward/mean": 1.3888888359069824,
      "rewards/reasoning_reward/std": 0.3134145140647888,
      "step": 755
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 284.0,
      "completions/max_terminated_length": 284.0,
      "completions/mean_length": 173.125,
      "completions/mean_terminated_length": 173.125,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.781799379524302,
      "grad_norm": 3.396721463922415,
      "kl": 0.06298828125,
      "learning_rate": 1.1397827261999793e-07,
      "loss": 0.0025,
      "num_tokens": 61853439.0,
      "reward": 1.1875,
      "reward_std": 0.4671441912651062,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.6395055651664734,
      "step": 756
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 258.0,
      "completions/max_terminated_length": 258.0,
      "completions/mean_length": 147.5,
      "completions/mean_terminated_length": 147.5,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.7828335056876939,
      "grad_norm": 4.325100386537247,
      "kl": 0.08447265625,
      "learning_rate": 1.1294789110898711e-07,
      "loss": 0.0034,
      "num_tokens": 61932395.0,
      "reward": 0.75,
      "reward_std": 0.5782498121261597,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.6255432367324829,
      "step": 757
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 272.0,
      "completions/max_terminated_length": 272.0,
      "completions/mean_length": 171.7916717529297,
      "completions/mean_terminated_length": 171.7916717529297,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.7838676318510859,
      "grad_norm": 4.370174208909052,
      "kl": 0.07275390625,
      "learning_rate": 1.119215948217756e-07,
      "loss": 0.0029,
      "num_tokens": 62009526.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.3493061661720276,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 758
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 266.0,
      "completions/max_terminated_length": 266.0,
      "completions/mean_length": 173.08334350585938,
      "completions/mean_terminated_length": 173.08334350585938,
      "completions/min_length": 121.0,
      "completions/min_terminated_length": 121.0,
      "epoch": 0.7849017580144778,
      "grad_norm": 2.9640064799173294,
      "kl": 0.07568359375,
      "learning_rate": 1.1089939459062602e-07,
      "loss": 0.003,
      "num_tokens": 62092712.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.6902530789375305,
      "step": 759
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 338.0,
      "completions/max_terminated_length": 338.0,
      "completions/mean_length": 181.25,
      "completions/mean_terminated_length": 181.25,
      "completions/min_length": 126.0,
      "completions/min_terminated_length": 126.0,
      "epoch": 0.7859358841778697,
      "grad_norm": 3.2949381510622544,
      "kl": 0.08251953125,
      "learning_rate": 1.0988130120456813e-07,
      "loss": 0.0033,
      "num_tokens": 62169958.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.7790277004241943,
      "step": 760
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 230.0,
      "completions/max_terminated_length": 230.0,
      "completions/mean_length": 154.75,
      "completions/mean_terminated_length": 154.75,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.7869700103412617,
      "grad_norm": 2.3992415178790885,
      "kl": 0.07373046875,
      "learning_rate": 1.088673254092849e-07,
      "loss": 0.003,
      "num_tokens": 62252120.0,
      "reward": 1.25,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.6079187393188477,
      "step": 761
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 274.0,
      "completions/max_terminated_length": 274.0,
      "completions/mean_length": 177.875,
      "completions/mean_terminated_length": 177.875,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.7880041365046536,
      "grad_norm": 3.3520352974736984,
      "kl": 0.042236328125,
      "learning_rate": 1.0785747790699978e-07,
      "loss": 0.0017,
      "num_tokens": 62333205.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 762
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 312.0,
      "completions/max_terminated_length": 312.0,
      "completions/mean_length": 161.83334350585938,
      "completions/mean_terminated_length": 161.83334350585938,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.7890382626680456,
      "grad_norm": 3.9881962321706217,
      "kl": 0.08740234375,
      "learning_rate": 1.0685176935636265e-07,
      "loss": 0.0035,
      "num_tokens": 62416633.0,
      "reward": 1.375,
      "reward_std": 0.31285393238067627,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.5160468220710754,
      "step": 763
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 185.0,
      "completions/max_terminated_length": 185.0,
      "completions/mean_length": 125.29167175292969,
      "completions/mean_terminated_length": 125.29167175292969,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.7900723888314375,
      "grad_norm": 4.570037175008805,
      "kl": 0.07177734375,
      "learning_rate": 1.0585021037233871e-07,
      "loss": 0.0029,
      "num_tokens": 62492632.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.43459486961364746,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.637022078037262,
      "step": 764
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 261.0,
      "completions/max_terminated_length": 261.0,
      "completions/mean_length": 170.83334350585938,
      "completions/mean_terminated_length": 170.83334350585938,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.7911065149948294,
      "grad_norm": 4.0290312319505395,
      "kl": 0.07763671875,
      "learning_rate": 1.0485281152609482e-07,
      "loss": 0.0031,
      "num_tokens": 62575660.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.36444199085235596,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.40824830532073975,
      "step": 765
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 157.4166717529297,
      "completions/mean_terminated_length": 157.4166717529297,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.7921406411582212,
      "grad_norm": 1.9636414153768984,
      "kl": 0.07568359375,
      "learning_rate": 1.0385958334488965e-07,
      "loss": 0.003,
      "num_tokens": 62654838.0,
      "reward": 0.375,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 766
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 156.4166717529297,
      "completions/mean_terminated_length": 156.4166717529297,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.7931747673216132,
      "grad_norm": 3.236809976550798,
      "kl": 0.09765625,
      "learning_rate": 1.0287053631196108e-07,
      "loss": 0.0039,
      "num_tokens": 62731304.0,
      "reward": 0.4583333432674408,
      "reward_std": 0.2842140197753906,
      "rewards/reasoning_reward/mean": 0.4583333432674408,
      "rewards/reasoning_reward/std": 0.6412736177444458,
      "step": 767
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 210.0,
      "completions/max_terminated_length": 210.0,
      "completions/mean_length": 151.33334350585938,
      "completions/mean_terminated_length": 151.33334350585938,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.7942088934850051,
      "grad_norm": 4.220315176240241,
      "kl": 0.08447265625,
      "learning_rate": 1.0188568086641614e-07,
      "loss": 0.0034,
      "num_tokens": 62813816.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.4039583206176758,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.5610387921333313,
      "step": 768
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 248.0,
      "completions/max_terminated_length": 248.0,
      "completions/mean_length": 175.9166717529297,
      "completions/mean_terminated_length": 175.9166717529297,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.795243019648397,
      "grad_norm": 3.8313849909839535,
      "kl": 0.05859375,
      "learning_rate": 1.0090502740312152e-07,
      "loss": 0.0023,
      "num_tokens": 62893494.0,
      "reward": 0.6875,
      "reward_std": 0.47950729727745056,
      "rewards/reasoning_reward/mean": 0.6875,
      "rewards/reasoning_reward/std": 0.4618605971336365,
      "step": 769
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 194.0,
      "completions/max_terminated_length": 194.0,
      "completions/mean_length": 144.4166717529297,
      "completions/mean_terminated_length": 144.4166717529297,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.796277145811789,
      "grad_norm": 4.064061665509299,
      "kl": 0.09033203125,
      "learning_rate": 9.992858627259237e-08,
      "loss": 0.0036,
      "num_tokens": 62975872.0,
      "reward": 1.125,
      "reward_std": 0.5049939155578613,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.6796738505363464,
      "step": 770
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 287.0,
      "completions/max_terminated_length": 287.0,
      "completions/mean_length": 169.45834350585938,
      "completions/mean_terminated_length": 169.45834350585938,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.7973112719751809,
      "grad_norm": 3.87391780953322,
      "kl": 0.07666015625,
      "learning_rate": 9.895636778088457e-08,
      "loss": 0.0031,
      "num_tokens": 63058859.0,
      "reward": 0.6875,
      "reward_std": 0.4851650893688202,
      "rewards/reasoning_reward/mean": 0.6875,
      "rewards/reasoning_reward/std": 0.7490936517715454,
      "step": 771
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 277.0,
      "completions/max_terminated_length": 277.0,
      "completions/mean_length": 142.5416717529297,
      "completions/mean_terminated_length": 142.5416717529297,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.7983453981385729,
      "grad_norm": 4.501546300705031,
      "kl": 0.0859375,
      "learning_rate": 9.798838218948468e-08,
      "loss": 0.0034,
      "num_tokens": 63135176.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.42156457901000977,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.7322785258293152,
      "step": 772
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 251.0,
      "completions/max_terminated_length": 251.0,
      "completions/mean_length": 156.08334350585938,
      "completions/mean_terminated_length": 156.08334350585938,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.7993795243019648,
      "grad_norm": 2.414414647480725,
      "kl": 0.055908203125,
      "learning_rate": 9.702463971520264e-08,
      "loss": 0.0022,
      "num_tokens": 63217306.0,
      "reward": 0.75,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 773
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 148.2916717529297,
      "completions/mean_terminated_length": 148.2916717529297,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.8004136504653567,
      "grad_norm": 4.229101545256886,
      "kl": 0.09716796875,
      "learning_rate": 9.606515053006347e-08,
      "loss": 0.0039,
      "num_tokens": 63305329.0,
      "reward": 1.375,
      "reward_std": 0.42645785212516785,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 774
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 115.79167175292969,
      "completions/mean_terminated_length": 115.79167175292969,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.8014477766287487,
      "grad_norm": 4.537579562049641,
      "kl": 0.050537109375,
      "learning_rate": 9.510992476119962e-08,
      "loss": 0.002,
      "num_tokens": 63385604.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.4446708858013153,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 775
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 152.0,
      "completions/mean_terminated_length": 152.0,
      "completions/min_length": 69.0,
      "completions/min_terminated_length": 69.0,
      "epoch": 0.8024819027921406,
      "grad_norm": 0.2165151880931561,
      "kl": 0.06982421875,
      "learning_rate": 9.415897249074478e-08,
      "loss": 0.0028,
      "num_tokens": 63474276.0,
      "reward": 1.6666667461395264,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.6666666269302368,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 776
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 191.0,
      "completions/max_terminated_length": 191.0,
      "completions/mean_length": 145.20834350585938,
      "completions/mean_terminated_length": 145.20834350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.8035160289555325,
      "grad_norm": 2.756431222536208,
      "kl": 0.044189453125,
      "learning_rate": 9.321230375572681e-08,
      "loss": 0.0018,
      "num_tokens": 63553017.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 777
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 166.83334350585938,
      "completions/mean_terminated_length": 166.83334350585938,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.8045501551189245,
      "grad_norm": 3.3926466918323492,
      "kl": 0.048095703125,
      "learning_rate": 9.226992854796234e-08,
      "loss": 0.0019,
      "num_tokens": 63633093.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.28302299976348877,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.416485458612442,
      "step": 778
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 266.0,
      "completions/max_terminated_length": 266.0,
      "completions/mean_length": 164.375,
      "completions/mean_terminated_length": 164.375,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.8055842812823164,
      "grad_norm": 3.0968300481941142,
      "kl": 0.068359375,
      "learning_rate": 9.133185681395072e-08,
      "loss": 0.0027,
      "num_tokens": 63716078.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 779
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 209.0,
      "completions/max_terminated_length": 209.0,
      "completions/mean_length": 153.625,
      "completions/mean_terminated_length": 153.625,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.8066184074457083,
      "grad_norm": 3.196506746978936,
      "kl": 0.0810546875,
      "learning_rate": 9.03980984547697e-08,
      "loss": 0.0032,
      "num_tokens": 63798485.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.2644323706626892,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.3895137906074524,
      "step": 780
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 164.25,
      "completions/mean_terminated_length": 164.25,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.8076525336091003,
      "grad_norm": 3.6585350966268906,
      "kl": 0.048095703125,
      "learning_rate": 8.946866332597064e-08,
      "loss": 0.0019,
      "num_tokens": 63885723.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.31285393238067627,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.5694518685340881,
      "step": 781
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 158.45834350585938,
      "completions/mean_terminated_length": 158.45834350585938,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.8086866597724922,
      "grad_norm": 3.5060979679162885,
      "kl": 0.049560546875,
      "learning_rate": 8.854356123747392e-08,
      "loss": 0.002,
      "num_tokens": 63964702.0,
      "reward": 0.625,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 782
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 291.0,
      "completions/max_terminated_length": 291.0,
      "completions/mean_length": 174.6666717529297,
      "completions/mean_terminated_length": 174.6666717529297,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.8097207859358841,
      "grad_norm": 3.1782626992547836,
      "kl": 0.076171875,
      "learning_rate": 8.762280195346655e-08,
      "loss": 0.003,
      "num_tokens": 64047942.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "step": 783
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 219.0,
      "completions/max_terminated_length": 219.0,
      "completions/mean_length": 146.375,
      "completions/mean_terminated_length": 146.375,
      "completions/min_length": 87.0,
      "completions/min_terminated_length": 87.0,
      "epoch": 0.8107549120992761,
      "grad_norm": 3.2114912520386825,
      "kl": 0.08154296875,
      "learning_rate": 8.6706395192298e-08,
      "loss": 0.0033,
      "num_tokens": 64128223.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 784
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 220.0,
      "completions/max_terminated_length": 220.0,
      "completions/mean_length": 150.375,
      "completions/mean_terminated_length": 150.375,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.811789038262668,
      "grad_norm": 2.4667511559934585,
      "kl": 0.0576171875,
      "learning_rate": 8.579435062637863e-08,
      "loss": 0.0023,
      "num_tokens": 64212144.0,
      "reward": 0.875,
      "reward_std": 0.07715167850255966,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.6954823136329651,
      "step": 785
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 208.0,
      "completions/max_terminated_length": 208.0,
      "completions/mean_length": 140.375,
      "completions/mean_terminated_length": 140.375,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.81282316442606,
      "grad_norm": 2.527486671838808,
      "kl": 0.08642578125,
      "learning_rate": 8.488667788207642e-08,
      "loss": 0.0035,
      "num_tokens": 64296945.0,
      "reward": 1.3125,
      "reward_std": 0.13908717036247253,
      "rewards/reasoning_reward/mean": 1.3125,
      "rewards/reasoning_reward/std": 0.5479705333709717,
      "step": 786
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 215.0,
      "completions/max_terminated_length": 215.0,
      "completions/mean_length": 147.25,
      "completions/mean_terminated_length": 147.25,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.8138572905894519,
      "grad_norm": 4.580987567602128,
      "kl": 0.09521484375,
      "learning_rate": 8.398338653961673e-08,
      "loss": 0.0038,
      "num_tokens": 64384935.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.3478729724884033,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.8670706748962402,
      "step": 787
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 357.0,
      "completions/max_terminated_length": 357.0,
      "completions/mean_length": 190.0,
      "completions/mean_terminated_length": 190.0,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.8148914167528438,
      "grad_norm": 3.6226405946498326,
      "kl": 0.08935546875,
      "learning_rate": 8.30844861329798e-08,
      "loss": 0.0036,
      "num_tokens": 64462951.0,
      "reward": 1.152777910232544,
      "reward_std": 0.237970232963562,
      "rewards/reasoning_reward/mean": 1.1527777910232544,
      "rewards/reasoning_reward/std": 0.30263206362724304,
      "step": 788
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 230.0,
      "completions/max_terminated_length": 230.0,
      "completions/mean_length": 142.0,
      "completions/mean_terminated_length": 142.0,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.8159255429162358,
      "grad_norm": 4.446754103716307,
      "kl": 0.08447265625,
      "learning_rate": 8.218998614980132e-08,
      "loss": 0.0034,
      "num_tokens": 64539215.0,
      "reward": 0.375,
      "reward_std": 0.47419947385787964,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 789
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 292.0,
      "completions/max_terminated_length": 292.0,
      "completions/mean_length": 179.0416717529297,
      "completions/mean_terminated_length": 179.0416717529297,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.8169596690796277,
      "grad_norm": 3.921864271005609,
      "kl": 0.061279296875,
      "learning_rate": 8.12998960312718e-08,
      "loss": 0.0024,
      "num_tokens": 64622584.0,
      "reward": 1.3402776718139648,
      "reward_std": 0.38187703490257263,
      "rewards/reasoning_reward/mean": 1.3402776718139648,
      "rewards/reasoning_reward/std": 0.5804362297058105,
      "step": 790
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 281.0,
      "completions/max_terminated_length": 281.0,
      "completions/mean_length": 164.95834350585938,
      "completions/mean_terminated_length": 164.95834350585938,
      "completions/min_length": 82.0,
      "completions/min_terminated_length": 82.0,
      "epoch": 0.8179937952430196,
      "grad_norm": 4.340848550091104,
      "kl": 0.076171875,
      "learning_rate": 8.041422517203627e-08,
      "loss": 0.003,
      "num_tokens": 64711807.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.4710209369659424,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.6197240948677063,
      "step": 791
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 135.375,
      "completions/mean_terminated_length": 135.375,
      "completions/min_length": 70.0,
      "completions/min_terminated_length": 70.0,
      "epoch": 0.8190279214064116,
      "grad_norm": 2.649981760760686,
      "kl": 0.0537109375,
      "learning_rate": 7.953298292009658e-08,
      "loss": 0.0021,
      "num_tokens": 64792344.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.20412415266036987,
      "step": 792
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 221.0,
      "completions/max_terminated_length": 221.0,
      "completions/mean_length": 153.6666717529297,
      "completions/mean_terminated_length": 153.6666717529297,
      "completions/min_length": 57.0,
      "completions/min_terminated_length": 57.0,
      "epoch": 0.8200620475698035,
      "grad_norm": 3.2958211549645253,
      "kl": 0.06689453125,
      "learning_rate": 7.86561785767112e-08,
      "loss": 0.0027,
      "num_tokens": 64868632.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3020375669002533,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.3877657949924469,
      "step": 793
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 281.0,
      "completions/max_terminated_length": 281.0,
      "completions/mean_length": 144.875,
      "completions/mean_terminated_length": 144.875,
      "completions/min_length": 60.0,
      "completions/min_terminated_length": 60.0,
      "epoch": 0.8210961737331954,
      "grad_norm": 3.09060522662954,
      "kl": 0.0693359375,
      "learning_rate": 7.77838213962983e-08,
      "loss": 0.0028,
      "num_tokens": 64950501.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.3666771650314331,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.5450701713562012,
      "step": 794
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 144.4166717529297,
      "completions/mean_terminated_length": 144.4166717529297,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.8221302998965874,
      "grad_norm": 3.8226975362183455,
      "kl": 0.06787109375,
      "learning_rate": 7.691592058633694e-08,
      "loss": 0.0027,
      "num_tokens": 65028095.0,
      "reward": 0.9375,
      "reward_std": 0.35495084524154663,
      "rewards/reasoning_reward/mean": 0.9375,
      "rewards/reasoning_reward/std": 0.5379611253738403,
      "step": 795
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 183.0,
      "completions/max_terminated_length": 183.0,
      "completions/mean_length": 127.91667175292969,
      "completions/mean_terminated_length": 127.91667175292969,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.8231644260599793,
      "grad_norm": 2.3041894307690525,
      "kl": 0.061767578125,
      "learning_rate": 7.605248530727115e-08,
      "loss": 0.0025,
      "num_tokens": 65112941.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 796
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 258.0,
      "completions/max_terminated_length": 258.0,
      "completions/mean_length": 167.83334350585938,
      "completions/mean_terminated_length": 167.83334350585938,
      "completions/min_length": 66.0,
      "completions/min_terminated_length": 66.0,
      "epoch": 0.8241985522233712,
      "grad_norm": 4.091470007656279,
      "kl": 0.10986328125,
      "learning_rate": 7.519352467241197e-08,
      "loss": 0.0044,
      "num_tokens": 65195697.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.3794546127319336,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.6197241544723511,
      "step": 797
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 269.0,
      "completions/max_terminated_length": 269.0,
      "completions/mean_length": 159.0,
      "completions/mean_terminated_length": 159.0,
      "completions/min_length": 60.0,
      "completions/min_terminated_length": 60.0,
      "epoch": 0.8252326783867632,
      "grad_norm": 3.2064801888792744,
      "kl": 0.07958984375,
      "learning_rate": 7.433904774784216e-08,
      "loss": 0.0032,
      "num_tokens": 65275673.0,
      "reward": 1.076388955116272,
      "reward_std": 0.12554192543029785,
      "rewards/reasoning_reward/mean": 1.076388955116272,
      "rewards/reasoning_reward/std": 0.17706114053726196,
      "step": 798
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 191.0,
      "completions/max_terminated_length": 191.0,
      "completions/mean_length": 149.0416717529297,
      "completions/mean_terminated_length": 149.0416717529297,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.8262668045501551,
      "grad_norm": 2.331881060813706,
      "kl": 0.06396484375,
      "learning_rate": 7.348906355232027e-08,
      "loss": 0.0026,
      "num_tokens": 65353674.0,
      "reward": 0.7291666865348816,
      "reward_std": 0.08625819534063339,
      "rewards/reasoning_reward/mean": 0.7291666865348816,
      "rewards/reasoning_reward/std": 0.551266610622406,
      "step": 799
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 200.0,
      "completions/max_terminated_length": 200.0,
      "completions/mean_length": 149.0,
      "completions/mean_terminated_length": 149.0,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.827300930713547,
      "grad_norm": 3.5190409760967327,
      "kl": 0.09228515625,
      "learning_rate": 7.264358105718505e-08,
      "loss": 0.0037,
      "num_tokens": 65443242.0,
      "reward": 1.125,
      "reward_std": 0.367926687002182,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.4484272003173828,
      "step": 800
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 208.0,
      "completions/max_terminated_length": 208.0,
      "completions/mean_length": 135.83334350585938,
      "completions/mean_terminated_length": 135.83334350585938,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.828335056876939,
      "grad_norm": 2.332152031715114,
      "kl": 0.0703125,
      "learning_rate": 7.180260918626152e-08,
      "loss": 0.0028,
      "num_tokens": 65530702.0,
      "reward": 1.3125,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 1.3125,
      "rewards/reasoning_reward/std": 0.4618605971336365,
      "step": 801
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 379.0,
      "completions/max_terminated_length": 379.0,
      "completions/mean_length": 172.625,
      "completions/mean_terminated_length": 172.625,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.8293691830403309,
      "grad_norm": 3.0269063067385193,
      "kl": 0.057373046875,
      "learning_rate": 7.096615681576596e-08,
      "loss": 0.0023,
      "num_tokens": 65607861.0,
      "reward": 0.5625,
      "reward_std": 0.33768826723098755,
      "rewards/reasoning_reward/mean": 0.5625,
      "rewards/reasoning_reward/std": 0.5954993963241577,
      "step": 802
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 240.0,
      "completions/max_terminated_length": 240.0,
      "completions/mean_length": 168.6666717529297,
      "completions/mean_terminated_length": 168.6666717529297,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.8304033092037229,
      "grad_norm": 3.273127666339797,
      "kl": 0.055419921875,
      "learning_rate": 7.013423277421299e-08,
      "loss": 0.0022,
      "num_tokens": 65686197.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.1767766922712326,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.5985338091850281,
      "step": 803
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 275.0,
      "completions/max_terminated_length": 275.0,
      "completions/mean_length": 198.95834350585938,
      "completions/mean_terminated_length": 198.95834350585938,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.8314374353671148,
      "grad_norm": 3.550378331709515,
      "kl": 0.07373046875,
      "learning_rate": 6.93068458423216e-08,
      "loss": 0.0029,
      "num_tokens": 65774468.0,
      "reward": 0.9722222685813904,
      "reward_std": 0.3949992060661316,
      "rewards/reasoning_reward/mean": 0.9722222685813904,
      "rewards/reasoning_reward/std": 0.5660837888717651,
      "step": 804
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 157.5,
      "completions/mean_terminated_length": 157.5,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.8324715615305067,
      "grad_norm": 2.947936195797767,
      "kl": 0.06689453125,
      "learning_rate": 6.848400475292343e-08,
      "loss": 0.0027,
      "num_tokens": 65854024.0,
      "reward": 1.0486111640930176,
      "reward_std": 0.06924575567245483,
      "rewards/reasoning_reward/mean": 1.0486111640930176,
      "rewards/reasoning_reward/std": 0.13440841436386108,
      "step": 805
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 305.0,
      "completions/max_terminated_length": 305.0,
      "completions/mean_length": 183.58334350585938,
      "completions/mean_terminated_length": 183.58334350585938,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.8335056876938987,
      "grad_norm": 3.8760059999402343,
      "kl": 0.08544921875,
      "learning_rate": 6.766571819086941e-08,
      "loss": 0.0034,
      "num_tokens": 65936142.0,
      "reward": 1.0625,
      "reward_std": 0.44589531421661377,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.7658352255821228,
      "step": 806
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 382.0,
      "completions/max_terminated_length": 382.0,
      "completions/mean_length": 183.4166717529297,
      "completions/mean_terminated_length": 183.4166717529297,
      "completions/min_length": 71.0,
      "completions/min_terminated_length": 71.0,
      "epoch": 0.8345398138572906,
      "grad_norm": 3.429218077443575,
      "kl": 0.06787109375,
      "learning_rate": 6.685199479293929e-08,
      "loss": 0.0027,
      "num_tokens": 66027528.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.34140023589134216,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.8678538203239441,
      "step": 807
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 408.0,
      "completions/max_terminated_length": 408.0,
      "completions/mean_length": 203.125,
      "completions/mean_terminated_length": 203.125,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.8355739400206825,
      "grad_norm": 3.4009884082930597,
      "kl": 0.049072265625,
      "learning_rate": 6.604284314774983e-08,
      "loss": 0.002,
      "num_tokens": 66111411.0,
      "reward": 1.3819444179534912,
      "reward_std": 0.36571210622787476,
      "rewards/reasoning_reward/mean": 1.3819442987442017,
      "rewards/reasoning_reward/std": 0.5189155340194702,
      "step": 808
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 215.0,
      "completions/max_terminated_length": 215.0,
      "completions/mean_length": 129.45834350585938,
      "completions/mean_terminated_length": 129.45834350585938,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.8366080661840745,
      "grad_norm": 6.175187456794133,
      "kl": 0.1650390625,
      "learning_rate": 6.523827179566394e-08,
      "loss": 0.0066,
      "num_tokens": 66195886.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.48678088188171387,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.7019641399383545,
      "step": 809
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 201.0,
      "completions/max_terminated_length": 201.0,
      "completions/mean_length": 134.375,
      "completions/mean_terminated_length": 134.375,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.8376421923474664,
      "grad_norm": 3.893758481978179,
      "kl": 0.058837890625,
      "learning_rate": 6.443828922870127e-08,
      "loss": 0.0024,
      "num_tokens": 66277543.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.3506905436515808,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.6240935921669006,
      "step": 810
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 308.0,
      "completions/max_terminated_length": 308.0,
      "completions/mean_length": 152.08334350585938,
      "completions/mean_terminated_length": 152.08334350585938,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.8386763185108583,
      "grad_norm": 3.882217378171403,
      "kl": 0.08642578125,
      "learning_rate": 6.364290389044769e-08,
      "loss": 0.0035,
      "num_tokens": 66365673.0,
      "reward": 1.4444444179534912,
      "reward_std": 0.3385341763496399,
      "rewards/reasoning_reward/mean": 1.4444442987442017,
      "rewards/reasoning_reward/std": 0.5787431597709656,
      "step": 811
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 256.0,
      "completions/max_terminated_length": 256.0,
      "completions/mean_length": 137.5416717529297,
      "completions/mean_terminated_length": 138.3913116455078,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.8397104446742503,
      "grad_norm": 3.974211152800969,
      "kl": 0.06787109375,
      "learning_rate": 6.285212417596719e-08,
      "loss": 0.0027,
      "num_tokens": 66448998.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.4446708858013153,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.7172814607620239,
      "step": 812
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 254.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 159.83334350585938,
      "completions/mean_terminated_length": 159.83334350585938,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.8407445708376422,
      "grad_norm": 3.374272533210629,
      "kl": 0.0537109375,
      "learning_rate": 6.206595843171225e-08,
      "loss": 0.0021,
      "num_tokens": 66525802.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.4042079448699951,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.5413181781768799,
      "step": 813
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 309.0,
      "completions/max_terminated_length": 309.0,
      "completions/mean_length": 188.7916717529297,
      "completions/mean_terminated_length": 188.7916717529297,
      "completions/min_length": 130.0,
      "completions/min_terminated_length": 130.0,
      "epoch": 0.8417786970010341,
      "grad_norm": 2.707109875272163,
      "kl": 0.07421875,
      "learning_rate": 6.128441495543646e-08,
      "loss": 0.003,
      "num_tokens": 66605133.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 814
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 224.0,
      "completions/max_terminated_length": 224.0,
      "completions/mean_length": 159.0,
      "completions/mean_terminated_length": 159.0,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.8428128231644261,
      "grad_norm": 11.687611255036002,
      "kl": 0.46875,
      "learning_rate": 6.050750199610682e-08,
      "loss": 0.0189,
      "num_tokens": 66684349.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 815
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 263.0,
      "completions/max_terminated_length": 263.0,
      "completions/mean_length": 179.625,
      "completions/mean_terminated_length": 179.625,
      "completions/min_length": 128.0,
      "completions/min_terminated_length": 128.0,
      "epoch": 0.843846949327818,
      "grad_norm": 3.5189829393185375,
      "kl": 0.06298828125,
      "learning_rate": 5.973522775381618e-08,
      "loss": 0.0025,
      "num_tokens": 66762020.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.4261821210384369,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 816
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 245.0,
      "completions/max_terminated_length": 245.0,
      "completions/mean_length": 163.0,
      "completions/mean_terminated_length": 163.0,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.8448810754912099,
      "grad_norm": 3.111695745233412,
      "kl": 0.0693359375,
      "learning_rate": 5.896760037969739e-08,
      "loss": 0.0028,
      "num_tokens": 66846116.0,
      "reward": 0.7777777910232544,
      "reward_std": 0.2375655621290207,
      "rewards/reasoning_reward/mean": 0.7777777314186096,
      "rewards/reasoning_reward/std": 0.8493627309799194,
      "step": 817
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 202.0,
      "completions/max_terminated_length": 202.0,
      "completions/mean_length": 144.9166717529297,
      "completions/mean_terminated_length": 144.9166717529297,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.8459152016546019,
      "grad_norm": 2.7736990622484976,
      "kl": 0.054931640625,
      "learning_rate": 5.8204627975836696e-08,
      "loss": 0.0022,
      "num_tokens": 66924266.0,
      "reward": 0.625,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 818
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 291.0,
      "completions/max_terminated_length": 291.0,
      "completions/mean_length": 170.625,
      "completions/mean_terminated_length": 170.625,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.8469493278179938,
      "grad_norm": 4.60558678901362,
      "kl": 0.068359375,
      "learning_rate": 5.744631859518878e-08,
      "loss": 0.0027,
      "num_tokens": 67002865.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.2840898931026459,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.35098204016685486,
      "step": 819
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 211.0,
      "completions/max_terminated_length": 211.0,
      "completions/mean_length": 151.7916717529297,
      "completions/mean_terminated_length": 151.7916717529297,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.8479834539813857,
      "grad_norm": 3.9470723321035543,
      "kl": 0.0556640625,
      "learning_rate": 5.66926802414911e-08,
      "loss": 0.0022,
      "num_tokens": 67087588.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.4386448860168457,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.6780058741569519,
      "step": 820
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 160.6666717529297,
      "completions/mean_terminated_length": 160.6666717529297,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.8490175801447777,
      "grad_norm": 3.2433337512879303,
      "kl": 0.052978515625,
      "learning_rate": 5.594372086918009e-08,
      "loss": 0.0021,
      "num_tokens": 67166756.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.32520395517349243,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.47729232907295227,
      "step": 821
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 155.95834350585938,
      "completions/mean_terminated_length": 155.95834350585938,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.8500517063081696,
      "grad_norm": 2.8260741217314123,
      "kl": 0.07080078125,
      "learning_rate": 5.519944838330659e-08,
      "loss": 0.0028,
      "num_tokens": 67249443.0,
      "reward": 1.1875,
      "reward_std": 0.27053868770599365,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.4848240315914154,
      "step": 822
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 294.0,
      "completions/max_terminated_length": 294.0,
      "completions/mean_length": 164.58334350585938,
      "completions/mean_terminated_length": 164.58334350585938,
      "completions/min_length": 62.0,
      "completions/min_terminated_length": 62.0,
      "epoch": 0.8510858324715616,
      "grad_norm": 4.350268666915288,
      "kl": 0.06494140625,
      "learning_rate": 5.4459870639452897e-08,
      "loss": 0.0026,
      "num_tokens": 67327881.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.494476854801178,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.580089271068573,
      "step": 823
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 214.0,
      "completions/max_terminated_length": 214.0,
      "completions/mean_length": 151.20834350585938,
      "completions/mean_terminated_length": 151.20834350585938,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.8521199586349535,
      "grad_norm": 4.367925297416209,
      "kl": 0.09228515625,
      "learning_rate": 5.372499544364972e-08,
      "loss": 0.0037,
      "num_tokens": 67416214.0,
      "reward": 1.625,
      "reward_std": 0.4151468276977539,
      "rewards/reasoning_reward/mean": 1.625,
      "rewards/reasoning_reward/std": 0.42633703351020813,
      "step": 824
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 166.875,
      "completions/mean_terminated_length": 166.875,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.8531540847983454,
      "grad_norm": 3.804426970311705,
      "kl": 0.083984375,
      "learning_rate": 5.2994830552293365e-08,
      "loss": 0.0033,
      "num_tokens": 67502027.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.4114243686199188,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.7172815203666687,
      "step": 825
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 309.0,
      "completions/max_terminated_length": 309.0,
      "completions/mean_length": 164.6666717529297,
      "completions/mean_terminated_length": 164.6666717529297,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.8541882109617374,
      "grad_norm": 2.9609678472482113,
      "kl": 0.046875,
      "learning_rate": 5.2269383672064736e-08,
      "loss": 0.0019,
      "num_tokens": 67585763.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.26726123690605164,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.6197240948677063,
      "step": 826
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 209.0,
      "completions/max_terminated_length": 209.0,
      "completions/mean_length": 147.70834350585938,
      "completions/mean_terminated_length": 147.70834350585938,
      "completions/min_length": 72.0,
      "completions/min_terminated_length": 72.0,
      "epoch": 0.8552223371251293,
      "grad_norm": 3.152704440105035,
      "kl": 0.08837890625,
      "learning_rate": 5.154866245984696e-08,
      "loss": 0.0035,
      "num_tokens": 67663468.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.43070170283317566,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.5103103518486023,
      "step": 827
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 243.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 154.33334350585938,
      "completions/mean_terminated_length": 154.33334350585938,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.8562564632885212,
      "grad_norm": 0.17170167898547498,
      "kl": 0.0634765625,
      "learning_rate": 5.083267452264556e-08,
      "loss": 0.0025,
      "num_tokens": 67747188.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 828
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 210.0,
      "completions/max_terminated_length": 210.0,
      "completions/mean_length": 131.2916717529297,
      "completions/mean_terminated_length": 131.2916717529297,
      "completions/min_length": 86.0,
      "completions/min_terminated_length": 86.0,
      "epoch": 0.8572905894519132,
      "grad_norm": 0.18506989130036433,
      "kl": 0.053955078125,
      "learning_rate": 5.012142741750725e-08,
      "loss": 0.0022,
      "num_tokens": 67824483.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 829
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 208.0,
      "completions/max_terminated_length": 208.0,
      "completions/mean_length": 141.25,
      "completions/mean_terminated_length": 141.25,
      "completions/min_length": 73.0,
      "completions/min_terminated_length": 73.0,
      "epoch": 0.8583247156153051,
      "grad_norm": 4.548140603940961,
      "kl": 0.130859375,
      "learning_rate": 4.941492865144115e-08,
      "loss": 0.0052,
      "num_tokens": 67906433.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.41873571276664734,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.7144345045089722,
      "step": 830
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 165.0,
      "completions/mean_terminated_length": 165.0,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.859358841778697,
      "grad_norm": 3.507638044504915,
      "kl": 0.080078125,
      "learning_rate": 4.8713185681338477e-08,
      "loss": 0.0032,
      "num_tokens": 67994881.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.6240935325622559,
      "step": 831
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 379.0,
      "completions/max_terminated_length": 379.0,
      "completions/mean_length": 179.625,
      "completions/mean_terminated_length": 179.625,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.860392967942089,
      "grad_norm": 2.847799628996861,
      "kl": 0.06591796875,
      "learning_rate": 4.801620591389477e-08,
      "loss": 0.0026,
      "num_tokens": 68074736.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.08908708393573761,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 832
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 288.0,
      "completions/max_terminated_length": 288.0,
      "completions/mean_length": 156.5416717529297,
      "completions/mean_terminated_length": 156.5416717529297,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.8614270941054809,
      "grad_norm": 3.2176346783915846,
      "kl": 0.0439453125,
      "learning_rate": 4.7323996705531335e-08,
      "loss": 0.0018,
      "num_tokens": 68152653.0,
      "reward": 1.1875,
      "reward_std": 0.2041093111038208,
      "rewards/reasoning_reward/mean": 1.1875,
      "rewards/reasoning_reward/std": 0.4376940429210663,
      "step": 833
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 180.0,
      "completions/max_terminated_length": 180.0,
      "completions/mean_length": 122.91667175292969,
      "completions/mean_terminated_length": 122.91667175292969,
      "completions/min_length": 58.0,
      "completions/min_terminated_length": 58.0,
      "epoch": 0.8624612202688728,
      "grad_norm": 2.119171203323999,
      "kl": 0.05859375,
      "learning_rate": 4.6636565362317304e-08,
      "loss": 0.0023,
      "num_tokens": 68230691.0,
      "reward": 0.75,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 834
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 228.0,
      "completions/max_terminated_length": 228.0,
      "completions/mean_length": 151.4166717529297,
      "completions/mean_terminated_length": 151.4166717529297,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.8634953464322648,
      "grad_norm": 2.9688355068863923,
      "kl": 0.06591796875,
      "learning_rate": 4.59539191398931e-08,
      "loss": 0.0026,
      "num_tokens": 68307461.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.10681166499853134,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.2750164568424225,
      "step": 835
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 385.0,
      "completions/max_terminated_length": 385.0,
      "completions/mean_length": 183.70834350585938,
      "completions/mean_terminated_length": 183.70834350585938,
      "completions/min_length": 114.0,
      "completions/min_terminated_length": 114.0,
      "epoch": 0.8645294725956567,
      "grad_norm": 4.374431997930627,
      "kl": 0.10595703125,
      "learning_rate": 4.527606524339328e-08,
      "loss": 0.0043,
      "num_tokens": 68391278.0,
      "reward": 1.3958333730697632,
      "reward_std": 0.4973698854446411,
      "rewards/reasoning_reward/mean": 1.3958333730697632,
      "rewards/reasoning_reward/std": 0.5512666702270508,
      "step": 836
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 200.0,
      "completions/max_terminated_length": 200.0,
      "completions/mean_length": 117.33333587646484,
      "completions/mean_terminated_length": 117.33333587646484,
      "completions/min_length": 54.0,
      "completions/min_terminated_length": 54.0,
      "epoch": 0.8655635987590486,
      "grad_norm": 3.368684441355164,
      "kl": 0.12890625,
      "learning_rate": 4.4603010827371224e-08,
      "loss": 0.0052,
      "num_tokens": 68475710.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.21362332999706268,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.35864076018333435,
      "step": 837
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 185.0,
      "completions/max_terminated_length": 185.0,
      "completions/mean_length": 110.04167175292969,
      "completions/mean_terminated_length": 110.04167175292969,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.8665977249224406,
      "grad_norm": 0.4287350953005526,
      "kl": 0.060546875,
      "learning_rate": 4.393476299572263e-08,
      "loss": 0.0024,
      "num_tokens": 68552959.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 838
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 255.0,
      "completions/max_terminated_length": 255.0,
      "completions/mean_length": 172.4166717529297,
      "completions/mean_terminated_length": 172.4166717529297,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.8676318510858325,
      "grad_norm": 3.1249910497191165,
      "kl": 0.0546875,
      "learning_rate": 4.327132880161161e-08,
      "loss": 0.0022,
      "num_tokens": 68630249.0,
      "reward": 0.4791666865348816,
      "reward_std": 0.2041093111038208,
      "rewards/reasoning_reward/mean": 0.4791666567325592,
      "rewards/reasoning_reward/std": 0.6507381796836853,
      "step": 839
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 318.0,
      "completions/max_terminated_length": 318.0,
      "completions/mean_length": 168.70834350585938,
      "completions/mean_terminated_length": 168.70834350585938,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.8686659772492245,
      "grad_norm": 3.2822755480432844,
      "kl": 0.052490234375,
      "learning_rate": 4.261271524739524e-08,
      "loss": 0.0021,
      "num_tokens": 68714258.0,
      "reward": 1.4027776718139648,
      "reward_std": 0.2818480134010315,
      "rewards/reasoning_reward/mean": 1.4027776718139648,
      "rewards/reasoning_reward/std": 0.44482171535491943,
      "step": 840
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 241.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 139.0416717529297,
      "completions/mean_terminated_length": 139.0416717529297,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.8697001034126164,
      "grad_norm": 3.768686239797151,
      "kl": 0.052978515625,
      "learning_rate": 4.195892928455047e-08,
      "loss": 0.0021,
      "num_tokens": 68792851.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.2314550280570984,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.3877657949924469,
      "step": 841
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 188.0,
      "completions/max_terminated_length": 188.0,
      "completions/mean_length": 149.45834350585938,
      "completions/mean_terminated_length": 149.45834350585938,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.8707342295760083,
      "grad_norm": 2.9671230713429453,
      "kl": 0.043212890625,
      "learning_rate": 4.130997781360035e-08,
      "loss": 0.0017,
      "num_tokens": 68872318.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 842
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 265.0,
      "completions/max_terminated_length": 265.0,
      "completions/mean_length": 150.1666717529297,
      "completions/mean_terminated_length": 150.1666717529297,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.8717683557394003,
      "grad_norm": 3.480205852106029,
      "kl": 0.068359375,
      "learning_rate": 4.0665867684041013e-08,
      "loss": 0.0027,
      "num_tokens": 68947754.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.28302299976348877,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.416485458612442,
      "step": 843
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 289.0,
      "completions/max_terminated_length": 289.0,
      "completions/mean_length": 161.9166717529297,
      "completions/mean_terminated_length": 161.9166717529297,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.8728024819027922,
      "grad_norm": 2.661968506889756,
      "kl": 0.07373046875,
      "learning_rate": 4.002660569426997e-08,
      "loss": 0.003,
      "num_tokens": 69031840.0,
      "reward": 1.3541667461395264,
      "reward_std": 0.0589255653321743,
      "rewards/reasoning_reward/mean": 1.3541666269302368,
      "rewards/reasoning_reward/std": 0.47729232907295227,
      "step": 844
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 300.0,
      "completions/max_terminated_length": 300.0,
      "completions/mean_length": 189.875,
      "completions/mean_terminated_length": 189.26087951660156,
      "completions/min_length": 122.0,
      "completions/min_terminated_length": 122.0,
      "epoch": 0.8738366080661841,
      "grad_norm": 3.489151843177023,
      "kl": 0.06689453125,
      "learning_rate": 3.939219859151377e-08,
      "loss": 0.0027,
      "num_tokens": 69110189.0,
      "reward": 0.7847222685813904,
      "reward_std": 0.5546908974647522,
      "rewards/reasoning_reward/mean": 0.7847222685813904,
      "rewards/reasoning_reward/std": 0.699342668056488,
      "step": 845
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 225.0,
      "completions/max_terminated_length": 225.0,
      "completions/mean_length": 150.2916717529297,
      "completions/mean_terminated_length": 150.2916717529297,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.8748707342295761,
      "grad_norm": 3.1194409662404183,
      "kl": 0.0859375,
      "learning_rate": 3.876265307175714e-08,
      "loss": 0.0034,
      "num_tokens": 69188180.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 846
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 238.0,
      "completions/max_terminated_length": 238.0,
      "completions/mean_length": 160.95834350585938,
      "completions/mean_terminated_length": 160.95834350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.8759048603929679,
      "grad_norm": 4.060590754998399,
      "kl": 0.06396484375,
      "learning_rate": 3.813797577967209e-08,
      "loss": 0.0026,
      "num_tokens": 69265083.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.33108004927635193,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.5103103518486023,
      "step": 847
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 445.0,
      "completions/max_terminated_length": 445.0,
      "completions/mean_length": 197.4166717529297,
      "completions/mean_terminated_length": 197.4166717529297,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.8769389865563598,
      "grad_norm": 1.7203510455755477,
      "kl": 0.07275390625,
      "learning_rate": 3.751817330854806e-08,
      "loss": 0.0029,
      "num_tokens": 69346101.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.5500329732894897,
      "step": 848
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 216.0,
      "completions/max_terminated_length": 216.0,
      "completions/mean_length": 162.25,
      "completions/mean_terminated_length": 162.25,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.8779731127197518,
      "grad_norm": 3.1548182837694405,
      "kl": 0.0751953125,
      "learning_rate": 3.6903252200222e-08,
      "loss": 0.003,
      "num_tokens": 69430283.0,
      "reward": 0.375,
      "reward_std": 0.2314550280570984,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.47204458713531494,
      "step": 849
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 153.5,
      "completions/mean_terminated_length": 153.5,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.8790072388831437,
      "grad_norm": 3.229700554376665,
      "kl": 0.0751953125,
      "learning_rate": 3.6293218945009364e-08,
      "loss": 0.003,
      "num_tokens": 69508583.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.2041093111038208,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.5311833620071411,
      "step": 850
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 140.0416717529297,
      "completions/mean_terminated_length": 140.0416717529297,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.8800413650465356,
      "grad_norm": 3.6374367800290743,
      "kl": 0.08837890625,
      "learning_rate": 3.56880799816362e-08,
      "loss": 0.0035,
      "num_tokens": 69597664.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.252508282661438,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.8795173168182373,
      "step": 851
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 247.0,
      "completions/max_terminated_length": 247.0,
      "completions/mean_length": 151.58334350585938,
      "completions/mean_terminated_length": 151.58334350585938,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.8810754912099276,
      "grad_norm": 3.8095535645459964,
      "kl": 0.07373046875,
      "learning_rate": 3.50878416971701e-08,
      "loss": 0.0029,
      "num_tokens": 69679846.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.42002925276756287,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.8531165719032288,
      "step": 852
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 130.58334350585938,
      "completions/mean_terminated_length": 130.58334350585938,
      "completions/min_length": 71.0,
      "completions/min_terminated_length": 71.0,
      "epoch": 0.8821096173733195,
      "grad_norm": 2.5914771322358585,
      "kl": 0.06494140625,
      "learning_rate": 3.449251042695378e-08,
      "loss": 0.0026,
      "num_tokens": 69766932.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 853
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 309.0,
      "completions/max_terminated_length": 309.0,
      "completions/mean_length": 139.625,
      "completions/mean_terminated_length": 139.625,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.8831437435367114,
      "grad_norm": 2.3610444312496774,
      "kl": 0.07373046875,
      "learning_rate": 3.39020924545379e-08,
      "loss": 0.0029,
      "num_tokens": 69848611.0,
      "reward": 1.25,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.6079187393188477,
      "step": 854
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 160.0,
      "completions/mean_terminated_length": 160.0,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.8841778697001034,
      "grad_norm": 3.907652167180534,
      "kl": 0.07861328125,
      "learning_rate": 3.331659401161435e-08,
      "loss": 0.0032,
      "num_tokens": 69936947.0,
      "reward": 1.7083333730697632,
      "reward_std": 0.3897872865200043,
      "rewards/reasoning_reward/mean": 1.7083333730697632,
      "rewards/reasoning_reward/std": 0.4402732849121094,
      "step": 855
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 291.0,
      "completions/max_terminated_length": 291.0,
      "completions/mean_length": 163.5416717529297,
      "completions/mean_terminated_length": 163.5416717529297,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.8852119958634953,
      "grad_norm": 3.4271002980194485,
      "kl": 0.072265625,
      "learning_rate": 3.2736021277951055e-08,
      "loss": 0.0029,
      "num_tokens": 70016560.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.39000558853149414,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.5646597146987915,
      "step": 856
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 309.0,
      "completions/max_terminated_length": 241.0,
      "completions/mean_length": 168.875,
      "completions/mean_terminated_length": 162.78260803222656,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.8862461220268872,
      "grad_norm": 3.242036202397786,
      "kl": 0.10400390625,
      "learning_rate": 3.216038038132623e-08,
      "loss": 0.0042,
      "num_tokens": 70099381.0,
      "reward": 0.75,
      "reward_std": 0.3900056481361389,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.6079187393188477,
      "step": 857
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 124.16667175292969,
      "completions/mean_terminated_length": 124.16667175292969,
      "completions/min_length": 71.0,
      "completions/min_terminated_length": 71.0,
      "epoch": 0.8872802481902792,
      "grad_norm": 2.4008627320484064,
      "kl": 0.039306640625,
      "learning_rate": 3.1589677397464433e-08,
      "loss": 0.0016,
      "num_tokens": 70178001.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 858
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 211.0,
      "completions/max_terminated_length": 211.0,
      "completions/mean_length": 144.9166717529297,
      "completions/mean_terminated_length": 144.9166717529297,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.8883143743536711,
      "grad_norm": 2.3968224744059543,
      "kl": 0.06298828125,
      "learning_rate": 3.102391834997142e-08,
      "loss": 0.0025,
      "num_tokens": 70260223.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.20412413775920868,
      "step": 859
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 171.0,
      "completions/max_terminated_length": 171.0,
      "completions/mean_length": 124.16667175292969,
      "completions/mean_terminated_length": 124.16667175292969,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.889348500517063,
      "grad_norm": 3.3305458427955363,
      "kl": 0.08056640625,
      "learning_rate": 3.0463109210271566e-08,
      "loss": 0.0032,
      "num_tokens": 70343299.0,
      "reward": 1.2708333730697632,
      "reward_std": 0.30318546295166016,
      "rewards/reasoning_reward/mean": 1.2708333730697632,
      "rewards/reasoning_reward/std": 0.5893837809562683,
      "step": 860
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 163.33334350585938,
      "completions/mean_terminated_length": 163.33334350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.890382626680455,
      "grad_norm": 4.099541212176896,
      "kl": 0.06494140625,
      "learning_rate": 2.990725589754406e-08,
      "loss": 0.0026,
      "num_tokens": 70434139.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.46631526947021484,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.6197241544723511,
      "step": 861
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 210.0,
      "completions/max_terminated_length": 210.0,
      "completions/mean_length": 149.95834350585938,
      "completions/mean_terminated_length": 149.95834350585938,
      "completions/min_length": 59.0,
      "completions/min_terminated_length": 59.0,
      "epoch": 0.8914167528438469,
      "grad_norm": 2.2220524006143076,
      "kl": 0.0732421875,
      "learning_rate": 2.935636427866095e-08,
      "loss": 0.0029,
      "num_tokens": 70517314.0,
      "reward": 0.9513888359069824,
      "reward_std": 0.06924576312303543,
      "rewards/reasoning_reward/mean": 0.9513888359069824,
      "rewards/reasoning_reward/std": 0.7824759483337402,
      "step": 862
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 260.0,
      "completions/max_terminated_length": 260.0,
      "completions/mean_length": 161.58334350585938,
      "completions/mean_terminated_length": 161.58334350585938,
      "completions/min_length": 76.0,
      "completions/min_terminated_length": 76.0,
      "epoch": 0.8924508790072389,
      "grad_norm": 3.2564376835733837,
      "kl": 0.059814453125,
      "learning_rate": 2.881044016812506e-08,
      "loss": 0.0024,
      "num_tokens": 70593264.0,
      "reward": 0.4791666865348816,
      "reward_std": 0.3704721927642822,
      "rewards/reasoning_reward/mean": 0.4791666567325592,
      "rewards/reasoning_reward/std": 0.5610387921333313,
      "step": 863
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 157.0416717529297,
      "completions/mean_terminated_length": 157.0416717529297,
      "completions/min_length": 115.0,
      "completions/min_terminated_length": 115.0,
      "epoch": 0.8934850051706308,
      "grad_norm": 3.445585242026353,
      "kl": 0.07763671875,
      "learning_rate": 2.8269489328008433e-08,
      "loss": 0.0031,
      "num_tokens": 70677257.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.20693820714950562,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.550032913684845,
      "step": 864
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 193.0,
      "completions/max_terminated_length": 193.0,
      "completions/mean_length": 124.54167175292969,
      "completions/mean_terminated_length": 124.54167175292969,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.8945191313340227,
      "grad_norm": 3.26944024523484,
      "kl": 0.07421875,
      "learning_rate": 2.7733517467891822e-08,
      "loss": 0.003,
      "num_tokens": 70761230.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.21322892606258392,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.5893837213516235,
      "step": 865
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 143.08334350585938,
      "completions/mean_terminated_length": 143.08334350585938,
      "completions/min_length": 80.0,
      "completions/min_terminated_length": 80.0,
      "epoch": 0.8955532574974147,
      "grad_norm": 3.6817862651467532,
      "kl": 0.08154296875,
      "learning_rate": 2.720253024480418e-08,
      "loss": 0.0033,
      "num_tokens": 70855008.0,
      "reward": 1.6458333730697632,
      "reward_std": 0.35878798365592957,
      "rewards/reasoning_reward/mean": 1.6458333730697632,
      "rewards/reasoning_reward/std": 0.5208514332771301,
      "step": 866
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 268.0,
      "completions/max_terminated_length": 268.0,
      "completions/mean_length": 147.875,
      "completions/mean_terminated_length": 147.875,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.8965873836608066,
      "grad_norm": 3.9516108836506296,
      "kl": 0.06982421875,
      "learning_rate": 2.6676533263163103e-08,
      "loss": 0.0028,
      "num_tokens": 70931629.0,
      "reward": 0.6041666865348816,
      "reward_std": 0.4130779206752777,
      "rewards/reasoning_reward/mean": 0.6041666865348816,
      "rewards/reasoning_reward/std": 0.5893837213516235,
      "step": 867
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 302.0,
      "completions/max_terminated_length": 263.0,
      "completions/mean_length": 150.0416717529297,
      "completions/mean_terminated_length": 143.43478393554688,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.8976215098241985,
      "grad_norm": 5.483549708220734,
      "kl": 0.21484375,
      "learning_rate": 2.6155532074715548e-08,
      "loss": 0.0086,
      "num_tokens": 71011766.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.46288391947746277,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 868
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 237.0,
      "completions/max_terminated_length": 237.0,
      "completions/mean_length": 138.08334350585938,
      "completions/mean_terminated_length": 138.08334350585938,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.8986556359875905,
      "grad_norm": 3.8988737222080285,
      "kl": 0.07373046875,
      "learning_rate": 2.5639532178479417e-08,
      "loss": 0.0029,
      "num_tokens": 71096920.0,
      "reward": 1.25,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.6079187393188477,
      "step": 869
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 244.0,
      "completions/max_terminated_length": 244.0,
      "completions/mean_length": 154.70834350585938,
      "completions/mean_terminated_length": 154.70834350585938,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.8996897621509824,
      "grad_norm": 3.83680323474309,
      "kl": 0.044921875,
      "learning_rate": 2.512853902068529e-08,
      "loss": 0.0018,
      "num_tokens": 71174081.0,
      "reward": 0.875,
      "reward_std": 0.46288391947746277,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 870
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 222.0,
      "completions/max_terminated_length": 222.0,
      "completions/mean_length": 133.33334350585938,
      "completions/mean_terminated_length": 133.33334350585938,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.9007238883143743,
      "grad_norm": 4.143284467335127,
      "kl": 0.08740234375,
      "learning_rate": 2.462255799471913e-08,
      "loss": 0.0035,
      "num_tokens": 71258481.0,
      "reward": 1.4583333730697632,
      "reward_std": 0.42645785212516785,
      "rewards/reasoning_reward/mean": 1.4583333730697632,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 871
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 194.0,
      "completions/max_terminated_length": 194.0,
      "completions/mean_length": 141.9166717529297,
      "completions/mean_terminated_length": 141.9166717529297,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.9017580144777663,
      "grad_norm": 2.618618934323824,
      "kl": 0.115234375,
      "learning_rate": 2.412159444106543e-08,
      "loss": 0.0046,
      "num_tokens": 71340479.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.21362332999706268,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.6240935921669006,
      "step": 872
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 168.0,
      "completions/max_terminated_length": 168.0,
      "completions/mean_length": 113.25,
      "completions/mean_terminated_length": 113.25,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.9027921406411582,
      "grad_norm": 2.942245207754384,
      "kl": 0.06396484375,
      "learning_rate": 2.3625653647250388e-08,
      "loss": 0.0026,
      "num_tokens": 71425653.0,
      "reward": 1.125,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.5366967916488647,
      "step": 873
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 198.0,
      "completions/max_terminated_length": 198.0,
      "completions/mean_length": 149.58334350585938,
      "completions/mean_terminated_length": 149.58334350585938,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.9038262668045501,
      "grad_norm": 3.0454811913733706,
      "kl": 0.099609375,
      "learning_rate": 2.3134740847786715e-08,
      "loss": 0.004,
      "num_tokens": 71513515.0,
      "reward": 1.5416667461395264,
      "reward_std": 0.20693820714950562,
      "rewards/reasoning_reward/mean": 1.5416666269302368,
      "rewards/reasoning_reward/std": 0.550032913684845,
      "step": 874
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 317.0,
      "completions/max_terminated_length": 317.0,
      "completions/mean_length": 179.33334350585938,
      "completions/mean_terminated_length": 179.33334350585938,
      "completions/min_length": 111.0,
      "completions/min_terminated_length": 111.0,
      "epoch": 0.9048603929679421,
      "grad_norm": 2.930325510601697,
      "kl": 0.054931640625,
      "learning_rate": 2.2648861224117856e-08,
      "loss": 0.0022,
      "num_tokens": 71591131.0,
      "reward": 1.3541667461395264,
      "reward_std": 0.3433460593223572,
      "rewards/reasoning_reward/mean": 1.3541666269302368,
      "rewards/reasoning_reward/std": 0.47729232907295227,
      "step": 875
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 298.0,
      "completions/max_terminated_length": 298.0,
      "completions/mean_length": 179.20834350585938,
      "completions/mean_terminated_length": 179.20834350585938,
      "completions/min_length": 120.0,
      "completions/min_terminated_length": 120.0,
      "epoch": 0.905894519131334,
      "grad_norm": 4.432380095813641,
      "kl": 0.0673828125,
      "learning_rate": 2.2168019904563683e-08,
      "loss": 0.0027,
      "num_tokens": 71668936.0,
      "reward": 0.5833333730697632,
      "reward_std": 0.3900056481361389,
      "rewards/reasoning_reward/mean": 0.5833333134651184,
      "rewards/reasoning_reward/std": 0.5036101341247559,
      "step": 876
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 227.0,
      "completions/max_terminated_length": 227.0,
      "completions/mean_length": 136.125,
      "completions/mean_terminated_length": 136.125,
      "completions/min_length": 53.0,
      "completions/min_terminated_length": 53.0,
      "epoch": 0.9069286452947259,
      "grad_norm": 3.138554136523989,
      "kl": 0.072265625,
      "learning_rate": 2.1692221964266123e-08,
      "loss": 0.0029,
      "num_tokens": 71752219.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.3535533845424652,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 877
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 228.0,
      "completions/max_terminated_length": 228.0,
      "completions/mean_length": 146.875,
      "completions/mean_terminated_length": 146.875,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.9079627714581179,
      "grad_norm": 2.9483481427069043,
      "kl": 0.044677734375,
      "learning_rate": 2.122147242513578e-08,
      "loss": 0.0018,
      "num_tokens": 71832360.0,
      "reward": 0.7083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.7083333134651184,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 878
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 185.0,
      "completions/max_terminated_length": 185.0,
      "completions/mean_length": 150.08334350585938,
      "completions/mean_terminated_length": 150.08334350585938,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.9089968976215098,
      "grad_norm": 3.1073607500565905,
      "kl": 0.048828125,
      "learning_rate": 2.0755776255798718e-08,
      "loss": 0.002,
      "num_tokens": 71916330.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.39000558853149414,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.637022078037262,
      "step": 879
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 288.0,
      "completions/max_terminated_length": 288.0,
      "completions/mean_length": 175.5416717529297,
      "completions/mean_terminated_length": 175.5416717529297,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.9100310237849017,
      "grad_norm": 3.8150186805318964,
      "kl": 0.06884765625,
      "learning_rate": 2.0295138371544228e-08,
      "loss": 0.0028,
      "num_tokens": 71994855.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.46288391947746277,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.7506036162376404,
      "step": 880
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 160.0,
      "completions/max_terminated_length": 160.0,
      "completions/mean_length": 107.95833587646484,
      "completions/mean_terminated_length": 107.95833587646484,
      "completions/min_length": 70.0,
      "completions/min_terminated_length": 70.0,
      "epoch": 0.9110651499482937,
      "grad_norm": 2.4581793568003634,
      "kl": 0.05078125,
      "learning_rate": 1.9839563634272972e-08,
      "loss": 0.002,
      "num_tokens": 72073422.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.20412415266036987,
      "step": 881
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 156.0416717529297,
      "completions/mean_terminated_length": 156.0416717529297,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 0.9120992761116856,
      "grad_norm": 2.6154409593707686,
      "kl": 0.0732421875,
      "learning_rate": 1.938905685244513e-08,
      "loss": 0.0029,
      "num_tokens": 72157559.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.24800793826580048,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5882299542427063,
      "step": 882
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 136.95834350585938,
      "completions/mean_terminated_length": 136.95834350585938,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.9131334022750776,
      "grad_norm": 3.5680063997463867,
      "kl": 0.07763671875,
      "learning_rate": 1.8943622781030564e-08,
      "loss": 0.0031,
      "num_tokens": 72236238.0,
      "reward": 0.2916666865348816,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.2916666567325592,
      "rewards/reasoning_reward/std": 0.4643056094646454,
      "step": 883
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 282.0,
      "completions/max_terminated_length": 282.0,
      "completions/mean_length": 149.45834350585938,
      "completions/mean_terminated_length": 149.45834350585938,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.9141675284384695,
      "grad_norm": 3.2270543081927117,
      "kl": 0.072265625,
      "learning_rate": 1.850326612145775e-08,
      "loss": 0.0029,
      "num_tokens": 72320209.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.323208749294281,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.6753286719322205,
      "step": 884
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 255.0,
      "completions/max_terminated_length": 255.0,
      "completions/mean_length": 169.83334350585938,
      "completions/mean_terminated_length": 169.83334350585938,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.9152016546018614,
      "grad_norm": 4.260735191605596,
      "kl": 0.0556640625,
      "learning_rate": 1.8067991521564852e-08,
      "loss": 0.0022,
      "num_tokens": 72399093.0,
      "reward": 0.8958333730697632,
      "reward_std": 0.3960779905319214,
      "rewards/reasoning_reward/mean": 0.8958333134651184,
      "rewards/reasoning_reward/std": 0.7220015525817871,
      "step": 885
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 358.0,
      "completions/max_terminated_length": 358.0,
      "completions/mean_length": 178.625,
      "completions/mean_terminated_length": 178.625,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.9162357807652534,
      "grad_norm": 3.3552268374877303,
      "kl": 0.076171875,
      "learning_rate": 1.7637803575550115e-08,
      "loss": 0.003,
      "num_tokens": 72476500.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.5009793043136597,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.601929247379303,
      "step": 886
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 224.0,
      "completions/max_terminated_length": 224.0,
      "completions/mean_length": 147.33334350585938,
      "completions/mean_terminated_length": 147.33334350585938,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.9172699069286453,
      "grad_norm": 2.054383529313265,
      "kl": 0.0517578125,
      "learning_rate": 1.7212706823923674e-08,
      "loss": 0.0021,
      "num_tokens": 72554484.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 887
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 265.0,
      "completions/max_terminated_length": 265.0,
      "completions/mean_length": 156.125,
      "completions/mean_terminated_length": 156.125,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.9183040330920372,
      "grad_norm": 4.1619103374208235,
      "kl": 0.08837890625,
      "learning_rate": 1.6792705753459757e-08,
      "loss": 0.0035,
      "num_tokens": 72632119.0,
      "reward": 0.875,
      "reward_std": 0.5383754968643188,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.5757792592048645,
      "step": 888
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 152.95834350585938,
      "completions/mean_terminated_length": 152.95834350585938,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.9193381592554292,
      "grad_norm": 2.5645639037488097,
      "kl": 0.06494140625,
      "learning_rate": 1.6377804797148788e-08,
      "loss": 0.0026,
      "num_tokens": 72718998.0,
      "reward": 1.2291667461395264,
      "reward_std": 0.2041093111038208,
      "rewards/reasoning_reward/mean": 1.2291666269302368,
      "rewards/reasoning_reward/std": 0.48854637145996094,
      "step": 889
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 291.0,
      "completions/max_terminated_length": 291.0,
      "completions/mean_length": 152.5416717529297,
      "completions/mean_terminated_length": 152.5416717529297,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.9203722854188211,
      "grad_norm": 3.725569433437225,
      "kl": 0.07177734375,
      "learning_rate": 1.596800833415135e-08,
      "loss": 0.0029,
      "num_tokens": 72796523.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.2616034746170044,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.4871538281440735,
      "step": 890
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 322.0,
      "completions/max_terminated_length": 322.0,
      "completions/mean_length": 191.5416717529297,
      "completions/mean_terminated_length": 191.5416717529297,
      "completions/min_length": 97.0,
      "completions/min_terminated_length": 97.0,
      "epoch": 0.921406411582213,
      "grad_norm": 2.795098631521861,
      "kl": 0.0556640625,
      "learning_rate": 1.5563320689751192e-08,
      "loss": 0.0022,
      "num_tokens": 72874368.0,
      "reward": 0.8402777910232544,
      "reward_std": 0.38484740257263184,
      "rewards/reasoning_reward/mean": 0.8402777314186096,
      "rewards/reasoning_reward/std": 0.7824759483337402,
      "step": 891
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 299.0,
      "completions/max_terminated_length": 299.0,
      "completions/mean_length": 165.58334350585938,
      "completions/mean_terminated_length": 165.58334350585938,
      "completions/min_length": 90.0,
      "completions/min_terminated_length": 90.0,
      "epoch": 0.922440537745605,
      "grad_norm": 2.3803367727934823,
      "kl": 0.0634765625,
      "learning_rate": 1.5163746135310186e-08,
      "loss": 0.0025,
      "num_tokens": 72953078.0,
      "reward": 0.875,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 892
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.04166666666666663,
      "completions/max_length": 355.0,
      "completions/max_terminated_length": 295.0,
      "completions/mean_length": 193.33334350585938,
      "completions/mean_terminated_length": 186.30435180664062,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.9234746639089969,
      "grad_norm": 3.5618288413220407,
      "kl": 0.057861328125,
      "learning_rate": 1.4769288888222985e-08,
      "loss": 0.0023,
      "num_tokens": 73031438.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3917974829673767,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.5089774131774902,
      "step": 893
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 234.0,
      "completions/max_terminated_length": 234.0,
      "completions/mean_length": 154.0416717529297,
      "completions/mean_terminated_length": 154.0416717529297,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.9245087900723888,
      "grad_norm": 4.201487410221962,
      "kl": 0.11279296875,
      "learning_rate": 1.4379953111872456e-08,
      "loss": 0.0045,
      "num_tokens": 73114015.0,
      "reward": 0.625,
      "reward_std": 0.5280382037162781,
      "rewards/reasoning_reward/mean": 0.625,
      "rewards/reasoning_reward/std": 0.5943574905395508,
      "step": 894
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 321.0,
      "completions/max_terminated_length": 321.0,
      "completions/mean_length": 190.75,
      "completions/mean_terminated_length": 190.75,
      "completions/min_length": 126.0,
      "completions/min_terminated_length": 126.0,
      "epoch": 0.9255429162357808,
      "grad_norm": 1.774495070242477,
      "kl": 0.083984375,
      "learning_rate": 1.3995742915585806e-08,
      "loss": 0.0034,
      "num_tokens": 73203993.0,
      "reward": 1.5208333730697632,
      "reward_std": 0.13908717036247253,
      "rewards/reasoning_reward/mean": 1.5208333730697632,
      "rewards/reasoning_reward/std": 0.47729232907295227,
      "step": 895
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 268.0,
      "completions/max_terminated_length": 268.0,
      "completions/mean_length": 174.7916717529297,
      "completions/mean_terminated_length": 174.7916717529297,
      "completions/min_length": 126.0,
      "completions/min_terminated_length": 126.0,
      "epoch": 0.9265770423991727,
      "grad_norm": 1.9576728453570886,
      "kl": 0.0517578125,
      "learning_rate": 1.3616662354591356e-08,
      "loss": 0.0021,
      "num_tokens": 73284876.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 896
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 366.0,
      "completions/max_terminated_length": 366.0,
      "completions/mean_length": 186.625,
      "completions/mean_terminated_length": 186.625,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.9276111685625646,
      "grad_norm": 3.2466683150458957,
      "kl": 0.07373046875,
      "learning_rate": 1.3242715429975515e-08,
      "loss": 0.003,
      "num_tokens": 73365827.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.7019641399383545,
      "step": 897
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 239.0,
      "completions/max_terminated_length": 239.0,
      "completions/mean_length": 160.9166717529297,
      "completions/mean_terminated_length": 160.9166717529297,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.9286452947259566,
      "grad_norm": 3.1708164450347276,
      "kl": 0.09814453125,
      "learning_rate": 1.2873906088640474e-08,
      "loss": 0.0039,
      "num_tokens": 73443857.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.3116035461425781,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.46576645970344543,
      "step": 898
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 226.0,
      "completions/max_terminated_length": 226.0,
      "completions/mean_length": 172.25,
      "completions/mean_terminated_length": 172.25,
      "completions/min_length": 93.0,
      "completions/min_terminated_length": 93.0,
      "epoch": 0.9296794208893485,
      "grad_norm": 2.0259653414483583,
      "kl": 0.052001953125,
      "learning_rate": 1.251023822326308e-08,
      "loss": 0.0021,
      "num_tokens": 73524719.0,
      "reward": 0.375,
      "reward_std": 0.07715167850255966,
      "rewards/reasoning_reward/mean": 0.375,
      "rewards/reasoning_reward/std": 0.5565811395645142,
      "step": 899
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 260.0,
      "completions/max_terminated_length": 260.0,
      "completions/mean_length": 161.5416717529297,
      "completions/mean_terminated_length": 161.5416717529297,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.9307135470527405,
      "grad_norm": 3.3236866100891342,
      "kl": 0.051025390625,
      "learning_rate": 1.2151715672252983e-08,
      "loss": 0.002,
      "num_tokens": 73601612.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.30416232347488403,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 900
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 201.0,
      "completions/max_terminated_length": 201.0,
      "completions/mean_length": 147.6666717529297,
      "completions/mean_terminated_length": 147.6666717529297,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.9317476732161324,
      "grad_norm": 2.423969131690449,
      "kl": 0.07470703125,
      "learning_rate": 1.179834221971282e-08,
      "loss": 0.003,
      "num_tokens": 73686812.0,
      "reward": 1.2916667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.2916666269302368,
      "rewards/reasoning_reward/std": 0.550032913684845,
      "step": 901
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 140.1666717529297,
      "completions/mean_terminated_length": 140.1666717529297,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.9327817993795243,
      "grad_norm": 3.842346880234536,
      "kl": 0.0615234375,
      "learning_rate": 1.145012159539771e-08,
      "loss": 0.0025,
      "num_tokens": 73762864.0,
      "reward": 0.6041666865348816,
      "reward_std": 0.4130779504776001,
      "rewards/reasoning_reward/mean": 0.6041666865348816,
      "rewards/reasoning_reward/std": 0.6753286719322205,
      "step": 902
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 262.0,
      "completions/max_terminated_length": 262.0,
      "completions/mean_length": 142.08334350585938,
      "completions/mean_terminated_length": 142.08334350585938,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 0.9338159255429163,
      "grad_norm": 3.166928125853971,
      "kl": 0.10595703125,
      "learning_rate": 1.110705747467644e-08,
      "loss": 0.0043,
      "num_tokens": 73844906.0,
      "reward": 0.9652777910232544,
      "reward_std": 0.2400643527507782,
      "rewards/reasoning_reward/mean": 0.9652777314186096,
      "rewards/reasoning_reward/std": 0.7420743107795715,
      "step": 903
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 223.0,
      "completions/max_terminated_length": 223.0,
      "completions/mean_length": 163.1666717529297,
      "completions/mean_terminated_length": 163.1666717529297,
      "completions/min_length": 73.0,
      "completions/min_terminated_length": 73.0,
      "epoch": 0.9348500517063082,
      "grad_norm": 4.500549858418397,
      "kl": 0.1015625,
      "learning_rate": 1.076915347849211e-08,
      "loss": 0.0041,
      "num_tokens": 73922094.0,
      "reward": 1.0,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.2948839068412781,
      "step": 904
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 196.0,
      "completions/max_terminated_length": 196.0,
      "completions/mean_length": 142.6666717529297,
      "completions/mean_terminated_length": 142.6666717529297,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.9358841778697001,
      "grad_norm": 4.169667991174683,
      "kl": 0.10302734375,
      "learning_rate": 1.0436413173324387e-08,
      "loss": 0.0041,
      "num_tokens": 74015702.0,
      "reward": 1.5416667461395264,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.5416666269302368,
      "rewards/reasoning_reward/std": 0.6412736773490906,
      "step": 905
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 298.0,
      "completions/max_terminated_length": 298.0,
      "completions/mean_length": 170.58334350585938,
      "completions/mean_terminated_length": 170.58334350585938,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.9369183040330921,
      "grad_norm": 4.512863571654002,
      "kl": 0.07861328125,
      "learning_rate": 1.0108840071151648e-08,
      "loss": 0.0031,
      "num_tokens": 74091884.0,
      "reward": 1.0625,
      "reward_std": 0.6556500792503357,
      "rewards/reasoning_reward/mean": 1.0625,
      "rewards/reasoning_reward/std": 0.6964584589004517,
      "step": 906
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 263.0,
      "completions/max_terminated_length": 263.0,
      "completions/mean_length": 161.0,
      "completions/mean_terminated_length": 161.0,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.937952430196484,
      "grad_norm": 2.964181511090769,
      "kl": 0.09423828125,
      "learning_rate": 9.786437629413669e-09,
      "loss": 0.0038,
      "num_tokens": 74174724.0,
      "reward": 1.4166667461395264,
      "reward_std": 0.1346571445465088,
      "rewards/reasoning_reward/mean": 1.4166666269302368,
      "rewards/reasoning_reward/std": 0.39927470684051514,
      "step": 907
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 247.0,
      "completions/max_terminated_length": 247.0,
      "completions/mean_length": 159.5,
      "completions/mean_terminated_length": 159.5,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.9389865563598759,
      "grad_norm": 0.17194578664988994,
      "kl": 0.06396484375,
      "learning_rate": 9.469209250975774e-09,
      "loss": 0.0026,
      "num_tokens": 74255520.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 908
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 258.0,
      "completions/max_terminated_length": 258.0,
      "completions/mean_length": 164.5,
      "completions/mean_terminated_length": 164.5,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.9400206825232679,
      "grad_norm": 3.278300391654624,
      "kl": 0.055419921875,
      "learning_rate": 9.157158284092248e-09,
      "loss": 0.0022,
      "num_tokens": 74340668.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.4082186222076416,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 909
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 153.5,
      "completions/mean_terminated_length": 153.5,
      "completions/min_length": 96.0,
      "completions/min_terminated_length": 96.0,
      "epoch": 0.9410548086866598,
      "grad_norm": 3.3566487108649325,
      "kl": 0.055908203125,
      "learning_rate": 8.850288022371478e-09,
      "loss": 0.0022,
      "num_tokens": 74419544.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148510992527008,
      "step": 910
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 166.0,
      "completions/max_terminated_length": 166.0,
      "completions/mean_length": 123.66667175292969,
      "completions/mean_terminated_length": 123.66667175292969,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.9420889348500517,
      "grad_norm": 2.6002648351979984,
      "kl": 0.0478515625,
      "learning_rate": 8.548601704740754e-09,
      "loss": 0.0019,
      "num_tokens": 74500136.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.4148511290550232,
      "step": 911
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 149.875,
      "completions/mean_terminated_length": 149.875,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.9431230610134437,
      "grad_norm": 3.0400992365987864,
      "kl": 0.044189453125,
      "learning_rate": 8.25210251541264e-09,
      "loss": 0.0018,
      "num_tokens": 74576941.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.29602527618408203,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 912
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 254.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 159.25,
      "completions/mean_terminated_length": 159.25,
      "completions/min_length": 57.0,
      "completions/min_terminated_length": 57.0,
      "epoch": 0.9441571871768356,
      "grad_norm": 3.3459982295899047,
      "kl": 0.0830078125,
      "learning_rate": 7.960793583850767e-09,
      "loss": 0.0033,
      "num_tokens": 74667907.0,
      "reward": 1.3194445371627808,
      "reward_std": 0.38191652297973633,
      "rewards/reasoning_reward/mean": 1.3194445371627808,
      "rewards/reasoning_reward/std": 0.5580258965492249,
      "step": 913
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 200.0,
      "completions/max_terminated_length": 200.0,
      "completions/mean_length": 147.20834350585938,
      "completions/mean_terminated_length": 147.20834350585938,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.9451913133402275,
      "grad_norm": 3.428054229956691,
      "kl": 0.08642578125,
      "learning_rate": 7.674677984737255e-09,
      "loss": 0.0035,
      "num_tokens": 74750248.0,
      "reward": 1.1666667461395264,
      "reward_std": 0.3535081148147583,
      "rewards/reasoning_reward/mean": 1.1666666269302368,
      "rewards/reasoning_reward/std": 0.4340573847293854,
      "step": 914
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 213.0,
      "completions/max_terminated_length": 213.0,
      "completions/mean_length": 154.20834350585938,
      "completions/mean_terminated_length": 154.20834350585938,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.9462254395036195,
      "grad_norm": 1.9872121372484721,
      "kl": 0.042724609375,
      "learning_rate": 7.393758737940126e-09,
      "loss": 0.0017,
      "num_tokens": 74829717.0,
      "reward": 0.9583333730697632,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 0.9583333134651184,
      "rewards/reasoning_reward/std": 0.20412415266036987,
      "step": 915
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 132.0,
      "completions/mean_terminated_length": 132.0,
      "completions/min_length": 79.0,
      "completions/min_terminated_length": 79.0,
      "epoch": 0.9472595656670114,
      "grad_norm": 0.19738482494940646,
      "kl": 0.0498046875,
      "learning_rate": 7.1180388084811635e-09,
      "loss": 0.002,
      "num_tokens": 74909541.0,
      "reward": 0.6666666865348816,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 0.6666666865348816,
      "rewards/reasoning_reward/std": 0.4815434217453003,
      "step": 916
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 207.0,
      "completions/max_terminated_length": 207.0,
      "completions/mean_length": 156.08334350585938,
      "completions/mean_terminated_length": 156.08334350585938,
      "completions/min_length": 99.0,
      "completions/min_terminated_length": 99.0,
      "epoch": 0.9482936918304034,
      "grad_norm": 3.2992755635594575,
      "kl": 0.07373046875,
      "learning_rate": 6.847521106505105e-09,
      "loss": 0.0029,
      "num_tokens": 74987655.0,
      "reward": 1.277777910232544,
      "reward_std": 0.33984312415122986,
      "rewards/reasoning_reward/mean": 1.2777777910232544,
      "rewards/reasoning_reward/std": 0.478187620639801,
      "step": 917
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 218.0,
      "completions/max_terminated_length": 218.0,
      "completions/mean_length": 158.75,
      "completions/mean_terminated_length": 158.75,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.9493278179937953,
      "grad_norm": 4.365967227862626,
      "kl": 0.07666015625,
      "learning_rate": 6.582208487248497e-09,
      "loss": 0.0031,
      "num_tokens": 75070441.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.24043500423431396,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.5582062602043152,
      "step": 918
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 170.4166717529297,
      "completions/mean_terminated_length": 170.4166717529297,
      "completions/min_length": 92.0,
      "completions/min_terminated_length": 92.0,
      "epoch": 0.9503619441571872,
      "grad_norm": 2.937333755573114,
      "kl": 0.07470703125,
      "learning_rate": 6.322103751009833e-09,
      "loss": 0.003,
      "num_tokens": 75154867.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.2644323706626892,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.642332136631012,
      "step": 919
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 236.0,
      "completions/max_terminated_length": 236.0,
      "completions/mean_length": 165.875,
      "completions/mean_terminated_length": 165.875,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.9513960703205792,
      "grad_norm": 2.975341812717674,
      "kl": 0.06591796875,
      "learning_rate": 6.067209643119908e-09,
      "loss": 0.0026,
      "num_tokens": 75245144.0,
      "reward": 1.2569444179534912,
      "reward_std": 0.3012464940547943,
      "rewards/reasoning_reward/mean": 1.2569444179534912,
      "rewards/reasoning_reward/std": 0.8708637952804565,
      "step": 920
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 325.0,
      "completions/max_terminated_length": 325.0,
      "completions/mean_length": 160.20834350585938,
      "completions/mean_terminated_length": 160.20834350585938,
      "completions/min_length": 56.0,
      "completions/min_terminated_length": 56.0,
      "epoch": 0.9524301964839711,
      "grad_norm": 4.142950589572342,
      "kl": 0.06640625,
      "learning_rate": 5.817528853912735e-09,
      "loss": 0.0027,
      "num_tokens": 75324685.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 921
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 300.0,
      "completions/max_terminated_length": 300.0,
      "completions/mean_length": 170.58334350585938,
      "completions/mean_terminated_length": 170.58334350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.953464322647363,
      "grad_norm": 3.416223841212196,
      "kl": 0.0537109375,
      "learning_rate": 5.573064018697393e-09,
      "loss": 0.0021,
      "num_tokens": 75404139.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.3268197476863861,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 922
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 362.0,
      "completions/max_terminated_length": 362.0,
      "completions/mean_length": 187.6666717529297,
      "completions/mean_terminated_length": 187.6666717529297,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.954498448810755,
      "grad_norm": 2.3838014741617886,
      "kl": 0.07275390625,
      "learning_rate": 5.333817717729894e-09,
      "loss": 0.0029,
      "num_tokens": 75481619.0,
      "reward": 1.625,
      "reward_std": 0.07715167850255966,
      "rewards/reasoning_reward/mean": 1.625,
      "rewards/reasoning_reward/std": 0.47204458713531494,
      "step": 923
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 235.0,
      "completions/max_terminated_length": 235.0,
      "completions/mean_length": 132.7916717529297,
      "completions/mean_terminated_length": 132.7916717529297,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 0.9555325749741469,
      "grad_norm": 4.457264325946904,
      "kl": 0.08251953125,
      "learning_rate": 5.099792476186249e-09,
      "loss": 0.0033,
      "num_tokens": 75563334.0,
      "reward": 1.125,
      "reward_std": 0.47920867800712585,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.5160468220710754,
      "step": 924
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 170.20834350585938,
      "completions/mean_terminated_length": 170.20834350585938,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.9565667011375388,
      "grad_norm": 5.647062715557137,
      "kl": 0.09619140625,
      "learning_rate": 4.870990764135552e-09,
      "loss": 0.0039,
      "num_tokens": 75647859.0,
      "reward": 0.75,
      "reward_std": 0.510651707649231,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.5316095352172852,
      "step": 925
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 184.0,
      "completions/max_terminated_length": 184.0,
      "completions/mean_length": 126.04167175292969,
      "completions/mean_terminated_length": 126.04167175292969,
      "completions/min_length": 74.0,
      "completions/min_terminated_length": 74.0,
      "epoch": 0.9576008273009308,
      "grad_norm": 3.442181606982195,
      "kl": 0.055908203125,
      "learning_rate": 4.647414996514276e-09,
      "loss": 0.0022,
      "num_tokens": 75723900.0,
      "reward": 1.3958333730697632,
      "reward_std": 0.23144195973873138,
      "rewards/reasoning_reward/mean": 1.3958333730697632,
      "rewards/reasoning_reward/std": 0.4164854884147644,
      "step": 926
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 234.0,
      "completions/max_terminated_length": 234.0,
      "completions/mean_length": 137.83334350585938,
      "completions/mean_terminated_length": 137.83334350585938,
      "completions/min_length": 75.0,
      "completions/min_terminated_length": 75.0,
      "epoch": 0.9586349534643226,
      "grad_norm": 2.5733431583919484,
      "kl": 0.0498046875,
      "learning_rate": 4.429067533100294e-09,
      "loss": 0.002,
      "num_tokens": 75802360.0,
      "reward": 0.7916666865348816,
      "reward_std": 0.3053751587867737,
      "rewards/reasoning_reward/mean": 0.7916666865348816,
      "rewards/reasoning_reward/std": 0.7790276408195496,
      "step": 927
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 277.0,
      "completions/max_terminated_length": 277.0,
      "completions/mean_length": 166.9166717529297,
      "completions/mean_terminated_length": 166.9166717529297,
      "completions/min_length": 73.0,
      "completions/min_terminated_length": 73.0,
      "epoch": 0.9596690796277145,
      "grad_norm": 3.220240618399756,
      "kl": 0.08544921875,
      "learning_rate": 4.2159506784884e-09,
      "loss": 0.0034,
      "num_tokens": 75885278.0,
      "reward": 1.3125,
      "reward_std": 0.22466278076171875,
      "rewards/reasoning_reward/mean": 1.3125,
      "rewards/reasoning_reward/std": 0.4848240315914154,
      "step": 928
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 295.0,
      "completions/max_terminated_length": 295.0,
      "completions/mean_length": 174.08334350585938,
      "completions/mean_terminated_length": 174.08334350585938,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.9607032057911065,
      "grad_norm": 3.6058446603983922,
      "kl": 0.061279296875,
      "learning_rate": 4.0080666820657135e-09,
      "loss": 0.0024,
      "num_tokens": 75970624.0,
      "reward": 0.9548611044883728,
      "reward_std": 0.18729600310325623,
      "rewards/reasoning_reward/mean": 0.9548611044883728,
      "rewards/reasoning_reward/std": 0.6269795298576355,
      "step": 929
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 252.0,
      "completions/max_terminated_length": 252.0,
      "completions/mean_length": 173.5,
      "completions/mean_terminated_length": 173.5,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.9617373319544984,
      "grad_norm": 2.476675970867201,
      "kl": 0.041748046875,
      "learning_rate": 3.805417737988148e-09,
      "loss": 0.0017,
      "num_tokens": 76049588.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.19416078925132751,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.5299029350280762,
      "step": 930
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 243.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 146.7916717529297,
      "completions/mean_terminated_length": 146.7916717529297,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.9627714581178903,
      "grad_norm": 3.988164632708771,
      "kl": 0.12158203125,
      "learning_rate": 3.6080059851570366e-09,
      "loss": 0.0049,
      "num_tokens": 76137775.0,
      "reward": 1.375,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.375,
      "rewards/reasoning_reward/std": 0.494535356760025,
      "step": 931
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 212.0,
      "completions/max_terminated_length": 212.0,
      "completions/mean_length": 159.5416717529297,
      "completions/mean_terminated_length": 159.5416717529297,
      "completions/min_length": 119.0,
      "completions/min_terminated_length": 119.0,
      "epoch": 0.9638055842812823,
      "grad_norm": 1.9405645993516027,
      "kl": 0.04052734375,
      "learning_rate": 3.415833507196764e-09,
      "loss": 0.0016,
      "num_tokens": 76215860.0,
      "reward": 0.5833333730697632,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.5833333134651184,
      "rewards/reasoning_reward/std": 0.5036101937294006,
      "step": 932
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.08333333333333337,
      "completions/max_length": 355.0,
      "completions/max_terminated_length": 355.0,
      "completions/mean_length": 177.08334350585938,
      "completions/mean_terminated_length": 168.68182373046875,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.9648397104446742,
      "grad_norm": 1.732646593438274,
      "kl": 0.045654296875,
      "learning_rate": 3.2289023324325592e-09,
      "loss": 0.0018,
      "num_tokens": 76291502.0,
      "reward": 0.75,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 933
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 214.0,
      "completions/max_terminated_length": 214.0,
      "completions/mean_length": 150.2916717529297,
      "completions/mean_terminated_length": 150.2916717529297,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.9658738366080661,
      "grad_norm": 2.1551686750168,
      "kl": 0.045166015625,
      "learning_rate": 3.0472144338692386e-09,
      "loss": 0.0018,
      "num_tokens": 76368853.0,
      "reward": 0.875,
      "reward_std": 0.17251639068126678,
      "rewards/reasoning_reward/mean": 0.875,
      "rewards/reasoning_reward/std": 0.337831974029541,
      "step": 934
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 292.0,
      "completions/max_terminated_length": 292.0,
      "completions/mean_length": 161.0416717529297,
      "completions/mean_terminated_length": 161.0416717529297,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 0.9669079627714581,
      "grad_norm": 4.651808954793325,
      "kl": 0.064453125,
      "learning_rate": 2.8707717291704405e-09,
      "loss": 0.0026,
      "num_tokens": 76447278.0,
      "reward": 0.5416666865348816,
      "reward_std": 0.42201346158981323,
      "rewards/reasoning_reward/mean": 0.5416666865348816,
      "rewards/reasoning_reward/std": 0.4871537983417511,
      "step": 935
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 254.0,
      "completions/max_terminated_length": 254.0,
      "completions/mean_length": 151.70834350585938,
      "completions/mean_terminated_length": 151.70834350585938,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.96794208893485,
      "grad_norm": 0.17710222219951927,
      "kl": 0.05078125,
      "learning_rate": 2.6995760806381994e-09,
      "loss": 0.002,
      "num_tokens": 76526415.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 936
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 249.0,
      "completions/max_terminated_length": 249.0,
      "completions/mean_length": 163.2916717529297,
      "completions/mean_terminated_length": 163.2916717529297,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.9689762150982419,
      "grad_norm": 4.476052750730717,
      "kl": 0.10498046875,
      "learning_rate": 2.5336292951933513e-09,
      "loss": 0.0042,
      "num_tokens": 76603662.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.4981882572174072,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.5800893306732178,
      "step": 937
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 273.0,
      "completions/max_terminated_length": 273.0,
      "completions/mean_length": 178.0416717529297,
      "completions/mean_terminated_length": 178.0416717529297,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.9700103412616339,
      "grad_norm": 3.9771878211904466,
      "kl": 0.0654296875,
      "learning_rate": 2.372933124356602e-09,
      "loss": 0.0026,
      "num_tokens": 76681175.0,
      "reward": 0.8541666865348816,
      "reward_std": 0.5118739604949951,
      "rewards/reasoning_reward/mean": 0.8541666865348816,
      "rewards/reasoning_reward/std": 0.6507381796836853,
      "step": 938
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 174.0,
      "completions/max_terminated_length": 174.0,
      "completions/mean_length": 125.5,
      "completions/mean_terminated_length": 125.5,
      "completions/min_length": 98.0,
      "completions/min_terminated_length": 98.0,
      "epoch": 0.9710444674250258,
      "grad_norm": 0.8056357466789215,
      "kl": 0.08740234375,
      "learning_rate": 2.2174892642298215e-09,
      "loss": 0.0035,
      "num_tokens": 76766011.0,
      "reward": 1.3333333730697632,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.3333333730697632,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 939
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 197.0,
      "completions/max_terminated_length": 197.0,
      "completions/mean_length": 132.83334350585938,
      "completions/mean_terminated_length": 132.83334350585938,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.9720785935884177,
      "grad_norm": 2.873333533007421,
      "kl": 0.076171875,
      "learning_rate": 2.0672993554783356e-09,
      "loss": 0.003,
      "num_tokens": 76851615.0,
      "reward": 1.0416667461395264,
      "reward_std": 0.1178511306643486,
      "rewards/reasoning_reward/mean": 1.0416666269302368,
      "rewards/reasoning_reward/std": 0.20412413775920868,
      "step": 940
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 279.0,
      "completions/max_terminated_length": 279.0,
      "completions/mean_length": 167.20834350585938,
      "completions/mean_terminated_length": 167.20834350585938,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.9731127197518097,
      "grad_norm": 2.6371055405899173,
      "kl": 0.060546875,
      "learning_rate": 1.9223649833133847e-09,
      "loss": 0.0024,
      "num_tokens": 76931660.0,
      "reward": 0.6527777910232544,
      "reward_std": 0.35072192549705505,
      "rewards/reasoning_reward/mean": 0.6527777314186096,
      "rewards/reasoning_reward/std": 0.533687949180603,
      "step": 941
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 385.0,
      "completions/max_terminated_length": 385.0,
      "completions/mean_length": 171.375,
      "completions/mean_terminated_length": 171.375,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.9741468459152016,
      "grad_norm": 4.0770969322540385,
      "kl": 0.10791015625,
      "learning_rate": 1.782687677475747e-09,
      "loss": 0.0043,
      "num_tokens": 77014581.0,
      "reward": 1.0347223281860352,
      "reward_std": 0.5191335678100586,
      "rewards/reasoning_reward/mean": 1.0347222089767456,
      "rewards/reasoning_reward/std": 0.6862682700157166,
      "step": 942
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 193.0,
      "completions/max_terminated_length": 193.0,
      "completions/mean_length": 136.20834350585938,
      "completions/mean_terminated_length": 136.20834350585938,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 0.9751809720785936,
      "grad_norm": 3.5475163041115847,
      "kl": 0.091796875,
      "learning_rate": 1.6482689122191418e-09,
      "loss": 0.0037,
      "num_tokens": 77096658.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.24056154489517212,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.6672325730323792,
      "step": 943
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 256.0,
      "completions/max_terminated_length": 256.0,
      "completions/mean_length": 164.625,
      "completions/mean_terminated_length": 164.625,
      "completions/min_length": 84.0,
      "completions/min_terminated_length": 84.0,
      "epoch": 0.9762150982419855,
      "grad_norm": 2.9003959757657087,
      "kl": 0.05078125,
      "learning_rate": 1.5191101062950186e-09,
      "loss": 0.002,
      "num_tokens": 77177521.0,
      "reward": 0.9791666865348816,
      "reward_std": 0.2041093111038208,
      "rewards/reasoning_reward/mean": 0.9791666865348816,
      "rewards/reasoning_reward/std": 0.7442411780357361,
      "step": 944
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 177.125,
      "completions/mean_terminated_length": 177.125,
      "completions/min_length": 67.0,
      "completions/min_terminated_length": 67.0,
      "epoch": 0.9772492244053774,
      "grad_norm": 2.6967137886335943,
      "kl": 0.052490234375,
      "learning_rate": 1.3952126229375693e-09,
      "loss": 0.0021,
      "num_tokens": 77257004.0,
      "reward": 0.75,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.75,
      "rewards/reasoning_reward/std": 0.5316095352172852,
      "step": 945
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 201.0,
      "completions/max_terminated_length": 201.0,
      "completions/mean_length": 144.4166717529297,
      "completions/mean_terminated_length": 144.4166717529297,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 0.9782833505687694,
      "grad_norm": 3.0293411969549453,
      "kl": 0.055908203125,
      "learning_rate": 1.2765777698490188e-09,
      "loss": 0.0022,
      "num_tokens": 77339542.0,
      "reward": 0.9166666865348816,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.9166666865348816,
      "rewards/reasoning_reward/std": 0.28232985734939575,
      "step": 946
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 161.9166717529297,
      "completions/mean_terminated_length": 161.9166717529297,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.9793174767321613,
      "grad_norm": 3.148361986897516,
      "kl": 0.044677734375,
      "learning_rate": 1.163206799186245e-09,
      "loss": 0.0018,
      "num_tokens": 77423660.0,
      "reward": 1.2083333730697632,
      "reward_std": 0.2721545100212097,
      "rewards/reasoning_reward/mean": 1.2083333730697632,
      "rewards/reasoning_reward/std": 0.5089773535728455,
      "step": 947
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 206.0,
      "completions/max_terminated_length": 206.0,
      "completions/mean_length": 134.20834350585938,
      "completions/mean_terminated_length": 134.20834350585938,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.9803516028955532,
      "grad_norm": 2.1261463422153004,
      "kl": 0.056396484375,
      "learning_rate": 1.0551009075471795e-09,
      "loss": 0.0023,
      "num_tokens": 77509713.0,
      "reward": 1.25,
      "reward_std": 0.15430335700511932,
      "rewards/reasoning_reward/mean": 1.25,
      "rewards/reasoning_reward/std": 0.4423258602619171,
      "step": 948
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 211.0,
      "completions/max_terminated_length": 211.0,
      "completions/mean_length": 122.29167175292969,
      "completions/mean_terminated_length": 122.29167175292969,
      "completions/min_length": 83.0,
      "completions/min_terminated_length": 83.0,
      "epoch": 0.9813857290589452,
      "grad_norm": 2.9707407010412505,
      "kl": 0.0703125,
      "learning_rate": 9.522612359585402e-10,
      "loss": 0.0028,
      "num_tokens": 77594632.0,
      "reward": 1.1458333730697632,
      "reward_std": 0.16517187654972076,
      "rewards/reasoning_reward/mean": 1.1458333730697632,
      "rewards/reasoning_reward/std": 0.7144345045089722,
      "step": 949
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 134.45834350585938,
      "completions/mean_terminated_length": 134.45834350585938,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.9824198552223371,
      "grad_norm": 3.7149348047540585,
      "kl": 0.060302734375,
      "learning_rate": 8.546888698634513e-10,
      "loss": 0.0024,
      "num_tokens": 77671787.0,
      "reward": 1.4375,
      "reward_std": 0.24232356250286102,
      "rewards/reasoning_reward/mean": 1.4375,
      "rewards/reasoning_reward/std": 0.47348156571388245,
      "step": 950
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 295.0,
      "completions/max_terminated_length": 295.0,
      "completions/mean_length": 175.25,
      "completions/mean_terminated_length": 175.25,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.983453981385729,
      "grad_norm": 3.0064934874302542,
      "kl": 0.06640625,
      "learning_rate": 7.623848391102305e-10,
      "loss": 0.0027,
      "num_tokens": 77750801.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.30860671401023865,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 951
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 199.0,
      "completions/max_terminated_length": 199.0,
      "completions/mean_length": 147.4166717529297,
      "completions/mean_terminated_length": 147.4166717529297,
      "completions/min_length": 85.0,
      "completions/min_terminated_length": 85.0,
      "epoch": 0.984488107549121,
      "grad_norm": 4.044111150534412,
      "kl": 0.0732421875,
      "learning_rate": 6.753501179413423e-10,
      "loss": 0.0029,
      "num_tokens": 77836395.0,
      "reward": 1.125,
      "reward_std": 0.36751919984817505,
      "rewards/reasoning_reward/mean": 1.125,
      "rewards/reasoning_reward/std": 0.6634888052940369,
      "step": 952
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 146.0,
      "completions/mean_terminated_length": 146.0,
      "completions/min_length": 81.0,
      "completions/min_terminated_length": 81.0,
      "epoch": 0.9855222337125129,
      "grad_norm": 3.045502458871555,
      "kl": 0.0859375,
      "learning_rate": 5.935856249833504e-10,
      "loss": 0.0034,
      "num_tokens": 77925307.0,
      "reward": 1.3541667461395264,
      "reward_std": 0.33324795961380005,
      "rewards/reasoning_reward/mean": 1.3541666269302368,
      "rewards/reasoning_reward/std": 0.47729232907295227,
      "step": 953
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 211.0,
      "completions/max_terminated_length": 211.0,
      "completions/mean_length": 154.70834350585938,
      "completions/mean_terminated_length": 154.70834350585938,
      "completions/min_length": 88.0,
      "completions/min_terminated_length": 88.0,
      "epoch": 0.9865563598759048,
      "grad_norm": 0.2945822889932008,
      "kl": 0.052978515625,
      "learning_rate": 5.170922232369257e-10,
      "loss": 0.0021,
      "num_tokens": 78002764.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 954
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 167.0,
      "completions/max_terminated_length": 167.0,
      "completions/mean_length": 130.75,
      "completions/mean_terminated_length": 130.75,
      "completions/min_length": 100.0,
      "completions/min_terminated_length": 100.0,
      "epoch": 0.9875904860392968,
      "grad_norm": 0.1833079894818197,
      "kl": 0.046630859375,
      "learning_rate": 4.4587072006796455e-10,
      "loss": 0.0019,
      "num_tokens": 78081614.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 955
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/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": 152.75,
      "completions/mean_terminated_length": 152.75,
      "completions/min_length": 78.0,
      "completions/min_terminated_length": 78.0,
      "epoch": 0.9886246122026887,
      "grad_norm": 3.4542538672410337,
      "kl": 0.051513671875,
      "learning_rate": 3.7992186719892907e-10,
      "loss": 0.0021,
      "num_tokens": 78160656.0,
      "reward": 0.3333333432674408,
      "reward_std": 0.2357022613286972,
      "rewards/reasoning_reward/mean": 0.3333333432674408,
      "rewards/reasoning_reward/std": 0.4815433919429779,
      "step": 956
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 314.0,
      "completions/max_terminated_length": 314.0,
      "completions/mean_length": 182.625,
      "completions/mean_terminated_length": 182.625,
      "completions/min_length": 123.0,
      "completions/min_terminated_length": 123.0,
      "epoch": 0.9896587383660806,
      "grad_norm": 3.6021539388001673,
      "kl": 0.10205078125,
      "learning_rate": 3.1924636070107535e-10,
      "loss": 0.0041,
      "num_tokens": 78238399.0,
      "reward": 0.7708333730697632,
      "reward_std": 0.31726133823394775,
      "rewards/reasoning_reward/mean": 0.7708333134651184,
      "rewards/reasoning_reward/std": 0.416485458612442,
      "step": 957
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 243.0,
      "completions/max_terminated_length": 243.0,
      "completions/mean_length": 159.9166717529297,
      "completions/mean_terminated_length": 159.9166717529297,
      "completions/min_length": 77.0,
      "completions/min_terminated_length": 77.0,
      "epoch": 0.9906928645294726,
      "grad_norm": 3.4288036998953846,
      "kl": 0.0771484375,
      "learning_rate": 2.6384484098690427e-10,
      "loss": 0.0031,
      "num_tokens": 78317653.0,
      "reward": 1.3958333730697632,
      "reward_std": 0.29116004705429077,
      "rewards/reasoning_reward/mean": 1.3958333730697632,
      "rewards/reasoning_reward/std": 0.4657664895057678,
      "step": 958
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 250.0,
      "completions/max_terminated_length": 250.0,
      "completions/mean_length": 157.6666717529297,
      "completions/mean_terminated_length": 157.6666717529297,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.9917269906928645,
      "grad_norm": 3.8235718441583564,
      "kl": 0.09716796875,
      "learning_rate": 2.1371789280355547e-10,
      "loss": 0.0039,
      "num_tokens": 78400205.0,
      "reward": 1.1041667461395264,
      "reward_std": 0.3190067708492279,
      "rewards/reasoning_reward/mean": 1.1041666269302368,
      "rewards/reasoning_reward/std": 0.8072924017906189,
      "step": 959
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 232.0,
      "completions/max_terminated_length": 232.0,
      "completions/mean_length": 145.7916717529297,
      "completions/mean_terminated_length": 145.7916717529297,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.9927611168562565,
      "grad_norm": 3.9856327424940687,
      "kl": 0.083984375,
      "learning_rate": 1.6886604522659e-10,
      "loss": 0.0034,
      "num_tokens": 78481128.0,
      "reward": 1.0,
      "reward_std": 0.5582748055458069,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.7939992547035217,
      "step": 960
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 297.0,
      "completions/max_terminated_length": 297.0,
      "completions/mean_length": 163.9166717529297,
      "completions/mean_terminated_length": 163.9166717529297,
      "completions/min_length": 71.0,
      "completions/min_terminated_length": 71.0,
      "epoch": 0.9937952430196484,
      "grad_norm": 0.2081115188693481,
      "kl": 0.06591796875,
      "learning_rate": 1.292897716542729e-10,
      "loss": 0.0026,
      "num_tokens": 78560702.0,
      "reward": 1.0,
      "reward_std": 0.0,
      "rewards/reasoning_reward/mean": 1.0,
      "rewards/reasoning_reward/std": 0.0,
      "step": 961
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 201.0,
      "completions/max_terminated_length": 201.0,
      "completions/mean_length": 168.70834350585938,
      "completions/mean_terminated_length": 168.70834350585938,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.9948293691830403,
      "grad_norm": 2.780255312640307,
      "kl": 0.062255859375,
      "learning_rate": 9.498948980291021e-11,
      "loss": 0.0025,
      "num_tokens": 78643983.0,
      "reward": 1.3125,
      "reward_std": 0.20665977895259857,
      "rewards/reasoning_reward/mean": 1.3125,
      "rewards/reasoning_reward/std": 0.355469673871994,
      "step": 962
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 281.0,
      "completions/max_terminated_length": 281.0,
      "completions/mean_length": 149.75,
      "completions/mean_terminated_length": 149.75,
      "completions/min_length": 57.0,
      "completions/min_terminated_length": 57.0,
      "epoch": 0.9958634953464323,
      "grad_norm": 3.0502524409622622,
      "kl": 0.046142578125,
      "learning_rate": 6.59655617020749e-11,
      "loss": 0.0018,
      "num_tokens": 78724921.0,
      "reward": 1.0208333730697632,
      "reward_std": 0.2965203523635864,
      "rewards/reasoning_reward/mean": 1.0208333730697632,
      "rewards/reasoning_reward/std": 0.6507381200790405,
      "step": 963
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 226.0,
      "completions/max_terminated_length": 226.0,
      "completions/mean_length": 132.08334350585938,
      "completions/mean_terminated_length": 132.08334350585938,
      "completions/min_length": 89.0,
      "completions/min_terminated_length": 89.0,
      "epoch": 0.9968976215098242,
      "grad_norm": 2.2883893243678246,
      "kl": 0.10888671875,
      "learning_rate": 4.221829369094321e-11,
      "loss": 0.0043,
      "num_tokens": 78809019.0,
      "reward": 0.8333333730697632,
      "reward_std": 0.17817416787147522,
      "rewards/reasoning_reward/mean": 0.8333333134651184,
      "rewards/reasoning_reward/std": 0.3806934952735901,
      "step": 964
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 203.0,
      "completions/max_terminated_length": 203.0,
      "completions/mean_length": 140.95834350585938,
      "completions/mean_terminated_length": 140.95834350585938,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 0.9979317476732161,
      "grad_norm": 3.6433341964307564,
      "kl": 0.08447265625,
      "learning_rate": 2.374793641518602e-11,
      "loss": 0.0034,
      "num_tokens": 78883898.0,
      "reward": 0.7013888359069824,
      "reward_std": 0.3162800669670105,
      "rewards/reasoning_reward/mean": 0.7013888359069824,
      "rewards/reasoning_reward/std": 0.7222802639007568,
      "step": 965
    },
    {
      "clip_ratio/high_max": 0.0,
      "clip_ratio/high_mean": 0.0,
      "clip_ratio/low_mean": 0.0,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0,
      "completions/clipped_ratio": 0.0,
      "completions/max_length": 446.0,
      "completions/max_terminated_length": 446.0,
      "completions/mean_length": 190.4499969482422,
      "completions/mean_terminated_length": 190.4499969482422,
      "completions/min_length": 114.0,
      "completions/min_terminated_length": 114.0,
      "epoch": 0.9989658738366081,
      "grad_norm": 2.642107476748316,
      "kl": 0.05419921875,
      "learning_rate": 1.0554684824137794e-11,
      "loss": 0.0024,
      "num_tokens": 78951363.0,
      "reward": 1.0833333730697632,
      "reward_std": 0.2903675436973572,
      "rewards/reasoning_reward/mean": 1.0833333730697632,
      "rewards/reasoning_reward/std": 0.7172815203666687,
      "step": 966
    },
    {
      "epoch": 0.9989658738366081,
      "step": 966,
      "total_flos": 0.0,
      "train_loss": 0.002488073633344592,
      "train_runtime": 159825.8329,
      "train_samples_per_second": 0.018,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 967,
  "num_input_tokens_seen": 78951363,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}