{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.73724117818606,
  "eval_steps": 500,
  "global_step": 1600,
  "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": 0.0200892857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3119.0,
      "completions/mean_length": 606.5625,
      "completions/mean_terminated_length": 535.0250854492188,
      "completions/min_length": 22.0,
      "completions/min_terminated_length": 22.0,
      "epoch": 0.0023330417031204435,
      "grad_norm": 0.139825239777565,
      "learning_rate": 1e-06,
      "loss": -0.0053,
      "num_tokens": 556256.0,
      "reward": 0.5424107313156128,
      "reward_std": 0.24488291144371033,
      "rewards/verify_math_reward/mean": 0.5424107313156128,
      "rewards/verify_math_reward/std": 0.4984763562679291,
      "step": 1
    },
    {
      "clip_ratio/high_max": 0.0022986409749137238,
      "clip_ratio/high_mean": 0.0010865736549021676,
      "clip_ratio/low_mean": 0.0006070504496165086,
      "clip_ratio/low_min": 3.928235310013406e-05,
      "clip_ratio/region_mean": 0.0016936241154326126,
      "epoch": 0.004666083406240887,
      "grad_norm": 0.13357040286064148,
      "learning_rate": 1e-06,
      "loss": -0.0052,
      "step": 2
    },
    {
      "clip_ratio/high_max": 0.002638528043462429,
      "clip_ratio/high_mean": 0.0011571372342586983,
      "clip_ratio/low_mean": 0.0006701715410599718,
      "clip_ratio/low_min": 7.16715467206086e-05,
      "clip_ratio/region_mean": 0.00182730880624149,
      "epoch": 0.00699912510936133,
      "grad_norm": 0.12927649915218353,
      "learning_rate": 1e-06,
      "loss": -0.0053,
      "step": 3
    },
    {
      "clip_ratio/high_max": 0.002463254946633242,
      "clip_ratio/high_mean": 0.0010911843601206783,
      "clip_ratio/low_mean": 0.0006269629561757029,
      "clip_ratio/low_min": 5.3640959777112585e-05,
      "clip_ratio/region_mean": 0.0017181472649099305,
      "epoch": 0.009332166812481774,
      "grad_norm": 0.13325555622577667,
      "learning_rate": 1e-06,
      "loss": -0.0053,
      "step": 4
    },
    {
      "clip_ratio/high_max": 0.002511580994905671,
      "clip_ratio/high_mean": 0.0010959685314446688,
      "clip_ratio/low_mean": 0.0007915860005596187,
      "clip_ratio/low_min": 5.303173657011939e-05,
      "clip_ratio/region_mean": 0.0018875545429182239,
      "completions/clipped_ratio": 0.010044642857142905,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2607.0,
      "completions/mean_length": 590.552490234375,
      "completions/mean_terminated_length": 554.9841918945312,
      "completions/min_length": 68.0,
      "completions/min_terminated_length": 68.0,
      "epoch": 0.011665208515602217,
      "grad_norm": 0.1198158785700798,
      "learning_rate": 1e-06,
      "loss": 0.0037,
      "num_tokens": 1135783.0,
      "reward": 0.4832589626312256,
      "reward_std": 0.2370942384004593,
      "rewards/verify_math_reward/mean": 0.4832589328289032,
      "rewards/verify_math_reward/std": 0.4999987483024597,
      "step": 5
    },
    {
      "clip_ratio/high_max": 0.0023428853455698118,
      "clip_ratio/high_mean": 0.0009512929136690218,
      "clip_ratio/low_mean": 0.0005539788216992747,
      "clip_ratio/low_min": 3.876696337101748e-05,
      "clip_ratio/region_mean": 0.0015052717571961693,
      "epoch": 0.01399825021872266,
      "grad_norm": 0.12222807109355927,
      "learning_rate": 1e-06,
      "loss": 0.0036,
      "step": 6
    },
    {
      "clip_ratio/high_max": 0.0023241781964316033,
      "clip_ratio/high_mean": 0.0010337102903577033,
      "clip_ratio/low_mean": 0.0005319816227711271,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001565691884025,
      "epoch": 0.016331291921843103,
      "grad_norm": 0.1279895007610321,
      "learning_rate": 1e-06,
      "loss": 0.0036,
      "step": 7
    },
    {
      "clip_ratio/high_max": 0.002724015103012789,
      "clip_ratio/high_mean": 0.0011034075978386682,
      "clip_ratio/low_mean": 0.0006094714681239566,
      "clip_ratio/low_min": 2.932166989921825e-05,
      "clip_ratio/region_mean": 0.0017128790204878896,
      "epoch": 0.018664333624963548,
      "grad_norm": 0.12145639955997467,
      "learning_rate": 1e-06,
      "loss": 0.0036,
      "step": 8
    },
    {
      "clip_ratio/high_max": 0.0023459133226424456,
      "clip_ratio/high_mean": 0.0009634899925003992,
      "clip_ratio/low_mean": 0.0006493575965578202,
      "clip_ratio/low_min": 1.1285662367299665e-05,
      "clip_ratio/region_mean": 0.0016128476345329545,
      "completions/clipped_ratio": 0.012276785714285698,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2327.0,
      "completions/mean_length": 581.685302734375,
      "completions/mean_terminated_length": 538.0045166015625,
      "completions/min_length": 24.0,
      "completions/min_terminated_length": 24.0,
      "epoch": 0.02099737532808399,
      "grad_norm": 0.11929132789373398,
      "learning_rate": 1e-06,
      "loss": 0.0097,
      "num_tokens": 1710461.0,
      "reward": 0.5290178656578064,
      "reward_std": 0.21602025628089905,
      "rewards/verify_math_reward/mean": 0.5290178656578064,
      "rewards/verify_math_reward/std": 0.49943605065345764,
      "step": 9
    },
    {
      "clip_ratio/high_max": 0.00253376059117727,
      "clip_ratio/high_mean": 0.0010959464925690554,
      "clip_ratio/low_mean": 0.0009177182091661962,
      "clip_ratio/low_min": 6.172536632220726e-05,
      "clip_ratio/region_mean": 0.0020136647290200926,
      "epoch": 0.023330417031204434,
      "grad_norm": 0.11741513758897781,
      "learning_rate": 1e-06,
      "loss": 0.0096,
      "step": 10
    },
    {
      "clip_ratio/high_max": 0.0021746221245848574,
      "clip_ratio/high_mean": 0.0009437728404009249,
      "clip_ratio/low_mean": 0.0006222142146725673,
      "clip_ratio/low_min": 3.810313319263514e-05,
      "clip_ratio/region_mean": 0.0015659870696254075,
      "epoch": 0.025663458734324875,
      "grad_norm": 0.12099823355674744,
      "learning_rate": 1e-06,
      "loss": 0.0096,
      "step": 11
    },
    {
      "clip_ratio/high_max": 0.002284212321683299,
      "clip_ratio/high_mean": 0.0009622946818126366,
      "clip_ratio/low_mean": 0.0006222454194357852,
      "clip_ratio/low_min": 3.2773188650025986e-05,
      "clip_ratio/region_mean": 0.0015845401139813475,
      "epoch": 0.02799650043744532,
      "grad_norm": 0.12343762069940567,
      "learning_rate": 1e-06,
      "loss": 0.0096,
      "step": 12
    },
    {
      "clip_ratio/high_max": 0.002133540387148969,
      "clip_ratio/high_mean": 0.0008776721952017397,
      "clip_ratio/low_mean": 0.0004960685146215837,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013737407025473658,
      "completions/clipped_ratio": 0.014508928571428603,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4078.0,
      "completions/mean_length": 587.7221069335938,
      "completions/mean_terminated_length": 536.0713500976562,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 0.03032954214056576,
      "grad_norm": 0.1237315759062767,
      "learning_rate": 1e-06,
      "loss": -0.008,
      "num_tokens": 2263244.0,
      "reward": 0.6004464626312256,
      "reward_std": 0.19208844006061554,
      "rewards/verify_math_reward/mean": 0.6004464030265808,
      "rewards/verify_math_reward/std": 0.49008017778396606,
      "step": 13
    },
    {
      "clip_ratio/high_max": 0.002390783491136972,
      "clip_ratio/high_mean": 0.0009201841166941449,
      "clip_ratio/low_mean": 0.0005904453855691827,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015106295177247375,
      "epoch": 0.032662583843686206,
      "grad_norm": 0.1241251677274704,
      "learning_rate": 1e-06,
      "loss": -0.0081,
      "step": 14
    },
    {
      "clip_ratio/high_max": 0.00265116490481887,
      "clip_ratio/high_mean": 0.0010003927927755285,
      "clip_ratio/low_mean": 0.0005856096286152024,
      "clip_ratio/low_min": 1.5866971807554364e-05,
      "clip_ratio/region_mean": 0.0015860024504945613,
      "epoch": 0.03499562554680665,
      "grad_norm": 0.1218414381146431,
      "learning_rate": 1e-06,
      "loss": -0.0081,
      "step": 15
    },
    {
      "clip_ratio/high_max": 0.002377994002017658,
      "clip_ratio/high_mean": 0.0008695639698999003,
      "clip_ratio/low_mean": 0.0005416954845713917,
      "clip_ratio/low_min": 1.2747297660098411e-05,
      "clip_ratio/region_mean": 0.0014112594762991648,
      "epoch": 0.037328667249927096,
      "grad_norm": 0.12389780580997467,
      "learning_rate": 1e-06,
      "loss": -0.0081,
      "step": 16
    },
    {
      "clip_ratio/high_max": 0.0024106820128508843,
      "clip_ratio/high_mean": 0.0009968766898964532,
      "clip_ratio/low_mean": 0.0006063756100047613,
      "clip_ratio/low_min": 1.5118529518076684e-05,
      "clip_ratio/region_mean": 0.0016032523271860555,
      "completions/clipped_ratio": 0.006696428571428603,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2943.0,
      "completions/mean_length": 569.2154541015625,
      "completions/mean_terminated_length": 545.4393310546875,
      "completions/min_length": 19.0,
      "completions/min_terminated_length": 19.0,
      "epoch": 0.039661708953047534,
      "grad_norm": 0.12719471752643585,
      "learning_rate": 1e-06,
      "loss": 0.0134,
      "num_tokens": 2832253.0,
      "reward": 0.5446428656578064,
      "reward_std": 0.2242865115404129,
      "rewards/verify_math_reward/mean": 0.5446428656578064,
      "rewards/verify_math_reward/std": 0.49828118085861206,
      "step": 17
    },
    {
      "clip_ratio/high_max": 0.0025036680308403447,
      "clip_ratio/high_mean": 0.0010271042483509518,
      "clip_ratio/low_mean": 0.0006537355602631578,
      "clip_ratio/low_min": 2.4149921955540776e-05,
      "clip_ratio/region_mean": 0.0016808397995191626,
      "epoch": 0.04199475065616798,
      "grad_norm": 0.12676291167736053,
      "learning_rate": 1e-06,
      "loss": 0.0134,
      "step": 18
    },
    {
      "clip_ratio/high_max": 0.0027959811995970085,
      "clip_ratio/high_mean": 0.0011196248633495998,
      "clip_ratio/low_mean": 0.000717350643753889,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018369754980085418,
      "epoch": 0.04432779235928842,
      "grad_norm": 0.12739385664463043,
      "learning_rate": 1e-06,
      "loss": 0.0133,
      "step": 19
    },
    {
      "clip_ratio/high_max": 0.0029148588073439896,
      "clip_ratio/high_mean": 0.001127249219280202,
      "clip_ratio/low_mean": 0.0007854968207539059,
      "clip_ratio/low_min": 6.304252019617707e-05,
      "clip_ratio/region_mean": 0.0019127460473100655,
      "epoch": 0.04666083406240887,
      "grad_norm": 0.12382540851831436,
      "learning_rate": 1e-06,
      "loss": 0.0132,
      "step": 20
    },
    {
      "clip_ratio/high_max": 0.0019592746248235926,
      "clip_ratio/high_mean": 0.0008812815285637043,
      "clip_ratio/low_mean": 0.0007547873065050226,
      "clip_ratio/low_min": 7.309357533813454e-05,
      "clip_ratio/region_mean": 0.0016360688241547905,
      "completions/clipped_ratio": 0.017857142857142905,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3173.0,
      "completions/mean_length": 622.6663208007812,
      "completions/mean_terminated_length": 559.5147705078125,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 0.048993875765529306,
      "grad_norm": 0.13165442645549774,
      "learning_rate": 1e-06,
      "loss": -0.0024,
      "num_tokens": 3415050.0,
      "reward": 0.5245535969734192,
      "reward_std": 0.24107471108436584,
      "rewards/verify_math_reward/mean": 0.5245535969734192,
      "rewards/verify_math_reward/std": 0.4996756613254547,
      "step": 21
    },
    {
      "clip_ratio/high_max": 0.0024175399157684296,
      "clip_ratio/high_mean": 0.0009496703860349953,
      "clip_ratio/low_mean": 0.00075606488280755,
      "clip_ratio/low_min": 5.7472104344924446e-05,
      "clip_ratio/region_mean": 0.0017057352742995135,
      "epoch": 0.05132691746864975,
      "grad_norm": 0.13229161500930786,
      "learning_rate": 1e-06,
      "loss": -0.0024,
      "step": 22
    },
    {
      "clip_ratio/high_max": 0.0022421761314035393,
      "clip_ratio/high_mean": 0.0009404981101397425,
      "clip_ratio/low_mean": 0.0007850584315747255,
      "clip_ratio/low_min": 6.583995582332136e-05,
      "clip_ratio/region_mean": 0.0017255565471714363,
      "epoch": 0.053659959171770195,
      "grad_norm": 0.12714257836341858,
      "learning_rate": 1e-06,
      "loss": -0.0026,
      "step": 23
    },
    {
      "clip_ratio/high_max": 0.002484742122760508,
      "clip_ratio/high_mean": 0.0010370668096584268,
      "clip_ratio/low_mean": 0.0007963930911500938,
      "clip_ratio/low_min": 6.353132994263433e-05,
      "clip_ratio/region_mean": 0.0018334599080844782,
      "epoch": 0.05599300087489064,
      "grad_norm": 0.13083310425281525,
      "learning_rate": 1e-06,
      "loss": -0.0026,
      "step": 24
    },
    {
      "clip_ratio/high_max": 0.002489036211045459,
      "clip_ratio/high_mean": 0.0010009248289861716,
      "clip_ratio/low_mean": 0.0006285366598604014,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016294614688376896,
      "completions/clipped_ratio": 0.012276785714285698,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4078.0,
      "completions/mean_length": 580.6439819335938,
      "completions/mean_terminated_length": 536.9503173828125,
      "completions/min_length": 108.0,
      "completions/min_terminated_length": 108.0,
      "epoch": 0.058326042578011085,
      "grad_norm": 0.12958918511867523,
      "learning_rate": 1e-06,
      "loss": 0.0058,
      "num_tokens": 3988707.0,
      "reward": 0.5892857313156128,
      "reward_std": 0.20095311105251312,
      "rewards/verify_math_reward/mean": 0.5892857313156128,
      "rewards/verify_math_reward/std": 0.49223825335502625,
      "step": 25
    },
    {
      "clip_ratio/high_max": 0.0024675652748555876,
      "clip_ratio/high_mean": 0.0009995967520808335,
      "clip_ratio/low_mean": 0.0006661116531176958,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001665708430664381,
      "epoch": 0.06065908428113152,
      "grad_norm": 0.12610486149787903,
      "learning_rate": 1e-06,
      "loss": 0.0056,
      "step": 26
    },
    {
      "clip_ratio/high_max": 0.002277686129673384,
      "clip_ratio/high_mean": 0.0009735188032209408,
      "clip_ratio/low_mean": 0.0006767354534531478,
      "clip_ratio/low_min": 3.548448148649186e-05,
      "clip_ratio/region_mean": 0.001650254249398131,
      "epoch": 0.06299212598425197,
      "grad_norm": 0.1267174482345581,
      "learning_rate": 1e-06,
      "loss": 0.0057,
      "step": 27
    },
    {
      "clip_ratio/high_max": 0.002439399620925542,
      "clip_ratio/high_mean": 0.0010170852947339881,
      "clip_ratio/low_mean": 0.0007562009886896703,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001773286254319828,
      "epoch": 0.06532516768737241,
      "grad_norm": 0.12486453354358673,
      "learning_rate": 1e-06,
      "loss": 0.0055,
      "step": 28
    },
    {
      "clip_ratio/high_max": 0.0021370626272982918,
      "clip_ratio/high_mean": 0.0008809797818685183,
      "clip_ratio/low_mean": 0.0005185944191907765,
      "clip_ratio/low_min": 3.953835221182089e-05,
      "clip_ratio/region_mean": 0.0013995742046972737,
      "completions/clipped_ratio": 0.021205357142857095,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3560.0,
      "completions/mean_length": 678.6060791015625,
      "completions/mean_terminated_length": 604.5689697265625,
      "completions/min_length": 94.0,
      "completions/min_terminated_length": 94.0,
      "epoch": 0.06765820939049286,
      "grad_norm": 0.11682160198688507,
      "learning_rate": 1e-06,
      "loss": 0.0046,
      "num_tokens": 4593306.0,
      "reward": 0.5457589626312256,
      "reward_std": 0.21774594485759735,
      "rewards/verify_math_reward/mean": 0.5457589030265808,
      "rewards/verify_math_reward/std": 0.4981797933578491,
      "step": 29
    },
    {
      "clip_ratio/high_max": 0.0021843241120222956,
      "clip_ratio/high_mean": 0.0009300625351897907,
      "clip_ratio/low_mean": 0.0006210761221154826,
      "clip_ratio/low_min": 4.4221876123629045e-05,
      "clip_ratio/region_mean": 0.0015511386664002202,
      "epoch": 0.0699912510936133,
      "grad_norm": 0.1146954670548439,
      "learning_rate": 1e-06,
      "loss": 0.0045,
      "step": 30
    },
    {
      "clip_ratio/high_max": 0.0024472579316352494,
      "clip_ratio/high_mean": 0.0010522642951400485,
      "clip_ratio/low_mean": 0.0005727457664761459,
      "clip_ratio/low_min": 3.6030963201483246e-05,
      "clip_ratio/region_mean": 0.001625010023417417,
      "epoch": 0.07232429279673375,
      "grad_norm": 0.1153184175491333,
      "learning_rate": 1e-06,
      "loss": 0.0044,
      "step": 31
    },
    {
      "clip_ratio/high_max": 0.0019284382869955152,
      "clip_ratio/high_mean": 0.000992046516330447,
      "clip_ratio/low_mean": 0.0006042511531632044,
      "clip_ratio/low_min": 2.522551039874088e-05,
      "clip_ratio/region_mean": 0.0015962977122399025,
      "epoch": 0.07465733449985419,
      "grad_norm": 0.1235361248254776,
      "learning_rate": 1e-06,
      "loss": 0.0044,
      "step": 32
    },
    {
      "clip_ratio/high_max": 0.0021176241425564513,
      "clip_ratio/high_mean": 0.0009372366621391848,
      "clip_ratio/low_mean": 0.0005595719358097995,
      "clip_ratio/low_min": 6.470445987361018e-05,
      "clip_ratio/region_mean": 0.0014968086179578677,
      "completions/clipped_ratio": 0.030133928571428603,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4067.0,
      "completions/mean_length": 700.318115234375,
      "completions/mean_terminated_length": 594.8135986328125,
      "completions/min_length": 147.0,
      "completions/min_terminated_length": 147.0,
      "epoch": 0.07699037620297462,
      "grad_norm": 0.12378786504268646,
      "learning_rate": 1e-06,
      "loss": -0.0102,
      "num_tokens": 5199175.0,
      "reward": 0.5412946939468384,
      "reward_std": 0.21545571088790894,
      "rewards/verify_math_reward/mean": 0.5412946343421936,
      "rewards/verify_math_reward/std": 0.49857014417648315,
      "step": 33
    },
    {
      "clip_ratio/high_max": 0.0022196359786903486,
      "clip_ratio/high_mean": 0.0009039962114911759,
      "clip_ratio/low_mean": 0.000660695368424058,
      "clip_ratio/low_min": 3.265412669861689e-05,
      "clip_ratio/region_mean": 0.0015646915635443293,
      "epoch": 0.07932341790609507,
      "grad_norm": 0.125546395778656,
      "learning_rate": 1e-06,
      "loss": -0.0103,
      "step": 34
    },
    {
      "clip_ratio/high_max": 0.0023568391989101656,
      "clip_ratio/high_mean": 0.0009934375430020737,
      "clip_ratio/low_mean": 0.0006520127171825152,
      "clip_ratio/low_min": 4.448789968591882e-05,
      "clip_ratio/region_mean": 0.0016454502547276206,
      "epoch": 0.08165645960921551,
      "grad_norm": 0.11966123431921005,
      "learning_rate": 1e-06,
      "loss": -0.0104,
      "step": 35
    },
    {
      "clip_ratio/high_max": 0.002292566980031552,
      "clip_ratio/high_mean": 0.0009935368743754225,
      "clip_ratio/low_mean": 0.0006416244887077482,
      "clip_ratio/low_min": 7.534917676821351e-05,
      "clip_ratio/region_mean": 0.0016351613521692343,
      "epoch": 0.08398950131233596,
      "grad_norm": 0.12310697883367538,
      "learning_rate": 1e-06,
      "loss": -0.0103,
      "step": 36
    },
    {
      "clip_ratio/high_max": 0.00229012560157571,
      "clip_ratio/high_mean": 0.0009375477602588944,
      "clip_ratio/low_mean": 0.0007024908227322157,
      "clip_ratio/low_min": 2.9890166842960753e-05,
      "clip_ratio/region_mean": 0.0016400385648012161,
      "completions/clipped_ratio": 0.014508928571428603,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2962.0,
      "completions/mean_length": 629.1998291015625,
      "completions/mean_terminated_length": 578.15966796875,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 0.0863225430154564,
      "grad_norm": 0.12611764669418335,
      "learning_rate": 1e-06,
      "loss": 0.0144,
      "num_tokens": 5804930.0,
      "reward": 0.5290178656578064,
      "reward_std": 0.23063203692436218,
      "rewards/verify_math_reward/mean": 0.5290178656578064,
      "rewards/verify_math_reward/std": 0.49943605065345764,
      "step": 37
    },
    {
      "clip_ratio/high_max": 0.0026505103378440253,
      "clip_ratio/high_mean": 0.000995517555566039,
      "clip_ratio/low_mean": 0.000749046572309453,
      "clip_ratio/low_min": 5.4068157623987645e-05,
      "clip_ratio/region_mean": 0.0017445641424274072,
      "epoch": 0.08865558471857685,
      "grad_norm": 0.12353263795375824,
      "learning_rate": 1e-06,
      "loss": 0.0143,
      "step": 38
    },
    {
      "clip_ratio/high_max": 0.002375702024437487,
      "clip_ratio/high_mean": 0.0009863118175417185,
      "clip_ratio/low_mean": 0.0008022430338314734,
      "clip_ratio/low_min": 3.6729577914229594e-05,
      "clip_ratio/region_mean": 0.0017885548804770224,
      "epoch": 0.09098862642169729,
      "grad_norm": 0.12338662147521973,
      "learning_rate": 1e-06,
      "loss": 0.0143,
      "step": 39
    },
    {
      "clip_ratio/high_max": 0.0024909446510719135,
      "clip_ratio/high_mean": 0.0010256626637783484,
      "clip_ratio/low_mean": 0.000834449481772026,
      "clip_ratio/low_min": 6.956895595067181e-05,
      "clip_ratio/region_mean": 0.001860112141002901,
      "epoch": 0.09332166812481774,
      "grad_norm": 0.12385766208171844,
      "learning_rate": 1e-06,
      "loss": 0.0142,
      "step": 40
    },
    {
      "clip_ratio/high_max": 0.001583784669492161,
      "clip_ratio/high_mean": 0.0005768877708760556,
      "clip_ratio/low_mean": 0.0005286252780933864,
      "clip_ratio/low_min": 1.4124293556960765e-05,
      "clip_ratio/region_mean": 0.0011055130635213573,
      "completions/clipped_ratio": 0.017857142857142905,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2063.0,
      "completions/mean_length": 667.3471069335938,
      "completions/mean_terminated_length": 605.0079345703125,
      "completions/min_length": 130.0,
      "completions/min_terminated_length": 130.0,
      "epoch": 0.09565470982793818,
      "grad_norm": 0.1104072779417038,
      "learning_rate": 1e-06,
      "loss": 0.0185,
      "num_tokens": 6421681.0,
      "reward": 0.5066964626312256,
      "reward_std": 0.1720547378063202,
      "rewards/verify_math_reward/mean": 0.5066964030265808,
      "rewards/verify_math_reward/std": 0.5002344250679016,
      "step": 41
    },
    {
      "clip_ratio/high_max": 0.0020037883587065153,
      "clip_ratio/high_mean": 0.0006965057691559196,
      "clip_ratio/low_mean": 0.0006347013313643401,
      "clip_ratio/low_min": 2.8372424822009634e-05,
      "clip_ratio/region_mean": 0.001331207105977228,
      "epoch": 0.09798775153105861,
      "grad_norm": 0.11040699481964111,
      "learning_rate": 1e-06,
      "loss": 0.0184,
      "step": 42
    },
    {
      "clip_ratio/high_max": 0.0018374437386228237,
      "clip_ratio/high_mean": 0.0006500711151602445,
      "clip_ratio/low_mean": 0.0005802347404824104,
      "clip_ratio/low_min": 1.0425354048493318e-05,
      "clip_ratio/region_mean": 0.001230305842909729,
      "epoch": 0.10032079323417906,
      "grad_norm": 0.10948064178228378,
      "learning_rate": 1e-06,
      "loss": 0.0184,
      "step": 43
    },
    {
      "clip_ratio/high_max": 0.0018857809591281693,
      "clip_ratio/high_mean": 0.0006436912299250253,
      "clip_ratio/low_mean": 0.000669299080982455,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013129903072695015,
      "epoch": 0.1026538349372995,
      "grad_norm": 0.1097576767206192,
      "learning_rate": 1e-06,
      "loss": 0.0183,
      "step": 44
    },
    {
      "clip_ratio/high_max": 0.001930827418618719,
      "clip_ratio/high_mean": 0.0006981309613820486,
      "clip_ratio/low_mean": 0.0005967152856101166,
      "clip_ratio/low_min": 2.5625256967032328e-05,
      "clip_ratio/region_mean": 0.001294846246310044,
      "completions/clipped_ratio": 0.030133928571428603,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3901.0,
      "completions/mean_length": 659.1183471679688,
      "completions/mean_terminated_length": 552.333740234375,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.10498687664041995,
      "grad_norm": 0.11341482400894165,
      "learning_rate": 1e-06,
      "loss": 0.0021,
      "num_tokens": 7004139.0,
      "reward": 0.5401785969734192,
      "reward_std": 0.15878842771053314,
      "rewards/verify_math_reward/mean": 0.5401785969734192,
      "rewards/verify_math_reward/std": 0.49866142868995667,
      "step": 45
    },
    {
      "clip_ratio/high_max": 0.002184344320994569,
      "clip_ratio/high_mean": 0.000801586273155408,
      "clip_ratio/low_mean": 0.0006106818127591396,
      "clip_ratio/low_min": 1.4501159967039712e-05,
      "clip_ratio/region_mean": 0.0014122680768196005,
      "epoch": 0.10731991834354039,
      "grad_norm": 0.1139429360628128,
      "learning_rate": 1e-06,
      "loss": 0.0021,
      "step": 46
    },
    {
      "clip_ratio/high_max": 0.0022679470203001983,
      "clip_ratio/high_mean": 0.000745011353501468,
      "clip_ratio/low_mean": 0.0006523218762595206,
      "clip_ratio/low_min": 1.2773350135830697e-05,
      "clip_ratio/region_mean": 0.0013973332061141264,
      "epoch": 0.10965296004666084,
      "grad_norm": 0.11152958869934082,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "step": 47
    },
    {
      "clip_ratio/high_max": 0.0024709537283342797,
      "clip_ratio/high_mean": 0.0008462621262879111,
      "clip_ratio/low_mean": 0.0006355218320095446,
      "clip_ratio/low_min": 3.843788363155909e-05,
      "clip_ratio/region_mean": 0.001481783951021498,
      "epoch": 0.11198600174978128,
      "grad_norm": 0.11159171164035797,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "step": 48
    },
    {
      "clip_ratio/high_max": 0.002606435209600022,
      "clip_ratio/high_mean": 0.0010781428645714186,
      "clip_ratio/low_mean": 0.000551244026837594,
      "clip_ratio/low_min": 8.372309730475536e-05,
      "clip_ratio/region_mean": 0.0016293868975481018,
      "completions/clipped_ratio": 0.016741071428571397,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3906.0,
      "completions/mean_length": 655.7846069335938,
      "completions/mean_terminated_length": 597.211181640625,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.11431904345290173,
      "grad_norm": 0.13207431137561798,
      "learning_rate": 1e-06,
      "loss": -0.0008,
      "num_tokens": 7618274.0,
      "reward": 0.5558035969734192,
      "reward_std": 0.22808194160461426,
      "rewards/verify_math_reward/mean": 0.5558035969734192,
      "rewards/verify_math_reward/std": 0.49715369939804077,
      "step": 49
    },
    {
      "clip_ratio/high_max": 0.002573193338321289,
      "clip_ratio/high_mean": 0.0010621982528391527,
      "clip_ratio/low_mean": 0.000621154951659264,
      "clip_ratio/low_min": 4.7855231059656944e-05,
      "clip_ratio/region_mean": 0.0016833531844895333,
      "epoch": 0.11665208515602217,
      "grad_norm": 0.13091568648815155,
      "learning_rate": 1e-06,
      "loss": -0.0009,
      "step": 50
    },
    {
      "clip_ratio/high_max": 0.002612927622976713,
      "clip_ratio/high_mean": 0.0010625724717101548,
      "clip_ratio/low_mean": 0.0007023778762231814,
      "clip_ratio/low_min": 5.2263617362768855e-05,
      "clip_ratio/region_mean": 0.0017649503679422196,
      "epoch": 0.1189851268591426,
      "grad_norm": 0.13183654844760895,
      "learning_rate": 1e-06,
      "loss": -0.0009,
      "step": 51
    },
    {
      "clip_ratio/high_max": 0.0026154524603043683,
      "clip_ratio/high_mean": 0.001107633943320252,
      "clip_ratio/low_mean": 0.00068232952071412,
      "clip_ratio/low_min": 3.851271958410507e-05,
      "clip_ratio/region_mean": 0.0017899634622153826,
      "epoch": 0.12131816856226305,
      "grad_norm": 0.12829582393169403,
      "learning_rate": 1e-06,
      "loss": -0.001,
      "step": 52
    },
    {
      "clip_ratio/high_max": 0.002519420573662501,
      "clip_ratio/high_mean": 0.0010384653105575126,
      "clip_ratio/low_mean": 0.0005523251293197973,
      "clip_ratio/low_min": 3.0062635232752655e-05,
      "clip_ratio/region_mean": 0.0015907904598861933,
      "completions/clipped_ratio": 0.016741071428571397,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3275.0,
      "completions/mean_length": 595.1596069335938,
      "completions/mean_terminated_length": 535.553955078125,
      "completions/min_length": 143.0,
      "completions/min_terminated_length": 143.0,
      "epoch": 0.12365121026538349,
      "grad_norm": 0.12745356559753418,
      "learning_rate": 1e-06,
      "loss": -0.0042,
      "num_tokens": 8186209.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.22529542446136475,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 53
    },
    {
      "clip_ratio/high_max": 0.002658939854882192,
      "clip_ratio/high_mean": 0.00113034357491415,
      "clip_ratio/low_mean": 0.0006279497829382308,
      "clip_ratio/low_min": 6.613486675632885e-05,
      "clip_ratio/region_mean": 0.0017582933578523807,
      "epoch": 0.12598425196850394,
      "grad_norm": 0.12348710000514984,
      "learning_rate": 1e-06,
      "loss": -0.0042,
      "step": 54
    },
    {
      "clip_ratio/high_max": 0.0025012689584400505,
      "clip_ratio/high_mean": 0.001088208387955092,
      "clip_ratio/low_mean": 0.0006285139834290021,
      "clip_ratio/low_min": 3.1737179597257636e-05,
      "clip_ratio/region_mean": 0.0017167224104923662,
      "epoch": 0.1283172936716244,
      "grad_norm": 0.12301290035247803,
      "learning_rate": 1e-06,
      "loss": -0.0043,
      "step": 55
    },
    {
      "clip_ratio/high_max": 0.002401656427537091,
      "clip_ratio/high_mean": 0.001099151726521086,
      "clip_ratio/low_mean": 0.0007453204689227277,
      "clip_ratio/low_min": 5.498792961589061e-05,
      "clip_ratio/region_mean": 0.00184447223728057,
      "epoch": 0.13065033537474482,
      "grad_norm": 0.12256094813346863,
      "learning_rate": 1e-06,
      "loss": -0.0044,
      "step": 56
    },
    {
      "clip_ratio/high_max": 0.0018831756169674918,
      "clip_ratio/high_mean": 0.0007435029820044292,
      "clip_ratio/low_mean": 0.0005309441312419949,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012744471132464241,
      "completions/clipped_ratio": 0.011160714285714302,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3656.0,
      "completions/mean_length": 642.0803833007812,
      "completions/mean_terminated_length": 603.0971069335938,
      "completions/min_length": 173.0,
      "completions/min_terminated_length": 173.0,
      "epoch": 0.13298337707786526,
      "grad_norm": 0.1111573651432991,
      "learning_rate": 1e-06,
      "loss": 0.0096,
      "num_tokens": 8814017.0,
      "reward": 0.5848214626312256,
      "reward_std": 0.17400752007961273,
      "rewards/verify_math_reward/mean": 0.5848214030265808,
      "rewards/verify_math_reward/std": 0.49302801489830017,
      "step": 57
    },
    {
      "clip_ratio/high_max": 0.0020343345895526,
      "clip_ratio/high_mean": 0.0007646062240382889,
      "clip_ratio/low_mean": 0.0006124447772890562,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013770509904134087,
      "epoch": 0.13531641878098571,
      "grad_norm": 0.1069447249174118,
      "learning_rate": 1e-06,
      "loss": 0.0095,
      "step": 58
    },
    {
      "clip_ratio/high_max": 0.001814778457628563,
      "clip_ratio/high_mean": 0.0007152991438488243,
      "clip_ratio/low_mean": 0.000650829207188508,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013661283046531025,
      "epoch": 0.13764946048410615,
      "grad_norm": 0.10686612129211426,
      "learning_rate": 1e-06,
      "loss": 0.0095,
      "step": 59
    },
    {
      "clip_ratio/high_max": 0.002180582654546015,
      "clip_ratio/high_mean": 0.0008631766177131794,
      "clip_ratio/low_mean": 0.0006756882612535264,
      "clip_ratio/low_min": 1.304801662627142e-05,
      "clip_ratio/region_mean": 0.001538864893518621,
      "epoch": 0.1399825021872266,
      "grad_norm": 0.10535623878240585,
      "learning_rate": 1e-06,
      "loss": 0.0094,
      "step": 60
    },
    {
      "clip_ratio/high_max": 0.002108367465552874,
      "clip_ratio/high_mean": 0.0007626612023159396,
      "clip_ratio/low_mean": 0.0004933809123031097,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012560421346279327,
      "completions/clipped_ratio": 0.029017857142857095,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3878.0,
      "completions/mean_length": 694.1864013671875,
      "completions/mean_terminated_length": 592.5230102539062,
      "completions/min_length": 91.0,
      "completions/min_terminated_length": 91.0,
      "epoch": 0.14231554389034703,
      "grad_norm": 0.11304501444101334,
      "learning_rate": 1e-06,
      "loss": 0.0074,
      "num_tokens": 9420976.0,
      "reward": 0.5569196939468384,
      "reward_std": 0.1749839335680008,
      "rewards/verify_math_reward/mean": 0.5569196343421936,
      "rewards/verify_math_reward/std": 0.49702703952789307,
      "step": 61
    },
    {
      "clip_ratio/high_max": 0.0021938198697171174,
      "clip_ratio/high_mean": 0.0008237038055085577,
      "clip_ratio/low_mean": 0.000559093690753798,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013827974908053875,
      "epoch": 0.1446485855934675,
      "grad_norm": 0.10986457765102386,
      "learning_rate": 1e-06,
      "loss": 0.0074,
      "step": 62
    },
    {
      "clip_ratio/high_max": 0.002056484743661713,
      "clip_ratio/high_mean": 0.0008661098709126236,
      "clip_ratio/low_mean": 0.0005379781496230862,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014040879941603635,
      "epoch": 0.14698162729658792,
      "grad_norm": 0.11047877371311188,
      "learning_rate": 1e-06,
      "loss": 0.0073,
      "step": 63
    },
    {
      "clip_ratio/high_max": 0.002259606761072064,
      "clip_ratio/high_mean": 0.0008478759955323767,
      "clip_ratio/low_mean": 0.0005690079124178737,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014168839370540809,
      "epoch": 0.14931466899970838,
      "grad_norm": 0.10816746205091476,
      "learning_rate": 1e-06,
      "loss": 0.0072,
      "step": 64
    },
    {
      "clip_ratio/high_max": 0.002090176487399731,
      "clip_ratio/high_mean": 0.0008006530915736221,
      "clip_ratio/low_mean": 0.000709360796463443,
      "clip_ratio/low_min": 4.4256788896746e-05,
      "clip_ratio/region_mean": 0.001510013888037065,
      "completions/clipped_ratio": 0.0078125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2538.0,
      "completions/mean_length": 605.8460083007812,
      "completions/mean_terminated_length": 578.364501953125,
      "completions/min_length": 119.0,
      "completions/min_terminated_length": 119.0,
      "epoch": 0.15164771070282881,
      "grad_norm": 0.12577134370803833,
      "learning_rate": 1e-06,
      "loss": 0.0023,
      "num_tokens": 10020766.0,
      "reward": 0.5837053656578064,
      "reward_std": 0.21857714653015137,
      "rewards/verify_math_reward/mean": 0.5837053656578064,
      "rewards/verify_math_reward/std": 0.49321895837783813,
      "step": 65
    },
    {
      "clip_ratio/high_max": 0.002333878477656981,
      "clip_ratio/high_mean": 0.0009288478941016365,
      "clip_ratio/low_mean": 0.0007335979662457248,
      "clip_ratio/low_min": 2.1956602722639218e-05,
      "clip_ratio/region_mean": 0.0016624458621663507,
      "epoch": 0.15398075240594924,
      "grad_norm": 0.12191906571388245,
      "learning_rate": 1e-06,
      "loss": 0.0022,
      "step": 66
    },
    {
      "clip_ratio/high_max": 0.002172569580579875,
      "clip_ratio/high_mean": 0.0008620889511803398,
      "clip_ratio/low_mean": 0.0008169914053723915,
      "clip_ratio/low_min": 3.694098995765671e-05,
      "clip_ratio/region_mean": 0.00167908036019071,
      "epoch": 0.1563137941090697,
      "grad_norm": 0.12169674783945084,
      "learning_rate": 1e-06,
      "loss": 0.0021,
      "step": 67
    },
    {
      "clip_ratio/high_max": 0.0023263591720024124,
      "clip_ratio/high_mean": 0.0009342802095488878,
      "clip_ratio/low_mean": 0.0009031826884893235,
      "clip_ratio/low_min": 5.15037982040667e-05,
      "clip_ratio/region_mean": 0.0018374628780293278,
      "epoch": 0.15864683581219013,
      "grad_norm": 0.12440544366836548,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "step": 68
    },
    {
      "clip_ratio/high_max": 0.0018946796481031924,
      "clip_ratio/high_mean": 0.0008617473213234916,
      "clip_ratio/low_mean": 0.0006712405856887926,
      "clip_ratio/low_min": 2.0611231775546912e-05,
      "clip_ratio/region_mean": 0.0015329879242926836,
      "completions/clipped_ratio": 0.029017857142857095,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3917.0,
      "completions/mean_length": 706.7098388671875,
      "completions/mean_terminated_length": 605.4207153320312,
      "completions/min_length": 123.0,
      "completions/min_terminated_length": 123.0,
      "epoch": 0.1609798775153106,
      "grad_norm": 0.13245128095149994,
      "learning_rate": 1e-06,
      "loss": 0.0078,
      "num_tokens": 10628714.0,
      "reward": 0.559151828289032,
      "reward_std": 0.23240120708942413,
      "rewards/verify_math_reward/mean": 0.5591517686843872,
      "rewards/verify_math_reward/std": 0.496766060590744,
      "step": 69
    },
    {
      "clip_ratio/high_max": 0.0024722638409002684,
      "clip_ratio/high_mean": 0.0009684698052296881,
      "clip_ratio/low_mean": 0.0007125175779947313,
      "clip_ratio/low_min": 3.960115100198891e-05,
      "clip_ratio/region_mean": 0.0016809873850434087,
      "epoch": 0.16331291921843102,
      "grad_norm": 0.13081303238868713,
      "learning_rate": 1e-06,
      "loss": 0.0076,
      "step": 70
    },
    {
      "clip_ratio/high_max": 0.0022747882321709767,
      "clip_ratio/high_mean": 0.0009260399037884781,
      "clip_ratio/low_mean": 0.000759382968681166,
      "clip_ratio/low_min": 4.311801512812963e-05,
      "clip_ratio/region_mean": 0.0016854228488227818,
      "epoch": 0.16564596092155148,
      "grad_norm": 0.1290624439716339,
      "learning_rate": 1e-06,
      "loss": 0.0076,
      "step": 71
    },
    {
      "clip_ratio/high_max": 0.0023498402879340574,
      "clip_ratio/high_mean": 0.0009991134465963114,
      "clip_ratio/low_mean": 0.0007842375252948841,
      "clip_ratio/low_min": 4.8880713620746974e-05,
      "clip_ratio/region_mean": 0.0017833509882621001,
      "epoch": 0.1679790026246719,
      "grad_norm": 0.1262524425983429,
      "learning_rate": 1e-06,
      "loss": 0.0075,
      "step": 72
    },
    {
      "clip_ratio/high_max": 0.0027180942706763744,
      "clip_ratio/high_mean": 0.0011511004449857865,
      "clip_ratio/low_mean": 0.0006964103449718095,
      "clip_ratio/low_min": 3.421556357352529e-05,
      "clip_ratio/region_mean": 0.0018475107644917443,
      "completions/clipped_ratio": 0.029017857142857095,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3315.0,
      "completions/mean_length": 680.4319458007812,
      "completions/mean_terminated_length": 578.3574829101562,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 0.17031204432779237,
      "grad_norm": 0.13039037585258484,
      "learning_rate": 1e-06,
      "loss": -0.0199,
      "num_tokens": 11226341.0,
      "reward": 0.660714328289032,
      "reward_std": 0.23322537541389465,
      "rewards/verify_math_reward/mean": 0.6607142686843872,
      "rewards/verify_math_reward/std": 0.4737313687801361,
      "step": 73
    },
    {
      "clip_ratio/high_max": 0.0024872794601833448,
      "clip_ratio/high_mean": 0.0011051116780436132,
      "clip_ratio/low_mean": 0.0007465089693141636,
      "clip_ratio/low_min": 6.628324354096549e-05,
      "clip_ratio/region_mean": 0.0018516206691856496,
      "epoch": 0.1726450860309128,
      "grad_norm": 0.12996259331703186,
      "learning_rate": 1e-06,
      "loss": -0.0199,
      "step": 74
    },
    {
      "clip_ratio/high_max": 0.0028027194930473343,
      "clip_ratio/high_mean": 0.0012348731033853255,
      "clip_ratio/low_mean": 0.0008350134203283233,
      "clip_ratio/low_min": 6.279769513639621e-05,
      "clip_ratio/region_mean": 0.0020698865337180905,
      "epoch": 0.17497812773403323,
      "grad_norm": 0.1261185258626938,
      "learning_rate": 1e-06,
      "loss": -0.0201,
      "step": 75
    },
    {
      "clip_ratio/high_max": 0.0032098262090585195,
      "clip_ratio/high_mean": 0.0013133431639289483,
      "clip_ratio/low_mean": 0.0007700925461904262,
      "clip_ratio/low_min": 4.5335311369854026e-05,
      "clip_ratio/region_mean": 0.002083435691019986,
      "epoch": 0.1773111694371537,
      "grad_norm": 0.1239931732416153,
      "learning_rate": 1e-06,
      "loss": -0.0201,
      "step": 76
    },
    {
      "clip_ratio/high_max": 0.0019031432893825695,
      "clip_ratio/high_mean": 0.0009410108505107928,
      "clip_ratio/low_mean": 0.0006787779257138027,
      "clip_ratio/low_min": 5.3184265198069625e-05,
      "clip_ratio/region_mean": 0.0016197887598536909,
      "completions/clipped_ratio": 0.008928571428571397,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2877.0,
      "completions/mean_length": 593.59375,
      "completions/mean_terminated_length": 562.04052734375,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 0.17964421114027412,
      "grad_norm": 0.1269347369670868,
      "learning_rate": 1e-06,
      "loss": 0.0066,
      "num_tokens": 11816601.0,
      "reward": 0.6495535969734192,
      "reward_std": 0.21132300794124603,
      "rewards/verify_math_reward/mean": 0.6495535969734192,
      "rewards/verify_math_reward/std": 0.477376252412796,
      "step": 77
    },
    {
      "clip_ratio/high_max": 0.0020921051764162257,
      "clip_ratio/high_mean": 0.0009143478564510588,
      "clip_ratio/low_mean": 0.0007781102540320717,
      "clip_ratio/low_min": 4.562021877063671e-05,
      "clip_ratio/region_mean": 0.0016924581068451516,
      "epoch": 0.18197725284339458,
      "grad_norm": 0.1262543946504593,
      "learning_rate": 1e-06,
      "loss": 0.0064,
      "step": 78
    },
    {
      "clip_ratio/high_max": 0.00232102838344872,
      "clip_ratio/high_mean": 0.0010060116510430817,
      "clip_ratio/low_mean": 0.0008259733658633195,
      "clip_ratio/low_min": 9.44464636631892e-05,
      "clip_ratio/region_mean": 0.00183198502054438,
      "epoch": 0.184310294546515,
      "grad_norm": 0.12525974214076996,
      "learning_rate": 1e-06,
      "loss": 0.0064,
      "step": 79
    },
    {
      "clip_ratio/high_max": 0.002366886088566389,
      "clip_ratio/high_mean": 0.0010620286157063674,
      "clip_ratio/low_mean": 0.0008741377005208051,
      "clip_ratio/low_min": 7.721427664364455e-05,
      "clip_ratio/region_mean": 0.0019361662562005222,
      "epoch": 0.18664333624963547,
      "grad_norm": 0.1252414882183075,
      "learning_rate": 1e-06,
      "loss": 0.0063,
      "step": 80
    },
    {
      "clip_ratio/high_max": 0.0018435443125781603,
      "clip_ratio/high_mean": 0.0007656542757104035,
      "clip_ratio/low_mean": 0.0005714547769457567,
      "clip_ratio/low_min": 1.332338524662191e-05,
      "clip_ratio/region_mean": 0.0013371090353757609,
      "completions/clipped_ratio": 0.022321428571428603,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3419.0,
      "completions/mean_length": 646.0457763671875,
      "completions/mean_terminated_length": 567.2796630859375,
      "completions/min_length": 150.0,
      "completions/min_terminated_length": 150.0,
      "epoch": 0.1889763779527559,
      "grad_norm": 0.1250150203704834,
      "learning_rate": 1e-06,
      "loss": 0.0102,
      "num_tokens": 12395730.0,
      "reward": 0.598214328289032,
      "reward_std": 0.18430186808109283,
      "rewards/verify_math_reward/mean": 0.5982142686843872,
      "rewards/verify_math_reward/std": 0.49053287506103516,
      "step": 81
    },
    {
      "clip_ratio/high_max": 0.0019822753747575916,
      "clip_ratio/high_mean": 0.0008148421766236424,
      "clip_ratio/low_mean": 0.0005596559967671055,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013744982024945784,
      "epoch": 0.19130941965587636,
      "grad_norm": 0.12403535842895508,
      "learning_rate": 1e-06,
      "loss": 0.0101,
      "step": 82
    },
    {
      "clip_ratio/high_max": 0.001799621146346908,
      "clip_ratio/high_mean": 0.0008243617266998626,
      "clip_ratio/low_mean": 0.0006454474469137494,
      "clip_ratio/low_min": 5.0011256462312303e-05,
      "clip_ratio/region_mean": 0.001469809198169969,
      "epoch": 0.1936424613589968,
      "grad_norm": 0.12789423763751984,
      "learning_rate": 1e-06,
      "loss": 0.0101,
      "step": 83
    },
    {
      "clip_ratio/high_max": 0.0021637038080370985,
      "clip_ratio/high_mean": 0.0009658185390435392,
      "clip_ratio/low_mean": 0.000647538096018252,
      "clip_ratio/low_min": 2.5005628231156152e-05,
      "clip_ratio/region_mean": 0.0016133566678036004,
      "epoch": 0.19597550306211722,
      "grad_norm": 0.11844993382692337,
      "learning_rate": 1e-06,
      "loss": 0.0099,
      "step": 84
    },
    {
      "clip_ratio/high_max": 0.0016195920543395914,
      "clip_ratio/high_mean": 0.0007305502513190731,
      "clip_ratio/low_mean": 0.00042413621758896625,
      "clip_ratio/low_min": 2.5902445486281067e-05,
      "clip_ratio/region_mean": 0.0011546864770934917,
      "completions/clipped_ratio": 0.0189732142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3058.0,
      "completions/mean_length": 694.1875610351562,
      "completions/mean_terminated_length": 628.3958740234375,
      "completions/min_length": 139.0,
      "completions/min_terminated_length": 139.0,
      "epoch": 0.19830854476523768,
      "grad_norm": 0.10732998698949814,
      "learning_rate": 1e-06,
      "loss": 0.0246,
      "num_tokens": 13043522.0,
      "reward": 0.5290178656578064,
      "reward_std": 0.1879132241010666,
      "rewards/verify_math_reward/mean": 0.5290178656578064,
      "rewards/verify_math_reward/std": 0.49943605065345764,
      "step": 85
    },
    {
      "clip_ratio/high_max": 0.001977817406441318,
      "clip_ratio/high_mean": 0.0008252691604866413,
      "clip_ratio/low_mean": 0.0005512036532309139,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013764728064415976,
      "epoch": 0.2006415864683581,
      "grad_norm": 0.10541458427906036,
      "learning_rate": 1e-06,
      "loss": 0.0245,
      "step": 86
    },
    {
      "clip_ratio/high_max": 0.0018707476629060693,
      "clip_ratio/high_mean": 0.0008347092680196511,
      "clip_ratio/low_mean": 0.0005563870163314277,
      "clip_ratio/low_min": 2.986857907671947e-05,
      "clip_ratio/region_mean": 0.001391096335282782,
      "epoch": 0.20297462817147857,
      "grad_norm": 0.10650135576725006,
      "learning_rate": 1e-06,
      "loss": 0.0245,
      "step": 87
    },
    {
      "clip_ratio/high_max": 0.0019968881679233164,
      "clip_ratio/high_mean": 0.000803314738732297,
      "clip_ratio/low_mean": 0.0005893544366699643,
      "clip_ratio/low_min": 1.4934289538359735e-05,
      "clip_ratio/region_mean": 0.0013926691899541765,
      "epoch": 0.205307669874599,
      "grad_norm": 0.10585729777812958,
      "learning_rate": 1e-06,
      "loss": 0.0244,
      "step": 88
    },
    {
      "clip_ratio/high_max": 0.0017711450818751473,
      "clip_ratio/high_mean": 0.0007248113288369495,
      "clip_ratio/low_mean": 0.0005172562141524395,
      "clip_ratio/low_min": 1.4178765923134051e-05,
      "clip_ratio/region_mean": 0.0012420675193425268,
      "completions/clipped_ratio": 0.030133928571428603,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3155.0,
      "completions/mean_length": 707.794677734375,
      "completions/mean_terminated_length": 602.5224609375,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 0.20764071157771946,
      "grad_norm": 0.12869536876678467,
      "learning_rate": 1e-06,
      "loss": -0.0064,
      "num_tokens": 13651450.0,
      "reward": 0.637276828289032,
      "reward_std": 0.18129737675189972,
      "rewards/verify_math_reward/mean": 0.6372767686843872,
      "rewards/verify_math_reward/std": 0.481054425239563,
      "step": 89
    },
    {
      "clip_ratio/high_max": 0.0019154321453243028,
      "clip_ratio/high_mean": 0.0007747264771751361,
      "clip_ratio/low_mean": 0.0005645960791298421,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001339322541753063,
      "epoch": 0.2099737532808399,
      "grad_norm": 0.12464048713445663,
      "learning_rate": 1e-06,
      "loss": -0.0065,
      "step": 90
    },
    {
      "clip_ratio/high_max": 0.0019451008993200958,
      "clip_ratio/high_mean": 0.0008710344573046314,
      "clip_ratio/low_mean": 0.0006486630554718431,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015196975255094003,
      "epoch": 0.21230679498396035,
      "grad_norm": 0.12116258591413498,
      "learning_rate": 1e-06,
      "loss": -0.0066,
      "step": 91
    },
    {
      "clip_ratio/high_max": 0.002008242765441537,
      "clip_ratio/high_mean": 0.0008125443891913164,
      "clip_ratio/low_mean": 0.0007152098842198029,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001527754280687077,
      "epoch": 0.21463983668708078,
      "grad_norm": 0.11910553276538849,
      "learning_rate": 1e-06,
      "loss": -0.0067,
      "step": 92
    },
    {
      "clip_ratio/high_max": 0.0018941184753202833,
      "clip_ratio/high_mean": 0.0007535947170254076,
      "clip_ratio/low_mean": 0.0004500252680372796,
      "clip_ratio/low_min": 1.5907355191302486e-05,
      "clip_ratio/region_mean": 0.001203619995067129,
      "completions/clipped_ratio": 0.022321428571428603,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2767.0,
      "completions/mean_length": 705.9219360351562,
      "completions/mean_terminated_length": 628.5228271484375,
      "completions/min_length": 155.0,
      "completions/min_terminated_length": 155.0,
      "epoch": 0.2169728783902012,
      "grad_norm": 0.1125323548913002,
      "learning_rate": 1e-06,
      "loss": -0.0052,
      "num_tokens": 14286332.0,
      "reward": 0.5535714626312256,
      "reward_std": 0.19163475930690765,
      "rewards/verify_math_reward/mean": 0.5535714030265808,
      "rewards/verify_math_reward/std": 0.4973994791507721,
      "step": 93
    },
    {
      "clip_ratio/high_max": 0.0021611167867376935,
      "clip_ratio/high_mean": 0.0007815511544322362,
      "clip_ratio/low_mean": 0.0005014456273784162,
      "clip_ratio/low_min": 1.736834747134708e-05,
      "clip_ratio/region_mean": 0.0012829967672587372,
      "epoch": 0.21930592009332167,
      "grad_norm": 0.11015673726797104,
      "learning_rate": 1e-06,
      "loss": -0.0053,
      "step": 94
    },
    {
      "clip_ratio/high_max": 0.0020602601434802637,
      "clip_ratio/high_mean": 0.0008277097422251245,
      "clip_ratio/low_mean": 0.0005118875869811745,
      "clip_ratio/low_min": 2.3599455744260922e-05,
      "clip_ratio/region_mean": 0.0013395973219303414,
      "epoch": 0.2216389617964421,
      "grad_norm": 0.11136667430400848,
      "learning_rate": 1e-06,
      "loss": -0.0054,
      "step": 95
    },
    {
      "clip_ratio/high_max": 0.002315436089702416,
      "clip_ratio/high_mean": 0.0008609901287854882,
      "clip_ratio/low_mean": 0.0005434525282907998,
      "clip_ratio/low_min": 1.5907355191302486e-05,
      "clip_ratio/region_mean": 0.0014044426607142668,
      "epoch": 0.22397200349956256,
      "grad_norm": 0.10752148181200027,
      "learning_rate": 1e-06,
      "loss": -0.0054,
      "step": 96
    },
    {
      "clip_ratio/high_max": 0.0016625657699478325,
      "clip_ratio/high_mean": 0.0006201890655574971,
      "clip_ratio/low_mean": 0.000499386642331956,
      "clip_ratio/low_min": 1.2033115126541816e-05,
      "clip_ratio/region_mean": 0.0011195757197128842,
      "completions/clipped_ratio": 0.0424107142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2853.0,
      "completions/mean_length": 797.1551513671875,
      "completions/mean_terminated_length": 651.052490234375,
      "completions/min_length": 129.0,
      "completions/min_terminated_length": 129.0,
      "epoch": 0.226305045202683,
      "grad_norm": 0.10692703723907471,
      "learning_rate": 1e-06,
      "loss": -0.0092,
      "num_tokens": 14930167.0,
      "reward": 0.5457589626312256,
      "reward_std": 0.17107722163200378,
      "rewards/verify_math_reward/mean": 0.5457589030265808,
      "rewards/verify_math_reward/std": 0.4981797933578491,
      "step": 97
    },
    {
      "clip_ratio/high_max": 0.0018763612533803098,
      "clip_ratio/high_mean": 0.0007112914427125361,
      "clip_ratio/low_mean": 0.0005139210170455044,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00122521246157703,
      "epoch": 0.22863808690580345,
      "grad_norm": 0.10555337369441986,
      "learning_rate": 1e-06,
      "loss": -0.0092,
      "step": 98
    },
    {
      "clip_ratio/high_max": 0.0016491648311784957,
      "clip_ratio/high_mean": 0.0006668064925179351,
      "clip_ratio/low_mean": 0.0006113593735790346,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001278165840631118,
      "epoch": 0.23097112860892388,
      "grad_norm": 0.1043100655078888,
      "learning_rate": 1e-06,
      "loss": -0.0093,
      "step": 99
    },
    {
      "clip_ratio/high_max": 0.00191963902398129,
      "clip_ratio/high_mean": 0.0007225982008094434,
      "clip_ratio/low_mean": 0.0005822008433824521,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001304799039644422,
      "epoch": 0.23330417031204434,
      "grad_norm": 0.10567686706781387,
      "learning_rate": 1e-06,
      "loss": -0.0093,
      "step": 100
    },
    {
      "clip_ratio/high_max": 0.0016583305387030123,
      "clip_ratio/high_mean": 0.0007518018992414,
      "clip_ratio/low_mean": 0.0007701156791881658,
      "clip_ratio/low_min": 5.871514713362558e-05,
      "clip_ratio/region_mean": 0.0015219175402307883,
      "completions/clipped_ratio": 0.0267857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3472.0,
      "completions/mean_length": 699.763427734375,
      "completions/mean_terminated_length": 606.2889404296875,
      "completions/min_length": 175.0,
      "completions/min_terminated_length": 175.0,
      "epoch": 0.23563721201516477,
      "grad_norm": 0.13025298714637756,
      "learning_rate": 1e-06,
      "loss": 0.0202,
      "num_tokens": 15546323.0,
      "reward": 0.5245535969734192,
      "reward_std": 0.21591150760650635,
      "rewards/verify_math_reward/mean": 0.5245535969734192,
      "rewards/verify_math_reward/std": 0.4996756911277771,
      "step": 101
    },
    {
      "clip_ratio/high_max": 0.002065653068711981,
      "clip_ratio/high_mean": 0.0008601816971349763,
      "clip_ratio/low_mean": 0.0007794047523930203,
      "clip_ratio/low_min": 7.620711767231114e-05,
      "clip_ratio/region_mean": 0.0016395864731748588,
      "epoch": 0.2379702537182852,
      "grad_norm": 0.12589409947395325,
      "learning_rate": 1e-06,
      "loss": 0.0201,
      "step": 102
    },
    {
      "clip_ratio/high_max": 0.0020022314784000628,
      "clip_ratio/high_mean": 0.0008255693137471098,
      "clip_ratio/low_mean": 0.0007917809125501662,
      "clip_ratio/low_min": 7.215025198092917e-05,
      "clip_ratio/region_mean": 0.0016173501862795092,
      "epoch": 0.24030329542140566,
      "grad_norm": 0.1242484375834465,
      "learning_rate": 1e-06,
      "loss": 0.02,
      "step": 103
    },
    {
      "clip_ratio/high_max": 0.0019035753612115514,
      "clip_ratio/high_mean": 0.0009137180004472611,
      "clip_ratio/low_mean": 0.0008719772231415845,
      "clip_ratio/low_min": 0.00010575910619081696,
      "clip_ratio/region_mean": 0.0017856952326837927,
      "epoch": 0.2426363371245261,
      "grad_norm": 0.1210545226931572,
      "learning_rate": 1e-06,
      "loss": 0.0199,
      "step": 104
    },
    {
      "clip_ratio/high_max": 0.0022454316713265143,
      "clip_ratio/high_mean": 0.000870076099090511,
      "clip_ratio/low_mean": 0.0005405405645433348,
      "clip_ratio/low_min": 1.1241007086937316e-05,
      "clip_ratio/region_mean": 0.0014106166454439517,
      "completions/clipped_ratio": 0.03125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3474.0,
      "completions/mean_length": 702.9375610351562,
      "completions/mean_terminated_length": 593.48388671875,
      "completions/min_length": 180.0,
      "completions/min_terminated_length": 180.0,
      "epoch": 0.24496937882764655,
      "grad_norm": 0.13943475484848022,
      "learning_rate": 1e-06,
      "loss": -0.0111,
      "num_tokens": 16149587.0,
      "reward": 0.582589328289032,
      "reward_std": 0.20993182063102722,
      "rewards/verify_math_reward/mean": 0.5825892686843872,
      "rewards/verify_math_reward/std": 0.4934072494506836,
      "step": 105
    },
    {
      "clip_ratio/high_max": 0.0022179896768648177,
      "clip_ratio/high_mean": 0.0009223711076629115,
      "clip_ratio/low_mean": 0.0006832732597104041,
      "clip_ratio/low_min": 7.951450606924482e-05,
      "clip_ratio/region_mean": 0.0016056443710112944,
      "epoch": 0.24730242053076698,
      "grad_norm": 0.1324169635772705,
      "learning_rate": 1e-06,
      "loss": -0.0113,
      "step": 106
    },
    {
      "clip_ratio/high_max": 0.0023009943761280738,
      "clip_ratio/high_mean": 0.00090966447169194,
      "clip_ratio/low_mean": 0.0006601435015909374,
      "clip_ratio/low_min": 4.010021802969277e-05,
      "clip_ratio/region_mean": 0.001569807980558835,
      "epoch": 0.24963546223388744,
      "grad_norm": 0.13063472509384155,
      "learning_rate": 1e-06,
      "loss": -0.0114,
      "step": 107
    },
    {
      "clip_ratio/high_max": 0.0026666615303838626,
      "clip_ratio/high_mean": 0.0010358955296396744,
      "clip_ratio/low_mean": 0.0007489463259844342,
      "clip_ratio/low_min": 3.096954242209904e-05,
      "clip_ratio/region_mean": 0.0017848418283392675,
      "epoch": 0.25196850393700787,
      "grad_norm": 0.12885810434818268,
      "learning_rate": 1e-06,
      "loss": -0.0115,
      "step": 108
    },
    {
      "clip_ratio/high_max": 0.0020536949050438125,
      "clip_ratio/high_mean": 0.0008293388964375481,
      "clip_ratio/low_mean": 0.0006642993957939325,
      "clip_ratio/low_min": 5.731327473768033e-05,
      "clip_ratio/region_mean": 0.00149363829405047,
      "completions/clipped_ratio": 0.016741071428571397,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3800.0,
      "completions/mean_length": 657.0457763671875,
      "completions/mean_terminated_length": 598.4937744140625,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 0.2543015456401283,
      "grad_norm": 0.13001294434070587,
      "learning_rate": 1e-06,
      "loss": 0.0058,
      "num_tokens": 16766852.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.18979185819625854,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514803290367126,
      "step": 109
    },
    {
      "clip_ratio/high_max": 0.0018642976283445023,
      "clip_ratio/high_mean": 0.000758285681513371,
      "clip_ratio/low_mean": 0.0006911922246217728,
      "clip_ratio/low_min": 6.702373775624437e-05,
      "clip_ratio/region_mean": 0.0014494779243250377,
      "epoch": 0.2566345873432488,
      "grad_norm": 0.1258811205625534,
      "learning_rate": 1e-06,
      "loss": 0.0057,
      "step": 110
    },
    {
      "clip_ratio/high_max": 0.002164812249247916,
      "clip_ratio/high_mean": 0.0008751272252993658,
      "clip_ratio/low_mean": 0.0007665609100513393,
      "clip_ratio/low_min": 5.334565867087804e-05,
      "clip_ratio/region_mean": 0.001641688148083631,
      "epoch": 0.2589676290463692,
      "grad_norm": 0.12354160845279694,
      "learning_rate": 1e-06,
      "loss": 0.0057,
      "step": 111
    },
    {
      "clip_ratio/high_max": 0.0021916895493632182,
      "clip_ratio/high_mean": 0.0008869685170793673,
      "clip_ratio/low_mean": 0.0008263004147011088,
      "clip_ratio/low_min": 6.131671671028016e-05,
      "clip_ratio/region_mean": 0.0017132688881247304,
      "epoch": 0.26130067074948965,
      "grad_norm": 0.12259406596422195,
      "learning_rate": 1e-06,
      "loss": 0.0055,
      "step": 112
    },
    {
      "clip_ratio/high_max": 0.001697391977359075,
      "clip_ratio/high_mean": 0.0006851519665360684,
      "clip_ratio/low_mean": 0.0006891176126373466,
      "clip_ratio/low_min": 7.1352720624418e-05,
      "clip_ratio/region_mean": 0.0013742695373366587,
      "completions/clipped_ratio": 0.0267857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3655.0,
      "completions/mean_length": 711.0535888671875,
      "completions/mean_terminated_length": 617.889892578125,
      "completions/min_length": 139.0,
      "completions/min_terminated_length": 139.0,
      "epoch": 0.2636337124526101,
      "grad_norm": 0.12998297810554504,
      "learning_rate": 1e-06,
      "loss": -0.0156,
      "num_tokens": 17394636.0,
      "reward": 0.5223214626312256,
      "reward_std": 0.20102617144584656,
      "rewards/verify_math_reward/mean": 0.5223214030265808,
      "rewards/verify_math_reward/std": 0.49978047609329224,
      "step": 113
    },
    {
      "clip_ratio/high_max": 0.0017535710612719413,
      "clip_ratio/high_mean": 0.0006981436126807239,
      "clip_ratio/low_mean": 0.0008300278659589821,
      "clip_ratio/low_min": 0.0001619695585759473,
      "clip_ratio/region_mean": 0.0015281714950106107,
      "epoch": 0.2659667541557305,
      "grad_norm": 0.12897886335849762,
      "learning_rate": 1e-06,
      "loss": -0.0158,
      "step": 114
    },
    {
      "clip_ratio/high_max": 0.0020468809198064264,
      "clip_ratio/high_mean": 0.0008466970393783413,
      "clip_ratio/low_mean": 0.0008826452158245957,
      "clip_ratio/low_min": 0.00015975156929926015,
      "clip_ratio/region_mean": 0.0017293422424700111,
      "epoch": 0.268299795858851,
      "grad_norm": 0.12687930464744568,
      "learning_rate": 1e-06,
      "loss": -0.0158,
      "step": 115
    },
    {
      "clip_ratio/high_max": 0.002170760475564748,
      "clip_ratio/high_mean": 0.0008199298881663708,
      "clip_ratio/low_mean": 0.000861517079101759,
      "clip_ratio/low_min": 0.00012403995697241044,
      "clip_ratio/region_mean": 0.001681446927250363,
      "epoch": 0.27063283756197143,
      "grad_norm": 0.12233294546604156,
      "learning_rate": 1e-06,
      "loss": -0.0159,
      "step": 116
    },
    {
      "clip_ratio/high_max": 0.0023675355914747342,
      "clip_ratio/high_mean": 0.0009463731294090394,
      "clip_ratio/low_mean": 0.0006463080644607544,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001592681190231815,
      "completions/clipped_ratio": 0.0189732142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3038.0,
      "completions/mean_length": 641.9732666015625,
      "completions/mean_terminated_length": 575.1717529296875,
      "completions/min_length": 163.0,
      "completions/min_terminated_length": 163.0,
      "epoch": 0.27296587926509186,
      "grad_norm": 0.1370822638273239,
      "learning_rate": 1e-06,
      "loss": 0.0036,
      "num_tokens": 17995540.0,
      "reward": 0.5859375,
      "reward_std": 0.2263127714395523,
      "rewards/verify_math_reward/mean": 0.5859375,
      "rewards/verify_math_reward/std": 0.4928344786167145,
      "step": 117
    },
    {
      "clip_ratio/high_max": 0.0023808866244507954,
      "clip_ratio/high_mean": 0.0010071201832033694,
      "clip_ratio/low_mean": 0.0006476468661276158,
      "clip_ratio/low_min": 1.4744043255632278e-05,
      "clip_ratio/region_mean": 0.0016547670675208792,
      "epoch": 0.2752989209682123,
      "grad_norm": 0.13685859739780426,
      "learning_rate": 1e-06,
      "loss": 0.0036,
      "step": 118
    },
    {
      "clip_ratio/high_max": 0.002364542982832063,
      "clip_ratio/high_mean": 0.0010847648300114088,
      "clip_ratio/low_mean": 0.0007723511116637383,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018571159744169563,
      "epoch": 0.2776319626713328,
      "grad_norm": 0.13253997266292572,
      "learning_rate": 1e-06,
      "loss": 0.0034,
      "step": 119
    },
    {
      "clip_ratio/high_max": 0.002691040266654454,
      "clip_ratio/high_mean": 0.0010794019362947438,
      "clip_ratio/low_mean": 0.0008650881063658744,
      "clip_ratio/low_min": 4.414796330820536e-05,
      "clip_ratio/region_mean": 0.0019444899953668937,
      "epoch": 0.2799650043744532,
      "grad_norm": 0.13256323337554932,
      "learning_rate": 1e-06,
      "loss": 0.0033,
      "step": 120
    },
    {
      "clip_ratio/high_max": 0.0018208038454758935,
      "clip_ratio/high_mean": 0.0007685111413593404,
      "clip_ratio/low_mean": 0.0005416209351096768,
      "clip_ratio/low_min": 5.682390292349737e-05,
      "clip_ratio/region_mean": 0.0013101320873829536,
      "completions/clipped_ratio": 0.0323660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3962.0,
      "completions/mean_length": 776.3292846679688,
      "completions/mean_terminated_length": 665.2906494140625,
      "completions/min_length": 174.0,
      "completions/min_terminated_length": 174.0,
      "epoch": 0.28229804607757364,
      "grad_norm": 0.11596595495939255,
      "learning_rate": 1e-06,
      "loss": -0.0029,
      "num_tokens": 18657227.0,
      "reward": 0.5613839626312256,
      "reward_std": 0.20169946551322937,
      "rewards/verify_math_reward/mean": 0.5613839030265808,
      "rewards/verify_math_reward/std": 0.496494859457016,
      "step": 121
    },
    {
      "clip_ratio/high_max": 0.0019638911999209085,
      "clip_ratio/high_mean": 0.0007903282944425882,
      "clip_ratio/low_mean": 0.0006332460270641604,
      "clip_ratio/low_min": 4.578135758492863e-05,
      "clip_ratio/region_mean": 0.0014235743146855384,
      "epoch": 0.28463108778069407,
      "grad_norm": 0.11225331574678421,
      "learning_rate": 1e-06,
      "loss": -0.003,
      "step": 122
    },
    {
      "clip_ratio/high_max": 0.0020143422298133373,
      "clip_ratio/high_mean": 0.0007923571538412943,
      "clip_ratio/low_mean": 0.0006825453601777554,
      "clip_ratio/low_min": 6.15502985965577e-05,
      "clip_ratio/region_mean": 0.0014749025249329861,
      "epoch": 0.2869641294838145,
      "grad_norm": 0.11058894544839859,
      "learning_rate": 1e-06,
      "loss": -0.003,
      "step": 123
    },
    {
      "clip_ratio/high_max": 0.00207445933483541,
      "clip_ratio/high_mean": 0.0008512227577739395,
      "clip_ratio/low_mean": 0.0006609621759707807,
      "clip_ratio/low_min": 5.902328211959684e-05,
      "clip_ratio/region_mean": 0.0015121849428396672,
      "epoch": 0.289297171186935,
      "grad_norm": 0.11066972464323044,
      "learning_rate": 1e-06,
      "loss": -0.0031,
      "step": 124
    },
    {
      "clip_ratio/high_max": 0.0021815118452650495,
      "clip_ratio/high_mean": 0.0008497404996887781,
      "clip_ratio/low_mean": 0.0004772272868649452,
      "clip_ratio/low_min": 2.0509113710431848e-05,
      "clip_ratio/region_mean": 0.001326967823843006,
      "completions/clipped_ratio": 0.03125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2720.0,
      "completions/mean_length": 705.1797485351562,
      "completions/mean_terminated_length": 595.7984008789062,
      "completions/min_length": 135.0,
      "completions/min_terminated_length": 135.0,
      "epoch": 0.2916302128900554,
      "grad_norm": 0.13661257922649384,
      "learning_rate": 1e-06,
      "loss": 0.0138,
      "num_tokens": 19265220.0,
      "reward": 0.609375,
      "reward_std": 0.18528781831264496,
      "rewards/verify_math_reward/mean": 0.609375,
      "rewards/verify_math_reward/std": 0.48816296458244324,
      "step": 125
    },
    {
      "clip_ratio/high_max": 0.0021811188526044134,
      "clip_ratio/high_mean": 0.0008807081148916041,
      "clip_ratio/low_mean": 0.0005780017290817341,
      "clip_ratio/low_min": 2.8875028874608688e-05,
      "clip_ratio/region_mean": 0.0014587098012270872,
      "epoch": 0.29396325459317585,
      "grad_norm": 0.1319681853055954,
      "learning_rate": 1e-06,
      "loss": 0.0137,
      "step": 126
    },
    {
      "clip_ratio/high_max": 0.002105122650391422,
      "clip_ratio/high_mean": 0.0008940856241679285,
      "clip_ratio/low_mean": 0.0006709501121804351,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015650357236154377,
      "epoch": 0.2962962962962963,
      "grad_norm": 0.14205984771251678,
      "learning_rate": 1e-06,
      "loss": 0.0136,
      "step": 127
    },
    {
      "clip_ratio/high_max": 0.0024559525918448344,
      "clip_ratio/high_mean": 0.0009362378386867931,
      "clip_ratio/low_mean": 0.0007356516161962645,
      "clip_ratio/low_min": 3.0419239919865504e-05,
      "clip_ratio/region_mean": 0.0016718894694349729,
      "epoch": 0.29862933799941677,
      "grad_norm": 0.13000363111495972,
      "learning_rate": 1e-06,
      "loss": 0.0135,
      "step": 128
    },
    {
      "clip_ratio/high_max": 0.0019030739968002308,
      "clip_ratio/high_mean": 0.0006965993607082055,
      "clip_ratio/low_mean": 0.0006350407384161372,
      "clip_ratio/low_min": 7.164487760746852e-05,
      "clip_ratio/region_mean": 0.0013316400873009115,
      "completions/clipped_ratio": 0.021205357142857095,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3387.0,
      "completions/mean_length": 629.5814819335938,
      "completions/mean_terminated_length": 554.4822998046875,
      "completions/min_length": 162.0,
      "completions/min_terminated_length": 162.0,
      "epoch": 0.3009623797025372,
      "grad_norm": 0.13182252645492554,
      "learning_rate": 1e-06,
      "loss": -0.0005,
      "num_tokens": 19838037.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.16630510985851288,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 129
    },
    {
      "clip_ratio/high_max": 0.0019440464311628602,
      "clip_ratio/high_mean": 0.0007428180806527962,
      "clip_ratio/low_mean": 0.0006763731698811171,
      "clip_ratio/low_min": 4.770672876475146e-05,
      "clip_ratio/region_mean": 0.0014191912378009874,
      "epoch": 0.30329542140565763,
      "grad_norm": 0.12787501513957977,
      "learning_rate": 1e-06,
      "loss": -0.0006,
      "step": 130
    },
    {
      "clip_ratio/high_max": 0.0021150591092009563,
      "clip_ratio/high_mean": 0.0008707102069820394,
      "clip_ratio/low_mean": 0.0007152069083531387,
      "clip_ratio/low_min": 4.3578196709859185e-05,
      "clip_ratio/region_mean": 0.0015859171362535562,
      "epoch": 0.30562846310877806,
      "grad_norm": 0.12346763163805008,
      "learning_rate": 1e-06,
      "loss": -0.0008,
      "step": 131
    },
    {
      "clip_ratio/high_max": 0.002172308690205682,
      "clip_ratio/high_mean": 0.0008703252351551782,
      "clip_ratio/low_mean": 0.0008561224449294969,
      "clip_ratio/low_min": 6.992353792156791e-05,
      "clip_ratio/region_mean": 0.0017264476628042758,
      "epoch": 0.3079615048118985,
      "grad_norm": 0.12661373615264893,
      "learning_rate": 1e-06,
      "loss": -0.0008,
      "step": 132
    },
    {
      "clip_ratio/high_max": 0.0016295224777422845,
      "clip_ratio/high_mean": 0.0006557831056852592,
      "clip_ratio/low_mean": 0.0005799791524623288,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012357622545096092,
      "completions/clipped_ratio": 0.0189732142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3400.0,
      "completions/mean_length": 681.9006958007812,
      "completions/mean_terminated_length": 615.8713989257812,
      "completions/min_length": 144.0,
      "completions/min_terminated_length": 144.0,
      "epoch": 0.310294546515019,
      "grad_norm": 0.12366899102926254,
      "learning_rate": 1e-06,
      "loss": 0.0032,
      "num_tokens": 20462780.0,
      "reward": 0.5725446939468384,
      "reward_std": 0.18144823610782623,
      "rewards/verify_math_reward/mean": 0.5725446343421936,
      "rewards/verify_math_reward/std": 0.49498558044433594,
      "step": 133
    },
    {
      "clip_ratio/high_max": 0.0019445515772531508,
      "clip_ratio/high_mean": 0.0007925075506136636,
      "clip_ratio/low_mean": 0.0006260709888010751,
      "clip_ratio/low_min": 2.2393407562049106e-05,
      "clip_ratio/region_mean": 0.0014185785184963606,
      "epoch": 0.3126275882181394,
      "grad_norm": 0.12053171545267105,
      "learning_rate": 1e-06,
      "loss": 0.0031,
      "step": 134
    },
    {
      "clip_ratio/high_max": 0.002057051740848692,
      "clip_ratio/high_mean": 0.000796252378677309,
      "clip_ratio/low_mean": 0.0006900128737470368,
      "clip_ratio/low_min": 2.3946360670379363e-05,
      "clip_ratio/region_mean": 0.0014862652569718193,
      "epoch": 0.31496062992125984,
      "grad_norm": 0.11586639285087585,
      "learning_rate": 1e-06,
      "loss": 0.003,
      "step": 135
    },
    {
      "clip_ratio/high_max": 0.00205574548090226,
      "clip_ratio/high_mean": 0.0008225631409004563,
      "clip_ratio/low_mean": 0.0007366543359239586,
      "clip_ratio/low_min": 2.2393407562049106e-05,
      "clip_ratio/region_mean": 0.0015592174604535103,
      "epoch": 0.31729367162438027,
      "grad_norm": 0.11688009649515152,
      "learning_rate": 1e-06,
      "loss": 0.0029,
      "step": 136
    },
    {
      "clip_ratio/high_max": 0.002024527595494874,
      "clip_ratio/high_mean": 0.0008082161057245685,
      "clip_ratio/low_mean": 0.0005800189078399853,
      "clip_ratio/low_min": 1.8047934645437635e-05,
      "clip_ratio/region_mean": 0.0013882350067433435,
      "completions/clipped_ratio": 0.0323660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3794.0,
      "completions/mean_length": 740.0647583007812,
      "completions/mean_terminated_length": 627.8131103515625,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 0.31962671332750076,
      "grad_norm": 0.13833001255989075,
      "learning_rate": 1e-06,
      "loss": -0.0078,
      "num_tokens": 21090750.0,
      "reward": 0.6171875,
      "reward_std": 0.19227458536624908,
      "rewards/verify_math_reward/mean": 0.6171875,
      "rewards/verify_math_reward/std": 0.4863446056842804,
      "step": 137
    },
    {
      "clip_ratio/high_max": 0.0022252232884056866,
      "clip_ratio/high_mean": 0.0008493184159306111,
      "clip_ratio/low_mean": 0.0006523041329273838,
      "clip_ratio/low_min": 2.575105281721335e-05,
      "clip_ratio/region_mean": 0.0015016225625004154,
      "epoch": 0.3219597550306212,
      "grad_norm": 0.1322672963142395,
      "learning_rate": 1e-06,
      "loss": -0.0079,
      "step": 138
    },
    {
      "clip_ratio/high_max": 0.0025541731010889634,
      "clip_ratio/high_mean": 0.0009663519540481502,
      "clip_ratio/low_mean": 0.0007721306756138802,
      "clip_ratio/low_min": 4.925636767438846e-05,
      "clip_ratio/region_mean": 0.0017384826496709138,
      "epoch": 0.3242927967337416,
      "grad_norm": 0.13279646635055542,
      "learning_rate": 1e-06,
      "loss": -0.0081,
      "step": 139
    },
    {
      "clip_ratio/high_max": 0.0024515801342204213,
      "clip_ratio/high_mean": 0.0009114635831792839,
      "clip_ratio/low_mean": 0.0007970291981109767,
      "clip_ratio/low_min": 8.409956171817612e-05,
      "clip_ratio/region_mean": 0.0017084927821997553,
      "epoch": 0.32662583843686205,
      "grad_norm": 0.13724446296691895,
      "learning_rate": 1e-06,
      "loss": -0.0081,
      "step": 140
    },
    {
      "clip_ratio/high_max": 0.0021759426381322555,
      "clip_ratio/high_mean": 0.0009506453116046032,
      "clip_ratio/low_mean": 0.0007584911727462895,
      "clip_ratio/low_min": 6.67031317789224e-05,
      "clip_ratio/region_mean": 0.0017091364570660517,
      "completions/clipped_ratio": 0.03125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3260.0,
      "completions/mean_length": 750.388427734375,
      "completions/mean_terminated_length": 642.4654541015625,
      "completions/min_length": 200.0,
      "completions/min_terminated_length": 200.0,
      "epoch": 0.3289588801399825,
      "grad_norm": 0.14134296774864197,
      "learning_rate": 1e-06,
      "loss": -0.0084,
      "num_tokens": 21741234.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.2296549528837204,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514803290367126,
      "step": 141
    },
    {
      "clip_ratio/high_max": 0.0024316191629623063,
      "clip_ratio/high_mean": 0.0009980091144825565,
      "clip_ratio/low_mean": 0.0008398861700698035,
      "clip_ratio/low_min": 2.4851598027453292e-05,
      "clip_ratio/region_mean": 0.0018378952809143811,
      "epoch": 0.33129192184310297,
      "grad_norm": 0.1454172432422638,
      "learning_rate": 1e-06,
      "loss": -0.0085,
      "step": 142
    },
    {
      "clip_ratio/high_max": 0.0026167565447394736,
      "clip_ratio/high_mean": 0.0010993376454280224,
      "clip_ratio/low_mean": 0.0009133251132880105,
      "clip_ratio/low_min": 9.388228863826953e-05,
      "clip_ratio/region_mean": 0.002012662727793213,
      "epoch": 0.3336249635462234,
      "grad_norm": 0.13567112386226654,
      "learning_rate": 1e-06,
      "loss": -0.0086,
      "step": 143
    },
    {
      "clip_ratio/high_max": 0.0025247305602533743,
      "clip_ratio/high_mean": 0.001162609776656609,
      "clip_ratio/low_mean": 0.0009580292135069612,
      "clip_ratio/low_min": 2.1999296222929843e-05,
      "clip_ratio/region_mean": 0.002120639015629422,
      "epoch": 0.3359580052493438,
      "grad_norm": 0.1326664388179779,
      "learning_rate": 1e-06,
      "loss": -0.0087,
      "step": 144
    },
    {
      "clip_ratio/high_max": 0.0018576670699985698,
      "clip_ratio/high_mean": 0.0008139792898873566,
      "clip_ratio/low_mean": 0.000670039120450383,
      "clip_ratio/low_min": 2.8118673071730882e-05,
      "clip_ratio/region_mean": 0.0014840184157947078,
      "completions/clipped_ratio": 0.0267857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2907.0,
      "completions/mean_length": 681.7734375,
      "completions/mean_terminated_length": 587.8038940429688,
      "completions/min_length": 151.0,
      "completions/min_terminated_length": 151.0,
      "epoch": 0.33829104695246426,
      "grad_norm": 0.14912009239196777,
      "learning_rate": 1e-06,
      "loss": -0.007,
      "num_tokens": 22330527.0,
      "reward": 0.5758928656578064,
      "reward_std": 0.21725811064243317,
      "rewards/verify_math_reward/mean": 0.5758928656578064,
      "rewards/verify_math_reward/std": 0.49448272585868835,
      "step": 145
    },
    {
      "clip_ratio/high_max": 0.0025272179118474014,
      "clip_ratio/high_mean": 0.0010477910109329969,
      "clip_ratio/low_mean": 0.0007653703396499623,
      "clip_ratio/low_min": 5.663632146024611e-05,
      "clip_ratio/region_mean": 0.0018131613687728532,
      "epoch": 0.34062408865558474,
      "grad_norm": 0.14827348291873932,
      "learning_rate": 1e-06,
      "loss": -0.0072,
      "step": 146
    },
    {
      "clip_ratio/high_max": 0.002495536347851157,
      "clip_ratio/high_mean": 0.0010808328224811703,
      "clip_ratio/low_mean": 0.0009262939474865561,
      "clip_ratio/low_min": 3.505966014927253e-05,
      "clip_ratio/region_mean": 0.0020071267208550125,
      "epoch": 0.3429571303587052,
      "grad_norm": 0.1410999745130539,
      "learning_rate": 1e-06,
      "loss": -0.0073,
      "step": 147
    },
    {
      "clip_ratio/high_max": 0.0025222120893886313,
      "clip_ratio/high_mean": 0.0010745862746261992,
      "clip_ratio/low_mean": 0.0010007284636230906,
      "clip_ratio/low_min": 9.22423450901988e-05,
      "clip_ratio/region_mean": 0.0020753147182404064,
      "epoch": 0.3452901720618256,
      "grad_norm": 0.13900841772556305,
      "learning_rate": 1e-06,
      "loss": -0.0073,
      "step": 148
    },
    {
      "clip_ratio/high_max": 0.001856691567809321,
      "clip_ratio/high_mean": 0.0007320679142139852,
      "clip_ratio/low_mean": 0.00043941574040218256,
      "clip_ratio/low_min": 1.2037750821036752e-05,
      "clip_ratio/region_mean": 0.001171483687357977,
      "completions/clipped_ratio": 0.044642857142857095,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2857.0,
      "completions/mean_length": 756.950927734375,
      "completions/mean_terminated_length": 600.9205322265625,
      "completions/min_length": 134.0,
      "completions/min_terminated_length": 134.0,
      "epoch": 0.34762321376494604,
      "grad_norm": 0.131646066904068,
      "learning_rate": 1e-06,
      "loss": -0.0145,
      "num_tokens": 22934251.0,
      "reward": 0.5881696939468384,
      "reward_std": 0.1741565614938736,
      "rewards/verify_math_reward/mean": 0.5881696343421936,
      "rewards/verify_math_reward/std": 0.4924395978450775,
      "step": 149
    },
    {
      "clip_ratio/high_max": 0.0019837734871543944,
      "clip_ratio/high_mean": 0.000780138074333081,
      "clip_ratio/low_mean": 0.0005641352909151465,
      "clip_ratio/low_min": 1.2037750821036752e-05,
      "clip_ratio/region_mean": 0.0013442733616102487,
      "epoch": 0.34995625546806647,
      "grad_norm": 0.1266884207725525,
      "learning_rate": 1e-06,
      "loss": -0.0147,
      "step": 150
    },
    {
      "clip_ratio/high_max": 0.0021597502782242373,
      "clip_ratio/high_mean": 0.0008289266588690225,
      "clip_ratio/low_mean": 0.0006222497831913643,
      "clip_ratio/low_min": 1.230072848557029e-05,
      "clip_ratio/region_mean": 0.0014511764529743232,
      "epoch": 0.35228929717118695,
      "grad_norm": 0.12174595147371292,
      "learning_rate": 1e-06,
      "loss": -0.0148,
      "step": 151
    },
    {
      "clip_ratio/high_max": 0.002154038018488791,
      "clip_ratio/high_mean": 0.0008754699047130998,
      "clip_ratio/low_mean": 0.0007365460114669986,
      "clip_ratio/low_min": 2.460145697114058e-05,
      "clip_ratio/region_mean": 0.00161201594164595,
      "epoch": 0.3546223388743074,
      "grad_norm": 0.12020301818847656,
      "learning_rate": 1e-06,
      "loss": -0.0148,
      "step": 152
    },
    {
      "clip_ratio/high_max": 0.0022428504744311795,
      "clip_ratio/high_mean": 0.0008285467338282615,
      "clip_ratio/low_mean": 0.0007076413589857111,
      "clip_ratio/low_min": 1.1980065210082103e-05,
      "clip_ratio/region_mean": 0.0015361880941782147,
      "completions/clipped_ratio": 0.0357142857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3907.0,
      "completions/mean_length": 724.7801513671875,
      "completions/mean_terminated_length": 599.920166015625,
      "completions/min_length": 135.0,
      "completions/min_terminated_length": 135.0,
      "epoch": 0.3569553805774278,
      "grad_norm": 0.1613074690103531,
      "learning_rate": 1e-06,
      "loss": -0.016,
      "num_tokens": 23548446.0,
      "reward": 0.5613839626312256,
      "reward_std": 0.19839511811733246,
      "rewards/verify_math_reward/mean": 0.5613839030265808,
      "rewards/verify_math_reward/std": 0.496494859457016,
      "step": 153
    },
    {
      "clip_ratio/high_max": 0.002339055143238511,
      "clip_ratio/high_mean": 0.0009058997620741138,
      "clip_ratio/low_mean": 0.0006424031689675758,
      "clip_ratio/low_min": 4.490345145313768e-05,
      "clip_ratio/region_mean": 0.001548302905575838,
      "epoch": 0.35928842228054825,
      "grad_norm": 0.15127702057361603,
      "learning_rate": 1e-06,
      "loss": -0.0161,
      "step": 154
    },
    {
      "clip_ratio/high_max": 0.0029553086278610863,
      "clip_ratio/high_mean": 0.0010807084327097982,
      "clip_ratio/low_mean": 0.0008939045674196677,
      "clip_ratio/low_min": 3.684923558466835e-05,
      "clip_ratio/region_mean": 0.001974613027414307,
      "epoch": 0.36162146398366873,
      "grad_norm": 0.1411028504371643,
      "learning_rate": 1e-06,
      "loss": -0.0163,
      "step": 155
    },
    {
      "clip_ratio/high_max": 0.0028567789049702697,
      "clip_ratio/high_mean": 0.0010759717879409436,
      "clip_ratio/low_mean": 0.0010138653688045451,
      "clip_ratio/low_min": 8.965483993961243e-05,
      "clip_ratio/region_mean": 0.002089837238600012,
      "epoch": 0.36395450568678916,
      "grad_norm": 0.13891728222370148,
      "learning_rate": 1e-06,
      "loss": -0.0164,
      "step": 156
    },
    {
      "clip_ratio/high_max": 0.0019052833595196716,
      "clip_ratio/high_mean": 0.0007558979359600926,
      "clip_ratio/low_mean": 0.0005618926061288221,
      "clip_ratio/low_min": 1.3661202501680236e-05,
      "clip_ratio/region_mean": 0.0013177905493648723,
      "completions/clipped_ratio": 0.030133928571428603,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3188.0,
      "completions/mean_length": 712.7745971679688,
      "completions/mean_terminated_length": 607.6571044921875,
      "completions/min_length": 190.0,
      "completions/min_terminated_length": 190.0,
      "epoch": 0.3662875473899096,
      "grad_norm": 0.1499905288219452,
      "learning_rate": 1e-06,
      "loss": -0.0182,
      "num_tokens": 24178852.0,
      "reward": 0.5625,
      "reward_std": 0.18693754076957703,
      "rewards/verify_math_reward/mean": 0.5625,
      "rewards/verify_math_reward/std": 0.49635544419288635,
      "step": 157
    },
    {
      "clip_ratio/high_max": 0.0022266593587119132,
      "clip_ratio/high_mean": 0.0008535192591807572,
      "clip_ratio/low_mean": 0.0005987178856230457,
      "clip_ratio/low_min": 7.914327306934865e-05,
      "clip_ratio/region_mean": 0.0014522371420753188,
      "epoch": 0.36862058909303,
      "grad_norm": 0.14097832143306732,
      "learning_rate": 1e-06,
      "loss": -0.0184,
      "step": 158
    },
    {
      "clip_ratio/high_max": 0.0020058465452166274,
      "clip_ratio/high_mean": 0.0008499585565004963,
      "clip_ratio/low_mean": 0.0007910596796136815,
      "clip_ratio/low_min": 9.284633597417269e-05,
      "clip_ratio/region_mean": 0.001641018214286305,
      "epoch": 0.37095363079615046,
      "grad_norm": 0.13096092641353607,
      "learning_rate": 1e-06,
      "loss": -0.0185,
      "step": 159
    },
    {
      "clip_ratio/high_max": 0.0022000707976985723,
      "clip_ratio/high_mean": 0.0009491592900303658,
      "clip_ratio/low_mean": 0.0007415883028443204,
      "clip_ratio/low_min": 8.971292754722526e-05,
      "clip_ratio/region_mean": 0.0016907475728658028,
      "epoch": 0.37328667249927094,
      "grad_norm": 0.13571088016033173,
      "learning_rate": 1e-06,
      "loss": -0.0186,
      "step": 160
    },
    {
      "clip_ratio/high_max": 0.0017179410206153989,
      "clip_ratio/high_mean": 0.0006355128489303752,
      "clip_ratio/low_mean": 0.0005784768000012264,
      "clip_ratio/low_min": 2.933319137810031e-05,
      "clip_ratio/region_mean": 0.0012139896389271598,
      "completions/clipped_ratio": 0.0513392857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3016.0,
      "completions/mean_length": 831.2489013671875,
      "completions/mean_terminated_length": 654.5682373046875,
      "completions/min_length": 206.0,
      "completions/min_terminated_length": 206.0,
      "epoch": 0.3756197142023914,
      "grad_norm": 0.13254326581954956,
      "learning_rate": 1e-06,
      "loss": -0.0197,
      "num_tokens": 24837019.0,
      "reward": 0.5212053656578064,
      "reward_std": 0.1868615597486496,
      "rewards/verify_math_reward/mean": 0.5212053656578064,
      "rewards/verify_math_reward/std": 0.49982914328575134,
      "step": 161
    },
    {
      "clip_ratio/high_max": 0.002329044058569707,
      "clip_ratio/high_mean": 0.0008827270321489777,
      "clip_ratio/low_mean": 0.0006735311708325753,
      "clip_ratio/low_min": 1.187310044770129e-05,
      "clip_ratio/region_mean": 0.0015562582266284153,
      "epoch": 0.3779527559055118,
      "grad_norm": 0.1271568387746811,
      "learning_rate": 1e-06,
      "loss": -0.0199,
      "step": 162
    },
    {
      "clip_ratio/high_max": 0.0021151488945179153,
      "clip_ratio/high_mean": 0.0008055039343162207,
      "clip_ratio/low_mean": 0.0008194172060029814,
      "clip_ratio/low_min": 6.695355841657147e-05,
      "clip_ratio/region_mean": 0.0016249211257672869,
      "epoch": 0.38028579760863224,
      "grad_norm": 0.12323304265737534,
      "learning_rate": 1e-06,
      "loss": -0.02,
      "step": 163
    },
    {
      "clip_ratio/high_max": 0.0023609396230312996,
      "clip_ratio/high_mean": 0.0008703189359948738,
      "clip_ratio/low_mean": 0.0008447565996902995,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017150755229522474,
      "epoch": 0.3826188393117527,
      "grad_norm": 0.11949825286865234,
      "learning_rate": 1e-06,
      "loss": -0.02,
      "step": 164
    },
    {
      "clip_ratio/high_max": 0.0016120658874569926,
      "clip_ratio/high_mean": 0.0006476519338320941,
      "clip_ratio/low_mean": 0.0004724546288343845,
      "clip_ratio/low_min": 2.183024844271131e-05,
      "clip_ratio/region_mean": 0.0011201065572095104,
      "completions/clipped_ratio": 0.0357142857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3809.0,
      "completions/mean_length": 746.591552734375,
      "completions/mean_terminated_length": 622.5393676757812,
      "completions/min_length": 163.0,
      "completions/min_terminated_length": 163.0,
      "epoch": 0.38495188101487315,
      "grad_norm": 0.1378088742494583,
      "learning_rate": 1e-06,
      "loss": -0.0108,
      "num_tokens": 25466093.0,
      "reward": 0.53125,
      "reward_std": 0.17765209078788757,
      "rewards/verify_math_reward/mean": 0.53125,
      "rewards/verify_math_reward/std": 0.4993011951446533,
      "step": 165
    },
    {
      "clip_ratio/high_max": 0.0019100851532130037,
      "clip_ratio/high_mean": 0.0007693097859373665,
      "clip_ratio/low_mean": 0.0006120883754192619,
      "clip_ratio/low_min": 1.1038502634619363e-05,
      "clip_ratio/region_mean": 0.0013813981604471337,
      "epoch": 0.3872849227179936,
      "grad_norm": 0.13167965412139893,
      "learning_rate": 1e-06,
      "loss": -0.0109,
      "step": 166
    },
    {
      "clip_ratio/high_max": 0.0020490221177169587,
      "clip_ratio/high_mean": 0.0007186526363511803,
      "clip_ratio/low_mean": 0.0005681742277374724,
      "clip_ratio/low_min": 1.5794794308021665e-05,
      "clip_ratio/region_mean": 0.0012868268458987586,
      "epoch": 0.389617964421114,
      "grad_norm": 0.13167418539524078,
      "learning_rate": 1e-06,
      "loss": -0.011,
      "step": 167
    },
    {
      "clip_ratio/high_max": 0.0022456580627476797,
      "clip_ratio/high_mean": 0.0008425064788752934,
      "clip_ratio/low_mean": 0.000783710835094098,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001626217275770614,
      "epoch": 0.39195100612423445,
      "grad_norm": 0.1238115131855011,
      "learning_rate": 1e-06,
      "loss": -0.0111,
      "step": 168
    },
    {
      "clip_ratio/high_max": 0.002102460159221664,
      "clip_ratio/high_mean": 0.0007984250751178479,
      "clip_ratio/low_mean": 0.0005976143966108793,
      "clip_ratio/low_min": 1.253258506039856e-05,
      "clip_ratio/region_mean": 0.0013960394826426636,
      "completions/clipped_ratio": 0.0323660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3499.0,
      "completions/mean_length": 747.599365234375,
      "completions/mean_terminated_length": 635.5997314453125,
      "completions/min_length": 95.0,
      "completions/min_terminated_length": 95.0,
      "epoch": 0.39428404782735493,
      "grad_norm": 0.14102819561958313,
      "learning_rate": 1e-06,
      "loss": 0.0138,
      "num_tokens": 26103294.0,
      "reward": 0.6015625,
      "reward_std": 0.20320719480514526,
      "rewards/verify_math_reward/mean": 0.6015625,
      "rewards/verify_math_reward/std": 0.48984986543655396,
      "step": 169
    },
    {
      "clip_ratio/high_max": 0.002048708603979321,
      "clip_ratio/high_mean": 0.000856896496770787,
      "clip_ratio/low_mean": 0.0007294035949598765,
      "clip_ratio/low_min": 2.252387821499724e-05,
      "clip_ratio/region_mean": 0.0015863000808167271,
      "epoch": 0.39661708953047536,
      "grad_norm": 0.13521471619606018,
      "learning_rate": 1e-06,
      "loss": 0.0136,
      "step": 170
    },
    {
      "clip_ratio/high_max": 0.0022176205384312198,
      "clip_ratio/high_mean": 0.0008204550431401003,
      "clip_ratio/low_mean": 0.00080624979455024,
      "clip_ratio/low_min": 1.025935671350453e-05,
      "clip_ratio/region_mean": 0.0016267048486042768,
      "epoch": 0.3989501312335958,
      "grad_norm": 0.132305309176445,
      "learning_rate": 1e-06,
      "loss": 0.0135,
      "step": 171
    },
    {
      "clip_ratio/high_max": 0.002695338276680559,
      "clip_ratio/high_mean": 0.0010746097032097168,
      "clip_ratio/low_mean": 0.0008721355006855447,
      "clip_ratio/low_min": 3.679356450447813e-05,
      "clip_ratio/region_mean": 0.0019467451784294099,
      "epoch": 0.4012831729367162,
      "grad_norm": 0.12492279708385468,
      "learning_rate": 1e-06,
      "loss": 0.0134,
      "step": 172
    },
    {
      "clip_ratio/high_max": 0.0018454058672432438,
      "clip_ratio/high_mean": 0.0007141997180042381,
      "clip_ratio/low_mean": 0.0004877425571976346,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012019422865705565,
      "completions/clipped_ratio": 0.0279017857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2615.0,
      "completions/mean_length": 666.3002319335938,
      "completions/mean_terminated_length": 567.8587646484375,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 0.4036162146398367,
      "grad_norm": 0.14558936655521393,
      "learning_rate": 1e-06,
      "loss": -0.0112,
      "num_tokens": 26684091.0,
      "reward": 0.6194196939468384,
      "reward_std": 0.17836888134479523,
      "rewards/verify_math_reward/mean": 0.6194196343421936,
      "rewards/verify_math_reward/std": 0.48580074310302734,
      "step": 173
    },
    {
      "clip_ratio/high_max": 0.0020503912965068594,
      "clip_ratio/high_mean": 0.0008501818829245167,
      "clip_ratio/low_mean": 0.0005826316228194628,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001432813543942757,
      "epoch": 0.40594925634295714,
      "grad_norm": 0.13812094926834106,
      "learning_rate": 1e-06,
      "loss": -0.0113,
      "step": 174
    },
    {
      "clip_ratio/high_max": 0.0021088762587169185,
      "clip_ratio/high_mean": 0.000818512820842443,
      "clip_ratio/low_mean": 0.0007027353531157132,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001521248159406241,
      "epoch": 0.4082822980460776,
      "grad_norm": 0.13025124371051788,
      "learning_rate": 1e-06,
      "loss": -0.0114,
      "step": 175
    },
    {
      "clip_ratio/high_max": 0.0024314813636010513,
      "clip_ratio/high_mean": 0.0009654291134211235,
      "clip_ratio/low_mean": 0.0008212085613195086,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017866376874735579,
      "epoch": 0.410615339749198,
      "grad_norm": 0.1292300820350647,
      "learning_rate": 1e-06,
      "loss": -0.0115,
      "step": 176
    },
    {
      "clip_ratio/high_max": 0.0019029411487281322,
      "clip_ratio/high_mean": 0.0007594223316118587,
      "clip_ratio/low_mean": 0.0005212702262724633,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001280692555155838,
      "completions/clipped_ratio": 0.056919642857142905,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3785.0,
      "completions/mean_length": 846.482177734375,
      "completions/mean_terminated_length": 650.357421875,
      "completions/min_length": 131.0,
      "completions/min_terminated_length": 131.0,
      "epoch": 0.41294838145231844,
      "grad_norm": 0.14513805508613586,
      "learning_rate": 1e-06,
      "loss": -0.0296,
      "num_tokens": 27327035.0,
      "reward": 0.5915178656578064,
      "reward_std": 0.195994034409523,
      "rewards/verify_math_reward/mean": 0.5915178656578064,
      "rewards/verify_math_reward/std": 0.49182769656181335,
      "step": 177
    },
    {
      "clip_ratio/high_max": 0.0019132906018057838,
      "clip_ratio/high_mean": 0.0008263411073130555,
      "clip_ratio/low_mean": 0.0006613379118789453,
      "clip_ratio/low_min": 4.798547797690844e-05,
      "clip_ratio/region_mean": 0.0014876790010021068,
      "epoch": 0.4152814231554389,
      "grad_norm": 0.13300026953220367,
      "learning_rate": 1e-06,
      "loss": -0.0297,
      "step": 178
    },
    {
      "clip_ratio/high_max": 0.002202157396823168,
      "clip_ratio/high_mean": 0.0009413544539711438,
      "clip_ratio/low_mean": 0.0007774288296786835,
      "clip_ratio/low_min": 2.999307071149815e-05,
      "clip_ratio/region_mean": 0.0017187832781928591,
      "epoch": 0.41761446485855935,
      "grad_norm": 0.13014237582683563,
      "learning_rate": 1e-06,
      "loss": -0.0298,
      "step": 179
    },
    {
      "clip_ratio/high_max": 0.0022671789629384875,
      "clip_ratio/high_mean": 0.0009622640500310808,
      "clip_ratio/low_mean": 0.0008495632300764555,
      "clip_ratio/low_min": 6.966615637793439e-05,
      "clip_ratio/region_mean": 0.0018118272419087589,
      "epoch": 0.4199475065616798,
      "grad_norm": 0.127271369099617,
      "learning_rate": 1e-06,
      "loss": -0.0299,
      "step": 180
    },
    {
      "clip_ratio/high_max": 0.0017390164030075539,
      "clip_ratio/high_mean": 0.000820291770651238,
      "clip_ratio/low_mean": 0.0005455044865811942,
      "clip_ratio/low_min": 1.134713147621369e-05,
      "clip_ratio/region_mean": 0.0013657962263096124,
      "completions/clipped_ratio": 0.041294642857142905,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3679.0,
      "completions/mean_length": 823.5189819335938,
      "completions/mean_terminated_length": 682.5623168945312,
      "completions/min_length": 181.0,
      "completions/min_terminated_length": 181.0,
      "epoch": 0.4222805482648002,
      "grad_norm": 0.13623927533626556,
      "learning_rate": 1e-06,
      "loss": -0.0081,
      "num_tokens": 28001804.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.20613498985767365,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 181
    },
    {
      "clip_ratio/high_max": 0.0019057195422647055,
      "clip_ratio/high_mean": 0.0009002240440167952,
      "clip_ratio/low_mean": 0.0007010096378508024,
      "clip_ratio/low_min": 2.528395452827681e-05,
      "clip_ratio/region_mean": 0.0016012336964195129,
      "epoch": 0.4246135899679207,
      "grad_norm": 0.1361505389213562,
      "learning_rate": 1e-06,
      "loss": -0.0082,
      "step": 182
    },
    {
      "clip_ratio/high_max": 0.0020468775073823053,
      "clip_ratio/high_mean": 0.000950065310462378,
      "clip_ratio/low_mean": 0.0007298784039448947,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001679943688941421,
      "epoch": 0.42694663167104113,
      "grad_norm": 0.16369010508060455,
      "learning_rate": 1e-06,
      "loss": -0.0083,
      "step": 183
    },
    {
      "clip_ratio/high_max": 0.0023009092692518607,
      "clip_ratio/high_mean": 0.0010173881219088798,
      "clip_ratio/low_mean": 0.0009192527759296354,
      "clip_ratio/low_min": 2.4488183044013567e-05,
      "clip_ratio/region_mean": 0.0019366408450878225,
      "epoch": 0.42927967337416156,
      "grad_norm": 0.12531021237373352,
      "learning_rate": 1e-06,
      "loss": -0.0084,
      "step": 184
    },
    {
      "clip_ratio/high_max": 0.001886182013549842,
      "clip_ratio/high_mean": 0.0008456229570583673,
      "clip_ratio/low_mean": 0.0006428303222492104,
      "clip_ratio/low_min": 4.2322684748796746e-05,
      "clip_ratio/region_mean": 0.0014884532793075778,
      "completions/clipped_ratio": 0.0479910714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4070.0,
      "completions/mean_length": 827.0301513671875,
      "completions/mean_terminated_length": 662.2402954101562,
      "completions/min_length": 132.0,
      "completions/min_terminated_length": 132.0,
      "epoch": 0.431612715077282,
      "grad_norm": 0.15010322630405426,
      "learning_rate": 1e-06,
      "loss": -0.0112,
      "num_tokens": 28651591.0,
      "reward": 0.5803571939468384,
      "reward_std": 0.21241775155067444,
      "rewards/verify_math_reward/mean": 0.5803571343421936,
      "rewards/verify_math_reward/std": 0.4937761127948761,
      "step": 185
    },
    {
      "clip_ratio/high_max": 0.0018301246491319034,
      "clip_ratio/high_mean": 0.0008700616945134243,
      "clip_ratio/low_mean": 0.0007397959761874517,
      "clip_ratio/low_min": 4.726601855509216e-05,
      "clip_ratio/region_mean": 0.001609857652510982,
      "epoch": 0.4339457567804024,
      "grad_norm": 0.14428773522377014,
      "learning_rate": 1e-06,
      "loss": -0.0114,
      "step": 186
    },
    {
      "clip_ratio/high_max": 0.0021695718314731494,
      "clip_ratio/high_mean": 0.0010226818194496445,
      "clip_ratio/low_mean": 0.0008679816110088723,
      "clip_ratio/low_min": 5.975446765660308e-05,
      "clip_ratio/region_mean": 0.0018906634286395274,
      "epoch": 0.4362787984835229,
      "grad_norm": 0.14368951320648193,
      "learning_rate": 1e-06,
      "loss": -0.0115,
      "step": 187
    },
    {
      "clip_ratio/high_max": 0.0022048446626286022,
      "clip_ratio/high_mean": 0.0010107983544003218,
      "clip_ratio/low_mean": 0.0009332002828159602,
      "clip_ratio/low_min": 0.00012687695470958715,
      "clip_ratio/region_mean": 0.001943998628121335,
      "epoch": 0.43861184018664334,
      "grad_norm": 0.13704383373260498,
      "learning_rate": 1e-06,
      "loss": -0.0116,
      "step": 188
    },
    {
      "clip_ratio/high_max": 0.0022271957059274428,
      "clip_ratio/high_mean": 0.0008861481364874635,
      "clip_ratio/low_mean": 0.0005503760521605727,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014365241913765203,
      "completions/clipped_ratio": 0.060267857142857095,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2837.0,
      "completions/mean_length": 823.7098388671875,
      "completions/mean_terminated_length": 613.8480224609375,
      "completions/min_length": 146.0,
      "completions/min_terminated_length": 146.0,
      "epoch": 0.4409448818897638,
      "grad_norm": 0.1536131203174591,
      "learning_rate": 1e-06,
      "loss": -0.0198,
      "num_tokens": 29263099.0,
      "reward": 0.5479910969734192,
      "reward_std": 0.1965574473142624,
      "rewards/verify_math_reward/mean": 0.5479910969734192,
      "rewards/verify_math_reward/std": 0.49796950817108154,
      "step": 189
    },
    {
      "clip_ratio/high_max": 0.0025035149592440575,
      "clip_ratio/high_mean": 0.0010717627228586935,
      "clip_ratio/low_mean": 0.0006566193624166772,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017283820998272859,
      "epoch": 0.4432779235928842,
      "grad_norm": 0.1450500190258026,
      "learning_rate": 1e-06,
      "loss": -0.0199,
      "step": 190
    },
    {
      "clip_ratio/high_max": 0.002502236246073153,
      "clip_ratio/high_mean": 0.0010475442395545542,
      "clip_ratio/low_mean": 0.0008652243286633166,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019127685518469661,
      "epoch": 0.4456109652960047,
      "grad_norm": 0.14346785843372345,
      "learning_rate": 1e-06,
      "loss": -0.0201,
      "step": 191
    },
    {
      "clip_ratio/high_max": 0.0028825998524553142,
      "clip_ratio/high_mean": 0.0011423403193475679,
      "clip_ratio/low_mean": 0.0009022599560921662,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002044600318185985,
      "epoch": 0.4479440069991251,
      "grad_norm": 0.13926590979099274,
      "learning_rate": 1e-06,
      "loss": -0.0202,
      "step": 192
    },
    {
      "clip_ratio/high_max": 0.0017455844244977925,
      "clip_ratio/high_mean": 0.0007080485738697462,
      "clip_ratio/low_mean": 0.0004908380342385499,
      "clip_ratio/low_min": 1.2672343473241199e-05,
      "clip_ratio/region_mean": 0.0011988865880994126,
      "completions/clipped_ratio": 0.0546875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3357.0,
      "completions/mean_length": 821.0848388671875,
      "completions/mean_terminated_length": 631.6268920898438,
      "completions/min_length": 209.0,
      "completions/min_terminated_length": 209.0,
      "epoch": 0.45027704870224555,
      "grad_norm": 0.15021604299545288,
      "learning_rate": 1e-06,
      "loss": -0.0227,
      "num_tokens": 29881103.0,
      "reward": 0.527901828289032,
      "reward_std": 0.17585225403308868,
      "rewards/verify_math_reward/mean": 0.5279017686843872,
      "rewards/verify_math_reward/std": 0.49949970841407776,
      "step": 193
    },
    {
      "clip_ratio/high_max": 0.002242021197162103,
      "clip_ratio/high_mean": 0.000786731641710503,
      "clip_ratio/low_mean": 0.0006091638706493541,
      "clip_ratio/low_min": 5.467925439006649e-05,
      "clip_ratio/region_mean": 0.0013958955023554154,
      "epoch": 0.452610090405366,
      "grad_norm": 0.13321596384048462,
      "learning_rate": 1e-06,
      "loss": -0.0229,
      "step": 194
    },
    {
      "clip_ratio/high_max": 0.002216878390754573,
      "clip_ratio/high_mean": 0.0008677324040036183,
      "clip_ratio/low_mean": 0.0007575850731882383,
      "clip_ratio/low_min": 2.8709254365821835e-05,
      "clip_ratio/region_mean": 0.001625317454454489,
      "epoch": 0.4549431321084864,
      "grad_norm": 0.12558101117610931,
      "learning_rate": 1e-06,
      "loss": -0.0231,
      "step": 195
    },
    {
      "clip_ratio/high_max": 0.0021615528021357022,
      "clip_ratio/high_mean": 0.0008111449260468362,
      "clip_ratio/low_mean": 0.0008866158423188608,
      "clip_ratio/low_min": 6.265901265578577e-05,
      "clip_ratio/region_mean": 0.0016977607738226652,
      "epoch": 0.4572761738116069,
      "grad_norm": 0.12740087509155273,
      "learning_rate": 1e-06,
      "loss": -0.0231,
      "step": 196
    },
    {
      "clip_ratio/high_max": 0.002232287883089157,
      "clip_ratio/high_mean": 0.0009673337535787141,
      "clip_ratio/low_mean": 0.0005202810407354264,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001487614783400204,
      "completions/clipped_ratio": 0.078125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3615.0,
      "completions/mean_length": 888.4888916015625,
      "completions/mean_terminated_length": 616.6658935546875,
      "completions/min_length": 151.0,
      "completions/min_terminated_length": 151.0,
      "epoch": 0.45960921551472733,
      "grad_norm": 0.15869812667369843,
      "learning_rate": 1e-06,
      "loss": -0.0224,
      "num_tokens": 30493005.0,
      "reward": 0.5814732313156128,
      "reward_std": 0.191481813788414,
      "rewards/verify_math_reward/mean": 0.5814732313156128,
      "rewards/verify_math_reward/std": 0.4935929775238037,
      "step": 197
    },
    {
      "clip_ratio/high_max": 0.0022578981042897794,
      "clip_ratio/high_mean": 0.001041828054439975,
      "clip_ratio/low_mean": 0.0005890234124308336,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016308514786942396,
      "epoch": 0.46194225721784776,
      "grad_norm": 0.140419140458107,
      "learning_rate": 1e-06,
      "loss": -0.0225,
      "step": 198
    },
    {
      "clip_ratio/high_max": 0.0025456565563217737,
      "clip_ratio/high_mean": 0.0011061316454288317,
      "clip_ratio/low_mean": 0.0007613937241330859,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001867525355919497,
      "epoch": 0.4642752989209682,
      "grad_norm": 0.1353747397661209,
      "learning_rate": 1e-06,
      "loss": -0.0226,
      "step": 199
    },
    {
      "clip_ratio/high_max": 0.0024561806458223145,
      "clip_ratio/high_mean": 0.0011150241025461582,
      "clip_ratio/low_mean": 0.0008858814271661686,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020009055187983904,
      "epoch": 0.4666083406240887,
      "grad_norm": 0.13788971304893494,
      "learning_rate": 1e-06,
      "loss": -0.0228,
      "step": 200
    },
    {
      "clip_ratio/high_max": 0.0020815222633245867,
      "clip_ratio/high_mean": 0.0008058053899731021,
      "clip_ratio/low_mean": 0.0007494486671930645,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015552540426142514,
      "completions/clipped_ratio": 0.0714285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3908.0,
      "completions/mean_length": 894.0234985351562,
      "completions/mean_terminated_length": 647.7175903320312,
      "completions/min_length": 165.0,
      "completions/min_terminated_length": 165.0,
      "epoch": 0.4689413823272091,
      "grad_norm": 0.16608601808547974,
      "learning_rate": 1e-06,
      "loss": -0.0029,
      "num_tokens": 31113114.0,
      "reward": 0.606026828289032,
      "reward_std": 0.19621339440345764,
      "rewards/verify_math_reward/mean": 0.6060267686843872,
      "rewards/verify_math_reward/std": 0.48890194296836853,
      "step": 201
    },
    {
      "clip_ratio/high_max": 0.0024526705165044405,
      "clip_ratio/high_mean": 0.0009260617298423313,
      "clip_ratio/low_mean": 0.0008891882062016521,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001815249961509835,
      "epoch": 0.47127442403032954,
      "grad_norm": 0.15323638916015625,
      "learning_rate": 1e-06,
      "loss": -0.0031,
      "step": 202
    },
    {
      "clip_ratio/high_max": 0.0025142339000012726,
      "clip_ratio/high_mean": 0.0010169709958063322,
      "clip_ratio/low_mean": 0.0009712312239571474,
      "clip_ratio/low_min": 1.633986903470941e-05,
      "clip_ratio/region_mean": 0.0019882022024830803,
      "epoch": 0.47360746573345,
      "grad_norm": 0.14349707961082458,
      "learning_rate": 1e-06,
      "loss": -0.0033,
      "step": 203
    },
    {
      "clip_ratio/high_max": 0.002481605850334745,
      "clip_ratio/high_mean": 0.0010075547861561063,
      "clip_ratio/low_mean": 0.001105919905967312,
      "clip_ratio/low_min": 3.267973806941882e-05,
      "clip_ratio/region_mean": 0.0021134746784809977,
      "epoch": 0.4759405074365704,
      "grad_norm": 0.14359316229820251,
      "learning_rate": 1e-06,
      "loss": -0.0033,
      "step": 204
    },
    {
      "clip_ratio/high_max": 0.0021208442267379723,
      "clip_ratio/high_mean": 0.0008455073002551217,
      "clip_ratio/low_mean": 0.000589847798437404,
      "clip_ratio/low_min": 1.098997745430097e-05,
      "clip_ratio/region_mean": 0.0014353550868690945,
      "completions/clipped_ratio": 0.0647321428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3849.0,
      "completions/mean_length": 836.6183471679688,
      "completions/mean_terminated_length": 611.0286865234375,
      "completions/min_length": 164.0,
      "completions/min_terminated_length": 164.0,
      "epoch": 0.4782735491396909,
      "grad_norm": 0.1829909235239029,
      "learning_rate": 1e-06,
      "loss": -0.0118,
      "num_tokens": 31704844.0,
      "reward": 0.6350446939468384,
      "reward_std": 0.17585085332393646,
      "rewards/verify_math_reward/mean": 0.6350446343421936,
      "rewards/verify_math_reward/std": 0.481686532497406,
      "step": 205
    },
    {
      "clip_ratio/high_max": 0.00222759220196167,
      "clip_ratio/high_mean": 0.0009623481091693975,
      "clip_ratio/low_mean": 0.0005849398330610711,
      "clip_ratio/low_min": 2.197995490860194e-05,
      "clip_ratio/region_mean": 0.0015472879240405746,
      "epoch": 0.4806065908428113,
      "grad_norm": 0.1492559164762497,
      "learning_rate": 1e-06,
      "loss": -0.0119,
      "step": 206
    },
    {
      "clip_ratio/high_max": 0.0026327979649067856,
      "clip_ratio/high_mean": 0.0010532176293054363,
      "clip_ratio/low_mean": 0.0006820712005719543,
      "clip_ratio/low_min": 2.197995490860194e-05,
      "clip_ratio/region_mean": 0.00173528883169638,
      "epoch": 0.48293963254593175,
      "grad_norm": 0.14454133808612823,
      "learning_rate": 1e-06,
      "loss": -0.0121,
      "step": 207
    },
    {
      "clip_ratio/high_max": 0.0027388943490223028,
      "clip_ratio/high_mean": 0.001097951564588584,
      "clip_ratio/low_mean": 0.0009181690720652114,
      "clip_ratio/low_min": 1.605858233233448e-05,
      "clip_ratio/region_mean": 0.002016120655753184,
      "epoch": 0.4852726742490522,
      "grad_norm": 0.13695617020130157,
      "learning_rate": 1e-06,
      "loss": -0.0122,
      "step": 208
    },
    {
      "clip_ratio/high_max": 0.0018147952578146942,
      "clip_ratio/high_mean": 0.0006773728041480354,
      "clip_ratio/low_mean": 0.0006203562052178313,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012977289879927412,
      "completions/clipped_ratio": 0.0546875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3284.0,
      "completions/mean_length": 784.1239013671875,
      "completions/mean_terminated_length": 592.5277099609375,
      "completions/min_length": 133.0,
      "completions/min_terminated_length": 133.0,
      "epoch": 0.48760571595217267,
      "grad_norm": 0.15552891790866852,
      "learning_rate": 1e-06,
      "loss": 0.0023,
      "num_tokens": 32292467.0,
      "reward": 0.5848214626312256,
      "reward_std": 0.17130544781684875,
      "rewards/verify_math_reward/mean": 0.5848214030265808,
      "rewards/verify_math_reward/std": 0.49302801489830017,
      "step": 209
    },
    {
      "clip_ratio/high_max": 0.0021525007468881086,
      "clip_ratio/high_mean": 0.0008598787881055614,
      "clip_ratio/low_mean": 0.0007220445304483292,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015819232976355124,
      "epoch": 0.4899387576552931,
      "grad_norm": 0.15443633496761322,
      "learning_rate": 1e-06,
      "loss": 0.0022,
      "step": 210
    },
    {
      "clip_ratio/high_max": 0.0023462451645173132,
      "clip_ratio/high_mean": 0.0009156435035038157,
      "clip_ratio/low_mean": 0.000828188593914092,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017438321083318442,
      "epoch": 0.49227179935841353,
      "grad_norm": 0.1460455358028412,
      "learning_rate": 1e-06,
      "loss": 0.002,
      "step": 211
    },
    {
      "clip_ratio/high_max": 0.002625642860948574,
      "clip_ratio/high_mean": 0.0009293550920119742,
      "clip_ratio/low_mean": 0.0009462321759201586,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018755872806650586,
      "epoch": 0.49460484106153396,
      "grad_norm": 0.1429179459810257,
      "learning_rate": 1e-06,
      "loss": 0.0019,
      "step": 212
    },
    {
      "clip_ratio/high_max": 0.0020073093110113405,
      "clip_ratio/high_mean": 0.000759914846639731,
      "clip_ratio/low_mean": 0.000462331128801452,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012222459736221936,
      "completions/clipped_ratio": 0.0513392857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3670.0,
      "completions/mean_length": 747.1752319335938,
      "completions/mean_terminated_length": 565.9447021484375,
      "completions/min_length": 118.0,
      "completions/min_terminated_length": 118.0,
      "epoch": 0.4969378827646544,
      "grad_norm": 0.15830470621585846,
      "learning_rate": 1e-06,
      "loss": -0.027,
      "num_tokens": 32863360.0,
      "reward": 0.621651828289032,
      "reward_std": 0.1690923273563385,
      "rewards/verify_math_reward/mean": 0.6216517686843872,
      "rewards/verify_math_reward/std": 0.4852459728717804,
      "step": 213
    },
    {
      "clip_ratio/high_max": 0.0022311388893285766,
      "clip_ratio/high_mean": 0.000997254599496955,
      "clip_ratio/low_mean": 0.0006108485558797838,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016081031499197707,
      "epoch": 0.4992709244677749,
      "grad_norm": 0.1517784148454666,
      "learning_rate": 1e-06,
      "loss": -0.0272,
      "step": 214
    },
    {
      "clip_ratio/high_max": 0.0021470280553330667,
      "clip_ratio/high_mean": 0.0008577115968364524,
      "clip_ratio/low_mean": 0.0007024550213827752,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001560166598210344,
      "epoch": 0.5016039661708953,
      "grad_norm": 0.13804104924201965,
      "learning_rate": 1e-06,
      "loss": -0.0273,
      "step": 215
    },
    {
      "clip_ratio/high_max": 0.0028051892440998927,
      "clip_ratio/high_mean": 0.0010142452265426982,
      "clip_ratio/low_mean": 0.0007994899351615459,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018137351871700957,
      "epoch": 0.5039370078740157,
      "grad_norm": 0.13298538327217102,
      "learning_rate": 1e-06,
      "loss": -0.0274,
      "step": 216
    },
    {
      "clip_ratio/high_max": 0.001884418226836715,
      "clip_ratio/high_mean": 0.0008349796298716683,
      "clip_ratio/low_mean": 0.0004762308617500821,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013112104643369094,
      "completions/clipped_ratio": 0.0401785714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2688.0,
      "completions/mean_length": 732.818115234375,
      "completions/mean_terminated_length": 592.03369140625,
      "completions/min_length": 145.0,
      "completions/min_terminated_length": 145.0,
      "epoch": 0.5062700495771362,
      "grad_norm": 0.15719586610794067,
      "learning_rate": 1e-06,
      "loss": -0.0224,
      "num_tokens": 33466901.0,
      "reward": 0.609375,
      "reward_std": 0.18490804731845856,
      "rewards/verify_math_reward/mean": 0.609375,
      "rewards/verify_math_reward/std": 0.48816296458244324,
      "step": 217
    },
    {
      "clip_ratio/high_max": 0.0024248765403172,
      "clip_ratio/high_mean": 0.0009681946321506985,
      "clip_ratio/low_mean": 0.0005426817997431499,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001510876463726163,
      "epoch": 0.5086030912802566,
      "grad_norm": 0.17562437057495117,
      "learning_rate": 1e-06,
      "loss": -0.0226,
      "step": 218
    },
    {
      "clip_ratio/high_max": 0.0024607518353150226,
      "clip_ratio/high_mean": 0.0010107837515533902,
      "clip_ratio/low_mean": 0.000817063302747556,
      "clip_ratio/low_min": 1.133478417614242e-05,
      "clip_ratio/region_mean": 0.0018278470524819568,
      "epoch": 0.510936132983377,
      "grad_norm": 0.13347984850406647,
      "learning_rate": 1e-06,
      "loss": -0.0227,
      "step": 219
    },
    {
      "clip_ratio/high_max": 0.0026186130417045206,
      "clip_ratio/high_mean": 0.0010899415683525149,
      "clip_ratio/low_mean": 0.0009008191700559109,
      "clip_ratio/low_min": 2.7310465156915598e-05,
      "clip_ratio/region_mean": 0.0019907606911147013,
      "epoch": 0.5132691746864976,
      "grad_norm": 0.13431869447231293,
      "learning_rate": 1e-06,
      "loss": -0.0228,
      "step": 220
    },
    {
      "clip_ratio/high_max": 0.001676053161645541,
      "clip_ratio/high_mean": 0.0005422348167485325,
      "clip_ratio/low_mean": 0.0005290484946272045,
      "clip_ratio/low_min": 2.7246001081948634e-05,
      "clip_ratio/region_mean": 0.0010712833209254313,
      "completions/clipped_ratio": 0.0479910714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2962.0,
      "completions/mean_length": 773.9029541015625,
      "completions/mean_terminated_length": 606.4349365234375,
      "completions/min_length": 102.0,
      "completions/min_terminated_length": 102.0,
      "epoch": 0.515602216389618,
      "grad_norm": 0.1542210876941681,
      "learning_rate": 1e-06,
      "loss": -0.0053,
      "num_tokens": 34075598.0,
      "reward": 0.515625,
      "reward_std": 0.16360442340373993,
      "rewards/verify_math_reward/mean": 0.515625,
      "rewards/verify_math_reward/std": 0.5000349283218384,
      "step": 221
    },
    {
      "clip_ratio/high_max": 0.0019124368445773143,
      "clip_ratio/high_mean": 0.0006477261977124726,
      "clip_ratio/low_mean": 0.000581258345846436,
      "clip_ratio/low_min": 3.816493881458882e-05,
      "clip_ratio/region_mean": 0.0012289845508348662,
      "epoch": 0.5179352580927384,
      "grad_norm": 0.14712415635585785,
      "learning_rate": 1e-06,
      "loss": -0.0054,
      "step": 222
    },
    {
      "clip_ratio/high_max": 0.002208602600148879,
      "clip_ratio/high_mean": 0.000756114433897892,
      "clip_ratio/low_mean": 0.0008703372177478741,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016264516452793032,
      "epoch": 0.5202682997958589,
      "grad_norm": 0.13447299599647522,
      "learning_rate": 1e-06,
      "loss": -0.0056,
      "step": 223
    },
    {
      "clip_ratio/high_max": 0.002081741469737608,
      "clip_ratio/high_mean": 0.0007299962762772338,
      "clip_ratio/low_mean": 0.0009493385659880005,
      "clip_ratio/low_min": 4.9603173465584405e-05,
      "clip_ratio/region_mean": 0.0016793348258943297,
      "epoch": 0.5226013414989793,
      "grad_norm": 0.13474909961223602,
      "learning_rate": 1e-06,
      "loss": -0.0056,
      "step": 224
    },
    {
      "clip_ratio/high_max": 0.0017302923006354831,
      "clip_ratio/high_mean": 0.0007095045375535847,
      "clip_ratio/low_mean": 0.0005966503613308305,
      "clip_ratio/low_min": 3.647505218395963e-05,
      "clip_ratio/region_mean": 0.0013061549034318887,
      "completions/clipped_ratio": 0.049107142857142905,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2674.0,
      "completions/mean_length": 749.3158569335938,
      "completions/mean_terminated_length": 576.482421875,
      "completions/min_length": 133.0,
      "completions/min_terminated_length": 133.0,
      "epoch": 0.5249343832020997,
      "grad_norm": 0.16695386171340942,
      "learning_rate": 1e-06,
      "loss": -0.0156,
      "num_tokens": 34653513.0,
      "reward": 0.5892857313156128,
      "reward_std": 0.17799869179725647,
      "rewards/verify_math_reward/mean": 0.5892857313156128,
      "rewards/verify_math_reward/std": 0.49223825335502625,
      "step": 225
    },
    {
      "clip_ratio/high_max": 0.002407783606031444,
      "clip_ratio/high_mean": 0.0008509271483490011,
      "clip_ratio/low_mean": 0.0007258261230163043,
      "clip_ratio/low_min": 1.885938399937004e-05,
      "clip_ratio/region_mean": 0.0015767532750032842,
      "epoch": 0.5272674249052202,
      "grad_norm": 0.1526087522506714,
      "learning_rate": 1e-06,
      "loss": -0.0158,
      "step": 226
    },
    {
      "clip_ratio/high_max": 0.0023025913833407685,
      "clip_ratio/high_mean": 0.00088137284910772,
      "clip_ratio/low_mean": 0.000892662950718659,
      "clip_ratio/low_min": 4.7148459998425096e-05,
      "clip_ratio/region_mean": 0.0017740357870934531,
      "epoch": 0.5296004666083406,
      "grad_norm": 0.14756852388381958,
      "learning_rate": 1e-06,
      "loss": -0.016,
      "step": 227
    },
    {
      "clip_ratio/high_max": 0.002523296330764424,
      "clip_ratio/high_mean": 0.000958762337177177,
      "clip_ratio/low_mean": 0.000998503339360468,
      "clip_ratio/low_min": 9.42969199968502e-06,
      "clip_ratio/region_mean": 0.001957265689270571,
      "epoch": 0.531933508311461,
      "grad_norm": 0.14524367451667786,
      "learning_rate": 1e-06,
      "loss": -0.0161,
      "step": 228
    },
    {
      "clip_ratio/high_max": 0.0017749884609656874,
      "clip_ratio/high_mean": 0.0006844742229077383,
      "clip_ratio/low_mean": 0.0005758627803515992,
      "clip_ratio/low_min": 2.6606637220538687e-05,
      "clip_ratio/region_mean": 0.0012603369905264117,
      "completions/clipped_ratio": 0.044642857142857095,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3527.0,
      "completions/mean_length": 783.2924194335938,
      "completions/mean_terminated_length": 628.4929809570312,
      "completions/min_length": 176.0,
      "completions/min_terminated_length": 176.0,
      "epoch": 0.5342665500145816,
      "grad_norm": 0.15385381877422333,
      "learning_rate": 1e-06,
      "loss": -0.02,
      "num_tokens": 35281367.0,
      "reward": 0.5524553656578064,
      "reward_std": 0.1795709878206253,
      "rewards/verify_math_reward/mean": 0.5524553656578064,
      "rewards/verify_math_reward/std": 0.49751853942871094,
      "step": 229
    },
    {
      "clip_ratio/high_max": 0.002248025863082148,
      "clip_ratio/high_mean": 0.0007510431696573505,
      "clip_ratio/low_mean": 0.0006982658669585362,
      "clip_ratio/low_min": 4.0437347706756555e-05,
      "clip_ratio/region_mean": 0.0014493090311589185,
      "epoch": 0.536599591717702,
      "grad_norm": 0.14346085488796234,
      "learning_rate": 1e-06,
      "loss": -0.0203,
      "step": 230
    },
    {
      "clip_ratio/high_max": 0.002111926947691245,
      "clip_ratio/high_mean": 0.0009057822080649203,
      "clip_ratio/low_mean": 0.0007865216102800332,
      "clip_ratio/low_min": 3.9903771721583325e-05,
      "clip_ratio/region_mean": 0.001692303834715858,
      "epoch": 0.5389326334208224,
      "grad_norm": 0.13601410388946533,
      "learning_rate": 1e-06,
      "loss": -0.0203,
      "step": 231
    },
    {
      "clip_ratio/high_max": 0.0023360310660791583,
      "clip_ratio/high_mean": 0.0008537830599379959,
      "clip_ratio/low_mean": 0.0009508076145721134,
      "clip_ratio/low_min": 5.0289281716686673e-05,
      "clip_ratio/region_mean": 0.0018045906108454801,
      "epoch": 0.5412656751239429,
      "grad_norm": 0.13489511609077454,
      "learning_rate": 1e-06,
      "loss": -0.0204,
      "step": 232
    },
    {
      "clip_ratio/high_max": 0.0014880389171594288,
      "clip_ratio/high_mean": 0.00048186242929659784,
      "clip_ratio/low_mean": 0.00047997832007240504,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0009618407475500135,
      "completions/clipped_ratio": 0.0457589285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3521.0,
      "completions/mean_length": 743.7467041015625,
      "completions/mean_terminated_length": 582.9953002929688,
      "completions/min_length": 125.0,
      "completions/min_terminated_length": 125.0,
      "epoch": 0.5435987168270633,
      "grad_norm": 0.13865098357200623,
      "learning_rate": 1e-06,
      "loss": -0.0151,
      "num_tokens": 35873972.0,
      "reward": 0.6071428656578064,
      "reward_std": 0.11953012645244598,
      "rewards/verify_math_reward/mean": 0.6071428656578064,
      "rewards/verify_math_reward/std": 0.48865827918052673,
      "step": 233
    },
    {
      "clip_ratio/high_max": 0.0019240232759329956,
      "clip_ratio/high_mean": 0.0006175566004458233,
      "clip_ratio/low_mean": 0.0005685450960299931,
      "clip_ratio/low_min": 4.293565507396124e-05,
      "clip_ratio/region_mean": 0.0011861016937473323,
      "epoch": 0.5459317585301837,
      "grad_norm": 0.13254310190677643,
      "learning_rate": 1e-06,
      "loss": -0.0152,
      "step": 234
    },
    {
      "clip_ratio/high_max": 0.0022041650518076494,
      "clip_ratio/high_mean": 0.0006964148251427105,
      "clip_ratio/low_mean": 0.0007112952880561352,
      "clip_ratio/low_min": 5.803156818728894e-05,
      "clip_ratio/region_mean": 0.0014077101077418774,
      "epoch": 0.5482648002333042,
      "grad_norm": 0.12212701886892319,
      "learning_rate": 1e-06,
      "loss": -0.0154,
      "step": 235
    },
    {
      "clip_ratio/high_max": 0.0022669806421617977,
      "clip_ratio/high_mean": 0.0007154976665333379,
      "clip_ratio/low_mean": 0.0007446610470651649,
      "clip_ratio/low_min": 4.724111931864172e-05,
      "clip_ratio/region_mean": 0.0014601587245124392,
      "epoch": 0.5505978419364246,
      "grad_norm": 0.1242891252040863,
      "learning_rate": 1e-06,
      "loss": -0.0154,
      "step": 236
    },
    {
      "clip_ratio/high_max": 0.001739940886182012,
      "clip_ratio/high_mean": 0.0006690709469694411,
      "clip_ratio/low_mean": 0.0006471121196227614,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013161830611352343,
      "completions/clipped_ratio": 0.0379464285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3510.0,
      "completions/mean_length": 718.7522583007812,
      "completions/mean_terminated_length": 585.5429077148438,
      "completions/min_length": 191.0,
      "completions/min_terminated_length": 191.0,
      "epoch": 0.552930883639545,
      "grad_norm": 0.19878889620304108,
      "learning_rate": 1e-06,
      "loss": 0.0002,
      "num_tokens": 36463718.0,
      "reward": 0.6283482313156128,
      "reward_std": 0.16938656568527222,
      "rewards/verify_math_reward/mean": 0.6283482313156128,
      "rewards/verify_math_reward/std": 0.4835159480571747,
      "step": 237
    },
    {
      "clip_ratio/high_max": 0.002143624980817549,
      "clip_ratio/high_mean": 0.0008554658179491526,
      "clip_ratio/low_mean": 0.0008386970785068115,
      "clip_ratio/low_min": 5.576920466410229e-05,
      "clip_ratio/region_mean": 0.00169416287826607,
      "epoch": 0.5552639253426656,
      "grad_norm": 0.15399928390979767,
      "learning_rate": 1e-06,
      "loss": -0.0,
      "step": 238
    },
    {
      "clip_ratio/high_max": 0.0023502816547988914,
      "clip_ratio/high_mean": 0.0009067850733117666,
      "clip_ratio/low_mean": 0.0009722007471282268,
      "clip_ratio/low_min": 5.104478077555541e-05,
      "clip_ratio/region_mean": 0.0018789857713272795,
      "epoch": 0.557596967045786,
      "grad_norm": 0.15232457220554352,
      "learning_rate": 1e-06,
      "loss": -0.0001,
      "step": 239
    },
    {
      "clip_ratio/high_max": 0.0023724731108814012,
      "clip_ratio/high_mean": 0.0009244451975973789,
      "clip_ratio/low_mean": 0.0010867376477108337,
      "clip_ratio/low_min": 4.5410639359033667e-05,
      "clip_ratio/region_mean": 0.0020111828271183185,
      "epoch": 0.5599300087489064,
      "grad_norm": 0.14790889620780945,
      "learning_rate": 1e-06,
      "loss": -0.0002,
      "step": 240
    },
    {
      "clip_ratio/high_max": 0.0016735052049625665,
      "clip_ratio/high_mean": 0.0006779307495889952,
      "clip_ratio/low_mean": 0.0006132294238341274,
      "clip_ratio/low_min": 2.033725832006894e-05,
      "clip_ratio/region_mean": 0.0012911601916130167,
      "completions/clipped_ratio": 0.0691964285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2889.0,
      "completions/mean_length": 889.2645263671875,
      "completions/mean_terminated_length": 650.8741455078125,
      "completions/min_length": 188.0,
      "completions/min_terminated_length": 188.0,
      "epoch": 0.5622630504520268,
      "grad_norm": 0.16441886126995087,
      "learning_rate": 1e-06,
      "loss": -0.0141,
      "num_tokens": 37090571.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.18403972685337067,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 241
    },
    {
      "clip_ratio/high_max": 0.0018577990049379878,
      "clip_ratio/high_mean": 0.0007666599976801081,
      "clip_ratio/low_mean": 0.0006816638942837017,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001448323866497958,
      "epoch": 0.5645960921551473,
      "grad_norm": 0.14850129187107086,
      "learning_rate": 1e-06,
      "loss": -0.0142,
      "step": 242
    },
    {
      "clip_ratio/high_max": 0.002126586354279425,
      "clip_ratio/high_mean": 0.0009070793985301862,
      "clip_ratio/low_mean": 0.0008444239938398823,
      "clip_ratio/low_min": 2.9691153940802906e-05,
      "clip_ratio/region_mean": 0.0017515034051029943,
      "epoch": 0.5669291338582677,
      "grad_norm": 0.13421630859375,
      "learning_rate": 1e-06,
      "loss": -0.0144,
      "step": 243
    },
    {
      "clip_ratio/high_max": 0.002174347777327057,
      "clip_ratio/high_mean": 0.0008673456795804668,
      "clip_ratio/low_mean": 0.0009673706917965319,
      "clip_ratio/low_min": 7.436059331666911e-05,
      "clip_ratio/region_mean": 0.001834716422308702,
      "epoch": 0.5692621755613881,
      "grad_norm": 0.132176011800766,
      "learning_rate": 1e-06,
      "loss": -0.0145,
      "step": 244
    },
    {
      "clip_ratio/high_max": 0.0028683393175015226,
      "clip_ratio/high_mean": 0.0010127493369509466,
      "clip_ratio/low_mean": 0.00075934248707199,
      "clip_ratio/low_min": 1.3736264008912258e-05,
      "clip_ratio/region_mean": 0.0017720918040140532,
      "completions/clipped_ratio": 0.0625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3895.0,
      "completions/mean_length": 881.755615234375,
      "completions/mean_terminated_length": 667.47265625,
      "completions/min_length": 185.0,
      "completions/min_terminated_length": 185.0,
      "epoch": 0.5715952172645086,
      "grad_norm": 0.18064840137958527,
      "learning_rate": 1e-06,
      "loss": -0.0202,
      "num_tokens": 37735216.0,
      "reward": 0.5412946939468384,
      "reward_std": 0.22469766438007355,
      "rewards/verify_math_reward/mean": 0.5412946343421936,
      "rewards/verify_math_reward/std": 0.49857014417648315,
      "step": 245
    },
    {
      "clip_ratio/high_max": 0.002810091638821177,
      "clip_ratio/high_mean": 0.0011130381317343563,
      "clip_ratio/low_mean": 0.0009102133517444599,
      "clip_ratio/low_min": 4.440503107616678e-05,
      "clip_ratio/region_mean": 0.0020232514580129646,
      "epoch": 0.573928258967629,
      "grad_norm": 0.1686396300792694,
      "learning_rate": 1e-06,
      "loss": -0.0204,
      "step": 246
    },
    {
      "clip_ratio/high_max": 0.003302433338831179,
      "clip_ratio/high_mean": 0.0012588584941113368,
      "clip_ratio/low_mean": 0.0010801408207044005,
      "clip_ratio/low_min": 5.4901300245546736e-05,
      "clip_ratio/region_mean": 0.0023389992784359492,
      "epoch": 0.5762613006707495,
      "grad_norm": 0.16458149254322052,
      "learning_rate": 1e-06,
      "loss": -0.0206,
      "step": 247
    },
    {
      "clip_ratio/high_max": 0.0036235110528650694,
      "clip_ratio/high_mean": 0.0013374627342273016,
      "clip_ratio/low_mean": 0.0011865439810208045,
      "clip_ratio/low_min": 0.00012475728908611927,
      "clip_ratio/region_mean": 0.002524006675230339,
      "epoch": 0.57859434237387,
      "grad_norm": 0.15335099399089813,
      "learning_rate": 1e-06,
      "loss": -0.0207,
      "step": 248
    },
    {
      "clip_ratio/high_max": 0.002065000957372831,
      "clip_ratio/high_mean": 0.0008984474952740129,
      "clip_ratio/low_mean": 0.00064315901727241,
      "clip_ratio/low_min": 1.1287700544926338e-05,
      "clip_ratio/region_mean": 0.0015416065143654123,
      "completions/clipped_ratio": 0.0580357142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3344.0,
      "completions/mean_length": 811.1272583007812,
      "completions/mean_terminated_length": 608.74169921875,
      "completions/min_length": 130.0,
      "completions/min_terminated_length": 130.0,
      "epoch": 0.5809273840769904,
      "grad_norm": 0.1811312586069107,
      "learning_rate": 1e-06,
      "loss": -0.0335,
      "num_tokens": 38327650.0,
      "reward": 0.6328125,
      "reward_std": 0.19808951020240784,
      "rewards/verify_math_reward/mean": 0.6328125,
      "rewards/verify_math_reward/std": 0.48230743408203125,
      "step": 249
    },
    {
      "clip_ratio/high_max": 0.002202488452894613,
      "clip_ratio/high_mean": 0.0009702885690785479,
      "clip_ratio/low_mean": 0.0008369570678041782,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018072456441586837,
      "epoch": 0.5832604257801108,
      "grad_norm": 0.17485851049423218,
      "learning_rate": 1e-06,
      "loss": -0.0337,
      "step": 250
    },
    {
      "clip_ratio/high_max": 0.0023402113292831928,
      "clip_ratio/high_mean": 0.0010030769644799875,
      "clip_ratio/low_mean": 0.0009955549285223242,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019986318657174706,
      "epoch": 0.5855934674832313,
      "grad_norm": 0.15953102707862854,
      "learning_rate": 1e-06,
      "loss": -0.0338,
      "step": 251
    },
    {
      "clip_ratio/high_max": 0.002859046173398383,
      "clip_ratio/high_mean": 0.0012013425111945253,
      "clip_ratio/low_mean": 0.001164675115433056,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023660176739213057,
      "epoch": 0.5879265091863517,
      "grad_norm": 0.15245676040649414,
      "learning_rate": 1e-06,
      "loss": -0.034,
      "step": 252
    },
    {
      "clip_ratio/high_max": 0.002454886576742865,
      "clip_ratio/high_mean": 0.0011808079943875782,
      "clip_ratio/low_mean": 0.0007260778120325995,
      "clip_ratio/low_min": 3.780636416195193e-05,
      "clip_ratio/region_mean": 0.001906885787320789,
      "completions/clipped_ratio": 0.0613839285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3406.0,
      "completions/mean_length": 845.630615234375,
      "completions/mean_terminated_length": 633.0618286132812,
      "completions/min_length": 132.0,
      "completions/min_terminated_length": 132.0,
      "epoch": 0.5902595508894721,
      "grad_norm": 0.19520410895347595,
      "learning_rate": 1e-06,
      "loss": -0.0342,
      "num_tokens": 38949823.0,
      "reward": 0.6316964626312256,
      "reward_std": 0.23886480927467346,
      "rewards/verify_math_reward/mean": 0.6316964030265808,
      "rewards/verify_math_reward/std": 0.4826137125492096,
      "step": 253
    },
    {
      "clip_ratio/high_max": 0.0029217097981018014,
      "clip_ratio/high_mean": 0.0012666904185607564,
      "clip_ratio/low_mean": 0.0009204373054672033,
      "clip_ratio/low_min": 7.578930308227427e-05,
      "clip_ratio/region_mean": 0.002187127734941896,
      "epoch": 0.5925925925925926,
      "grad_norm": 0.19551533460617065,
      "learning_rate": 1e-06,
      "loss": -0.0344,
      "step": 254
    },
    {
      "clip_ratio/high_max": 0.003028957042261027,
      "clip_ratio/high_mean": 0.0013827830480295233,
      "clip_ratio/low_mean": 0.0011081993379775668,
      "clip_ratio/low_min": 5.027349016017979e-05,
      "clip_ratio/region_mean": 0.0024909823696361855,
      "epoch": 0.594925634295713,
      "grad_norm": 0.1766534298658371,
      "learning_rate": 1e-06,
      "loss": -0.0346,
      "step": 255
    },
    {
      "clip_ratio/high_max": 0.0029570692277047783,
      "clip_ratio/high_mean": 0.001343818676105002,
      "clip_ratio/low_mean": 0.0012197565392852994,
      "clip_ratio/low_min": 5.0371709221508354e-05,
      "clip_ratio/region_mean": 0.002563575231761206,
      "epoch": 0.5972586759988335,
      "grad_norm": 0.16498038172721863,
      "learning_rate": 1e-06,
      "loss": -0.0347,
      "step": 256
    },
    {
      "clip_ratio/high_max": 0.0020004543803224806,
      "clip_ratio/high_mean": 0.0007045400034257909,
      "clip_ratio/low_mean": 0.0006450886503444053,
      "clip_ratio/low_min": 3.832004949799739e-05,
      "clip_ratio/region_mean": 0.0013496286410372704,
      "completions/clipped_ratio": 0.052455357142857095,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4033.0,
      "completions/mean_length": 819.6406860351562,
      "completions/mean_terminated_length": 638.2638549804688,
      "completions/min_length": 152.0,
      "completions/min_terminated_length": 152.0,
      "epoch": 0.599591717701954,
      "grad_norm": 0.174686998128891,
      "learning_rate": 1e-06,
      "loss": -0.0272,
      "num_tokens": 39589389.0,
      "reward": 0.5703125,
      "reward_std": 0.18776056170463562,
      "rewards/verify_math_reward/mean": 0.5703125,
      "rewards/verify_math_reward/std": 0.49530795216560364,
      "step": 257
    },
    {
      "clip_ratio/high_max": 0.0021911398507654667,
      "clip_ratio/high_mean": 0.0008066736936598318,
      "clip_ratio/low_mean": 0.000846038323288667,
      "clip_ratio/low_min": 2.730450069066137e-05,
      "clip_ratio/region_mean": 0.0016527120096725412,
      "epoch": 0.6019247594050744,
      "grad_norm": 0.15377922356128693,
      "learning_rate": 1e-06,
      "loss": -0.0274,
      "step": 258
    },
    {
      "clip_ratio/high_max": 0.0026595802191877738,
      "clip_ratio/high_mean": 0.0009453495586058125,
      "clip_ratio/low_mean": 0.0009764476362761343,
      "clip_ratio/low_min": 1.9160024748998694e-05,
      "clip_ratio/region_mean": 0.0019217972112528514,
      "epoch": 0.6042578011081948,
      "grad_norm": 0.14722417294979095,
      "learning_rate": 1e-06,
      "loss": -0.0276,
      "step": 259
    },
    {
      "clip_ratio/high_max": 0.002826516800269019,
      "clip_ratio/high_mean": 0.0009853889441728825,
      "clip_ratio/low_mean": 0.0011208343730686465,
      "clip_ratio/low_min": 2.603443499538116e-05,
      "clip_ratio/region_mean": 0.002106223335431423,
      "epoch": 0.6065908428113153,
      "grad_norm": 0.14746461808681488,
      "learning_rate": 1e-06,
      "loss": -0.0277,
      "step": 260
    },
    {
      "clip_ratio/high_max": 0.0017576406498847064,
      "clip_ratio/high_mean": 0.0005976478887532721,
      "clip_ratio/low_mean": 0.0004928772068524268,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0010905250819632784,
      "completions/clipped_ratio": 0.0558035714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3622.0,
      "completions/mean_length": 769.7455444335938,
      "completions/mean_terminated_length": 573.1583862304688,
      "completions/min_length": 174.0,
      "completions/min_terminated_length": 174.0,
      "epoch": 0.6089238845144357,
      "grad_norm": 0.16936379671096802,
      "learning_rate": 1e-06,
      "loss": -0.0272,
      "num_tokens": 40172817.0,
      "reward": 0.598214328289032,
      "reward_std": 0.1515413373708725,
      "rewards/verify_math_reward/mean": 0.5982142686843872,
      "rewards/verify_math_reward/std": 0.49053290486335754,
      "step": 261
    },
    {
      "clip_ratio/high_max": 0.002137923140253406,
      "clip_ratio/high_mean": 0.0007493480406992603,
      "clip_ratio/low_mean": 0.000567885364034737,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013172333929105662,
      "epoch": 0.6112569262175561,
      "grad_norm": 0.15487205982208252,
      "learning_rate": 1e-06,
      "loss": -0.0274,
      "step": 262
    },
    {
      "clip_ratio/high_max": 0.002313453282113187,
      "clip_ratio/high_mean": 0.0008186651230062125,
      "clip_ratio/low_mean": 0.0007745958864688873,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001593260996742174,
      "epoch": 0.6135899679206765,
      "grad_norm": 0.13710205256938934,
      "learning_rate": 1e-06,
      "loss": -0.0276,
      "step": 263
    },
    {
      "clip_ratio/high_max": 0.002353799540287582,
      "clip_ratio/high_mean": 0.0008824724536680151,
      "clip_ratio/low_mean": 0.0008264293965112302,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017089018401748035,
      "epoch": 0.615923009623797,
      "grad_norm": 0.13875557482242584,
      "learning_rate": 1e-06,
      "loss": -0.0276,
      "step": 264
    },
    {
      "clip_ratio/high_max": 0.0013810562340950128,
      "clip_ratio/high_mean": 0.0005493980743267457,
      "clip_ratio/low_mean": 0.0005788339622085914,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0011282320301688742,
      "completions/clipped_ratio": 0.0625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3763.0,
      "completions/mean_length": 908.1875610351562,
      "completions/mean_terminated_length": 695.6666870117188,
      "completions/min_length": 141.0,
      "completions/min_terminated_length": 141.0,
      "epoch": 0.6182560513269175,
      "grad_norm": 0.14903157949447632,
      "learning_rate": 1e-06,
      "loss": -0.0152,
      "num_tokens": 40852849.0,
      "reward": 0.535714328289032,
      "reward_std": 0.17731469869613647,
      "rewards/verify_math_reward/mean": 0.5357142686843872,
      "rewards/verify_math_reward/std": 0.4990014135837555,
      "step": 265
    },
    {
      "clip_ratio/high_max": 0.001720084110274911,
      "clip_ratio/high_mean": 0.0006810845243307995,
      "clip_ratio/low_mean": 0.000695541473760386,
      "clip_ratio/low_min": 6.149285763967782e-05,
      "clip_ratio/region_mean": 0.0013766260017291643,
      "epoch": 0.620589093030038,
      "grad_norm": 0.14406567811965942,
      "learning_rate": 1e-06,
      "loss": -0.0154,
      "step": 266
    },
    {
      "clip_ratio/high_max": 0.0020672293176176026,
      "clip_ratio/high_mean": 0.0007439389773935545,
      "clip_ratio/low_mean": 0.0008149237601173809,
      "clip_ratio/low_min": 6.404508076229831e-05,
      "clip_ratio/region_mean": 0.00155886275388184,
      "epoch": 0.6229221347331584,
      "grad_norm": 0.14669474959373474,
      "learning_rate": 1e-06,
      "loss": -0.0155,
      "step": 267
    },
    {
      "clip_ratio/high_max": 0.0019061784987570718,
      "clip_ratio/high_mean": 0.0007410540856653824,
      "clip_ratio/low_mean": 0.0009376859288749984,
      "clip_ratio/low_min": 6.939301329111913e-05,
      "clip_ratio/region_mean": 0.001678740001807455,
      "epoch": 0.6252551764362788,
      "grad_norm": 0.13039635121822357,
      "learning_rate": 1e-06,
      "loss": -0.0156,
      "step": 268
    },
    {
      "clip_ratio/high_max": 0.0018028199483524077,
      "clip_ratio/high_mean": 0.0006874569044157397,
      "clip_ratio/low_mean": 0.0005106005064590136,
      "clip_ratio/low_min": 2.9585884476546198e-05,
      "clip_ratio/region_mean": 0.0011980574017798062,
      "completions/clipped_ratio": 0.0770089285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3214.0,
      "completions/mean_length": 888.19091796875,
      "completions/mean_terminated_length": 620.5502319335938,
      "completions/min_length": 152.0,
      "completions/min_terminated_length": 152.0,
      "epoch": 0.6275882181393992,
      "grad_norm": 0.15618275105953217,
      "learning_rate": 1e-06,
      "loss": -0.0256,
      "num_tokens": 41449668.0,
      "reward": 0.5658482313156128,
      "reward_std": 0.1510535031557083,
      "rewards/verify_math_reward/mean": 0.5658482313156128,
      "rewards/verify_math_reward/std": 0.49592188000679016,
      "step": 269
    },
    {
      "clip_ratio/high_max": 0.0019446795340627432,
      "clip_ratio/high_mean": 0.0007613918933202513,
      "clip_ratio/low_mean": 0.0007282111346285092,
      "clip_ratio/low_min": 3.0350084671226796e-05,
      "clip_ratio/region_mean": 0.001489602989749983,
      "epoch": 0.6299212598425197,
      "grad_norm": 0.14463922381401062,
      "learning_rate": 1e-06,
      "loss": -0.0258,
      "step": 270
    },
    {
      "clip_ratio/high_max": 0.0021781206960440613,
      "clip_ratio/high_mean": 0.0008448699973087059,
      "clip_ratio/low_mean": 0.0008981303326436318,
      "clip_ratio/low_min": 8.13642363937106e-05,
      "clip_ratio/region_mean": 0.0017430003572371788,
      "epoch": 0.6322543015456401,
      "grad_norm": 0.1415518820285797,
      "learning_rate": 1e-06,
      "loss": -0.026,
      "step": 271
    },
    {
      "clip_ratio/high_max": 0.0020730302603624295,
      "clip_ratio/high_mean": 0.0008484179506922374,
      "clip_ratio/low_mean": 0.000955949011768098,
      "clip_ratio/low_min": 0.00013005232176510617,
      "clip_ratio/region_mean": 0.0018043669479084201,
      "epoch": 0.6345873432487605,
      "grad_norm": 0.13381057977676392,
      "learning_rate": 1e-06,
      "loss": -0.026,
      "step": 272
    },
    {
      "clip_ratio/high_max": 0.002148325598682277,
      "clip_ratio/high_mean": 0.0009207470575347543,
      "clip_ratio/low_mean": 0.0006785067789678578,
      "clip_ratio/low_min": 1.9409937522141263e-05,
      "clip_ratio/region_mean": 0.0015992538355931174,
      "completions/clipped_ratio": 0.0837053571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3847.0,
      "completions/mean_length": 948.7210083007812,
      "completions/mean_terminated_length": 661.2107543945312,
      "completions/min_length": 145.0,
      "completions/min_terminated_length": 145.0,
      "epoch": 0.636920384951881,
      "grad_norm": 0.20500075817108154,
      "learning_rate": 1e-06,
      "loss": -0.0056,
      "num_tokens": 42082890.0,
      "reward": 0.5368303656578064,
      "reward_std": 0.2144351601600647,
      "rewards/verify_math_reward/mean": 0.5368303656578064,
      "rewards/verify_math_reward/std": 0.49892017245292664,
      "step": 273
    },
    {
      "clip_ratio/high_max": 0.0024067259291769005,
      "clip_ratio/high_mean": 0.0010504097954253666,
      "clip_ratio/low_mean": 0.0008104904227366205,
      "clip_ratio/low_min": 1.625487675482873e-05,
      "clip_ratio/region_mean": 0.0018609002800076269,
      "epoch": 0.6392534266550015,
      "grad_norm": 0.18786416947841644,
      "learning_rate": 1e-06,
      "loss": -0.0058,
      "step": 274
    },
    {
      "clip_ratio/high_max": 0.0024999762536026537,
      "clip_ratio/high_mean": 0.0010430742440803442,
      "clip_ratio/low_mean": 0.0010387395932411891,
      "clip_ratio/low_min": 8.285888179671019e-05,
      "clip_ratio/region_mean": 0.0020818138364120387,
      "epoch": 0.6415864683581219,
      "grad_norm": 0.17506757378578186,
      "learning_rate": 1e-06,
      "loss": -0.006,
      "step": 275
    },
    {
      "clip_ratio/high_max": 0.0027001566559192725,
      "clip_ratio/high_mean": 0.001160879499366274,
      "clip_ratio/low_mean": 0.0014031115624675294,
      "clip_ratio/low_min": 0.0001541474302939605,
      "clip_ratio/region_mean": 0.0025639910527388565,
      "epoch": 0.6439195100612424,
      "grad_norm": 0.19004221260547638,
      "learning_rate": 1e-06,
      "loss": -0.0061,
      "step": 276
    },
    {
      "clip_ratio/high_max": 0.001982006178877782,
      "clip_ratio/high_mean": 0.0008011192785488674,
      "clip_ratio/low_mean": 0.0008978966488939477,
      "clip_ratio/low_min": 0.0001257303429156309,
      "clip_ratio/region_mean": 0.0016990159565466456,
      "completions/clipped_ratio": 0.0736607142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4062.0,
      "completions/mean_length": 922.2232666015625,
      "completions/mean_terminated_length": 669.8505859375,
      "completions/min_length": 185.0,
      "completions/min_terminated_length": 185.0,
      "epoch": 0.6462525517643628,
      "grad_norm": 0.2027897983789444,
      "learning_rate": 1e-06,
      "loss": -0.0333,
      "num_tokens": 42728906.0,
      "reward": 0.5569196939468384,
      "reward_std": 0.22837404906749725,
      "rewards/verify_math_reward/mean": 0.5569196343421936,
      "rewards/verify_math_reward/std": 0.49702703952789307,
      "step": 277
    },
    {
      "clip_ratio/high_max": 0.002469796600053087,
      "clip_ratio/high_mean": 0.0010389416602265555,
      "clip_ratio/low_mean": 0.001157696471636882,
      "clip_ratio/low_min": 0.0002515920023142826,
      "clip_ratio/region_mean": 0.00219663812458748,
      "epoch": 0.6485855934674832,
      "grad_norm": 0.18985936045646667,
      "learning_rate": 1e-06,
      "loss": -0.0335,
      "step": 278
    },
    {
      "clip_ratio/high_max": 0.002642264138557948,
      "clip_ratio/high_mean": 0.0011488028867461253,
      "clip_ratio/low_mean": 0.0012836683745263144,
      "clip_ratio/low_min": 0.00027828193560708314,
      "clip_ratio/region_mean": 0.0024324712867382914,
      "epoch": 0.6509186351706037,
      "grad_norm": 0.18310676515102386,
      "learning_rate": 1e-06,
      "loss": -0.0337,
      "step": 279
    },
    {
      "clip_ratio/high_max": 0.0028919992182636634,
      "clip_ratio/high_mean": 0.0011562096296984237,
      "clip_ratio/low_mean": 0.0015426580976054538,
      "clip_ratio/low_min": 0.0002890102396122529,
      "clip_ratio/region_mean": 0.002698867676372174,
      "epoch": 0.6532516768737241,
      "grad_norm": 0.16260650753974915,
      "learning_rate": 1e-06,
      "loss": -0.0338,
      "step": 280
    },
    {
      "clip_ratio/high_max": 0.00219465876580216,
      "clip_ratio/high_mean": 0.0007919146764834295,
      "clip_ratio/low_mean": 0.0006715007675666129,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014634154649684206,
      "completions/clipped_ratio": 0.0680803571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3350.0,
      "completions/mean_length": 842.8627319335938,
      "completions/mean_terminated_length": 605.2083740234375,
      "completions/min_length": 150.0,
      "completions/min_terminated_length": 150.0,
      "epoch": 0.6555847185768445,
      "grad_norm": 0.19941651821136475,
      "learning_rate": 1e-06,
      "loss": -0.0228,
      "num_tokens": 43333839.0,
      "reward": 0.5569196939468384,
      "reward_std": 0.17528702318668365,
      "rewards/verify_math_reward/mean": 0.5569196343421936,
      "rewards/verify_math_reward/std": 0.4970270097255707,
      "step": 281
    },
    {
      "clip_ratio/high_max": 0.0021507229139388073,
      "clip_ratio/high_mean": 0.0008451492731182952,
      "clip_ratio/low_mean": 0.0008469266977044754,
      "clip_ratio/low_min": 1.4895138519932516e-05,
      "clip_ratio/region_mean": 0.001692075948085403,
      "epoch": 0.657917760279965,
      "grad_norm": 0.16673003137111664,
      "learning_rate": 1e-06,
      "loss": -0.0229,
      "step": 282
    },
    {
      "clip_ratio/high_max": 0.0024777984945103526,
      "clip_ratio/high_mean": 0.0009478648280492052,
      "clip_ratio/low_mean": 0.001053386182320537,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002001250999455806,
      "epoch": 0.6602508019830855,
      "grad_norm": 0.15395739674568176,
      "learning_rate": 1e-06,
      "loss": -0.0232,
      "step": 283
    },
    {
      "clip_ratio/high_max": 0.00210989813786,
      "clip_ratio/high_mean": 0.0008852029659465188,
      "clip_ratio/low_mean": 0.001249679880857002,
      "clip_ratio/low_min": 1.4384349924512208e-05,
      "clip_ratio/region_mean": 0.0021348828231566586,
      "epoch": 0.6625838436862059,
      "grad_norm": 0.14514604210853577,
      "learning_rate": 1e-06,
      "loss": -0.0232,
      "step": 284
    },
    {
      "clip_ratio/high_max": 0.001999265434278641,
      "clip_ratio/high_mean": 0.000810526340501383,
      "clip_ratio/low_mean": 0.0009376847829116741,
      "clip_ratio/low_min": 6.594186834263382e-05,
      "clip_ratio/region_mean": 0.0017482111034041736,
      "completions/clipped_ratio": 0.0691964285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2510.0,
      "completions/mean_length": 875.3549194335938,
      "completions/mean_terminated_length": 635.9304809570312,
      "completions/min_length": 125.0,
      "completions/min_terminated_length": 125.0,
      "epoch": 0.6649168853893264,
      "grad_norm": 0.22797183692455292,
      "learning_rate": 1e-06,
      "loss": -0.0146,
      "num_tokens": 43961645.0,
      "reward": 0.5212053656578064,
      "reward_std": 0.2005843222141266,
      "rewards/verify_math_reward/mean": 0.5212053656578064,
      "rewards/verify_math_reward/std": 0.49982911348342896,
      "step": 285
    },
    {
      "clip_ratio/high_max": 0.002601971646072343,
      "clip_ratio/high_mean": 0.000978758151177317,
      "clip_ratio/low_mean": 0.0011493046149553265,
      "clip_ratio/low_min": 0.00012364785106910858,
      "clip_ratio/region_mean": 0.002128062733390834,
      "epoch": 0.6672499270924468,
      "grad_norm": 0.19644752144813538,
      "learning_rate": 1e-06,
      "loss": -0.0149,
      "step": 286
    },
    {
      "clip_ratio/high_max": 0.0029370293123065494,
      "clip_ratio/high_mean": 0.001147280399891315,
      "clip_ratio/low_mean": 0.0013407869519141968,
      "clip_ratio/low_min": 0.00013045562263869215,
      "clip_ratio/region_mean": 0.0024880673008738086,
      "epoch": 0.6695829687955672,
      "grad_norm": 0.17700296640396118,
      "learning_rate": 1e-06,
      "loss": -0.0151,
      "step": 287
    },
    {
      "clip_ratio/high_max": 0.0031012436520541087,
      "clip_ratio/high_mean": 0.0011367060978955124,
      "clip_ratio/low_mean": 0.0016410894750151783,
      "clip_ratio/low_min": 0.00017486737488070503,
      "clip_ratio/region_mean": 0.002777795525616966,
      "epoch": 0.6719160104986877,
      "grad_norm": 0.17286638915538788,
      "learning_rate": 1e-06,
      "loss": -0.0152,
      "step": 288
    },
    {
      "clip_ratio/high_max": 0.001724022728012642,
      "clip_ratio/high_mean": 0.0006826344042565324,
      "clip_ratio/low_mean": 0.0005667309815180488,
      "clip_ratio/low_min": 2.887398932216456e-05,
      "clip_ratio/region_mean": 0.0012493653921410441,
      "completions/clipped_ratio": 0.0770089285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3297.0,
      "completions/mean_length": 920.333740234375,
      "completions/mean_terminated_length": 655.3748779296875,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 0.6742490522018081,
      "grad_norm": 0.17062070965766907,
      "learning_rate": 1e-06,
      "loss": -0.0542,
      "num_tokens": 44600416.0,
      "reward": 0.5580357313156128,
      "reward_std": 0.16750861704349518,
      "rewards/verify_math_reward/mean": 0.5580357313156128,
      "rewards/verify_math_reward/std": 0.49689781665802,
      "step": 289
    },
    {
      "clip_ratio/high_max": 0.0020607160186045803,
      "clip_ratio/high_mean": 0.0008350994885404361,
      "clip_ratio/low_mean": 0.0006770236104784999,
      "clip_ratio/low_min": 1.2852148756792303e-05,
      "clip_ratio/region_mean": 0.0015121231153898407,
      "epoch": 0.6765820939049285,
      "grad_norm": 0.154007688164711,
      "learning_rate": 1e-06,
      "loss": -0.0543,
      "step": 290
    },
    {
      "clip_ratio/high_max": 0.002040704872342758,
      "clip_ratio/high_mean": 0.0009163299873762298,
      "clip_ratio/low_mean": 0.0008831634622765705,
      "clip_ratio/low_min": 3.047500467801001e-05,
      "clip_ratio/region_mean": 0.0017994934823946096,
      "epoch": 0.678915135608049,
      "grad_norm": 0.1436816304922104,
      "learning_rate": 1e-06,
      "loss": -0.0544,
      "step": 291
    },
    {
      "clip_ratio/high_max": 0.002354840806219727,
      "clip_ratio/high_mean": 0.0008853112758515636,
      "clip_ratio/low_mean": 0.0009733925799082499,
      "clip_ratio/low_min": 2.5704297513584606e-05,
      "clip_ratio/region_mean": 0.0018587038503028452,
      "epoch": 0.6812481773111695,
      "grad_norm": 0.13852904736995697,
      "learning_rate": 1e-06,
      "loss": -0.0545,
      "step": 292
    },
    {
      "clip_ratio/high_max": 0.0021498534333659336,
      "clip_ratio/high_mean": 0.0008024014623515541,
      "clip_ratio/low_mean": 0.0006454346130340127,
      "clip_ratio/low_min": 2.5375558834639378e-05,
      "clip_ratio/region_mean": 0.001447836053557694,
      "completions/clipped_ratio": 0.0926339285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3623.0,
      "completions/mean_length": 971.1328735351562,
      "completions/mean_terminated_length": 652.1119384765625,
      "completions/min_length": 164.0,
      "completions/min_terminated_length": 164.0,
      "epoch": 0.6835812190142899,
      "grad_norm": 0.20537430047988892,
      "learning_rate": 1e-06,
      "loss": -0.0384,
      "num_tokens": 45220287.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.18303194642066956,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 293
    },
    {
      "clip_ratio/high_max": 0.0021460836433107033,
      "clip_ratio/high_mean": 0.0008764559806877514,
      "clip_ratio/low_mean": 0.0008536525929230265,
      "clip_ratio/low_min": 8.626158432889497e-05,
      "clip_ratio/region_mean": 0.0017301085536018945,
      "epoch": 0.6859142607174104,
      "grad_norm": 0.18146011233329773,
      "learning_rate": 1e-06,
      "loss": -0.0386,
      "step": 294
    },
    {
      "clip_ratio/high_max": 0.002821837035298813,
      "clip_ratio/high_mean": 0.0010902132889896166,
      "clip_ratio/low_mean": 0.0010684805311029777,
      "clip_ratio/low_min": 5.1454062486300245e-05,
      "clip_ratio/region_mean": 0.0021586938310065307,
      "epoch": 0.6882473024205308,
      "grad_norm": 0.17304526269435883,
      "learning_rate": 1e-06,
      "loss": -0.0388,
      "step": 295
    },
    {
      "clip_ratio/high_max": 0.0025890369288390502,
      "clip_ratio/high_mean": 0.0010107452362717595,
      "clip_ratio/low_mean": 0.0011576586766750552,
      "clip_ratio/low_min": 7.048963198030833e-05,
      "clip_ratio/region_mean": 0.0021684039020328782,
      "epoch": 0.6905803441236512,
      "grad_norm": 0.15993866324424744,
      "learning_rate": 1e-06,
      "loss": -0.0389,
      "step": 296
    },
    {
      "clip_ratio/high_max": 0.002087494191073347,
      "clip_ratio/high_mean": 0.0007458662621502299,
      "clip_ratio/low_mean": 0.0006469466452472261,
      "clip_ratio/low_min": 2.270076856802916e-05,
      "clip_ratio/region_mean": 0.0013928129119449295,
      "completions/clipped_ratio": 0.0636160714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3791.0,
      "completions/mean_length": 866.2857666015625,
      "completions/mean_terminated_length": 646.8652954101562,
      "completions/min_length": 153.0,
      "completions/min_terminated_length": 153.0,
      "epoch": 0.6929133858267716,
      "grad_norm": 0.19377551972866058,
      "learning_rate": 1e-06,
      "loss": -0.0055,
      "num_tokens": 45848071.0,
      "reward": 0.5390625,
      "reward_std": 0.16830208897590637,
      "rewards/verify_math_reward/mean": 0.5390625,
      "rewards/verify_math_reward/std": 0.4987502098083496,
      "step": 297
    },
    {
      "clip_ratio/high_max": 0.002341042520129122,
      "clip_ratio/high_mean": 0.0008734871789783938,
      "clip_ratio/low_mean": 0.0007677624307689257,
      "clip_ratio/low_min": 3.7830393011972774e-05,
      "clip_ratio/region_mean": 0.0016412495970143937,
      "epoch": 0.6952464275298921,
      "grad_norm": 0.18440525233745575,
      "learning_rate": 1e-06,
      "loss": -0.0056,
      "step": 298
    },
    {
      "clip_ratio/high_max": 0.0025020054017659277,
      "clip_ratio/high_mean": 0.000964397990173893,
      "clip_ratio/low_mean": 0.0009638205920055043,
      "clip_ratio/low_min": 1.9797276763711125e-05,
      "clip_ratio/region_mean": 0.0019282185603515245,
      "epoch": 0.6975794692330125,
      "grad_norm": 0.1549079865217209,
      "learning_rate": 1e-06,
      "loss": -0.0058,
      "step": 299
    },
    {
      "clip_ratio/high_max": 0.002495607521268539,
      "clip_ratio/high_mean": 0.000916457383937086,
      "clip_ratio/low_mean": 0.0011305743719276506,
      "clip_ratio/low_min": 3.956953059969237e-05,
      "clip_ratio/region_mean": 0.002047031703114044,
      "epoch": 0.6999125109361329,
      "grad_norm": 0.15340441465377808,
      "learning_rate": 1e-06,
      "loss": -0.0059,
      "step": 300
    },
    {
      "clip_ratio/high_max": 0.002479487993696239,
      "clip_ratio/high_mean": 0.0010109548456966877,
      "clip_ratio/low_mean": 0.0007489804047509097,
      "clip_ratio/low_min": 8.222711767302826e-05,
      "clip_ratio/region_mean": 0.0017599352431716397,
      "completions/clipped_ratio": 0.1004464285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2059.0,
      "completions/mean_length": 1006.3047485351562,
      "completions/mean_terminated_length": 661.301513671875,
      "completions/min_length": 120.0,
      "completions/min_terminated_length": 120.0,
      "epoch": 0.7022455526392535,
      "grad_norm": 0.21473151445388794,
      "learning_rate": 1e-06,
      "loss": -0.0478,
      "num_tokens": 46474320.0,
      "reward": 0.543526828289032,
      "reward_std": 0.2144775092601776,
      "rewards/verify_math_reward/mean": 0.5435267686843872,
      "rewards/verify_math_reward/std": 0.49838000535964966,
      "step": 301
    },
    {
      "clip_ratio/high_max": 0.0028480310429586098,
      "clip_ratio/high_mean": 0.0011781237517425325,
      "clip_ratio/low_mean": 0.0008835742810333613,
      "clip_ratio/low_min": 9.115397187997587e-05,
      "clip_ratio/region_mean": 0.002061698047327809,
      "epoch": 0.7045785943423739,
      "grad_norm": 0.19146059453487396,
      "learning_rate": 1e-06,
      "loss": -0.048,
      "step": 302
    },
    {
      "clip_ratio/high_max": 0.0030311847804114223,
      "clip_ratio/high_mean": 0.0012647803814616054,
      "clip_ratio/low_mean": 0.0011280409526079893,
      "clip_ratio/low_min": 6.366150228132028e-05,
      "clip_ratio/region_mean": 0.0023928213558974676,
      "epoch": 0.7069116360454943,
      "grad_norm": 0.18539467453956604,
      "learning_rate": 1e-06,
      "loss": -0.0482,
      "step": 303
    },
    {
      "clip_ratio/high_max": 0.0030102770251687616,
      "clip_ratio/high_mean": 0.001217226883454714,
      "clip_ratio/low_mean": 0.0013476571002684068,
      "clip_ratio/low_min": 0.00012812625027436297,
      "clip_ratio/region_mean": 0.002564883994637057,
      "epoch": 0.7092446777486148,
      "grad_norm": 0.18372808396816254,
      "learning_rate": 1e-06,
      "loss": -0.0483,
      "step": 304
    },
    {
      "clip_ratio/high_max": 0.0018803218481480144,
      "clip_ratio/high_mean": 0.00078264167495945,
      "clip_ratio/low_mean": 0.0006800834380555898,
      "clip_ratio/low_min": 1.233471448358614e-05,
      "clip_ratio/region_mean": 0.0014627251330239233,
      "completions/clipped_ratio": 0.0848214285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2283.0,
      "completions/mean_length": 905.3359985351562,
      "completions/mean_terminated_length": 609.6158447265625,
      "completions/min_length": 143.0,
      "completions/min_terminated_length": 143.0,
      "epoch": 0.7115777194517352,
      "grad_norm": 0.2239924669265747,
      "learning_rate": 1e-06,
      "loss": -0.0251,
      "num_tokens": 47065829.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.1870116889476776,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514803290367126,
      "step": 305
    },
    {
      "clip_ratio/high_max": 0.0026517381411395036,
      "clip_ratio/high_mean": 0.0011003393010469154,
      "clip_ratio/low_mean": 0.0010205757225776324,
      "clip_ratio/low_min": 7.395359716610983e-05,
      "clip_ratio/region_mean": 0.002120915065461304,
      "epoch": 0.7139107611548556,
      "grad_norm": 0.2101132869720459,
      "learning_rate": 1e-06,
      "loss": -0.0254,
      "step": 306
    },
    {
      "clip_ratio/high_max": 0.0028270179318496957,
      "clip_ratio/high_mean": 0.0011196137056685984,
      "clip_ratio/low_mean": 0.0012128089510952123,
      "clip_ratio/low_min": 2.466942896717228e-05,
      "clip_ratio/region_mean": 0.0023324227513512596,
      "epoch": 0.7162438028579761,
      "grad_norm": 0.18428486585617065,
      "learning_rate": 1e-06,
      "loss": -0.0256,
      "step": 307
    },
    {
      "clip_ratio/high_max": 0.002363015402806923,
      "clip_ratio/high_mean": 0.0010658901901479112,
      "clip_ratio/low_mean": 0.0014643462345702574,
      "clip_ratio/low_min": 0.00010415140059194528,
      "clip_ratio/region_mean": 0.002530236371967476,
      "epoch": 0.7185768445610965,
      "grad_norm": 0.17178674042224884,
      "learning_rate": 1e-06,
      "loss": -0.0257,
      "step": 308
    },
    {
      "clip_ratio/high_max": 0.002385223771852907,
      "clip_ratio/high_mean": 0.0009904428261506837,
      "clip_ratio/low_mean": 0.0005621537156912382,
      "clip_ratio/low_min": 2.756753019639291e-05,
      "clip_ratio/region_mean": 0.0015525966045970563,
      "completions/clipped_ratio": 0.0848214285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2489.0,
      "completions/mean_length": 897.09716796875,
      "completions/mean_terminated_length": 600.6134033203125,
      "completions/min_length": 129.0,
      "completions/min_terminated_length": 129.0,
      "epoch": 0.7209098862642169,
      "grad_norm": 0.19375817477703094,
      "learning_rate": 1e-06,
      "loss": -0.0374,
      "num_tokens": 47644244.0,
      "reward": 0.6149553656578064,
      "reward_std": 0.1734083741903305,
      "rewards/verify_math_reward/mean": 0.6149553656578064,
      "rewards/verify_math_reward/std": 0.4868776500225067,
      "step": 309
    },
    {
      "clip_ratio/high_max": 0.0027656317324726842,
      "clip_ratio/high_mean": 0.0011467549702501856,
      "clip_ratio/low_mean": 0.0006837193341198144,
      "clip_ratio/low_min": 1.0453252798470203e-05,
      "clip_ratio/region_mean": 0.0018304742843611166,
      "epoch": 0.7232429279673375,
      "grad_norm": 0.1754893809556961,
      "learning_rate": 1e-06,
      "loss": -0.0375,
      "step": 310
    },
    {
      "clip_ratio/high_max": 0.002889984578359872,
      "clip_ratio/high_mean": 0.0012613694634637795,
      "clip_ratio/low_mean": 0.0009629708802094683,
      "clip_ratio/low_min": 3.538256623869529e-05,
      "clip_ratio/region_mean": 0.002224340358225163,
      "epoch": 0.7255759696704579,
      "grad_norm": 0.16185636818408966,
      "learning_rate": 1e-06,
      "loss": -0.0377,
      "step": 311
    },
    {
      "clip_ratio/high_max": 0.002871236007194966,
      "clip_ratio/high_mean": 0.0012253369859536178,
      "clip_ratio/low_mean": 0.001014525489154039,
      "clip_ratio/low_min": 1.799078927433584e-05,
      "clip_ratio/region_mean": 0.0022398624787456356,
      "epoch": 0.7279090113735783,
      "grad_norm": 0.17588123679161072,
      "learning_rate": 1e-06,
      "loss": -0.0378,
      "step": 312
    },
    {
      "clip_ratio/high_max": 0.002086147724185139,
      "clip_ratio/high_mean": 0.0008469103850075044,
      "clip_ratio/low_mean": 0.0008083435477601597,
      "clip_ratio/low_min": 1.1007397006324027e-05,
      "clip_ratio/region_mean": 0.0016552538945688866,
      "completions/clipped_ratio": 0.078125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3981.0,
      "completions/mean_length": 885.739990234375,
      "completions/mean_terminated_length": 613.6840209960938,
      "completions/min_length": 183.0,
      "completions/min_terminated_length": 183.0,
      "epoch": 0.7302420530766988,
      "grad_norm": 0.23492024838924408,
      "learning_rate": 1e-06,
      "loss": -0.0367,
      "num_tokens": 48240235.0,
      "reward": 0.5870535969734192,
      "reward_std": 0.20587676763534546,
      "rewards/verify_math_reward/mean": 0.5870535969734192,
      "rewards/verify_math_reward/std": 0.49263837933540344,
      "step": 313
    },
    {
      "clip_ratio/high_max": 0.002379067002038937,
      "clip_ratio/high_mean": 0.0010538723690842744,
      "clip_ratio/low_mean": 0.0010255752076773206,
      "clip_ratio/low_min": 5.036196489527356e-05,
      "clip_ratio/region_mean": 0.0020794475931324996,
      "epoch": 0.7325750947798192,
      "grad_norm": 0.20596159994602203,
      "learning_rate": 1e-06,
      "loss": -0.0369,
      "step": 314
    },
    {
      "clip_ratio/high_max": 0.002566054739872925,
      "clip_ratio/high_mean": 0.0010753861861303449,
      "clip_ratio/low_mean": 0.0012584659052663483,
      "clip_ratio/low_min": 1.940692527568899e-05,
      "clip_ratio/region_mean": 0.0023338521641562693,
      "epoch": 0.7349081364829396,
      "grad_norm": 0.18503864109516144,
      "learning_rate": 1e-06,
      "loss": -0.0371,
      "step": 315
    },
    {
      "clip_ratio/high_max": 0.0028267628003959544,
      "clip_ratio/high_mean": 0.0011631867128016893,
      "clip_ratio/low_mean": 0.001585173016792396,
      "clip_ratio/low_min": 8.646383503219113e-05,
      "clip_ratio/region_mean": 0.0027483597295940854,
      "epoch": 0.73724117818606,
      "grad_norm": 0.17676009237766266,
      "learning_rate": 1e-06,
      "loss": -0.0373,
      "step": 316
    },
    {
      "clip_ratio/high_max": 0.0020308344828663394,
      "clip_ratio/high_mean": 0.0008841200924507575,
      "clip_ratio/low_mean": 0.0006222654037628672,
      "clip_ratio/low_min": 3.280446890130406e-05,
      "clip_ratio/region_mean": 0.001506385437096469,
      "completions/clipped_ratio": 0.0725446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2776.0,
      "completions/mean_length": 837.7857666015625,
      "completions/mean_terminated_length": 582.931396484375,
      "completions/min_length": 134.0,
      "completions/min_terminated_length": 134.0,
      "epoch": 0.7395742198891805,
      "grad_norm": 0.2314973920583725,
      "learning_rate": 1e-06,
      "loss": 0.0017,
      "num_tokens": 48814699.0,
      "reward": 0.613839328289032,
      "reward_std": 0.19084171950817108,
      "rewards/verify_math_reward/mean": 0.6138392686843872,
      "rewards/verify_math_reward/std": 0.48714008927345276,
      "step": 317
    },
    {
      "clip_ratio/high_max": 0.0025161904195556417,
      "clip_ratio/high_mean": 0.0010388361515651923,
      "clip_ratio/low_mean": 0.0009122929877776187,
      "clip_ratio/low_min": 7.236033252411289e-05,
      "clip_ratio/region_mean": 0.0019511291611706838,
      "epoch": 0.7419072615923009,
      "grad_norm": 0.20519977807998657,
      "learning_rate": 1e-06,
      "loss": 0.0014,
      "step": 318
    },
    {
      "clip_ratio/high_max": 0.0024081753472273704,
      "clip_ratio/high_mean": 0.0009931490312737878,
      "clip_ratio/low_mean": 0.001269341999432072,
      "clip_ratio/low_min": 8.174411141226301e-05,
      "clip_ratio/region_mean": 0.0022624911143793724,
      "epoch": 0.7442403032954215,
      "grad_norm": 0.1907121241092682,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "step": 319
    },
    {
      "clip_ratio/high_max": 0.0027625623624771833,
      "clip_ratio/high_mean": 0.001064769570803037,
      "clip_ratio/low_mean": 0.0013623945651488611,
      "clip_ratio/low_min": 7.313470996450633e-05,
      "clip_ratio/region_mean": 0.002427164145046845,
      "epoch": 0.7465733449985419,
      "grad_norm": 0.18573161959648132,
      "learning_rate": 1e-06,
      "loss": 0.0012,
      "step": 320
    },
    {
      "clip_ratio/high_max": 0.0020261932440917008,
      "clip_ratio/high_mean": 0.0008740671855775872,
      "clip_ratio/low_mean": 0.0007327803323278204,
      "clip_ratio/low_min": 2.4659695554873906e-05,
      "clip_ratio/region_mean": 0.001606847537914291,
      "completions/clipped_ratio": 0.0959821428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2926.0,
      "completions/mean_length": 984.7779541015625,
      "completions/mean_terminated_length": 654.4506225585938,
      "completions/min_length": 142.0,
      "completions/min_terminated_length": 142.0,
      "epoch": 0.7489063867016623,
      "grad_norm": 0.22532038390636444,
      "learning_rate": 1e-06,
      "loss": -0.02,
      "num_tokens": 49440620.0,
      "reward": 0.5613839626312256,
      "reward_std": 0.18416057527065277,
      "rewards/verify_math_reward/mean": 0.5613839030265808,
      "rewards/verify_math_reward/std": 0.496494859457016,
      "step": 321
    },
    {
      "clip_ratio/high_max": 0.002825237992510665,
      "clip_ratio/high_mean": 0.0011012507529812865,
      "clip_ratio/low_mean": 0.0009591566486051306,
      "clip_ratio/low_min": 1.2329847777436953e-05,
      "clip_ratio/region_mean": 0.002060407350654714,
      "epoch": 0.7512394284047827,
      "grad_norm": 0.2141539305448532,
      "learning_rate": 1e-06,
      "loss": -0.0202,
      "step": 322
    },
    {
      "clip_ratio/high_max": 0.0029633525336976163,
      "clip_ratio/high_mean": 0.0011840704191854456,
      "clip_ratio/low_mean": 0.001118860669521382,
      "clip_ratio/low_min": 4.405286381370388e-05,
      "clip_ratio/region_mean": 0.002302931126905605,
      "epoch": 0.7535724701079032,
      "grad_norm": 0.17943425476551056,
      "learning_rate": 1e-06,
      "loss": -0.0204,
      "step": 323
    },
    {
      "clip_ratio/high_max": 0.0028357562187011354,
      "clip_ratio/high_mean": 0.001128319876443129,
      "clip_ratio/low_mean": 0.001372187805827707,
      "clip_ratio/low_min": 7.397909212158993e-05,
      "clip_ratio/region_mean": 0.0025005076968227513,
      "epoch": 0.7559055118110236,
      "grad_norm": 0.16282466053962708,
      "learning_rate": 1e-06,
      "loss": -0.0205,
      "step": 324
    },
    {
      "clip_ratio/high_max": 0.0017659160221228376,
      "clip_ratio/high_mean": 0.0007654340861336095,
      "clip_ratio/low_mean": 0.0006649163590282114,
      "clip_ratio/low_min": 1.1312216884107329e-05,
      "clip_ratio/region_mean": 0.001430350461305352,
      "completions/clipped_ratio": 0.0859375,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3992.0,
      "completions/mean_length": 940.239990234375,
      "completions/mean_terminated_length": 643.5446166992188,
      "completions/min_length": 145.0,
      "completions/min_terminated_length": 145.0,
      "epoch": 0.758238553514144,
      "grad_norm": 0.21520856022834778,
      "learning_rate": 1e-06,
      "loss": -0.035,
      "num_tokens": 50061851.0,
      "reward": 0.5424107313156128,
      "reward_std": 0.18437741696834564,
      "rewards/verify_math_reward/mean": 0.5424107313156128,
      "rewards/verify_math_reward/std": 0.4984763562679291,
      "step": 325
    },
    {
      "clip_ratio/high_max": 0.002330236129637342,
      "clip_ratio/high_mean": 0.0009415907279617386,
      "clip_ratio/low_mean": 0.000835203873066348,
      "clip_ratio/low_min": 3.393665247131139e-05,
      "clip_ratio/region_mean": 0.001776794575562235,
      "epoch": 0.7605715952172645,
      "grad_norm": 0.1842113435268402,
      "learning_rate": 1e-06,
      "loss": -0.0353,
      "step": 326
    },
    {
      "clip_ratio/high_max": 0.00230511437257519,
      "clip_ratio/high_mean": 0.001003508476060233,
      "clip_ratio/low_mean": 0.0010929000836767955,
      "clip_ratio/low_min": 1.1312216884107329e-05,
      "clip_ratio/region_mean": 0.0020964085560990497,
      "epoch": 0.7629046369203849,
      "grad_norm": 0.16316959261894226,
      "learning_rate": 1e-06,
      "loss": -0.0354,
      "step": 327
    },
    {
      "clip_ratio/high_max": 0.002567019429989159,
      "clip_ratio/high_mean": 0.000990610795270186,
      "clip_ratio/low_mean": 0.0012704540695267497,
      "clip_ratio/low_min": 9.153485734714195e-06,
      "clip_ratio/region_mean": 0.0022610649466514587,
      "epoch": 0.7652376786235054,
      "grad_norm": 0.204722598195076,
      "learning_rate": 1e-06,
      "loss": -0.0355,
      "step": 328
    },
    {
      "clip_ratio/high_max": 0.0017108206229750067,
      "clip_ratio/high_mean": 0.0006872556414236897,
      "clip_ratio/low_mean": 0.0005310801495852502,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012183357739559142,
      "completions/clipped_ratio": 0.0691964285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2780.0,
      "completions/mean_length": 852.4699096679688,
      "completions/mean_terminated_length": 611.3441162109375,
      "completions/min_length": 159.0,
      "completions/min_terminated_length": 159.0,
      "epoch": 0.7675707203266259,
      "grad_norm": 0.21647314727306366,
      "learning_rate": 1e-06,
      "loss": -0.026,
      "num_tokens": 50656928.0,
      "reward": 0.6395089626312256,
      "reward_std": 0.15737581253051758,
      "rewards/verify_math_reward/mean": 0.6395089030265808,
      "rewards/verify_math_reward/std": 0.4804111421108246,
      "step": 329
    },
    {
      "clip_ratio/high_max": 0.0020555987684929278,
      "clip_ratio/high_mean": 0.0007949248674776754,
      "clip_ratio/low_mean": 0.000770878323010038,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015658032170904335,
      "epoch": 0.7699037620297463,
      "grad_norm": 0.1775524616241455,
      "learning_rate": 1e-06,
      "loss": -0.0262,
      "step": 330
    },
    {
      "clip_ratio/high_max": 0.0025171050292556174,
      "clip_ratio/high_mean": 0.0009998566874855896,
      "clip_ratio/low_mean": 0.0008616442519269185,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018615009503264446,
      "epoch": 0.7722368037328667,
      "grad_norm": 0.16343437135219574,
      "learning_rate": 1e-06,
      "loss": -0.0264,
      "step": 331
    },
    {
      "clip_ratio/high_max": 0.002173786866478622,
      "clip_ratio/high_mean": 0.0009114055246755015,
      "clip_ratio/low_mean": 0.0010367860127189488,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019481915842334274,
      "epoch": 0.7745698454359872,
      "grad_norm": 0.15880723297595978,
      "learning_rate": 1e-06,
      "loss": -0.0264,
      "step": 332
    },
    {
      "clip_ratio/high_max": 0.0017265135975321755,
      "clip_ratio/high_mean": 0.0005917233247600961,
      "clip_ratio/low_mean": 0.0006106325026848936,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001202355808345601,
      "completions/clipped_ratio": 0.1227678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3840.0,
      "completions/mean_length": 1066.146240234375,
      "completions/mean_terminated_length": 642.120849609375,
      "completions/min_length": 181.0,
      "completions/min_terminated_length": 181.0,
      "epoch": 0.7769028871391076,
      "grad_norm": 0.21014513075351715,
      "learning_rate": 1e-06,
      "loss": -0.0184,
      "num_tokens": 51257251.0,
      "reward": 0.504464328289032,
      "reward_std": 0.1623242348432541,
      "rewards/verify_math_reward/mean": 0.5044642686843872,
      "rewards/verify_math_reward/std": 0.5002593398094177,
      "step": 333
    },
    {
      "clip_ratio/high_max": 0.0021170642612560187,
      "clip_ratio/high_mean": 0.0007432475922541926,
      "clip_ratio/low_mean": 0.0008848217221384402,
      "clip_ratio/low_min": 1.387963584420504e-05,
      "clip_ratio/region_mean": 0.0016280692943837494,
      "epoch": 0.779235928842228,
      "grad_norm": 0.18322604894638062,
      "learning_rate": 1e-06,
      "loss": -0.0186,
      "step": 334
    },
    {
      "clip_ratio/high_max": 0.0023809336016711313,
      "clip_ratio/high_mean": 0.0008279726753244177,
      "clip_ratio/low_mean": 0.0009638046303734882,
      "clip_ratio/low_min": 1.4501159967039712e-05,
      "clip_ratio/region_mean": 0.00179177729296498,
      "epoch": 0.7815689705453485,
      "grad_norm": 0.1649850606918335,
      "learning_rate": 1e-06,
      "loss": -0.0188,
      "step": 335
    },
    {
      "clip_ratio/high_max": 0.0023388637600874063,
      "clip_ratio/high_mean": 0.0008620175885880599,
      "clip_ratio/low_mean": 0.0011599286262935493,
      "clip_ratio/low_min": 5.6761591622489505e-05,
      "clip_ratio/region_mean": 0.0020219462167005986,
      "epoch": 0.7839020122484689,
      "grad_norm": 0.1583496779203415,
      "learning_rate": 1e-06,
      "loss": -0.0188,
      "step": 336
    },
    {
      "clip_ratio/high_max": 0.0019701817764143925,
      "clip_ratio/high_mean": 0.0008163482880263473,
      "clip_ratio/low_mean": 0.0004907682782686607,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013071165485598613,
      "completions/clipped_ratio": 0.1015625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3998.0,
      "completions/mean_length": 972.9855346679688,
      "completions/mean_terminated_length": 619.9490356445312,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 0.7862350539515894,
      "grad_norm": 0.2171715795993805,
      "learning_rate": 1e-06,
      "loss": -0.0345,
      "num_tokens": 51846038.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.15593409538269043,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514803290367126,
      "step": 337
    },
    {
      "clip_ratio/high_max": 0.002572505996795371,
      "clip_ratio/high_mean": 0.0009753500671649817,
      "clip_ratio/low_mean": 0.000732240172510501,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017075901851058006,
      "epoch": 0.7885680956547099,
      "grad_norm": 0.20398429036140442,
      "learning_rate": 1e-06,
      "loss": -0.0347,
      "step": 338
    },
    {
      "clip_ratio/high_max": 0.0028881516482215375,
      "clip_ratio/high_mean": 0.001134720796471811,
      "clip_ratio/low_mean": 0.0007902334382379195,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019249542820034549,
      "epoch": 0.7909011373578303,
      "grad_norm": 0.16836610436439514,
      "learning_rate": 1e-06,
      "loss": -0.0349,
      "step": 339
    },
    {
      "clip_ratio/high_max": 0.002433148052659817,
      "clip_ratio/high_mean": 0.0010081968639497063,
      "clip_ratio/low_mean": 0.001013724282529438,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002021921107370872,
      "epoch": 0.7932341790609507,
      "grad_norm": 0.15733319520950317,
      "learning_rate": 1e-06,
      "loss": -0.035,
      "step": 340
    },
    {
      "clip_ratio/high_max": 0.0019665661711769644,
      "clip_ratio/high_mean": 0.0006874319560665754,
      "clip_ratio/low_mean": 0.0005987316217215266,
      "clip_ratio/low_min": 5.924893321207492e-05,
      "clip_ratio/region_mean": 0.001286163580516586,
      "completions/clipped_ratio": 0.0703125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3553.0,
      "completions/mean_length": 863.029052734375,
      "completions/mean_terminated_length": 618.5186157226562,
      "completions/min_length": 125.0,
      "completions/min_terminated_length": 125.0,
      "epoch": 0.7955672207640712,
      "grad_norm": 0.22709819674491882,
      "learning_rate": 1e-06,
      "loss": -0.0126,
      "num_tokens": 52448128.0,
      "reward": 0.5625,
      "reward_std": 0.15225742757320404,
      "rewards/verify_math_reward/mean": 0.5625,
      "rewards/verify_math_reward/std": 0.49635544419288635,
      "step": 341
    },
    {
      "clip_ratio/high_max": 0.0019139994692523032,
      "clip_ratio/high_mean": 0.0007924420715426095,
      "clip_ratio/low_mean": 0.0008447009604424238,
      "clip_ratio/low_min": 0.00012417925518093398,
      "clip_ratio/region_mean": 0.0016371430319850333,
      "epoch": 0.7979002624671916,
      "grad_norm": 0.17402707040309906,
      "learning_rate": 1e-06,
      "loss": -0.0129,
      "step": 342
    },
    {
      "clip_ratio/high_max": 0.00228307512588799,
      "clip_ratio/high_mean": 0.0007988824791027582,
      "clip_ratio/low_mean": 0.0009502735165369813,
      "clip_ratio/low_min": 0.00010351143191655865,
      "clip_ratio/region_mean": 0.0017491560320195276,
      "epoch": 0.800233304170312,
      "grad_norm": 0.1693955808877945,
      "learning_rate": 1e-06,
      "loss": -0.013,
      "step": 343
    },
    {
      "clip_ratio/high_max": 0.0020274076487112325,
      "clip_ratio/high_mean": 0.0008058251578404452,
      "clip_ratio/low_mean": 0.0012059471646352904,
      "clip_ratio/low_min": 0.0001847325765993446,
      "clip_ratio/region_mean": 0.002011772339756135,
      "epoch": 0.8025663458734325,
      "grad_norm": 0.15824578702449799,
      "learning_rate": 1e-06,
      "loss": -0.0131,
      "step": 344
    },
    {
      "clip_ratio/high_max": 0.0024287665510200895,
      "clip_ratio/high_mean": 0.0010701709761633538,
      "clip_ratio/low_mean": 0.0005244867297733435,
      "clip_ratio/low_min": 4.60348637716379e-05,
      "clip_ratio/region_mean": 0.0015946576968417503,
      "completions/clipped_ratio": 0.0658482142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2066.0,
      "completions/mean_length": 801.6574096679688,
      "completions/mean_terminated_length": 569.4396362304688,
      "completions/min_length": 148.0,
      "completions/min_terminated_length": 148.0,
      "epoch": 0.8048993875765529,
      "grad_norm": 0.25556159019470215,
      "learning_rate": 1e-06,
      "loss": -0.0212,
      "num_tokens": 53005413.0,
      "reward": 0.6383928656578064,
      "reward_std": 0.18126347661018372,
      "rewards/verify_math_reward/mean": 0.6383928656578064,
      "rewards/verify_math_reward/std": 0.4807341694831848,
      "step": 345
    },
    {
      "clip_ratio/high_max": 0.0026033492176793516,
      "clip_ratio/high_mean": 0.0011321816746203694,
      "clip_ratio/low_mean": 0.0007589621236547828,
      "clip_ratio/low_min": 1.8463810192770325e-05,
      "clip_ratio/region_mean": 0.0018911437728093006,
      "epoch": 0.8072324292796734,
      "grad_norm": 0.2019054889678955,
      "learning_rate": 1e-06,
      "loss": -0.0214,
      "step": 346
    },
    {
      "clip_ratio/high_max": 0.0024528612120775506,
      "clip_ratio/high_mean": 0.0010744591600087006,
      "clip_ratio/low_mean": 0.0009249952199752443,
      "clip_ratio/low_min": 5.539143239730038e-05,
      "clip_ratio/region_mean": 0.001999454398173839,
      "epoch": 0.8095654709827939,
      "grad_norm": 0.18154841661453247,
      "learning_rate": 1e-06,
      "loss": -0.0215,
      "step": 347
    },
    {
      "clip_ratio/high_max": 0.002686970961804036,
      "clip_ratio/high_mean": 0.0011972458305535838,
      "clip_ratio/low_mean": 0.0011541733802005183,
      "clip_ratio/low_min": 0.00010741467849584296,
      "clip_ratio/region_mean": 0.002351419192564208,
      "epoch": 0.8118985126859143,
      "grad_norm": 0.179201140999794,
      "learning_rate": 1e-06,
      "loss": -0.0217,
      "step": 348
    },
    {
      "clip_ratio/high_max": 0.002198454505560221,
      "clip_ratio/high_mean": 0.0009481786855758401,
      "clip_ratio/low_mean": 0.0006049338808225002,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015531125500274356,
      "completions/clipped_ratio": 0.0825892857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3831.0,
      "completions/mean_length": 877.8605346679688,
      "completions/mean_terminated_length": 588.149658203125,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 0.8142315543890347,
      "grad_norm": 0.22538886964321136,
      "learning_rate": 1e-06,
      "loss": -0.0396,
      "num_tokens": 53568944.0,
      "reward": 0.6573660969734192,
      "reward_std": 0.1759275496006012,
      "rewards/verify_math_reward/mean": 0.6573660969734192,
      "rewards/verify_math_reward/std": 0.47485533356666565,
      "step": 349
    },
    {
      "clip_ratio/high_max": 0.0029960423926240765,
      "clip_ratio/high_mean": 0.001205331936944276,
      "clip_ratio/low_mean": 0.0009345814778498607,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021399134275270626,
      "epoch": 0.8165645960921551,
      "grad_norm": 0.25057047605514526,
      "learning_rate": 1e-06,
      "loss": -0.0398,
      "step": 350
    },
    {
      "clip_ratio/high_max": 0.002933300071163103,
      "clip_ratio/high_mean": 0.0012073634934495203,
      "clip_ratio/low_mean": 0.00103660965396557,
      "clip_ratio/low_min": 2.9634898965014145e-05,
      "clip_ratio/region_mean": 0.0022439731619670056,
      "epoch": 0.8188976377952756,
      "grad_norm": 0.18622159957885742,
      "learning_rate": 1e-06,
      "loss": -0.04,
      "step": 351
    },
    {
      "clip_ratio/high_max": 0.0032266734924633056,
      "clip_ratio/high_mean": 0.0013009330614295322,
      "clip_ratio/low_mean": 0.0013344781218620483,
      "clip_ratio/low_min": 2.133105772372801e-05,
      "clip_ratio/region_mean": 0.002635411190567538,
      "epoch": 0.821230679498396,
      "grad_norm": 0.17074429988861084,
      "learning_rate": 1e-06,
      "loss": -0.0402,
      "step": 352
    },
    {
      "clip_ratio/high_max": 0.0017971512352232821,
      "clip_ratio/high_mean": 0.0007026633556961315,
      "clip_ratio/low_mean": 0.0006032489382050699,
      "clip_ratio/low_min": 1.3160665730538312e-05,
      "clip_ratio/region_mean": 0.0013059123120910954,
      "completions/clipped_ratio": 0.0892857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3867.0,
      "completions/mean_length": 921.6998291015625,
      "completions/mean_terminated_length": 610.493896484375,
      "completions/min_length": 165.0,
      "completions/min_terminated_length": 165.0,
      "epoch": 0.8235637212015164,
      "grad_norm": 0.2515932023525238,
      "learning_rate": 1e-06,
      "loss": -0.0245,
      "num_tokens": 54168955.0,
      "reward": 0.5424107313156128,
      "reward_std": 0.15988317131996155,
      "rewards/verify_math_reward/mean": 0.5424107313156128,
      "rewards/verify_math_reward/std": 0.4984763562679291,
      "step": 353
    },
    {
      "clip_ratio/high_max": 0.0021717835843446665,
      "clip_ratio/high_mean": 0.0007233772030303953,
      "clip_ratio/low_mean": 0.0008419789210165618,
      "clip_ratio/low_min": 1.4545031262969133e-05,
      "clip_ratio/region_mean": 0.0015653561240469571,
      "epoch": 0.8258967629046369,
      "grad_norm": 0.1717970371246338,
      "learning_rate": 1e-06,
      "loss": -0.0247,
      "step": 354
    },
    {
      "clip_ratio/high_max": 0.0023595982711412944,
      "clip_ratio/high_mean": 0.000842728490169975,
      "clip_ratio/low_mean": 0.0009849208618106786,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018276493137818761,
      "epoch": 0.8282298046077574,
      "grad_norm": 0.1771852672100067,
      "learning_rate": 1e-06,
      "loss": -0.0248,
      "step": 355
    },
    {
      "clip_ratio/high_max": 0.00218829129516962,
      "clip_ratio/high_mean": 0.0008186861086869612,
      "clip_ratio/low_mean": 0.0012674419085669797,
      "clip_ratio/low_min": 1.3160665730538312e-05,
      "clip_ratio/region_mean": 0.0020861279699602164,
      "epoch": 0.8305628463108778,
      "grad_norm": 0.15224948525428772,
      "learning_rate": 1e-06,
      "loss": -0.0249,
      "step": 356
    },
    {
      "clip_ratio/high_max": 0.0019288021903776098,
      "clip_ratio/high_mean": 0.0007495696963815135,
      "clip_ratio/low_mean": 0.00046653347021674563,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012161031409050338,
      "completions/clipped_ratio": 0.0870535714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3902.0,
      "completions/mean_length": 918.7522583007812,
      "completions/mean_terminated_length": 615.7872924804688,
      "completions/min_length": 160.0,
      "completions/min_terminated_length": 160.0,
      "epoch": 0.8328958880139983,
      "grad_norm": 0.21693141758441925,
      "learning_rate": 1e-06,
      "loss": -0.0144,
      "num_tokens": 54760221.0,
      "reward": 0.5491071939468384,
      "reward_std": 0.1401536762714386,
      "rewards/verify_math_reward/mean": 0.5491071343421936,
      "rewards/verify_math_reward/std": 0.49786055088043213,
      "step": 357
    },
    {
      "clip_ratio/high_max": 0.0026613782247295603,
      "clip_ratio/high_mean": 0.0009070894393516937,
      "clip_ratio/low_mean": 0.0006486262627731776,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015557157057628501,
      "epoch": 0.8352289297171187,
      "grad_norm": 0.179405078291893,
      "learning_rate": 1e-06,
      "loss": -0.0146,
      "step": 358
    },
    {
      "clip_ratio/high_max": 0.0024907990446081385,
      "clip_ratio/high_mean": 0.0009263725514756516,
      "clip_ratio/low_mean": 0.0008601827576057985,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017865552872535773,
      "epoch": 0.8375619714202391,
      "grad_norm": 0.15576620399951935,
      "learning_rate": 1e-06,
      "loss": -0.0148,
      "step": 359
    },
    {
      "clip_ratio/high_max": 0.002331049137865193,
      "clip_ratio/high_mean": 0.0008176261253538541,
      "clip_ratio/low_mean": 0.0009222162534570089,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017398423842678312,
      "epoch": 0.8398950131233596,
      "grad_norm": 0.18607597053050995,
      "learning_rate": 1e-06,
      "loss": -0.0148,
      "step": 360
    },
    {
      "clip_ratio/high_max": 0.0020165567293588538,
      "clip_ratio/high_mean": 0.000853951276440057,
      "clip_ratio/low_mean": 0.0006930507770448457,
      "clip_ratio/low_min": 2.6749410608317703e-05,
      "clip_ratio/region_mean": 0.0015470020662178285,
      "completions/clipped_ratio": 0.078125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3452.0,
      "completions/mean_length": 865.521240234375,
      "completions/mean_terminated_length": 591.7518310546875,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 0.84222805482648,
      "grad_norm": 0.24153633415699005,
      "learning_rate": 1e-06,
      "loss": -0.0373,
      "num_tokens": 55336968.0,
      "reward": 0.5915178656578064,
      "reward_std": 0.1732928454875946,
      "rewards/verify_math_reward/mean": 0.5915178656578064,
      "rewards/verify_math_reward/std": 0.49182769656181335,
      "step": 361
    },
    {
      "clip_ratio/high_max": 0.002464430181134958,
      "clip_ratio/high_mean": 0.0010238850773021113,
      "clip_ratio/low_mean": 0.0009376328980579274,
      "clip_ratio/low_min": 2.2498199541587383e-05,
      "clip_ratio/region_mean": 0.0019615179335232824,
      "epoch": 0.8445610965296004,
      "grad_norm": 0.20365145802497864,
      "learning_rate": 1e-06,
      "loss": -0.0376,
      "step": 362
    },
    {
      "clip_ratio/high_max": 0.002556008934334386,
      "clip_ratio/high_mean": 0.0010945256326522212,
      "clip_ratio/low_mean": 0.001119626198487822,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002214151834778022,
      "epoch": 0.8468941382327209,
      "grad_norm": 0.18342220783233643,
      "learning_rate": 1e-06,
      "loss": -0.0377,
      "step": 363
    },
    {
      "clip_ratio/high_max": 0.0025188724284817,
      "clip_ratio/high_mean": 0.0009661511594458716,
      "clip_ratio/low_mean": 0.0013099270981911104,
      "clip_ratio/low_min": 2.2123893359093927e-05,
      "clip_ratio/region_mean": 0.002276078288559802,
      "epoch": 0.8492271799358414,
      "grad_norm": 0.183487206697464,
      "learning_rate": 1e-06,
      "loss": -0.0378,
      "step": 364
    },
    {
      "clip_ratio/high_max": 0.0023666196648264304,
      "clip_ratio/high_mean": 0.000941422698815586,
      "clip_ratio/low_mean": 0.0005456536673591472,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014870763588987757,
      "completions/clipped_ratio": 0.0725446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3743.0,
      "completions/mean_length": 881.25341796875,
      "completions/mean_terminated_length": 629.799072265625,
      "completions/min_length": 187.0,
      "completions/min_terminated_length": 187.0,
      "epoch": 0.8515602216389618,
      "grad_norm": 0.21291105449199677,
      "learning_rate": 1e-06,
      "loss": -0.0398,
      "num_tokens": 55947411.0,
      "reward": 0.6417410969734192,
      "reward_std": 0.16958087682724,
      "rewards/verify_math_reward/mean": 0.6417410969734192,
      "rewards/verify_math_reward/std": 0.47975653409957886,
      "step": 365
    },
    {
      "clip_ratio/high_max": 0.0023889943404356018,
      "clip_ratio/high_mean": 0.0010003773859352805,
      "clip_ratio/low_mean": 0.0007598972979394603,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001760274768457748,
      "epoch": 0.8538932633420823,
      "grad_norm": 0.18001306056976318,
      "learning_rate": 1e-06,
      "loss": -0.0401,
      "step": 366
    },
    {
      "clip_ratio/high_max": 0.002810595411574468,
      "clip_ratio/high_mean": 0.0011152851329825353,
      "clip_ratio/low_mean": 0.0008203240631701192,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019356091434019618,
      "epoch": 0.8562263050452027,
      "grad_norm": 0.16618499159812927,
      "learning_rate": 1e-06,
      "loss": -0.0402,
      "step": 367
    },
    {
      "clip_ratio/high_max": 0.002615925644931849,
      "clip_ratio/high_mean": 0.0010511315558687784,
      "clip_ratio/low_mean": 0.0010743256916612154,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021254572566249408,
      "epoch": 0.8585593467483231,
      "grad_norm": 0.1623554676771164,
      "learning_rate": 1e-06,
      "loss": -0.0403,
      "step": 368
    },
    {
      "clip_ratio/high_max": 0.002231936941825552,
      "clip_ratio/high_mean": 0.0008358096129086334,
      "clip_ratio/low_mean": 0.0006096504575907602,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014454600895987824,
      "completions/clipped_ratio": 0.09375,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3463.0,
      "completions/mean_length": 927.8984985351562,
      "completions/mean_terminated_length": 600.1637573242188,
      "completions/min_length": 145.0,
      "completions/min_terminated_length": 145.0,
      "epoch": 0.8608923884514436,
      "grad_norm": 0.2518787682056427,
      "learning_rate": 1e-06,
      "loss": -0.0199,
      "num_tokens": 56532280.0,
      "reward": 0.5680803656578064,
      "reward_std": 0.18975545465946198,
      "rewards/verify_math_reward/mean": 0.5680803656578064,
      "rewards/verify_math_reward/std": 0.4956200420856476,
      "step": 369
    },
    {
      "clip_ratio/high_max": 0.0022443662091973238,
      "clip_ratio/high_mean": 0.0009726381431391928,
      "clip_ratio/low_mean": 0.0009172410991595825,
      "clip_ratio/low_min": 2.8811271477025002e-05,
      "clip_ratio/region_mean": 0.0018898792513937224,
      "epoch": 0.863225430154564,
      "grad_norm": 0.21699672937393188,
      "learning_rate": 1e-06,
      "loss": -0.0203,
      "step": 370
    },
    {
      "clip_ratio/high_max": 0.002635184835526161,
      "clip_ratio/high_mean": 0.0010936449471046217,
      "clip_ratio/low_mean": 0.0010467112697369885,
      "clip_ratio/low_min": 3.5840941563947126e-05,
      "clip_ratio/region_mean": 0.00214035621684161,
      "epoch": 0.8655584718576844,
      "grad_norm": 0.18343502283096313,
      "learning_rate": 1e-06,
      "loss": -0.0204,
      "step": 371
    },
    {
      "clip_ratio/high_max": 0.002463686316332314,
      "clip_ratio/high_mean": 0.000991379471088294,
      "clip_ratio/low_mean": 0.0012978556478628889,
      "clip_ratio/low_min": 8.788790728431195e-05,
      "clip_ratio/region_mean": 0.0022892351262271404,
      "epoch": 0.8678915135608049,
      "grad_norm": 0.20154787600040436,
      "learning_rate": 1e-06,
      "loss": -0.0205,
      "step": 372
    },
    {
      "clip_ratio/high_max": 0.0015627559732820373,
      "clip_ratio/high_mean": 0.0007060485222609714,
      "clip_ratio/low_mean": 0.0005485578758452903,
      "clip_ratio/low_min": 9.521632819087245e-06,
      "clip_ratio/region_mean": 0.0012546063589979894,
      "completions/clipped_ratio": 0.0803571428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4008.0,
      "completions/mean_length": 925.9285888671875,
      "completions/mean_terminated_length": 648.9320678710938,
      "completions/min_length": 167.0,
      "completions/min_terminated_length": 167.0,
      "epoch": 0.8702245552639254,
      "grad_norm": 0.2302597016096115,
      "learning_rate": 1e-06,
      "loss": -0.0179,
      "num_tokens": 57143752.0,
      "reward": 0.5613839626312256,
      "reward_std": 0.17092610895633698,
      "rewards/verify_math_reward/mean": 0.5613839030265808,
      "rewards/verify_math_reward/std": 0.496494859457016,
      "step": 373
    },
    {
      "clip_ratio/high_max": 0.0020515899450401776,
      "clip_ratio/high_mean": 0.0007992923783604056,
      "clip_ratio/low_mean": 0.0007848022023608792,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015840945925447159,
      "epoch": 0.8725575969670458,
      "grad_norm": 0.17715898156166077,
      "learning_rate": 1e-06,
      "loss": -0.0181,
      "step": 374
    },
    {
      "clip_ratio/high_max": 0.0024287963824463077,
      "clip_ratio/high_mean": 0.0008912174562283326,
      "clip_ratio/low_mean": 0.0009563523071847158,
      "clip_ratio/low_min": 2.8564900276251137e-05,
      "clip_ratio/region_mean": 0.0018475697725079954,
      "epoch": 0.8748906386701663,
      "grad_norm": 0.245353102684021,
      "learning_rate": 1e-06,
      "loss": -0.0182,
      "step": 375
    },
    {
      "clip_ratio/high_max": 0.0021456541326188017,
      "clip_ratio/high_mean": 0.0008803438231552718,
      "clip_ratio/low_mean": 0.001216527227370534,
      "clip_ratio/low_min": 2.196836612711195e-05,
      "clip_ratio/region_mean": 0.002096871074172668,
      "epoch": 0.8772236803732867,
      "grad_norm": 0.15898433327674866,
      "learning_rate": 1e-06,
      "loss": -0.0184,
      "step": 376
    },
    {
      "clip_ratio/high_max": 0.002165385289117694,
      "clip_ratio/high_mean": 0.0008119576868921285,
      "clip_ratio/low_mean": 0.000668839120407938,
      "clip_ratio/low_min": 2.529340417822823e-05,
      "clip_ratio/region_mean": 0.0014807967891101725,
      "completions/clipped_ratio": 0.0915178571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3139.0,
      "completions/mean_length": 963.4910888671875,
      "completions/mean_terminated_length": 647.9312133789062,
      "completions/min_length": 130.0,
      "completions/min_terminated_length": 130.0,
      "epoch": 0.8795567220764071,
      "grad_norm": 0.21807657182216644,
      "learning_rate": 1e-06,
      "loss": -0.0364,
      "num_tokens": 57758664.0,
      "reward": 0.5412946939468384,
      "reward_std": 0.19512638449668884,
      "rewards/verify_math_reward/mean": 0.5412946343421936,
      "rewards/verify_math_reward/std": 0.49857014417648315,
      "step": 377
    },
    {
      "clip_ratio/high_max": 0.002510105274268426,
      "clip_ratio/high_mean": 0.001023588039970491,
      "clip_ratio/low_mean": 0.0008575574447604595,
      "clip_ratio/low_min": 3.056949663005071e-05,
      "clip_ratio/region_mean": 0.0018811454792739823,
      "epoch": 0.8818897637795275,
      "grad_norm": 0.24140609800815582,
      "learning_rate": 1e-06,
      "loss": -0.0367,
      "step": 378
    },
    {
      "clip_ratio/high_max": 0.00252076792094158,
      "clip_ratio/high_mean": 0.0010582318172964733,
      "clip_ratio/low_mean": 0.0010428122513985727,
      "clip_ratio/low_min": 3.6654876566899475e-05,
      "clip_ratio/region_mean": 0.0021010440614190884,
      "epoch": 0.884222805482648,
      "grad_norm": 0.1708415299654007,
      "learning_rate": 1e-06,
      "loss": -0.0368,
      "step": 379
    },
    {
      "clip_ratio/high_max": 0.0025287936223321594,
      "clip_ratio/high_mean": 0.0010501180331630167,
      "clip_ratio/low_mean": 0.0013154496082279366,
      "clip_ratio/low_min": 9.351067819807213e-05,
      "clip_ratio/region_mean": 0.002365567663218826,
      "epoch": 0.8865558471857684,
      "grad_norm": 0.17794418334960938,
      "learning_rate": 1e-06,
      "loss": -0.0369,
      "step": 380
    },
    {
      "clip_ratio/high_max": 0.002487028541509062,
      "clip_ratio/high_mean": 0.0010293005434505176,
      "clip_ratio/low_mean": 0.000717295255526551,
      "clip_ratio/low_min": 1.2608432371052913e-05,
      "clip_ratio/region_mean": 0.0017465958153479733,
      "completions/clipped_ratio": 0.0837053571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3377.0,
      "completions/mean_length": 919.2344360351562,
      "completions/mean_terminated_length": 629.0304565429688,
      "completions/min_length": 162.0,
      "completions/min_terminated_length": 162.0,
      "epoch": 0.8888888888888888,
      "grad_norm": 0.24101272225379944,
      "learning_rate": 1e-06,
      "loss": -0.0444,
      "num_tokens": 58358554.0,
      "reward": 0.5625,
      "reward_std": 0.2137964516878128,
      "rewards/verify_math_reward/mean": 0.5625,
      "rewards/verify_math_reward/std": 0.49635544419288635,
      "step": 381
    },
    {
      "clip_ratio/high_max": 0.002524464900488965,
      "clip_ratio/high_mean": 0.0012198094445921015,
      "clip_ratio/low_mean": 0.0009389289261889644,
      "clip_ratio/low_min": 1.2608432371052913e-05,
      "clip_ratio/region_mean": 0.002158738367143087,
      "epoch": 0.8912219305920094,
      "grad_norm": 0.20234255492687225,
      "learning_rate": 1e-06,
      "loss": -0.0447,
      "step": 382
    },
    {
      "clip_ratio/high_max": 0.0027121251841890626,
      "clip_ratio/high_mean": 0.0012514218615251593,
      "clip_ratio/low_mean": 0.0012125361117796274,
      "clip_ratio/low_min": 1.2014609637844842e-05,
      "clip_ratio/region_mean": 0.0024639579642098397,
      "epoch": 0.8935549722951298,
      "grad_norm": 0.18023191392421722,
      "learning_rate": 1e-06,
      "loss": -0.0449,
      "step": 383
    },
    {
      "clip_ratio/high_max": 0.0028525425732368603,
      "clip_ratio/high_mean": 0.0012203649239381775,
      "clip_ratio/low_mean": 0.0013325776926649269,
      "clip_ratio/low_min": 4.805843855137937e-05,
      "clip_ratio/region_mean": 0.0025529425765853375,
      "epoch": 0.8958880139982502,
      "grad_norm": 0.2880335748195648,
      "learning_rate": 1e-06,
      "loss": -0.0449,
      "step": 384
    },
    {
      "clip_ratio/high_max": 0.0023073976262821816,
      "clip_ratio/high_mean": 0.0009232878073817119,
      "clip_ratio/low_mean": 0.000623939338765922,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015472271697944961,
      "completions/clipped_ratio": 0.1082589285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3576.0,
      "completions/mean_length": 988.0011596679688,
      "completions/mean_terminated_length": 610.6846313476562,
      "completions/min_length": 179.0,
      "completions/min_terminated_length": 179.0,
      "epoch": 0.8982210557013707,
      "grad_norm": 0.22138522565364838,
      "learning_rate": 1e-06,
      "loss": -0.0393,
      "num_tokens": 58937787.0,
      "reward": 0.546875,
      "reward_std": 0.17461372911930084,
      "rewards/verify_math_reward/mean": 0.546875,
      "rewards/verify_math_reward/std": 0.4980759024620056,
      "step": 385
    },
    {
      "clip_ratio/high_max": 0.002377451295615174,
      "clip_ratio/high_mean": 0.000971550065514748,
      "clip_ratio/low_mean": 0.0008350397292815614,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001806589811167214,
      "epoch": 0.9005540974044911,
      "grad_norm": 0.21801231801509857,
      "learning_rate": 1e-06,
      "loss": -0.0396,
      "step": 386
    },
    {
      "clip_ratio/high_max": 0.0024675863969605416,
      "clip_ratio/high_mean": 0.001060649034116068,
      "clip_ratio/low_mean": 0.0009746981941134436,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020353472282295115,
      "epoch": 0.9028871391076115,
      "grad_norm": 0.18678408861160278,
      "learning_rate": 1e-06,
      "loss": -0.0397,
      "step": 387
    },
    {
      "clip_ratio/high_max": 0.002575601582066156,
      "clip_ratio/high_mean": 0.0010663571447366849,
      "clip_ratio/low_mean": 0.0012509522930486128,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002317309415957425,
      "epoch": 0.905220180810732,
      "grad_norm": 0.19864848256111145,
      "learning_rate": 1e-06,
      "loss": -0.0399,
      "step": 388
    },
    {
      "clip_ratio/high_max": 0.0014403713648789562,
      "clip_ratio/high_mean": 0.0005397231088863919,
      "clip_ratio/low_mean": 0.0004968672092218185,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0010365903108322527,
      "completions/clipped_ratio": 0.1383928571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3012.0,
      "completions/mean_length": 1146.665283203125,
      "completions/mean_terminated_length": 672.9378051757812,
      "completions/min_length": 181.0,
      "completions/min_terminated_length": 181.0,
      "epoch": 0.9075532225138524,
      "grad_norm": 0.2242993712425232,
      "learning_rate": 1e-06,
      "loss": -0.0327,
      "num_tokens": 59543223.0,
      "reward": 0.4933035969734192,
      "reward_std": 0.14199379086494446,
      "rewards/verify_math_reward/mean": 0.4933035671710968,
      "rewards/verify_math_reward/std": 0.5002344250679016,
      "step": 389
    },
    {
      "clip_ratio/high_max": 0.001978857944777701,
      "clip_ratio/high_mean": 0.0007474165613530204,
      "clip_ratio/low_mean": 0.0006779378245482803,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014253543886297848,
      "epoch": 0.9098862642169728,
      "grad_norm": 0.1980229914188385,
      "learning_rate": 1e-06,
      "loss": -0.033,
      "step": 390
    },
    {
      "clip_ratio/high_max": 0.0021036925827502273,
      "clip_ratio/high_mean": 0.0007958486821735278,
      "clip_ratio/low_mean": 0.0007948368183861021,
      "clip_ratio/low_min": 3.2445059332530946e-05,
      "clip_ratio/region_mean": 0.0015906855296634603,
      "epoch": 0.9122193059200934,
      "grad_norm": 0.18185946345329285,
      "learning_rate": 1e-06,
      "loss": -0.0331,
      "step": 391
    },
    {
      "clip_ratio/high_max": 0.0019272925237601157,
      "clip_ratio/high_mean": 0.0007416932021442335,
      "clip_ratio/low_mean": 0.0009933531619026326,
      "clip_ratio/low_min": 3.4594097087392583e-05,
      "clip_ratio/region_mean": 0.0017350463895127177,
      "epoch": 0.9145523476232138,
      "grad_norm": 0.16489240527153015,
      "learning_rate": 1e-06,
      "loss": -0.0331,
      "step": 392
    },
    {
      "clip_ratio/high_max": 0.0018829276377800852,
      "clip_ratio/high_mean": 0.0008267601297120564,
      "clip_ratio/low_mean": 0.0007134396828405443,
      "clip_ratio/low_min": 3.410331191844307e-05,
      "clip_ratio/region_mean": 0.0015401998243760318,
      "completions/clipped_ratio": 0.1171875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4007.0,
      "completions/mean_length": 1032.266845703125,
      "completions/mean_terminated_length": 625.5765380859375,
      "completions/min_length": 175.0,
      "completions/min_terminated_length": 175.0,
      "epoch": 0.9168853893263342,
      "grad_norm": 0.22167567908763885,
      "learning_rate": 1e-06,
      "loss": -0.0315,
      "num_tokens": 60137270.0,
      "reward": 0.5111607313156128,
      "reward_std": 0.17851904034614563,
      "rewards/verify_math_reward/mean": 0.5111607313156128,
      "rewards/verify_math_reward/std": 0.5001546144485474,
      "step": 393
    },
    {
      "clip_ratio/high_max": 0.002248465272714384,
      "clip_ratio/high_mean": 0.0009830765520746354,
      "clip_ratio/low_mean": 0.000986932851446909,
      "clip_ratio/low_min": 4.5183445763541386e-05,
      "clip_ratio/region_mean": 0.0019700093725987244,
      "epoch": 0.9192184310294547,
      "grad_norm": 0.21001669764518738,
      "learning_rate": 1e-06,
      "loss": -0.0318,
      "step": 394
    },
    {
      "clip_ratio/high_max": 0.0024827140441630036,
      "clip_ratio/high_mean": 0.0010166514584852848,
      "clip_ratio/low_mean": 0.00117380219126062,
      "clip_ratio/low_min": 5.6479304475942627e-05,
      "clip_ratio/region_mean": 0.0021904536624788307,
      "epoch": 0.9215514727325751,
      "grad_norm": 0.20626439154148102,
      "learning_rate": 1e-06,
      "loss": -0.0319,
      "step": 395
    },
    {
      "clip_ratio/high_max": 0.002260945359012112,
      "clip_ratio/high_mean": 0.0009840688053373015,
      "clip_ratio/low_mean": 0.0013305551474331878,
      "clip_ratio/low_min": 5.683884955942631e-05,
      "clip_ratio/region_mean": 0.002314623910933733,
      "epoch": 0.9238845144356955,
      "grad_norm": 0.18903161585330963,
      "learning_rate": 1e-06,
      "loss": -0.032,
      "step": 396
    },
    {
      "clip_ratio/high_max": 0.0024143254122463986,
      "clip_ratio/high_mean": 0.0008168222502717981,
      "clip_ratio/low_mean": 0.0005513674987014383,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001368189728964353,
      "completions/clipped_ratio": 0.1495535714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3614.0,
      "completions/mean_length": 1145.763427734375,
      "completions/mean_terminated_length": 626.9553833007812,
      "completions/min_length": 167.0,
      "completions/min_terminated_length": 167.0,
      "epoch": 0.926217556138816,
      "grad_norm": 0.22756241261959076,
      "learning_rate": 1e-06,
      "loss": -0.05,
      "num_tokens": 60705274.0,
      "reward": 0.5345982313156128,
      "reward_std": 0.156542107462883,
      "rewards/verify_math_reward/mean": 0.5345982313156128,
      "rewards/verify_math_reward/std": 0.499080091714859,
      "step": 397
    },
    {
      "clip_ratio/high_max": 0.002396913645497989,
      "clip_ratio/high_mean": 0.0009040411860041786,
      "clip_ratio/low_mean": 0.0007567752163595287,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016608163641649298,
      "epoch": 0.9285505978419364,
      "grad_norm": 0.1984955072402954,
      "learning_rate": 1e-06,
      "loss": -0.0502,
      "step": 398
    },
    {
      "clip_ratio/high_max": 0.002589438794529997,
      "clip_ratio/high_mean": 0.0009209879790432751,
      "clip_ratio/low_mean": 0.0009547004647174617,
      "clip_ratio/low_min": 1.715148209768813e-05,
      "clip_ratio/region_mean": 0.0018756884637696203,
      "epoch": 0.9308836395450568,
      "grad_norm": 0.18810021877288818,
      "learning_rate": 1e-06,
      "loss": -0.0503,
      "step": 399
    },
    {
      "clip_ratio/high_max": 0.0030086884726188146,
      "clip_ratio/high_mean": 0.0010029266068158904,
      "clip_ratio/low_mean": 0.001136030974521418,
      "clip_ratio/low_min": 3.430296419537626e-05,
      "clip_ratio/region_mean": 0.002138957632269012,
      "epoch": 0.9332166812481774,
      "grad_norm": 0.1831684708595276,
      "learning_rate": 1e-06,
      "loss": -0.0504,
      "step": 400
    },
    {
      "clip_ratio/high_max": 0.0022944446245674044,
      "clip_ratio/high_mean": 0.0009729608082125196,
      "clip_ratio/low_mean": 0.0006262494916882133,
      "clip_ratio/low_min": 2.6870164219872095e-05,
      "clip_ratio/region_mean": 0.001599210318090627,
      "completions/clipped_ratio": 0.0982142857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3530.0,
      "completions/mean_length": 968.2835083007812,
      "completions/mean_terminated_length": 627.6410522460938,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 0.9355497229512978,
      "grad_norm": 0.2524567246437073,
      "learning_rate": 1e-06,
      "loss": -0.0506,
      "num_tokens": 61297928.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.1988469660282135,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 401
    },
    {
      "clip_ratio/high_max": 0.0024275017785839736,
      "clip_ratio/high_mean": 0.0011202143268747022,
      "clip_ratio/low_mean": 0.0008954741042543901,
      "clip_ratio/low_min": 2.6870164219872095e-05,
      "clip_ratio/region_mean": 0.0020156884420430288,
      "epoch": 0.9378827646544182,
      "grad_norm": 0.19105035066604614,
      "learning_rate": 1e-06,
      "loss": -0.0508,
      "step": 402
    },
    {
      "clip_ratio/high_max": 0.0025642600739956833,
      "clip_ratio/high_mean": 0.0011850045411847532,
      "clip_ratio/low_mean": 0.0010663451394066215,
      "clip_ratio/low_min": 8.061048720264807e-05,
      "clip_ratio/region_mean": 0.0022513496660394594,
      "epoch": 0.9402158063575387,
      "grad_norm": 0.1914723962545395,
      "learning_rate": 1e-06,
      "loss": -0.0509,
      "step": 403
    },
    {
      "clip_ratio/high_max": 0.002771794876025524,
      "clip_ratio/high_mean": 0.0011238851147936657,
      "clip_ratio/low_mean": 0.0012456118020054419,
      "clip_ratio/low_min": 5.5915901612024754e-05,
      "clip_ratio/region_mean": 0.0023694968404015526,
      "epoch": 0.9425488480606591,
      "grad_norm": 0.1938973218202591,
      "learning_rate": 1e-06,
      "loss": -0.0511,
      "step": 404
    },
    {
      "clip_ratio/high_max": 0.002230028661870165,
      "clip_ratio/high_mean": 0.0007739496850263095,
      "clip_ratio/low_mean": 0.0006409202414943138,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00141486987558892,
      "completions/clipped_ratio": 0.0982142857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3528.0,
      "completions/mean_length": 962.7857666015625,
      "completions/mean_terminated_length": 621.5445556640625,
      "completions/min_length": 144.0,
      "completions/min_terminated_length": 144.0,
      "epoch": 0.9448818897637795,
      "grad_norm": 0.22243300080299377,
      "learning_rate": 1e-06,
      "loss": -0.0346,
      "num_tokens": 61900048.0,
      "reward": 0.5167410969734192,
      "reward_std": 0.155902698636055,
      "rewards/verify_math_reward/mean": 0.5167410969734192,
      "rewards/verify_math_reward/std": 0.4999987483024597,
      "step": 405
    },
    {
      "clip_ratio/high_max": 0.0027140029123984277,
      "clip_ratio/high_mean": 0.0009580726582498755,
      "clip_ratio/low_mean": 0.0008908036925276974,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018488763671484776,
      "epoch": 0.9472149314669,
      "grad_norm": 0.19398583471775055,
      "learning_rate": 1e-06,
      "loss": -0.0348,
      "step": 406
    },
    {
      "clip_ratio/high_max": 0.0027486858452903107,
      "clip_ratio/high_mean": 0.0010101074276462896,
      "clip_ratio/low_mean": 0.0010566887012828374,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020667961143772118,
      "epoch": 0.9495479731700204,
      "grad_norm": 0.17073506116867065,
      "learning_rate": 1e-06,
      "loss": -0.035,
      "step": 407
    },
    {
      "clip_ratio/high_max": 0.002703230216866359,
      "clip_ratio/high_mean": 0.0008891025681805331,
      "clip_ratio/low_mean": 0.0011867632620123914,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002075865850201808,
      "epoch": 0.9518810148731408,
      "grad_norm": 0.17584951221942902,
      "learning_rate": 1e-06,
      "loss": -0.035,
      "step": 408
    },
    {
      "clip_ratio/high_max": 0.0017499868990853429,
      "clip_ratio/high_mean": 0.0006895441420056159,
      "clip_ratio/low_mean": 0.0005366919976950157,
      "clip_ratio/low_min": 2.0552450223476626e-05,
      "clip_ratio/region_mean": 0.001226236159709515,
      "completions/clipped_ratio": 0.1127232142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3481.0,
      "completions/mean_length": 1060.05810546875,
      "completions/mean_terminated_length": 674.3597412109375,
      "completions/min_length": 153.0,
      "completions/min_terminated_length": 153.0,
      "epoch": 0.9542140565762613,
      "grad_norm": 0.21991465985774994,
      "learning_rate": 1e-06,
      "loss": -0.0314,
      "num_tokens": 62523628.0,
      "reward": 0.5178571939468384,
      "reward_std": 0.1661185324192047,
      "rewards/verify_math_reward/mean": 0.5178571343421936,
      "rewards/verify_math_reward/std": 0.4999600946903229,
      "step": 409
    },
    {
      "clip_ratio/high_max": 0.002041467319941148,
      "clip_ratio/high_mean": 0.0008404169966524933,
      "clip_ratio/low_mean": 0.0008108198853733484,
      "clip_ratio/low_min": 9.36329615797149e-06,
      "clip_ratio/region_mean": 0.0016512368674739264,
      "epoch": 0.9565470982793818,
      "grad_norm": 0.2058495134115219,
      "learning_rate": 1e-06,
      "loss": -0.0316,
      "step": 410
    },
    {
      "clip_ratio/high_max": 0.002102226069837343,
      "clip_ratio/high_mean": 0.000885735877091065,
      "clip_ratio/low_mean": 0.0010361661043134518,
      "clip_ratio/low_min": 3.745318463188596e-05,
      "clip_ratio/region_mean": 0.0019219019668526016,
      "epoch": 0.9588801399825022,
      "grad_norm": 0.16779199242591858,
      "learning_rate": 1e-06,
      "loss": -0.0318,
      "step": 411
    },
    {
      "clip_ratio/high_max": 0.002049068563792389,
      "clip_ratio/high_mean": 0.0008057550767262001,
      "clip_ratio/low_mean": 0.0011653085093712434,
      "clip_ratio/low_min": 1.872659231594298e-05,
      "clip_ratio/region_mean": 0.00197106359701138,
      "epoch": 0.9612131816856226,
      "grad_norm": 0.20040445029735565,
      "learning_rate": 1e-06,
      "loss": -0.0319,
      "step": 412
    },
    {
      "clip_ratio/high_max": 0.0021113881448400207,
      "clip_ratio/high_mean": 0.0008135189400491072,
      "clip_ratio/low_mean": 0.0006807227355238865,
      "clip_ratio/low_min": 5.56997511012014e-05,
      "clip_ratio/region_mean": 0.0014942416819394566,
      "completions/clipped_ratio": 0.109375,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3887.0,
      "completions/mean_length": 1005.1975708007812,
      "completions/mean_terminated_length": 625.6253051757812,
      "completions/min_length": 128.0,
      "completions/min_terminated_length": 128.0,
      "epoch": 0.9635462233887431,
      "grad_norm": 0.2439853399991989,
      "learning_rate": 1e-06,
      "loss": -0.0488,
      "num_tokens": 63115821.0,
      "reward": 0.5546875,
      "reward_std": 0.171687051653862,
      "rewards/verify_math_reward/mean": 0.5546875,
      "rewards/verify_math_reward/std": 0.4972778558731079,
      "step": 413
    },
    {
      "clip_ratio/high_max": 0.0029173197981435806,
      "clip_ratio/high_mean": 0.001055622829881031,
      "clip_ratio/low_mean": 0.0009044086964422604,
      "clip_ratio/low_min": 3.234574705857085e-05,
      "clip_ratio/region_mean": 0.001960031564522069,
      "epoch": 0.9658792650918635,
      "grad_norm": 0.212842658162117,
      "learning_rate": 1e-06,
      "loss": -0.0491,
      "step": 414
    },
    {
      "clip_ratio/high_max": 0.002813082253851462,
      "clip_ratio/high_mean": 0.0010721070702857105,
      "clip_ratio/low_mean": 0.0011335684757796116,
      "clip_ratio/low_min": 0.00010438599565532058,
      "clip_ratio/region_mean": 0.0022056755478843115,
      "epoch": 0.9682123067949839,
      "grad_norm": 0.18718186020851135,
      "learning_rate": 1e-06,
      "loss": -0.0492,
      "step": 415
    },
    {
      "clip_ratio/high_max": 0.002706474653678015,
      "clip_ratio/high_mean": 0.0010399333223176654,
      "clip_ratio/low_mean": 0.0013874257856514305,
      "clip_ratio/low_min": 0.00010292992737959139,
      "clip_ratio/region_mean": 0.0024273590970551595,
      "epoch": 0.9705453484981044,
      "grad_norm": 0.16945691406726837,
      "learning_rate": 1e-06,
      "loss": -0.0493,
      "step": 416
    },
    {
      "clip_ratio/high_max": 0.002532029800931923,
      "clip_ratio/high_mean": 0.00104540411484777,
      "clip_ratio/low_mean": 0.0006206272819326841,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001666031384957023,
      "completions/clipped_ratio": 0.0948660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3944.0,
      "completions/mean_length": 929.4297485351562,
      "completions/mean_terminated_length": 597.5449829101562,
      "completions/min_length": 149.0,
      "completions/min_terminated_length": 149.0,
      "epoch": 0.9728783902012248,
      "grad_norm": 0.2921268939971924,
      "learning_rate": 1e-06,
      "loss": -0.027,
      "num_tokens": 63688326.0,
      "reward": 0.5401785969734192,
      "reward_std": 0.19787125289440155,
      "rewards/verify_math_reward/mean": 0.5401785969734192,
      "rewards/verify_math_reward/std": 0.49866142868995667,
      "step": 417
    },
    {
      "clip_ratio/high_max": 0.002880187767004827,
      "clip_ratio/high_mean": 0.0012010327791358577,
      "clip_ratio/low_mean": 0.0008800021423667204,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020810349014936946,
      "epoch": 0.9752114319043453,
      "grad_norm": 0.22427064180374146,
      "learning_rate": 1e-06,
      "loss": -0.0273,
      "step": 418
    },
    {
      "clip_ratio/high_max": 0.002787976758554578,
      "clip_ratio/high_mean": 0.0011979122064076364,
      "clip_ratio/low_mean": 0.0011710269518516725,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002368939225561917,
      "epoch": 0.9775444736074658,
      "grad_norm": 0.22764481604099274,
      "learning_rate": 1e-06,
      "loss": -0.0275,
      "step": 419
    },
    {
      "clip_ratio/high_max": 0.0028928016909048893,
      "clip_ratio/high_mean": 0.0012647529365494847,
      "clip_ratio/low_mean": 0.0013513580543076387,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026161109999520704,
      "epoch": 0.9798775153105862,
      "grad_norm": 0.19134294986724854,
      "learning_rate": 1e-06,
      "loss": -0.0276,
      "step": 420
    },
    {
      "clip_ratio/high_max": 0.0018266616134496871,
      "clip_ratio/high_mean": 0.0007190967153292149,
      "clip_ratio/low_mean": 0.0005109191979499883,
      "clip_ratio/low_min": 1.6720170606276952e-05,
      "clip_ratio/region_mean": 0.001230015892360825,
      "completions/clipped_ratio": 0.1037946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3565.0,
      "completions/mean_length": 982.4832763671875,
      "completions/mean_terminated_length": 621.88916015625,
      "completions/min_length": 173.0,
      "completions/min_terminated_length": 173.0,
      "epoch": 0.9822105570137066,
      "grad_norm": 0.20579728484153748,
      "learning_rate": 1e-06,
      "loss": -0.0488,
      "num_tokens": 64269895.0,
      "reward": 0.574776828289032,
      "reward_std": 0.14571575820446014,
      "rewards/verify_math_reward/mean": 0.5747767686843872,
      "rewards/verify_math_reward/std": 0.49465295672416687,
      "step": 421
    },
    {
      "clip_ratio/high_max": 0.0022223303021746688,
      "clip_ratio/high_mean": 0.0008412895622313954,
      "clip_ratio/low_mean": 0.0006025078419042984,
      "clip_ratio/low_min": 1.1704119970090687e-05,
      "clip_ratio/region_mean": 0.0014437974132306408,
      "epoch": 0.9845435987168271,
      "grad_norm": 0.1702871322631836,
      "learning_rate": 1e-06,
      "loss": -0.0489,
      "step": 422
    },
    {
      "clip_ratio/high_max": 0.002629214672197122,
      "clip_ratio/high_mean": 0.0008907833580451552,
      "clip_ratio/low_mean": 0.000756228415411897,
      "clip_ratio/low_min": 4.724111931864172e-05,
      "clip_ratio/region_mean": 0.0016470117516291793,
      "epoch": 0.9868766404199475,
      "grad_norm": 0.18380087614059448,
      "learning_rate": 1e-06,
      "loss": -0.0491,
      "step": 423
    },
    {
      "clip_ratio/high_max": 0.0024758439176366664,
      "clip_ratio/high_mean": 0.0008649492338008713,
      "clip_ratio/low_mean": 0.0009179535700241104,
      "clip_ratio/low_min": 7.022471982054412e-05,
      "clip_ratio/region_mean": 0.0017829028074629605,
      "epoch": 0.9892096821230679,
      "grad_norm": 0.1679142862558365,
      "learning_rate": 1e-06,
      "loss": -0.0491,
      "step": 424
    },
    {
      "clip_ratio/high_max": 0.0023570269368065055,
      "clip_ratio/high_mean": 0.0010959362571156817,
      "clip_ratio/low_mean": 0.000697734321875032,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017936705917236395,
      "completions/clipped_ratio": 0.1116071428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3606.0,
      "completions/mean_length": 995.2154541015625,
      "completions/mean_terminated_length": 605.6696166992188,
      "completions/min_length": 147.0,
      "completions/min_terminated_length": 147.0,
      "epoch": 0.9915427238261884,
      "grad_norm": 0.310123085975647,
      "learning_rate": 1e-06,
      "loss": -0.0432,
      "num_tokens": 64848512.0,
      "reward": 0.5412946939468384,
      "reward_std": 0.20440296828746796,
      "rewards/verify_math_reward/mean": 0.5412946343421936,
      "rewards/verify_math_reward/std": 0.49857014417648315,
      "step": 425
    },
    {
      "clip_ratio/high_max": 0.0029505515594792087,
      "clip_ratio/high_mean": 0.0013292095172801055,
      "clip_ratio/low_mean": 0.001031558427712298,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023607679031556472,
      "epoch": 0.9938757655293088,
      "grad_norm": 0.25105151534080505,
      "learning_rate": 1e-06,
      "loss": -0.0435,
      "step": 426
    },
    {
      "clip_ratio/high_max": 0.0030129859660519287,
      "clip_ratio/high_mean": 0.0014149835515127052,
      "clip_ratio/low_mean": 0.001323643886280479,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002738627459621057,
      "epoch": 0.9962088072324293,
      "grad_norm": 0.2301369458436966,
      "learning_rate": 1e-06,
      "loss": -0.0438,
      "step": 427
    },
    {
      "clip_ratio/high_max": 0.0031693995551904663,
      "clip_ratio/high_mean": 0.001362594950478524,
      "clip_ratio/low_mean": 0.0015099316769919824,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0028725266965921037,
      "epoch": 0.9985418489355498,
      "grad_norm": 0.21407221257686615,
      "learning_rate": 1e-06,
      "loss": -0.0439,
      "step": 428
    },
    {
      "clip_ratio/high_max": 0.00188614493526984,
      "clip_ratio/high_mean": 0.000804677469204762,
      "clip_ratio/low_mean": 0.0006909416897542542,
      "clip_ratio/low_min": 1.371215421386296e-05,
      "clip_ratio/region_mean": 0.0014956191917008255,
      "completions/clipped_ratio": 0.0669642857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3768.0,
      "completions/mean_length": 850.3594360351562,
      "completions/mean_terminated_length": 617.4186401367188,
      "completions/min_length": 186.0,
      "completions/min_terminated_length": 186.0,
      "epoch": 1.0023330417031204,
      "grad_norm": 0.2333025187253952,
      "learning_rate": 1e-06,
      "loss": -0.0315,
      "num_tokens": 65455098.0,
      "reward": 0.6350446939468384,
      "reward_std": 0.17867198586463928,
      "rewards/verify_math_reward/mean": 0.6350446343421936,
      "rewards/verify_math_reward/std": 0.481686532497406,
      "step": 429
    },
    {
      "clip_ratio/high_max": 0.0023995939482119866,
      "clip_ratio/high_mean": 0.0010416395452921279,
      "clip_ratio/low_mean": 0.0009332010322395945,
      "clip_ratio/low_min": 4.335260018706322e-05,
      "clip_ratio/region_mean": 0.001974840590264648,
      "epoch": 1.0046660834062409,
      "grad_norm": 0.2440246194601059,
      "learning_rate": 1e-06,
      "loss": -0.0317,
      "step": 430
    },
    {
      "clip_ratio/high_max": 0.0022679026078549214,
      "clip_ratio/high_mean": 0.0009606390412955079,
      "clip_ratio/low_mean": 0.0011004831012542127,
      "clip_ratio/low_min": 1.7936576114152558e-05,
      "clip_ratio/region_mean": 0.0020611221698345616,
      "epoch": 1.0069991251093613,
      "grad_norm": 0.17937007546424866,
      "learning_rate": 1e-06,
      "loss": -0.0319,
      "step": 431
    },
    {
      "clip_ratio/high_max": 0.002265503804665059,
      "clip_ratio/high_mean": 0.0009249858303519432,
      "clip_ratio/low_mean": 0.0011825941983261146,
      "clip_ratio/low_min": 8.968288057076279e-06,
      "clip_ratio/region_mean": 0.0021075799959362485,
      "epoch": 1.0093321668124817,
      "grad_norm": 0.18658733367919922,
      "learning_rate": 1e-06,
      "loss": -0.032,
      "step": 432
    },
    {
      "clip_ratio/high_max": 0.002188250800827518,
      "clip_ratio/high_mean": 0.0007850914971641032,
      "clip_ratio/low_mean": 0.0006092657213230268,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013943572448624764,
      "completions/clipped_ratio": 0.0959821428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3200.0,
      "completions/mean_length": 925.9609985351562,
      "completions/mean_terminated_length": 589.388916015625,
      "completions/min_length": 172.0,
      "completions/min_terminated_length": 172.0,
      "epoch": 1.0116652085156022,
      "grad_norm": 0.24572688341140747,
      "learning_rate": 1e-06,
      "loss": -0.0442,
      "num_tokens": 66028951.0,
      "reward": 0.637276828289032,
      "reward_std": 0.15319034457206726,
      "rewards/verify_math_reward/mean": 0.6372767686843872,
      "rewards/verify_math_reward/std": 0.481054425239563,
      "step": 433
    },
    {
      "clip_ratio/high_max": 0.0027622638517641462,
      "clip_ratio/high_mean": 0.0010239535840810277,
      "clip_ratio/low_mean": 0.0007826659921192913,
      "clip_ratio/low_min": 2.5562372684362344e-05,
      "clip_ratio/region_mean": 0.0018066195843857713,
      "epoch": 1.0139982502187226,
      "grad_norm": 0.22620150446891785,
      "learning_rate": 1e-06,
      "loss": -0.0445,
      "step": 434
    },
    {
      "clip_ratio/high_max": 0.002733622441155603,
      "clip_ratio/high_mean": 0.0009858197190624196,
      "clip_ratio/low_mean": 0.0010387497350166086,
      "clip_ratio/low_min": 1.6297262845910154e-05,
      "clip_ratio/region_mean": 0.0020245694468030706,
      "epoch": 1.016331291921843,
      "grad_norm": 0.18662086129188538,
      "learning_rate": 1e-06,
      "loss": -0.0447,
      "step": 435
    },
    {
      "clip_ratio/high_max": 0.002920517210441176,
      "clip_ratio/high_mean": 0.001070248461473966,
      "clip_ratio/low_mean": 0.001208086943734088,
      "clip_ratio/low_min": 3.259452569182031e-05,
      "clip_ratio/region_mean": 0.00227833535609534,
      "epoch": 1.0186643336249634,
      "grad_norm": 0.19832929968833923,
      "learning_rate": 1e-06,
      "loss": -0.0448,
      "step": 436
    },
    {
      "clip_ratio/high_max": 0.00233198835485382,
      "clip_ratio/high_mean": 0.000919369622351951,
      "clip_ratio/low_mean": 0.0005849066842529282,
      "clip_ratio/low_min": 2.708247848204337e-05,
      "clip_ratio/region_mean": 0.0015042763370729517,
      "completions/clipped_ratio": 0.1339285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3678.0,
      "completions/mean_length": 1107.8035888671875,
      "completions/mean_terminated_length": 645.7113037109375,
      "completions/min_length": 131.0,
      "completions/min_terminated_length": 131.0,
      "epoch": 1.020997375328084,
      "grad_norm": 0.25459355115890503,
      "learning_rate": 1e-06,
      "loss": -0.0649,
      "num_tokens": 66623023.0,
      "reward": 0.5334821939468384,
      "reward_std": 0.1828376203775406,
      "rewards/verify_math_reward/mean": 0.5334821343421936,
      "rewards/verify_math_reward/std": 0.49915632605552673,
      "step": 437
    },
    {
      "clip_ratio/high_max": 0.0032297305806423537,
      "clip_ratio/high_mean": 0.0012008415415039053,
      "clip_ratio/low_mean": 0.0008006979960555327,
      "clip_ratio/low_min": 2.0398172637214884e-05,
      "clip_ratio/region_mean": 0.00200153960759053,
      "epoch": 1.0233304170312045,
      "grad_norm": 0.22269797325134277,
      "learning_rate": 1e-06,
      "loss": -0.0652,
      "step": 438
    },
    {
      "clip_ratio/high_max": 0.0029371954951784573,
      "clip_ratio/high_mean": 0.0011158537872688612,
      "clip_ratio/low_mean": 0.0009924400619638618,
      "clip_ratio/low_min": 5.387098099163268e-05,
      "clip_ratio/region_mean": 0.0021082937964820303,
      "epoch": 1.025663458734325,
      "grad_norm": 0.17895840108394623,
      "learning_rate": 1e-06,
      "loss": -0.0654,
      "step": 439
    },
    {
      "clip_ratio/high_max": 0.0031041721085784957,
      "clip_ratio/high_mean": 0.0011557863945199642,
      "clip_ratio/low_mean": 0.0011620121422311058,
      "clip_ratio/low_min": 6.109514288255014e-05,
      "clip_ratio/region_mean": 0.0023177985203801654,
      "epoch": 1.0279965004374454,
      "grad_norm": 0.1888708770275116,
      "learning_rate": 1e-06,
      "loss": -0.0655,
      "step": 440
    },
    {
      "clip_ratio/high_max": 0.002363800063903909,
      "clip_ratio/high_mean": 0.000765284659792087,
      "clip_ratio/low_mean": 0.0006764551799278706,
      "clip_ratio/low_min": 2.4149921955540776e-05,
      "clip_ratio/region_mean": 0.00144173986336682,
      "completions/clipped_ratio": 0.1205357142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3328.0,
      "completions/mean_length": 1028.1195068359375,
      "completions/mean_terminated_length": 607.6484375,
      "completions/min_length": 120.0,
      "completions/min_terminated_length": 120.0,
      "epoch": 1.0303295421405658,
      "grad_norm": 0.29041218757629395,
      "learning_rate": 1e-06,
      "loss": -0.0467,
      "num_tokens": 67206226.0,
      "reward": 0.5055803656578064,
      "reward_std": 0.1731078028678894,
      "rewards/verify_math_reward/mean": 0.5055803656578064,
      "rewards/verify_math_reward/std": 0.5002480745315552,
      "step": 441
    },
    {
      "clip_ratio/high_max": 0.0028002179751638323,
      "clip_ratio/high_mean": 0.0009567866181896534,
      "clip_ratio/low_mean": 0.0009720387351990212,
      "clip_ratio/low_min": 1.2074960977770388e-05,
      "clip_ratio/region_mean": 0.0019288253606646322,
      "epoch": 1.0326625838436863,
      "grad_norm": 0.21335922181606293,
      "learning_rate": 1e-06,
      "loss": -0.047,
      "step": 442
    },
    {
      "clip_ratio/high_max": 0.002605081615911331,
      "clip_ratio/high_mean": 0.0009598919823474716,
      "clip_ratio/low_mean": 0.001113256159442244,
      "clip_ratio/low_min": 1.4397603990801144e-05,
      "clip_ratio/region_mean": 0.002073148156341631,
      "epoch": 1.0349956255468067,
      "grad_norm": 0.1877458244562149,
      "learning_rate": 1e-06,
      "loss": -0.0472,
      "step": 443
    },
    {
      "clip_ratio/high_max": 0.002722404162341263,
      "clip_ratio/high_mean": 0.0009160343215626199,
      "clip_ratio/low_mean": 0.001470162289479049,
      "clip_ratio/low_min": 5.759041596320458e-05,
      "clip_ratio/region_mean": 0.0023861965819378383,
      "epoch": 1.0373286672499271,
      "grad_norm": 0.17515434324741364,
      "learning_rate": 1e-06,
      "loss": -0.0473,
      "step": 444
    },
    {
      "clip_ratio/high_max": 0.0018894981767516583,
      "clip_ratio/high_mean": 0.0007696930078964215,
      "clip_ratio/low_mean": 0.0004294440727790061,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0011991370738542173,
      "completions/clipped_ratio": 0.1261160714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3809.0,
      "completions/mean_length": 1033.2054443359375,
      "completions/mean_terminated_length": 591.1928100585938,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 1.0396617089530475,
      "grad_norm": 0.25584280490875244,
      "learning_rate": 1e-06,
      "loss": -0.0647,
      "num_tokens": 67763706.0,
      "reward": 0.6205357313156128,
      "reward_std": 0.16398167610168457,
      "rewards/verify_math_reward/mean": 0.6205357313156128,
      "rewards/verify_math_reward/std": 0.4855247139930725,
      "step": 445
    },
    {
      "clip_ratio/high_max": 0.0025248981764889322,
      "clip_ratio/high_mean": 0.0010226915183011442,
      "clip_ratio/low_mean": 0.0007047745557429153,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017274660640396178,
      "epoch": 1.041994750656168,
      "grad_norm": 0.20202475786209106,
      "learning_rate": 1e-06,
      "loss": -0.0649,
      "step": 446
    },
    {
      "clip_ratio/high_max": 0.0024538811994716525,
      "clip_ratio/high_mean": 0.0010547231504460797,
      "clip_ratio/low_mean": 0.0008309453060064698,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018856684473576024,
      "epoch": 1.0443277923592884,
      "grad_norm": 0.17896480858325958,
      "learning_rate": 1e-06,
      "loss": -0.0651,
      "step": 447
    },
    {
      "clip_ratio/high_max": 0.0026427094198879786,
      "clip_ratio/high_mean": 0.0010373942131991498,
      "clip_ratio/low_mean": 0.0009970391201932216,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020344333242974244,
      "epoch": 1.0466608340624088,
      "grad_norm": 0.17334654927253723,
      "learning_rate": 1e-06,
      "loss": -0.0652,
      "step": 448
    },
    {
      "clip_ratio/high_max": 0.0017308068145212019,
      "clip_ratio/high_mean": 0.0006800707551519736,
      "clip_ratio/low_mean": 0.0004515814352998859,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00113165218863287,
      "completions/clipped_ratio": 0.1149553571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3804.0,
      "completions/mean_length": 1005.1350708007812,
      "completions/mean_terminated_length": 603.6734008789062,
      "completions/min_length": 159.0,
      "completions/min_terminated_length": 159.0,
      "epoch": 1.0489938757655293,
      "grad_norm": 0.2161651849746704,
      "learning_rate": 1e-06,
      "loss": -0.0636,
      "num_tokens": 68333051.0,
      "reward": 0.5703125,
      "reward_std": 0.13084182143211365,
      "rewards/verify_math_reward/mean": 0.5703125,
      "rewards/verify_math_reward/std": 0.49530795216560364,
      "step": 449
    },
    {
      "clip_ratio/high_max": 0.0022399268855224364,
      "clip_ratio/high_mean": 0.0007800824223522795,
      "clip_ratio/low_mean": 0.0006723454164330178,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001452427837648429,
      "epoch": 1.0513269174686497,
      "grad_norm": 0.16871459782123566,
      "learning_rate": 1e-06,
      "loss": -0.0639,
      "step": 450
    },
    {
      "clip_ratio/high_max": 0.0026769223622977734,
      "clip_ratio/high_mean": 0.0008887108624549,
      "clip_ratio/low_mean": 0.000769434676385572,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016581455092818942,
      "epoch": 1.0536599591717701,
      "grad_norm": 0.17597126960754395,
      "learning_rate": 1e-06,
      "loss": -0.0639,
      "step": 451
    },
    {
      "clip_ratio/high_max": 0.0022117251392046455,
      "clip_ratio/high_mean": 0.0008225946094171377,
      "clip_ratio/low_mean": 0.0008988589370346745,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017214535691891797,
      "epoch": 1.0559930008748906,
      "grad_norm": 0.16394490003585815,
      "learning_rate": 1e-06,
      "loss": -0.064,
      "step": 452
    },
    {
      "clip_ratio/high_max": 0.001844191509007942,
      "clip_ratio/high_mean": 0.0006051035788914305,
      "clip_ratio/low_mean": 0.0005907940394536126,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0011958976356254425,
      "completions/clipped_ratio": 0.1305803571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2765.0,
      "completions/mean_length": 1068.271240234375,
      "completions/mean_terminated_length": 613.5289306640625,
      "completions/min_length": 170.0,
      "completions/min_terminated_length": 170.0,
      "epoch": 1.058326042578011,
      "grad_norm": 0.24267460405826569,
      "learning_rate": 1e-06,
      "loss": -0.055,
      "num_tokens": 68922438.0,
      "reward": 0.5189732313156128,
      "reward_std": 0.16856057941913605,
      "rewards/verify_math_reward/mean": 0.5189732313156128,
      "rewards/verify_math_reward/std": 0.49991893768310547,
      "step": 453
    },
    {
      "clip_ratio/high_max": 0.002777697729470674,
      "clip_ratio/high_mean": 0.0010148937635676702,
      "clip_ratio/low_mean": 0.0007792864762450336,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001794180272554513,
      "epoch": 1.0606590842811314,
      "grad_norm": 0.2198466807603836,
      "learning_rate": 1e-06,
      "loss": -0.0553,
      "step": 454
    },
    {
      "clip_ratio/high_max": 0.0028655557398451492,
      "clip_ratio/high_mean": 0.001007710066915024,
      "clip_ratio/low_mean": 0.0010238405156997032,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020315505607868545,
      "epoch": 1.0629921259842519,
      "grad_norm": 0.18871402740478516,
      "learning_rate": 1e-06,
      "loss": -0.0555,
      "step": 455
    },
    {
      "clip_ratio/high_max": 0.0023621154177817516,
      "clip_ratio/high_mean": 0.0009254372853320092,
      "clip_ratio/low_mean": 0.0011858456418849528,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021112828471814282,
      "epoch": 1.0653251676873725,
      "grad_norm": 0.1881953328847885,
      "learning_rate": 1e-06,
      "loss": -0.0556,
      "step": 456
    },
    {
      "clip_ratio/high_max": 0.0020752472410094924,
      "clip_ratio/high_mean": 0.0008432522045040969,
      "clip_ratio/low_mean": 0.0003859125345115899,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012291647362872027,
      "completions/clipped_ratio": 0.1071428571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2376.0,
      "completions/mean_length": 985.64404296875,
      "completions/mean_terminated_length": 612.4012451171875,
      "completions/min_length": 152.0,
      "completions/min_terminated_length": 152.0,
      "epoch": 1.067658209390493,
      "grad_norm": 0.2506106197834015,
      "learning_rate": 1e-06,
      "loss": -0.0379,
      "num_tokens": 69500271.0,
      "reward": 0.5558035969734192,
      "reward_std": 0.15361177921295166,
      "rewards/verify_math_reward/mean": 0.5558035969734192,
      "rewards/verify_math_reward/std": 0.49715372920036316,
      "step": 457
    },
    {
      "clip_ratio/high_max": 0.002430212880426552,
      "clip_ratio/high_mean": 0.0009447149423067458,
      "clip_ratio/low_mean": 0.0006093982910897466,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001554113237943966,
      "epoch": 1.0699912510936134,
      "grad_norm": 0.19296887516975403,
      "learning_rate": 1e-06,
      "loss": -0.0382,
      "step": 458
    },
    {
      "clip_ratio/high_max": 0.0026876567644649185,
      "clip_ratio/high_mean": 0.0010661713204171974,
      "clip_ratio/low_mean": 0.0007335922709899023,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017997636023210362,
      "epoch": 1.0723242927967338,
      "grad_norm": 0.21186336874961853,
      "learning_rate": 1e-06,
      "loss": -0.0383,
      "step": 459
    },
    {
      "clip_ratio/high_max": 0.0026717531436588615,
      "clip_ratio/high_mean": 0.0010539452923694625,
      "clip_ratio/low_mean": 0.0009490556058153743,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002003000867262017,
      "epoch": 1.0746573344998542,
      "grad_norm": 0.1976313292980194,
      "learning_rate": 1e-06,
      "loss": -0.0384,
      "step": 460
    },
    {
      "clip_ratio/high_max": 0.0017274421516049188,
      "clip_ratio/high_mean": 0.0006933596951057552,
      "clip_ratio/low_mean": 0.0006798601580157992,
      "clip_ratio/low_min": 4.327072747400962e-05,
      "clip_ratio/region_mean": 0.0013732198531215545,
      "completions/clipped_ratio": 0.0904017857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3699.0,
      "completions/mean_length": 932.0457763671875,
      "completions/mean_terminated_length": 617.5914306640625,
      "completions/min_length": 170.0,
      "completions/min_terminated_length": 170.0,
      "epoch": 1.0769903762029747,
      "grad_norm": 0.270966500043869,
      "learning_rate": 1e-06,
      "loss": -0.0645,
      "num_tokens": 70091448.0,
      "reward": 0.6383928656578064,
      "reward_std": 0.15631499886512756,
      "rewards/verify_math_reward/mean": 0.6383928656578064,
      "rewards/verify_math_reward/std": 0.4807341694831848,
      "step": 461
    },
    {
      "clip_ratio/high_max": 0.0022845575695100706,
      "clip_ratio/high_mean": 0.0008984553369373316,
      "clip_ratio/low_mean": 0.0008425280557275983,
      "clip_ratio/low_min": 2.961796326417243e-05,
      "clip_ratio/region_mean": 0.0017409833853889722,
      "epoch": 1.079323417906095,
      "grad_norm": 0.1994178146123886,
      "learning_rate": 1e-06,
      "loss": -0.0647,
      "step": 462
    },
    {
      "clip_ratio/high_max": 0.002211950828495901,
      "clip_ratio/high_mean": 0.0008875480789356516,
      "clip_ratio/low_mean": 0.0010095132511196425,
      "clip_ratio/low_min": 5.4537522373721004e-05,
      "clip_ratio/region_mean": 0.0018970613418787252,
      "epoch": 1.0816564596092155,
      "grad_norm": 0.18105797469615936,
      "learning_rate": 1e-06,
      "loss": -0.0648,
      "step": 463
    },
    {
      "clip_ratio/high_max": 0.0021518303838092834,
      "clip_ratio/high_mean": 0.0008788373797870008,
      "clip_ratio/low_mean": 0.0012122353618906345,
      "clip_ratio/low_min": 5.429171142168343e-05,
      "clip_ratio/region_mean": 0.002091072739858646,
      "epoch": 1.083989501312336,
      "grad_norm": 0.16402515769004822,
      "learning_rate": 1e-06,
      "loss": -0.065,
      "step": 464
    },
    {
      "clip_ratio/high_max": 0.0024228825350292027,
      "clip_ratio/high_mean": 0.0010354480000387412,
      "clip_ratio/low_mean": 0.0005089695328024391,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015444175223819911,
      "completions/clipped_ratio": 0.1015625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2188.0,
      "completions/mean_length": 959.919677734375,
      "completions/mean_terminated_length": 605.4061889648438,
      "completions/min_length": 172.0,
      "completions/min_terminated_length": 172.0,
      "epoch": 1.0863225430154564,
      "grad_norm": 0.28489306569099426,
      "learning_rate": 1e-06,
      "loss": -0.0359,
      "num_tokens": 70663920.0,
      "reward": 0.5948660969734192,
      "reward_std": 0.19050613045692444,
      "rewards/verify_math_reward/mean": 0.5948660969734192,
      "rewards/verify_math_reward/std": 0.49119213223457336,
      "step": 465
    },
    {
      "clip_ratio/high_max": 0.002697384203202091,
      "clip_ratio/high_mean": 0.0011934083595406264,
      "clip_ratio/low_mean": 0.0007661204526812071,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019595287885749713,
      "epoch": 1.0886555847185768,
      "grad_norm": 0.2708839774131775,
      "learning_rate": 1e-06,
      "loss": -0.0362,
      "step": 466
    },
    {
      "clip_ratio/high_max": 0.00271049264119938,
      "clip_ratio/high_mean": 0.00129466382713872,
      "clip_ratio/low_mean": 0.0009820834166021086,
      "clip_ratio/low_min": 2.2329402781906538e-05,
      "clip_ratio/region_mean": 0.002276747276482638,
      "epoch": 1.0909886264216972,
      "grad_norm": 0.20637908577919006,
      "learning_rate": 1e-06,
      "loss": -0.0364,
      "step": 467
    },
    {
      "clip_ratio/high_max": 0.002703095513425069,
      "clip_ratio/high_mean": 0.0012898474706162233,
      "clip_ratio/low_mean": 0.0012216093127790373,
      "clip_ratio/low_min": 2.2329402781906538e-05,
      "clip_ratio/region_mean": 0.002511456848878879,
      "epoch": 1.0933216681248177,
      "grad_norm": 0.2538661062717438,
      "learning_rate": 1e-06,
      "loss": -0.0365,
      "step": 468
    },
    {
      "clip_ratio/high_max": 0.002326559078937862,
      "clip_ratio/high_mean": 0.0008292814873129828,
      "clip_ratio/low_mean": 0.0004958757281201542,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013251571945147589,
      "completions/clipped_ratio": 0.1138392857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3780.0,
      "completions/mean_length": 1031.818115234375,
      "completions/mean_terminated_length": 638.1826171875,
      "completions/min_length": 177.0,
      "completions/min_terminated_length": 177.0,
      "epoch": 1.0956547098279381,
      "grad_norm": 0.23984180390834808,
      "learning_rate": 1e-06,
      "loss": -0.0455,
      "num_tokens": 71244989.0,
      "reward": 0.640625,
      "reward_std": 0.15570516884326935,
      "rewards/verify_math_reward/mean": 0.640625,
      "rewards/verify_math_reward/std": 0.48008525371551514,
      "step": 469
    },
    {
      "clip_ratio/high_max": 0.0030103095414233394,
      "clip_ratio/high_mean": 0.0009939269912138116,
      "clip_ratio/low_mean": 0.0006345219062495744,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016284488810924813,
      "epoch": 1.0979877515310585,
      "grad_norm": 0.20683689415454865,
      "learning_rate": 1e-06,
      "loss": -0.0456,
      "step": 470
    },
    {
      "clip_ratio/high_max": 0.0025437069343752228,
      "clip_ratio/high_mean": 0.0008786136968410574,
      "clip_ratio/low_mean": 0.0008079185627138941,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016865322904777713,
      "epoch": 1.100320793234179,
      "grad_norm": 0.18974076211452484,
      "learning_rate": 1e-06,
      "loss": -0.0458,
      "step": 471
    },
    {
      "clip_ratio/high_max": 0.0029568531972472556,
      "clip_ratio/high_mean": 0.0009815776320465375,
      "clip_ratio/low_mean": 0.000939180177738308,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019207578152418137,
      "epoch": 1.1026538349372994,
      "grad_norm": 0.18512237071990967,
      "learning_rate": 1e-06,
      "loss": -0.0458,
      "step": 472
    },
    {
      "clip_ratio/high_max": 0.0018132032637367956,
      "clip_ratio/high_mean": 0.0007385272347164573,
      "clip_ratio/low_mean": 0.000592039981711423,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013305672109709121,
      "completions/clipped_ratio": 0.0926339285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3465.0,
      "completions/mean_length": 925.0803833007812,
      "completions/mean_terminated_length": 601.35791015625,
      "completions/min_length": 104.0,
      "completions/min_terminated_length": 104.0,
      "epoch": 1.10498687664042,
      "grad_norm": 0.2671792209148407,
      "learning_rate": 1e-06,
      "loss": -0.047,
      "num_tokens": 71807517.0,
      "reward": 0.6261160969734192,
      "reward_std": 0.15503577888011932,
      "rewards/verify_math_reward/mean": 0.6261160969734192,
      "rewards/verify_math_reward/std": 0.48410359025001526,
      "step": 473
    },
    {
      "clip_ratio/high_max": 0.0022711845631420147,
      "clip_ratio/high_mean": 0.000970492787018884,
      "clip_ratio/low_mean": 0.0007535526510764612,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017240454253624193,
      "epoch": 1.1073199183435405,
      "grad_norm": 0.2036406397819519,
      "learning_rate": 1e-06,
      "loss": -0.0471,
      "step": 474
    },
    {
      "clip_ratio/high_max": 0.0024492949814884923,
      "clip_ratio/high_mean": 0.0009340437973150983,
      "clip_ratio/low_mean": 0.0009111156687140465,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018451594587531872,
      "epoch": 1.109652960046661,
      "grad_norm": 0.1986713707447052,
      "learning_rate": 1e-06,
      "loss": -0.0473,
      "step": 475
    },
    {
      "clip_ratio/high_max": 0.0025339678759337403,
      "clip_ratio/high_mean": 0.000984592094027903,
      "clip_ratio/low_mean": 0.0010988305184582714,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020834225942962803,
      "epoch": 1.1119860017497813,
      "grad_norm": 0.17704980075359344,
      "learning_rate": 1e-06,
      "loss": -0.0474,
      "step": 476
    },
    {
      "clip_ratio/high_max": 0.0026995309817721136,
      "clip_ratio/high_mean": 0.0010780071061162744,
      "clip_ratio/low_mean": 0.0005695656591342413,
      "clip_ratio/low_min": 9.328357919002883e-06,
      "clip_ratio/region_mean": 0.0016475728079967666,
      "completions/clipped_ratio": 0.1160714285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4095.0,
      "completions/mean_length": 1060.641845703125,
      "completions/mean_terminated_length": 662.059326171875,
      "completions/min_length": 127.0,
      "completions/min_terminated_length": 127.0,
      "epoch": 1.1143190434529018,
      "grad_norm": 0.26159003376960754,
      "learning_rate": 1e-06,
      "loss": -0.0551,
      "num_tokens": 72428508.0,
      "reward": 0.5703125,
      "reward_std": 0.2054987996816635,
      "rewards/verify_math_reward/mean": 0.5703125,
      "rewards/verify_math_reward/std": 0.49530795216560364,
      "step": 477
    },
    {
      "clip_ratio/high_max": 0.0029697629215661436,
      "clip_ratio/high_mean": 0.0012597079148690682,
      "clip_ratio/low_mean": 0.0008002120393939549,
      "clip_ratio/low_min": 2.153192508558277e-05,
      "clip_ratio/region_mean": 0.0020599199342541397,
      "epoch": 1.1166520851560222,
      "grad_norm": 0.21325941383838654,
      "learning_rate": 1e-06,
      "loss": -0.0554,
      "step": 478
    },
    {
      "clip_ratio/high_max": 0.002954839372250717,
      "clip_ratio/high_mean": 0.0012681778644036967,
      "clip_ratio/low_mean": 0.0009019180924951797,
      "clip_ratio/low_min": 1.0257672329316847e-05,
      "clip_ratio/region_mean": 0.0021700959332520142,
      "epoch": 1.1189851268591426,
      "grad_norm": 0.2004043459892273,
      "learning_rate": 1e-06,
      "loss": -0.0555,
      "step": 479
    },
    {
      "clip_ratio/high_max": 0.0030141370443743654,
      "clip_ratio/high_mean": 0.0012908005010103807,
      "clip_ratio/low_mean": 0.0012389729054120835,
      "clip_ratio/low_min": 3.0773018806939945e-05,
      "clip_ratio/region_mean": 0.002529773410060443,
      "epoch": 1.121318168562263,
      "grad_norm": 0.18501633405685425,
      "learning_rate": 1e-06,
      "loss": -0.0557,
      "step": 480
    },
    {
      "clip_ratio/high_max": 0.002389666642557131,
      "clip_ratio/high_mean": 0.0007511532421631273,
      "clip_ratio/low_mean": 0.0006754465671292564,
      "clip_ratio/low_min": 1.4688601368106902e-05,
      "clip_ratio/region_mean": 0.0014265998215705622,
      "completions/clipped_ratio": 0.1339285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3584.0,
      "completions/mean_length": 1110.560302734375,
      "completions/mean_terminated_length": 648.894287109375,
      "completions/min_length": 171.0,
      "completions/min_terminated_length": 171.0,
      "epoch": 1.1236512102653835,
      "grad_norm": 0.27346184849739075,
      "learning_rate": 1e-06,
      "loss": -0.0267,
      "num_tokens": 73029250.0,
      "reward": 0.5212053656578064,
      "reward_std": 0.15770326554775238,
      "rewards/verify_math_reward/mean": 0.5212053656578064,
      "rewards/verify_math_reward/std": 0.49982911348342896,
      "step": 481
    },
    {
      "clip_ratio/high_max": 0.0030613427225034684,
      "clip_ratio/high_mean": 0.0010165627209062222,
      "clip_ratio/low_mean": 0.0008935898658819497,
      "clip_ratio/low_min": 5.8754405472427607e-05,
      "clip_ratio/region_mean": 0.001910152604978066,
      "epoch": 1.125984251968504,
      "grad_norm": 0.2034171223640442,
      "learning_rate": 1e-06,
      "loss": -0.0271,
      "step": 482
    },
    {
      "clip_ratio/high_max": 0.0025401779857929796,
      "clip_ratio/high_mean": 0.0009724038245622069,
      "clip_ratio/low_mean": 0.0010596897463983623,
      "clip_ratio/low_min": 2.6282590624759905e-05,
      "clip_ratio/region_mean": 0.0020320935982454102,
      "epoch": 1.1283172936716244,
      "grad_norm": 0.1914975643157959,
      "learning_rate": 1e-06,
      "loss": -0.0272,
      "step": 483
    },
    {
      "clip_ratio/high_max": 0.003259218472521752,
      "clip_ratio/high_mean": 0.0010540255243540742,
      "clip_ratio/low_mean": 0.001212483823110233,
      "clip_ratio/low_min": 4.4065804104320705e-05,
      "clip_ratio/region_mean": 0.0022665093711111695,
      "epoch": 1.1306503353747448,
      "grad_norm": 0.22389191389083862,
      "learning_rate": 1e-06,
      "loss": -0.0273,
      "step": 484
    },
    {
      "clip_ratio/high_max": 0.0021428443360491656,
      "clip_ratio/high_mean": 0.0008868116710800678,
      "clip_ratio/low_mean": 0.0006069602713978384,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014937718988221604,
      "completions/clipped_ratio": 0.1037946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2847.0,
      "completions/mean_length": 958.5491333007812,
      "completions/mean_terminated_length": 595.1830444335938,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 1.1329833770778652,
      "grad_norm": 0.2494707852602005,
      "learning_rate": 1e-06,
      "loss": -0.0271,
      "num_tokens": 73585750.0,
      "reward": 0.640625,
      "reward_std": 0.16871324181556702,
      "rewards/verify_math_reward/mean": 0.640625,
      "rewards/verify_math_reward/std": 0.48008525371551514,
      "step": 485
    },
    {
      "clip_ratio/high_max": 0.002578438194177579,
      "clip_ratio/high_mean": 0.0010743611092038918,
      "clip_ratio/low_mean": 0.0008911755230656127,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019655366268125363,
      "epoch": 1.1353164187809857,
      "grad_norm": 0.2744096517562866,
      "learning_rate": 1e-06,
      "loss": -0.0273,
      "step": 486
    },
    {
      "clip_ratio/high_max": 0.003209933507605456,
      "clip_ratio/high_mean": 0.001222103772306582,
      "clip_ratio/low_mean": 0.0009726033858896699,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021947071436443366,
      "epoch": 1.137649460484106,
      "grad_norm": 0.29313361644744873,
      "learning_rate": 1e-06,
      "loss": -0.0275,
      "step": 487
    },
    {
      "clip_ratio/high_max": 0.0030149374433676712,
      "clip_ratio/high_mean": 0.001209986636240501,
      "clip_ratio/low_mean": 0.0013064717168163043,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0025164583494188264,
      "epoch": 1.1399825021872265,
      "grad_norm": 0.19453810155391693,
      "learning_rate": 1e-06,
      "loss": -0.0277,
      "step": 488
    },
    {
      "clip_ratio/high_max": 0.002624071013997309,
      "clip_ratio/high_mean": 0.0009458125041419407,
      "clip_ratio/low_mean": 0.0006403016905096592,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015861142237554304,
      "completions/clipped_ratio": 0.1227678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3931.0,
      "completions/mean_length": 1115.6239013671875,
      "completions/mean_terminated_length": 698.5228881835938,
      "completions/min_length": 151.0,
      "completions/min_terminated_length": 151.0,
      "epoch": 1.142315543890347,
      "grad_norm": 0.24084880948066711,
      "learning_rate": 1e-06,
      "loss": -0.0431,
      "num_tokens": 74224877.0,
      "reward": 0.5948660969734192,
      "reward_std": 0.1831425130367279,
      "rewards/verify_math_reward/mean": 0.5948660969734192,
      "rewards/verify_math_reward/std": 0.49119213223457336,
      "step": 489
    },
    {
      "clip_ratio/high_max": 0.0029585360607597977,
      "clip_ratio/high_mean": 0.0010731726724770851,
      "clip_ratio/low_mean": 0.0008471020973956911,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019202747862436809,
      "epoch": 1.1446485855934676,
      "grad_norm": 0.21253973245620728,
      "learning_rate": 1e-06,
      "loss": -0.0433,
      "step": 490
    },
    {
      "clip_ratio/high_max": 0.003574863643734716,
      "clip_ratio/high_mean": 0.0012558270209410693,
      "clip_ratio/low_mean": 0.0009229206416421221,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002178747636207845,
      "epoch": 1.1469816272965878,
      "grad_norm": 0.1913764625787735,
      "learning_rate": 1e-06,
      "loss": -0.0435,
      "step": 491
    },
    {
      "clip_ratio/high_max": 0.0031777483600308187,
      "clip_ratio/high_mean": 0.0012014315143460408,
      "clip_ratio/low_mean": 0.001132662691816222,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023340941916103475,
      "epoch": 1.1493146689997085,
      "grad_norm": 0.18061964213848114,
      "learning_rate": 1e-06,
      "loss": -0.0435,
      "step": 492
    },
    {
      "clip_ratio/high_max": 0.0020101828704355285,
      "clip_ratio/high_mean": 0.0007751140274194768,
      "clip_ratio/low_mean": 0.00044571058697329136,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012208246043883264,
      "completions/clipped_ratio": 0.1004464285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3894.0,
      "completions/mean_length": 1004.3839721679688,
      "completions/mean_terminated_length": 659.166259765625,
      "completions/min_length": 116.0,
      "completions/min_terminated_length": 116.0,
      "epoch": 1.151647710702829,
      "grad_norm": 0.22036553919315338,
      "learning_rate": 1e-06,
      "loss": -0.0327,
      "num_tokens": 74845677.0,
      "reward": 0.6350446939468384,
      "reward_std": 0.14948049187660217,
      "rewards/verify_math_reward/mean": 0.6350446343421936,
      "rewards/verify_math_reward/std": 0.4816865026950836,
      "step": 493
    },
    {
      "clip_ratio/high_max": 0.0022590553999179974,
      "clip_ratio/high_mean": 0.0009336157945654122,
      "clip_ratio/low_mean": 0.0006544762891280698,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001588092025485821,
      "epoch": 1.1539807524059493,
      "grad_norm": 0.18672379851341248,
      "learning_rate": 1e-06,
      "loss": -0.0329,
      "step": 494
    },
    {
      "clip_ratio/high_max": 0.0024380449067393783,
      "clip_ratio/high_mean": 0.0010128747944690986,
      "clip_ratio/low_mean": 0.0007495823338103946,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017624571337364614,
      "epoch": 1.1563137941090698,
      "grad_norm": 0.22866769134998322,
      "learning_rate": 1e-06,
      "loss": -0.033,
      "step": 495
    },
    {
      "clip_ratio/high_max": 0.0025246197037631646,
      "clip_ratio/high_mean": 0.0009747100557433441,
      "clip_ratio/low_mean": 0.000957716760240146,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019324268505442888,
      "epoch": 1.1586468358121902,
      "grad_norm": 0.16688209772109985,
      "learning_rate": 1e-06,
      "loss": -0.0331,
      "step": 496
    },
    {
      "clip_ratio/high_max": 0.002160491898393957,
      "clip_ratio/high_mean": 0.0007186983257270185,
      "clip_ratio/low_mean": 0.0007738161439192481,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014925144787412137,
      "completions/clipped_ratio": 0.1573660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3946.0,
      "completions/mean_length": 1202.20654296875,
      "completions/mean_terminated_length": 661.7761840820312,
      "completions/min_length": 185.0,
      "completions/min_terminated_length": 185.0,
      "epoch": 1.1609798775153106,
      "grad_norm": 0.24846456944942474,
      "learning_rate": 1e-06,
      "loss": -0.0506,
      "num_tokens": 75435662.0,
      "reward": 0.5234375,
      "reward_std": 0.13575772941112518,
      "rewards/verify_math_reward/mean": 0.5234375,
      "rewards/verify_math_reward/std": 0.49972933530807495,
      "step": 497
    },
    {
      "clip_ratio/high_max": 0.0026769454598252196,
      "clip_ratio/high_mean": 0.0008719501320229028,
      "clip_ratio/low_mean": 0.00085258467879612,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001724534831737401,
      "epoch": 1.163312919218431,
      "grad_norm": 0.22556813061237335,
      "learning_rate": 1e-06,
      "loss": -0.0507,
      "step": 498
    },
    {
      "clip_ratio/high_max": 0.003002273697347846,
      "clip_ratio/high_mean": 0.0009335558734164806,
      "clip_ratio/low_mean": 0.0011308597659080988,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020644156757043675,
      "epoch": 1.1656459609215515,
      "grad_norm": 0.18102315068244934,
      "learning_rate": 1e-06,
      "loss": -0.0509,
      "step": 499
    },
    {
      "clip_ratio/high_max": 0.0029013332896283828,
      "clip_ratio/high_mean": 0.0009207827542923042,
      "clip_ratio/low_mean": 0.0012869851234427188,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022077678513596766,
      "epoch": 1.167979002624672,
      "grad_norm": 0.16825012862682343,
      "learning_rate": 1e-06,
      "loss": -0.051,
      "step": 500
    },
    {
      "clip_ratio/high_max": 0.0017726223231875338,
      "clip_ratio/high_mean": 0.0006842501479695784,
      "clip_ratio/low_mean": 0.00048078535655804444,
      "clip_ratio/low_min": 8.508031896781176e-06,
      "clip_ratio/region_mean": 0.0011650355045276228,
      "completions/clipped_ratio": 0.1015625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3630.0,
      "completions/mean_length": 984.5201416015625,
      "completions/mean_terminated_length": 632.7875366210938,
      "completions/min_length": 164.0,
      "completions/min_terminated_length": 164.0,
      "epoch": 1.1703120443277923,
      "grad_norm": 0.24229595065116882,
      "learning_rate": 1e-06,
      "loss": -0.0404,
      "num_tokens": 76033536.0,
      "reward": 0.6104910969734192,
      "reward_std": 0.15480685234069824,
      "rewards/verify_math_reward/mean": 0.6104910969734192,
      "rewards/verify_math_reward/std": 0.48791125416755676,
      "step": 501
    },
    {
      "clip_ratio/high_max": 0.0024169529206119478,
      "clip_ratio/high_mean": 0.0010063247500511352,
      "clip_ratio/low_mean": 0.0006739567579643335,
      "clip_ratio/low_min": 2.0223264073138125e-05,
      "clip_ratio/region_mean": 0.0016802815152914263,
      "epoch": 1.1726450860309128,
      "grad_norm": 0.2216462790966034,
      "learning_rate": 1e-06,
      "loss": -0.0406,
      "step": 502
    },
    {
      "clip_ratio/high_max": 0.0023699274461250752,
      "clip_ratio/high_mean": 0.0009497327700955793,
      "clip_ratio/low_mean": 0.0008968771453510271,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018466099500074051,
      "epoch": 1.1749781277340332,
      "grad_norm": 0.1699136197566986,
      "learning_rate": 1e-06,
      "loss": -0.0408,
      "step": 503
    },
    {
      "clip_ratio/high_max": 0.002410533001238946,
      "clip_ratio/high_mean": 0.0009472103756706929,
      "clip_ratio/low_mean": 0.0010392735275672749,
      "clip_ratio/low_min": 2.691065674298443e-05,
      "clip_ratio/region_mean": 0.001986483890505042,
      "epoch": 1.1773111694371536,
      "grad_norm": 0.18011756241321564,
      "learning_rate": 1e-06,
      "loss": -0.0409,
      "step": 504
    },
    {
      "clip_ratio/high_max": 0.0017886991263367236,
      "clip_ratio/high_mean": 0.0007002090806054184,
      "clip_ratio/low_mean": 0.0007003072787483688,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001400516335706925,
      "completions/clipped_ratio": 0.1261160714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2693.0,
      "completions/mean_length": 1028.407470703125,
      "completions/mean_terminated_length": 585.702392578125,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 1.179644211140274,
      "grad_norm": 0.27300500869750977,
      "learning_rate": 1e-06,
      "loss": -0.0383,
      "num_tokens": 76574741.0,
      "reward": 0.6350446939468384,
      "reward_std": 0.16183637082576752,
      "rewards/verify_math_reward/mean": 0.6350446343421936,
      "rewards/verify_math_reward/std": 0.481686532497406,
      "step": 505
    },
    {
      "clip_ratio/high_max": 0.0026057456525450107,
      "clip_ratio/high_mean": 0.0010075661084556486,
      "clip_ratio/low_mean": 0.001023779108436429,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020313452623668127,
      "epoch": 1.1819772528433945,
      "grad_norm": 0.2584632933139801,
      "learning_rate": 1e-06,
      "loss": -0.0386,
      "step": 506
    },
    {
      "clip_ratio/high_max": 0.0027077028789790347,
      "clip_ratio/high_mean": 0.0010219967316515977,
      "clip_ratio/low_mean": 0.0011785405349655775,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022005372447893023,
      "epoch": 1.184310294546515,
      "grad_norm": 0.21871411800384521,
      "learning_rate": 1e-06,
      "loss": -0.0388,
      "step": 507
    },
    {
      "clip_ratio/high_max": 0.002569734613643959,
      "clip_ratio/high_mean": 0.0009420320166100282,
      "clip_ratio/low_mean": 0.0014475440202659229,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023895760532468557,
      "epoch": 1.1866433362496354,
      "grad_norm": 0.21576610207557678,
      "learning_rate": 1e-06,
      "loss": -0.0389,
      "step": 508
    },
    {
      "clip_ratio/high_max": 0.0026465314876986668,
      "clip_ratio/high_mean": 0.0011771513309213333,
      "clip_ratio/low_mean": 0.0005883481826458592,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017654995463090017,
      "completions/clipped_ratio": 0.1194196428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2788.0,
      "completions/mean_length": 1036.7489013671875,
      "completions/mean_terminated_length": 621.8694458007812,
      "completions/min_length": 117.0,
      "completions/min_terminated_length": 117.0,
      "epoch": 1.188976377952756,
      "grad_norm": 0.2937774658203125,
      "learning_rate": 1e-06,
      "loss": -0.0722,
      "num_tokens": 77157316.0,
      "reward": 0.5792410969734192,
      "reward_std": 0.19799211621284485,
      "rewards/verify_math_reward/mean": 0.5792410969734192,
      "rewards/verify_math_reward/std": 0.49395665526390076,
      "step": 509
    },
    {
      "clip_ratio/high_max": 0.0030839903265587054,
      "clip_ratio/high_mean": 0.0013798255131405313,
      "clip_ratio/low_mean": 0.0008560552560084034,
      "clip_ratio/low_min": 3.3068783523049206e-05,
      "clip_ratio/region_mean": 0.0022358807545970194,
      "epoch": 1.1913094196558764,
      "grad_norm": 0.25443971157073975,
      "learning_rate": 1e-06,
      "loss": -0.0725,
      "step": 510
    },
    {
      "clip_ratio/high_max": 0.0034403286554152146,
      "clip_ratio/high_mean": 0.0014830293293925934,
      "clip_ratio/low_mean": 0.001141271259257337,
      "clip_ratio/low_min": 4.0102662751451135e-05,
      "clip_ratio/region_mean": 0.0026243005704600364,
      "epoch": 1.1936424613589969,
      "grad_norm": 0.19449834525585175,
      "learning_rate": 1e-06,
      "loss": -0.0727,
      "step": 511
    },
    {
      "clip_ratio/high_max": 0.003032838176295627,
      "clip_ratio/high_mean": 0.001370453795971116,
      "clip_ratio/low_mean": 0.0011973431919614086,
      "clip_ratio/low_min": 7.449344411725178e-05,
      "clip_ratio/region_mean": 0.0025677970043034293,
      "epoch": 1.1959755030621173,
      "grad_norm": 0.20635251700878143,
      "learning_rate": 1e-06,
      "loss": -0.0728,
      "step": 512
    },
    {
      "clip_ratio/high_max": 0.0024143223017745186,
      "clip_ratio/high_mean": 0.0009924522710207384,
      "clip_ratio/low_mean": 0.0007360769614024321,
      "clip_ratio/low_min": 1.7293858036282472e-05,
      "clip_ratio/region_mean": 0.001728529248794075,
      "completions/clipped_ratio": 0.1171875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3750.0,
      "completions/mean_length": 1060.2254638671875,
      "completions/mean_terminated_length": 657.2465209960938,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 1.1983085447652377,
      "grad_norm": 0.2654973864555359,
      "learning_rate": 1e-06,
      "loss": -0.0526,
      "num_tokens": 77770286.0,
      "reward": 0.5680803656578064,
      "reward_std": 0.1995958387851715,
      "rewards/verify_math_reward/mean": 0.5680803656578064,
      "rewards/verify_math_reward/std": 0.4956200420856476,
      "step": 513
    },
    {
      "clip_ratio/high_max": 0.0024228927577496506,
      "clip_ratio/high_mean": 0.0010833551295945654,
      "clip_ratio/low_mean": 0.0010713286137615796,
      "clip_ratio/low_min": 9.454540486331098e-05,
      "clip_ratio/region_mean": 0.0021546837378991768,
      "epoch": 1.2006415864683582,
      "grad_norm": 0.22051560878753662,
      "learning_rate": 1e-06,
      "loss": -0.053,
      "step": 514
    },
    {
      "clip_ratio/high_max": 0.0030360161181306466,
      "clip_ratio/high_mean": 0.0011675423866108758,
      "clip_ratio/low_mean": 0.0012180573357909452,
      "clip_ratio/low_min": 0.00010423983803775627,
      "clip_ratio/region_mean": 0.002385599735134747,
      "epoch": 1.2029746281714786,
      "grad_norm": 0.20579476654529572,
      "learning_rate": 1e-06,
      "loss": -0.0531,
      "step": 515
    },
    {
      "clip_ratio/high_max": 0.002736599933996331,
      "clip_ratio/high_mean": 0.0011807907576439902,
      "clip_ratio/low_mean": 0.0014078295971557964,
      "clip_ratio/low_min": 9.662347838457208e-05,
      "clip_ratio/region_mean": 0.0025886203802656382,
      "epoch": 1.205307669874599,
      "grad_norm": 0.2896462082862854,
      "learning_rate": 1e-06,
      "loss": -0.0533,
      "step": 516
    },
    {
      "clip_ratio/high_max": 0.002434489084407687,
      "clip_ratio/high_mean": 0.001015359470329713,
      "clip_ratio/low_mean": 0.0005827424547533155,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001598101960553322,
      "completions/clipped_ratio": 0.1294642857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4044.0,
      "completions/mean_length": 1130.09375,
      "completions/mean_terminated_length": 689.01025390625,
      "completions/min_length": 123.0,
      "completions/min_terminated_length": 123.0,
      "epoch": 1.2076407115777195,
      "grad_norm": 0.46300458908081055,
      "learning_rate": 1e-06,
      "loss": -0.0369,
      "num_tokens": 78400858.0,
      "reward": 0.5658482313156128,
      "reward_std": 0.17829221487045288,
      "rewards/verify_math_reward/mean": 0.5658482313156128,
      "rewards/verify_math_reward/std": 0.49592188000679016,
      "step": 517
    },
    {
      "clip_ratio/high_max": 0.0025428009321331047,
      "clip_ratio/high_mean": 0.0011331242694723187,
      "clip_ratio/low_mean": 0.0012312159096836695,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002364340180065483,
      "epoch": 1.20997375328084,
      "grad_norm": 0.6013938784599304,
      "learning_rate": 1e-06,
      "loss": -0.0371,
      "step": 518
    },
    {
      "clip_ratio/high_max": 0.0031117841717787087,
      "clip_ratio/high_mean": 0.001198860463773599,
      "clip_ratio/low_mean": 0.002348972273466643,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0035478327226883266,
      "epoch": 1.2123067949839603,
      "grad_norm": 0.3071836829185486,
      "learning_rate": 1e-06,
      "loss": -0.0375,
      "step": 519
    },
    {
      "clip_ratio/high_max": 0.002819619156070985,
      "clip_ratio/high_mean": 0.0012652650621021166,
      "clip_ratio/low_mean": 0.002037731521340902,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.003302996585262008,
      "epoch": 1.2146398366870808,
      "grad_norm": 2.0215115547180176,
      "learning_rate": 1e-06,
      "loss": -0.0354,
      "step": 520
    },
    {
      "clip_ratio/high_max": 0.002378325239988044,
      "clip_ratio/high_mean": 0.0008747008068894502,
      "clip_ratio/low_mean": 0.0004809233500964183,
      "clip_ratio/low_min": 1.7477628716733307e-05,
      "clip_ratio/region_mean": 0.001355624190182425,
      "completions/clipped_ratio": 0.0814732142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3359.0,
      "completions/mean_length": 879.5569458007812,
      "completions/mean_terminated_length": 594.2588500976562,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 1.2169728783902012,
      "grad_norm": 0.29543742537498474,
      "learning_rate": 1e-06,
      "loss": -0.0413,
      "num_tokens": 78975085.0,
      "reward": 0.5892857313156128,
      "reward_std": 0.15221282839775085,
      "rewards/verify_math_reward/mean": 0.5892857313156128,
      "rewards/verify_math_reward/std": 0.49223825335502625,
      "step": 521
    },
    {
      "clip_ratio/high_max": 0.002576551472884603,
      "clip_ratio/high_mean": 0.0009847943820204819,
      "clip_ratio/low_mean": 0.0007131154156923003,
      "clip_ratio/low_min": 2.307124486833345e-05,
      "clip_ratio/region_mean": 0.0016979097708826885,
      "epoch": 1.2193059200933216,
      "grad_norm": 0.22369630634784698,
      "learning_rate": 1e-06,
      "loss": -0.0416,
      "step": 522
    },
    {
      "clip_ratio/high_max": 0.0028981081777601503,
      "clip_ratio/high_mean": 0.0010226434642390814,
      "clip_ratio/low_mean": 0.0008572595120313053,
      "clip_ratio/low_min": 3.495525743346661e-05,
      "clip_ratio/region_mean": 0.0018799030040099751,
      "epoch": 1.221638961796442,
      "grad_norm": 0.19105824828147888,
      "learning_rate": 1e-06,
      "loss": -0.0418,
      "step": 523
    },
    {
      "clip_ratio/high_max": 0.0028165603798697703,
      "clip_ratio/high_mean": 0.0010105444034707034,
      "clip_ratio/low_mean": 0.0009608435766494949,
      "clip_ratio/low_min": 2.4029219275689684e-05,
      "clip_ratio/region_mean": 0.0019713879737537354,
      "epoch": 1.2239720034995625,
      "grad_norm": 0.2389257699251175,
      "learning_rate": 1e-06,
      "loss": -0.0418,
      "step": 524
    },
    {
      "clip_ratio/high_max": 0.002592427917988971,
      "clip_ratio/high_mean": 0.0009008916185848648,
      "clip_ratio/low_mean": 0.00042396093704155646,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013248525792732835,
      "completions/clipped_ratio": 0.1082589285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3717.0,
      "completions/mean_length": 969.5078735351562,
      "completions/mean_terminated_length": 589.9461669921875,
      "completions/min_length": 163.0,
      "completions/min_terminated_length": 163.0,
      "epoch": 1.226305045202683,
      "grad_norm": 0.22978700697422028,
      "learning_rate": 1e-06,
      "loss": -0.0676,
      "num_tokens": 79531948.0,
      "reward": 0.6071428656578064,
      "reward_std": 0.151575967669487,
      "rewards/verify_math_reward/mean": 0.6071428656578064,
      "rewards/verify_math_reward/std": 0.48865827918052673,
      "step": 525
    },
    {
      "clip_ratio/high_max": 0.003308030150947161,
      "clip_ratio/high_mean": 0.0011520420393935638,
      "clip_ratio/low_mean": 0.0006933599233889254,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018454019736964256,
      "epoch": 1.2286380869058036,
      "grad_norm": 0.1991288810968399,
      "learning_rate": 1e-06,
      "loss": -0.0679,
      "step": 526
    },
    {
      "clip_ratio/high_max": 0.0037966473028063774,
      "clip_ratio/high_mean": 0.0012030818907078356,
      "clip_ratio/low_mean": 0.0008858901655912632,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002088972025376279,
      "epoch": 1.2309711286089238,
      "grad_norm": 0.1985888034105301,
      "learning_rate": 1e-06,
      "loss": -0.068,
      "step": 527
    },
    {
      "clip_ratio/high_max": 0.0035153744684066623,
      "clip_ratio/high_mean": 0.0011972285283263773,
      "clip_ratio/low_mean": 0.0010277893488819245,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002225017909950111,
      "epoch": 1.2333041703120444,
      "grad_norm": 0.18501132726669312,
      "learning_rate": 1e-06,
      "loss": -0.0681,
      "step": 528
    },
    {
      "clip_ratio/high_max": 0.002251866906590294,
      "clip_ratio/high_mean": 0.0007997916636668378,
      "clip_ratio/low_mean": 0.0006696858235955006,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014694774727104232,
      "completions/clipped_ratio": 0.1450892857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2754.0,
      "completions/mean_length": 1112.25341796875,
      "completions/mean_terminated_length": 605.8733520507812,
      "completions/min_length": 181.0,
      "completions/min_terminated_length": 181.0,
      "epoch": 1.2356372120151649,
      "grad_norm": 0.29438626766204834,
      "learning_rate": 1e-06,
      "loss": -0.0569,
      "num_tokens": 80086919.0,
      "reward": 0.5390625,
      "reward_std": 0.16435301303863525,
      "rewards/verify_math_reward/mean": 0.5390625,
      "rewards/verify_math_reward/std": 0.4987502098083496,
      "step": 529
    },
    {
      "clip_ratio/high_max": 0.0027898812768398784,
      "clip_ratio/high_mean": 0.0010612102978484472,
      "clip_ratio/low_mean": 0.0008457083604298532,
      "clip_ratio/low_min": 1.3377568393480033e-05,
      "clip_ratio/region_mean": 0.0019069186528213322,
      "epoch": 1.2379702537182853,
      "grad_norm": 0.22571726143360138,
      "learning_rate": 1e-06,
      "loss": -0.0572,
      "step": 530
    },
    {
      "clip_ratio/high_max": 0.0031235070564434864,
      "clip_ratio/high_mean": 0.0010874585277633741,
      "clip_ratio/low_mean": 0.0010993516680173343,
      "clip_ratio/low_min": 1.3377568393480033e-05,
      "clip_ratio/region_mean": 0.002186810153943952,
      "epoch": 1.2403032954214057,
      "grad_norm": 0.18684718012809753,
      "learning_rate": 1e-06,
      "loss": -0.0574,
      "step": 531
    },
    {
      "clip_ratio/high_max": 0.002868225252314005,
      "clip_ratio/high_mean": 0.0010452830683789216,
      "clip_ratio/low_mean": 0.0012705202316283248,
      "clip_ratio/low_min": 3.6909448681399226e-05,
      "clip_ratio/region_mean": 0.0023158032781793736,
      "epoch": 1.2426363371245261,
      "grad_norm": 0.20576512813568115,
      "learning_rate": 1e-06,
      "loss": -0.0574,
      "step": 532
    },
    {
      "clip_ratio/high_max": 0.0024350592066184618,
      "clip_ratio/high_mean": 0.0010799655137816444,
      "clip_ratio/low_mean": 0.0005255854721326614,
      "clip_ratio/low_min": 1.6657782907714136e-05,
      "clip_ratio/region_mean": 0.001605550991371274,
      "completions/clipped_ratio": 0.109375,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3299.0,
      "completions/mean_length": 976.3772583007812,
      "completions/mean_terminated_length": 593.2656860351562,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 1.2449693788276466,
      "grad_norm": 0.3275110423564911,
      "learning_rate": 1e-06,
      "loss": -0.0671,
      "num_tokens": 80641609.0,
      "reward": 0.6428571939468384,
      "reward_std": 0.18265356123447418,
      "rewards/verify_math_reward/mean": 0.6428571343421936,
      "rewards/verify_math_reward/std": 0.4794250428676605,
      "step": 533
    },
    {
      "clip_ratio/high_max": 0.0031015579297672957,
      "clip_ratio/high_mean": 0.0013122145355737302,
      "clip_ratio/low_mean": 0.000775938459810277,
      "clip_ratio/low_min": 2.710908665903844e-05,
      "clip_ratio/region_mean": 0.002088153036311269,
      "epoch": 1.247302420530767,
      "grad_norm": 0.281019926071167,
      "learning_rate": 1e-06,
      "loss": -0.0673,
      "step": 534
    },
    {
      "clip_ratio/high_max": 0.002716930066526402,
      "clip_ratio/high_mean": 0.0012175443043815903,
      "clip_ratio/low_mean": 0.0009726612188387662,
      "clip_ratio/low_min": 1.6657782907714136e-05,
      "clip_ratio/region_mean": 0.0021902055596001446,
      "epoch": 1.2496354622338874,
      "grad_norm": 0.20063097774982452,
      "learning_rate": 1e-06,
      "loss": -0.0675,
      "step": 535
    },
    {
      "clip_ratio/high_max": 0.0030896394382580183,
      "clip_ratio/high_mean": 0.0012307345277804416,
      "clip_ratio/low_mean": 0.0011316145337332273,
      "clip_ratio/low_min": 1.355454332951922e-05,
      "clip_ratio/region_mean": 0.0023623489978490397,
      "epoch": 1.2519685039370079,
      "grad_norm": 0.2279004156589508,
      "learning_rate": 1e-06,
      "loss": -0.0675,
      "step": 536
    },
    {
      "clip_ratio/high_max": 0.002269060983962845,
      "clip_ratio/high_mean": 0.0009541396066197194,
      "clip_ratio/low_mean": 0.0005999795075695147,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015541191351076122,
      "completions/clipped_ratio": 0.0970982142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3890.0,
      "completions/mean_length": 934.7801513671875,
      "completions/mean_terminated_length": 594.822021484375,
      "completions/min_length": 126.0,
      "completions/min_terminated_length": 126.0,
      "epoch": 1.2543015456401283,
      "grad_norm": 0.2848651111125946,
      "learning_rate": 1e-06,
      "loss": -0.0739,
      "num_tokens": 81206836.0,
      "reward": 0.6696428656578064,
      "reward_std": 0.17649206519126892,
      "rewards/verify_math_reward/mean": 0.6696428656578064,
      "rewards/verify_math_reward/std": 0.47060438990592957,
      "step": 537
    },
    {
      "clip_ratio/high_max": 0.00296455933857942,
      "clip_ratio/high_mean": 0.0012060076842317358,
      "clip_ratio/low_mean": 0.0008301700763695408,
      "clip_ratio/low_min": 1.385194991598837e-05,
      "clip_ratio/region_mean": 0.0020361777787911706,
      "epoch": 1.2566345873432487,
      "grad_norm": 0.3101758360862732,
      "learning_rate": 1e-06,
      "loss": -0.0741,
      "step": 538
    },
    {
      "clip_ratio/high_max": 0.0026383453950984403,
      "clip_ratio/high_mean": 0.0011584382482396904,
      "clip_ratio/low_mean": 0.0010574414545772015,
      "clip_ratio/low_min": 3.907471182174049e-05,
      "clip_ratio/region_mean": 0.0022158796709845774,
      "epoch": 1.2589676290463692,
      "grad_norm": 0.21931804716587067,
      "learning_rate": 1e-06,
      "loss": -0.0743,
      "step": 539
    },
    {
      "clip_ratio/high_max": 0.002541691101214383,
      "clip_ratio/high_mean": 0.001116355419071624,
      "clip_ratio/low_mean": 0.0012633948963411967,
      "clip_ratio/low_min": 2.770389983197674e-05,
      "clip_ratio/region_mean": 0.0023797503308742307,
      "epoch": 1.2613006707494896,
      "grad_norm": 0.3209487199783325,
      "learning_rate": 1e-06,
      "loss": -0.0744,
      "step": 540
    },
    {
      "clip_ratio/high_max": 0.0023388528879877413,
      "clip_ratio/high_mean": 0.0008931090860642144,
      "clip_ratio/low_mean": 0.0006677318378933705,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001560840908496175,
      "completions/clipped_ratio": 0.1127232142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3303.0,
      "completions/mean_length": 992.5625610351562,
      "completions/mean_terminated_length": 598.289306640625,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 1.26363371245261,
      "grad_norm": 0.24479156732559204,
      "learning_rate": 1e-06,
      "loss": -0.0438,
      "num_tokens": 81774076.0,
      "reward": 0.5770089626312256,
      "reward_std": 0.1669466197490692,
      "rewards/verify_math_reward/mean": 0.5770089030265808,
      "rewards/verify_math_reward/std": 0.4943099617958069,
      "step": 541
    },
    {
      "clip_ratio/high_max": 0.002695808510907227,
      "clip_ratio/high_mean": 0.0011374612786312355,
      "clip_ratio/low_mean": 0.0007252048162627034,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018626660930749495,
      "epoch": 1.2659667541557305,
      "grad_norm": 0.22449366748332977,
      "learning_rate": 1e-06,
      "loss": -0.0439,
      "step": 542
    },
    {
      "clip_ratio/high_max": 0.0026009647990576923,
      "clip_ratio/high_mean": 0.0010876012001972413,
      "clip_ratio/low_mean": 0.0010571295370027656,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021447307444759645,
      "epoch": 1.268299795858851,
      "grad_norm": 0.1874256134033203,
      "learning_rate": 1e-06,
      "loss": -0.0441,
      "step": 543
    },
    {
      "clip_ratio/high_max": 0.002826059189828811,
      "clip_ratio/high_mean": 0.0011576665710890666,
      "clip_ratio/low_mean": 0.0011849075144709786,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002342574080103077,
      "epoch": 1.2706328375619713,
      "grad_norm": 0.17332985997200012,
      "learning_rate": 1e-06,
      "loss": -0.0442,
      "step": 544
    },
    {
      "clip_ratio/high_max": 0.0022264186009124387,
      "clip_ratio/high_mean": 0.001017978900563321,
      "clip_ratio/low_mean": 0.000610701295954641,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016286801983369514,
      "completions/clipped_ratio": 0.125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2789.0,
      "completions/mean_length": 1057.01904296875,
      "completions/mean_terminated_length": 622.8787841796875,
      "completions/min_length": 142.0,
      "completions/min_terminated_length": 142.0,
      "epoch": 1.272965879265092,
      "grad_norm": 0.3083045184612274,
      "learning_rate": 1e-06,
      "loss": -0.05,
      "num_tokens": 82348637.0,
      "reward": 0.5770089626312256,
      "reward_std": 0.18761083483695984,
      "rewards/verify_math_reward/mean": 0.5770089030265808,
      "rewards/verify_math_reward/std": 0.4943099319934845,
      "step": 545
    },
    {
      "clip_ratio/high_max": 0.0033035823944373988,
      "clip_ratio/high_mean": 0.0012792956767953,
      "clip_ratio/low_mean": 0.0010433476818434428,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023226433695526794,
      "epoch": 1.2752989209682122,
      "grad_norm": 0.2574704885482788,
      "learning_rate": 1e-06,
      "loss": -0.0503,
      "step": 546
    },
    {
      "clip_ratio/high_max": 0.0032511619647266343,
      "clip_ratio/high_mean": 0.0012776143485098146,
      "clip_ratio/low_mean": 0.0011831686570076272,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002460783020069357,
      "epoch": 1.2776319626713328,
      "grad_norm": 0.23422737419605255,
      "learning_rate": 1e-06,
      "loss": -0.0504,
      "step": 547
    },
    {
      "clip_ratio/high_max": 0.002982414713187609,
      "clip_ratio/high_mean": 0.0011638829091680236,
      "clip_ratio/low_mean": 0.001405524933943525,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002569407799455803,
      "epoch": 1.2799650043744533,
      "grad_norm": 0.2075558751821518,
      "learning_rate": 1e-06,
      "loss": -0.0506,
      "step": 548
    },
    {
      "clip_ratio/high_max": 0.0018218839504697826,
      "clip_ratio/high_mean": 0.0006480422534878016,
      "clip_ratio/low_mean": 0.0006002505042488337,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012482927522796672,
      "completions/clipped_ratio": 0.1450892857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3967.0,
      "completions/mean_length": 1145.2366943359375,
      "completions/mean_terminated_length": 644.454345703125,
      "completions/min_length": 171.0,
      "completions/min_terminated_length": 171.0,
      "epoch": 1.2822980460775737,
      "grad_norm": 0.23727191984653473,
      "learning_rate": 1e-06,
      "loss": -0.0327,
      "num_tokens": 82934601.0,
      "reward": 0.5502232313156128,
      "reward_std": 0.13722196221351624,
      "rewards/verify_math_reward/mean": 0.5502232313156128,
      "rewards/verify_math_reward/std": 0.49774909019470215,
      "step": 549
    },
    {
      "clip_ratio/high_max": 0.0020135488175583305,
      "clip_ratio/high_mean": 0.000720536571861885,
      "clip_ratio/low_mean": 0.000795816269601346,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015163528587436303,
      "epoch": 1.2846310877806941,
      "grad_norm": 0.20960785448551178,
      "learning_rate": 1e-06,
      "loss": -0.0328,
      "step": 550
    },
    {
      "clip_ratio/high_max": 0.0021363516370911384,
      "clip_ratio/high_mean": 0.0007712140754847496,
      "clip_ratio/low_mean": 0.001017558195599122,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017887722860905342,
      "epoch": 1.2869641294838146,
      "grad_norm": 0.1890256106853485,
      "learning_rate": 1e-06,
      "loss": -0.0329,
      "step": 551
    },
    {
      "clip_ratio/high_max": 0.002385011419391958,
      "clip_ratio/high_mean": 0.0007658743115825928,
      "clip_ratio/low_mean": 0.0012119299499318004,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019778042478719726,
      "epoch": 1.289297171186935,
      "grad_norm": 0.16167958080768585,
      "learning_rate": 1e-06,
      "loss": -0.033,
      "step": 552
    },
    {
      "clip_ratio/high_max": 0.0025781944350455888,
      "clip_ratio/high_mean": 0.001074661959137302,
      "clip_ratio/low_mean": 0.0006483473480329849,
      "clip_ratio/low_min": 1.0751763511507306e-05,
      "clip_ratio/region_mean": 0.0017230093289981596,
      "completions/clipped_ratio": 0.1618303571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3884.0,
      "completions/mean_length": 1220.677490234375,
      "completions/mean_terminated_length": 665.52197265625,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 1.2916302128900554,
      "grad_norm": 0.2890176475048065,
      "learning_rate": 1e-06,
      "loss": -0.0684,
      "num_tokens": 83530192.0,
      "reward": 0.515625,
      "reward_std": 0.19622044265270233,
      "rewards/verify_math_reward/mean": 0.515625,
      "rewards/verify_math_reward/std": 0.5000349283218384,
      "step": 553
    },
    {
      "clip_ratio/high_max": 0.0029420962018775754,
      "clip_ratio/high_mean": 0.0012198284202895593,
      "clip_ratio/low_mean": 0.000763745676522376,
      "clip_ratio/low_min": 2.1606361769954674e-05,
      "clip_ratio/region_mean": 0.0019835740604321472,
      "epoch": 1.2939632545931758,
      "grad_norm": 0.24971313774585724,
      "learning_rate": 1e-06,
      "loss": -0.0685,
      "step": 554
    },
    {
      "clip_ratio/high_max": 0.0034426571655785665,
      "clip_ratio/high_mean": 0.001382934980938444,
      "clip_ratio/low_mean": 0.001083816277969163,
      "clip_ratio/low_min": 1.0751763511507306e-05,
      "clip_ratio/region_mean": 0.0024667512843734585,
      "epoch": 1.2962962962962963,
      "grad_norm": 0.21363192796707153,
      "learning_rate": 1e-06,
      "loss": -0.0688,
      "step": 555
    },
    {
      "clip_ratio/high_max": 0.003283892227045726,
      "clip_ratio/high_mean": 0.0013199301065469626,
      "clip_ratio/low_mean": 0.0012499797667260282,
      "clip_ratio/low_min": 2.1244051822577603e-05,
      "clip_ratio/region_mean": 0.0025699098987388425,
      "epoch": 1.2986293379994167,
      "grad_norm": 0.22296246886253357,
      "learning_rate": 1e-06,
      "loss": -0.0689,
      "step": 556
    },
    {
      "clip_ratio/high_max": 0.001826815896492917,
      "clip_ratio/high_mean": 0.0006240852071641712,
      "clip_ratio/low_mean": 0.0005783656320090813,
      "clip_ratio/low_min": 9.656983820605092e-06,
      "clip_ratio/region_mean": 0.0012024508323520422,
      "completions/clipped_ratio": 0.1517857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3930.0,
      "completions/mean_length": 1182.9554443359375,
      "completions/mean_terminated_length": 661.6737060546875,
      "completions/min_length": 178.0,
      "completions/min_terminated_length": 178.0,
      "epoch": 1.3009623797025371,
      "grad_norm": 0.2738368511199951,
      "learning_rate": 1e-06,
      "loss": -0.054,
      "num_tokens": 84126120.0,
      "reward": 0.5089285969734192,
      "reward_std": 0.1451537311077118,
      "rewards/verify_math_reward/mean": 0.5089285969734192,
      "rewards/verify_math_reward/std": 0.5001994967460632,
      "step": 557
    },
    {
      "clip_ratio/high_max": 0.0022162531822687015,
      "clip_ratio/high_mean": 0.0007980015361681581,
      "clip_ratio/low_mean": 0.0007888012933108257,
      "clip_ratio/low_min": 1.0867674973269459e-05,
      "clip_ratio/region_mean": 0.0015868028131080791,
      "epoch": 1.3032954214056576,
      "grad_norm": 0.2366650104522705,
      "learning_rate": 1e-06,
      "loss": -0.0543,
      "step": 558
    },
    {
      "clip_ratio/high_max": 0.0025047057497431524,
      "clip_ratio/high_mean": 0.0008618533920525806,
      "clip_ratio/low_mean": 0.001016271615299047,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001878125018265564,
      "epoch": 1.305628463108778,
      "grad_norm": 0.21774974465370178,
      "learning_rate": 1e-06,
      "loss": -0.0544,
      "step": 559
    },
    {
      "clip_ratio/high_max": 0.0023650196781090926,
      "clip_ratio/high_mean": 0.0008386210593016585,
      "clip_ratio/low_mean": 0.0012069545991835184,
      "clip_ratio/low_min": 6.15359385847114e-05,
      "clip_ratio/region_mean": 0.0020455756784940604,
      "epoch": 1.3079615048118984,
      "grad_norm": 0.18776483833789825,
      "learning_rate": 1e-06,
      "loss": -0.0545,
      "step": 560
    },
    {
      "clip_ratio/high_max": 0.002210908722190652,
      "clip_ratio/high_mean": 0.000765536697144853,
      "clip_ratio/low_mean": 0.000620098414401582,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013856350979040144,
      "completions/clipped_ratio": 0.1696428571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4046.0,
      "completions/mean_length": 1280.943115234375,
      "completions/mean_terminated_length": 705.8239135742188,
      "completions/min_length": 149.0,
      "completions/min_terminated_length": 149.0,
      "epoch": 1.3102945465150189,
      "grad_norm": 0.24883843958377838,
      "learning_rate": 1e-06,
      "loss": -0.0668,
      "num_tokens": 84745181.0,
      "reward": 0.5167410969734192,
      "reward_std": 0.15518662333488464,
      "rewards/verify_math_reward/mean": 0.5167410969734192,
      "rewards/verify_math_reward/std": 0.4999987483024597,
      "step": 561
    },
    {
      "clip_ratio/high_max": 0.0027370128518668935,
      "clip_ratio/high_mean": 0.0010297183680449962,
      "clip_ratio/low_mean": 0.0008404065083595924,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018701249246078078,
      "epoch": 1.3126275882181395,
      "grad_norm": 0.19404339790344238,
      "learning_rate": 1e-06,
      "loss": -0.067,
      "step": 562
    },
    {
      "clip_ratio/high_max": 0.002672463200724451,
      "clip_ratio/high_mean": 0.0009291456617575022,
      "clip_ratio/low_mean": 0.0010373450859333389,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001966490781342145,
      "epoch": 1.3149606299212597,
      "grad_norm": 0.2075706124305725,
      "learning_rate": 1e-06,
      "loss": -0.0672,
      "step": 563
    },
    {
      "clip_ratio/high_max": 0.0024174200952984393,
      "clip_ratio/high_mean": 0.0009365485079797509,
      "clip_ratio/low_mean": 0.0011528040340635926,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002089352550683543,
      "epoch": 1.3172936716243804,
      "grad_norm": 0.173910990357399,
      "learning_rate": 1e-06,
      "loss": -0.0672,
      "step": 564
    },
    {
      "clip_ratio/high_max": 0.002502191076928284,
      "clip_ratio/high_mean": 0.0008761862118262798,
      "clip_ratio/low_mean": 0.0008554907271900447,
      "clip_ratio/low_min": 1.070755570253823e-05,
      "clip_ratio/region_mean": 0.0017316769881290384,
      "completions/clipped_ratio": 0.1774553571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4043.0,
      "completions/mean_length": 1279.318115234375,
      "completions/mean_terminated_length": 671.6485595703125,
      "completions/min_length": 181.0,
      "completions/min_terminated_length": 181.0,
      "epoch": 1.3196267133275008,
      "grad_norm": 0.2694673240184784,
      "learning_rate": 1e-06,
      "loss": -0.0587,
      "num_tokens": 85333594.0,
      "reward": 0.4955357313156128,
      "reward_std": 0.16450457274913788,
      "rewards/verify_math_reward/mean": 0.4955357015132904,
      "rewards/verify_math_reward/std": 0.500259280204773,
      "step": 565
    },
    {
      "clip_ratio/high_max": 0.0036288581541157328,
      "clip_ratio/high_mean": 0.0013075614733679686,
      "clip_ratio/low_mean": 0.001007825490887626,
      "clip_ratio/low_min": 2.141511140507646e-05,
      "clip_ratio/region_mean": 0.0023153869624366052,
      "epoch": 1.3219597550306212,
      "grad_norm": 0.27903032302856445,
      "learning_rate": 1e-06,
      "loss": -0.0589,
      "step": 566
    },
    {
      "clip_ratio/high_max": 0.0031396789054269902,
      "clip_ratio/high_mean": 0.0010936912440229207,
      "clip_ratio/low_mean": 0.0011650597607513191,
      "clip_ratio/low_min": 1.2437811165000312e-05,
      "clip_ratio/region_mean": 0.002258751028421102,
      "epoch": 1.3242927967337417,
      "grad_norm": 0.22430910170078278,
      "learning_rate": 1e-06,
      "loss": -0.0592,
      "step": 567
    },
    {
      "clip_ratio/high_max": 0.0029859956121072173,
      "clip_ratio/high_mean": 0.0010755592793429969,
      "clip_ratio/low_mean": 0.0014766222884645686,
      "clip_ratio/low_min": 3.731343167601153e-05,
      "clip_ratio/region_mean": 0.0025521815186948515,
      "epoch": 1.326625838436862,
      "grad_norm": 0.20644904673099518,
      "learning_rate": 1e-06,
      "loss": -0.0593,
      "step": 568
    },
    {
      "clip_ratio/high_max": 0.0024077963025774807,
      "clip_ratio/high_mean": 0.0009524892739136703,
      "clip_ratio/low_mean": 0.0007456240709871054,
      "clip_ratio/low_min": 2.887827122322051e-05,
      "clip_ratio/region_mean": 0.0016981133230729029,
      "completions/clipped_ratio": 0.1741071428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4002.0,
      "completions/mean_length": 1243.03466796875,
      "completions/mean_terminated_length": 641.5986938476562,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 1.3289588801399825,
      "grad_norm": 0.3015615940093994,
      "learning_rate": 1e-06,
      "loss": -0.0808,
      "num_tokens": 85899945.0,
      "reward": 0.5011160969734192,
      "reward_std": 0.19895894825458527,
      "rewards/verify_math_reward/mean": 0.5011160969734192,
      "rewards/verify_math_reward/std": 0.5002779960632324,
      "step": 569
    },
    {
      "clip_ratio/high_max": 0.003500665996398311,
      "clip_ratio/high_mean": 0.0012464672872738447,
      "clip_ratio/low_mean": 0.0010257575977448141,
      "clip_ratio/low_min": 0.00012356816660030745,
      "clip_ratio/region_mean": 0.002272224875923712,
      "epoch": 1.331291921843103,
      "grad_norm": 0.24528682231903076,
      "learning_rate": 1e-06,
      "loss": -0.0811,
      "step": 570
    },
    {
      "clip_ratio/high_max": 0.0038307572322082706,
      "clip_ratio/high_mean": 0.0013789985459879972,
      "clip_ratio/low_mean": 0.0012182924365333747,
      "clip_ratio/low_min": 0.00011780763452406973,
      "clip_ratio/region_mean": 0.002597291022539139,
      "epoch": 1.3336249635462234,
      "grad_norm": 0.208794504404068,
      "learning_rate": 1e-06,
      "loss": -0.0814,
      "step": 571
    },
    {
      "clip_ratio/high_max": 0.0031483758939430118,
      "clip_ratio/high_mean": 0.0011578826415643562,
      "clip_ratio/low_mean": 0.0014242229008232243,
      "clip_ratio/low_min": 0.00011548066322575323,
      "clip_ratio/region_mean": 0.002582105516921729,
      "epoch": 1.3359580052493438,
      "grad_norm": 0.2046995311975479,
      "learning_rate": 1e-06,
      "loss": -0.0814,
      "step": 572
    },
    {
      "clip_ratio/high_max": 0.0023555047664558515,
      "clip_ratio/high_mean": 0.0008531126677553402,
      "clip_ratio/low_mean": 0.0005902969119233603,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014434096083277836,
      "completions/clipped_ratio": 0.1752232142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3835.0,
      "completions/mean_length": 1285.040283203125,
      "completions/mean_terminated_length": 687.8538208007812,
      "completions/min_length": 187.0,
      "completions/min_terminated_length": 187.0,
      "epoch": 1.3382910469524643,
      "grad_norm": 0.25017184019088745,
      "learning_rate": 1e-06,
      "loss": -0.0753,
      "num_tokens": 86504485.0,
      "reward": 0.5256696939468384,
      "reward_std": 0.17757472395896912,
      "rewards/verify_math_reward/mean": 0.5256696343421936,
      "rewards/verify_math_reward/std": 0.4996195137500763,
      "step": 573
    },
    {
      "clip_ratio/high_max": 0.0028082246353733353,
      "clip_ratio/high_mean": 0.0010514356654311996,
      "clip_ratio/low_mean": 0.0007836565755496849,
      "clip_ratio/low_min": 4.149826600041706e-05,
      "clip_ratio/region_mean": 0.0018350922473473474,
      "epoch": 1.3406240886555847,
      "grad_norm": 0.22260355949401855,
      "learning_rate": 1e-06,
      "loss": -0.0755,
      "step": 574
    },
    {
      "clip_ratio/high_max": 0.0034009277733275667,
      "clip_ratio/high_mean": 0.0011136180037283339,
      "clip_ratio/low_mean": 0.001009279998470447,
      "clip_ratio/low_min": 5.6302465964108706e-05,
      "clip_ratio/region_mean": 0.0021228979603620246,
      "epoch": 1.3429571303587051,
      "grad_norm": 0.2014068365097046,
      "learning_rate": 1e-06,
      "loss": -0.0758,
      "step": 575
    },
    {
      "clip_ratio/high_max": 0.0031681032451160718,
      "clip_ratio/high_mean": 0.0011202151526958914,
      "clip_ratio/low_mean": 0.0011742266851797467,
      "clip_ratio/low_min": 6.622609544137958e-05,
      "clip_ratio/region_mean": 0.0022944418524275534,
      "epoch": 1.3452901720618256,
      "grad_norm": 0.2062731385231018,
      "learning_rate": 1e-06,
      "loss": -0.0758,
      "step": 576
    },
    {
      "clip_ratio/high_max": 0.00190445312182419,
      "clip_ratio/high_mean": 0.0007755427886877442,
      "clip_ratio/low_mean": 0.0005309752777975518,
      "clip_ratio/low_min": 1.972231075342279e-05,
      "clip_ratio/region_mean": 0.0013065180792182218,
      "completions/clipped_ratio": 0.1852678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3200.0,
      "completions/mean_length": 1232.1640625,
      "completions/mean_terminated_length": 580.9356079101562,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 1.347623213764946,
      "grad_norm": 0.30050161480903625,
      "learning_rate": 1e-06,
      "loss": -0.0672,
      "num_tokens": 87015224.0,
      "reward": 0.559151828289032,
      "reward_std": 0.15045617520809174,
      "rewards/verify_math_reward/mean": 0.5591517686843872,
      "rewards/verify_math_reward/std": 0.496766060590744,
      "step": 577
    },
    {
      "clip_ratio/high_max": 0.002414862574369181,
      "clip_ratio/high_mean": 0.0009709475780255161,
      "clip_ratio/low_mean": 0.0007054359193716664,
      "clip_ratio/low_min": 1.972231075342279e-05,
      "clip_ratio/region_mean": 0.001676383486483246,
      "epoch": 1.3499562554680664,
      "grad_norm": 0.23338362574577332,
      "learning_rate": 1e-06,
      "loss": -0.0674,
      "step": 578
    },
    {
      "clip_ratio/high_max": 0.0023903432229417376,
      "clip_ratio/high_mean": 0.0009181600580632221,
      "clip_ratio/low_mean": 0.0009727856850076932,
      "clip_ratio/low_min": 3.944462150684558e-05,
      "clip_ratio/region_mean": 0.0018909457576228306,
      "epoch": 1.352289297171187,
      "grad_norm": 0.21429312229156494,
      "learning_rate": 1e-06,
      "loss": -0.0675,
      "step": 579
    },
    {
      "clip_ratio/high_max": 0.0023471144886570983,
      "clip_ratio/high_mean": 0.0009947415128408466,
      "clip_ratio/low_mean": 0.0010963070253637852,
      "clip_ratio/low_min": 3.944462150684558e-05,
      "clip_ratio/region_mean": 0.0020910485764034092,
      "epoch": 1.3546223388743073,
      "grad_norm": 0.19566044211387634,
      "learning_rate": 1e-06,
      "loss": -0.0676,
      "step": 580
    },
    {
      "clip_ratio/high_max": 0.0024408697499893606,
      "clip_ratio/high_mean": 0.000832719366371748,
      "clip_ratio/low_mean": 0.0005201030976422771,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013528224553738255,
      "completions/clipped_ratio": 0.2087053571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4034.0,
      "completions/mean_length": 1369.0457763671875,
      "completions/mean_terminated_length": 649.8067626953125,
      "completions/min_length": 143.0,
      "completions/min_terminated_length": 143.0,
      "epoch": 1.356955380577428,
      "grad_norm": 0.2526213526725769,
      "learning_rate": 1e-06,
      "loss": -0.0483,
      "num_tokens": 87579321.0,
      "reward": 0.535714328289032,
      "reward_std": 0.1345217078924179,
      "rewards/verify_math_reward/mean": 0.5357142686843872,
      "rewards/verify_math_reward/std": 0.4990014135837555,
      "step": 581
    },
    {
      "clip_ratio/high_max": 0.0030322665261337534,
      "clip_ratio/high_mean": 0.0010206833885604283,
      "clip_ratio/low_mean": 0.0007519806695199804,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017726640435284935,
      "epoch": 1.3592884222805481,
      "grad_norm": 0.25519487261772156,
      "learning_rate": 1e-06,
      "loss": -0.0485,
      "step": 582
    },
    {
      "clip_ratio/high_max": 0.003515000549668912,
      "clip_ratio/high_mean": 0.0011885368476214353,
      "clip_ratio/low_mean": 0.0009024976052387501,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002091034493787447,
      "epoch": 1.3616214639836688,
      "grad_norm": 0.4738200902938843,
      "learning_rate": 1e-06,
      "loss": -0.0487,
      "step": 583
    },
    {
      "clip_ratio/high_max": 0.003024574962182669,
      "clip_ratio/high_mean": 0.0009989466548177006,
      "clip_ratio/low_mean": 0.0010811063275468769,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002080053040117491,
      "epoch": 1.3639545056867892,
      "grad_norm": 0.2041773796081543,
      "learning_rate": 1e-06,
      "loss": -0.0488,
      "step": 584
    },
    {
      "clip_ratio/high_max": 0.002207355821155943,
      "clip_ratio/high_mean": 0.0008487261347909225,
      "clip_ratio/low_mean": 0.00045651231687315885,
      "clip_ratio/low_min": 1.7443482647649944e-05,
      "clip_ratio/region_mean": 0.0013052384383627214,
      "completions/clipped_ratio": 0.2042410714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4059.0,
      "completions/mean_length": 1383.12841796875,
      "completions/mean_terminated_length": 686.8372802734375,
      "completions/min_length": 126.0,
      "completions/min_terminated_length": 126.0,
      "epoch": 1.3662875473899097,
      "grad_norm": 0.28928208351135254,
      "learning_rate": 1e-06,
      "loss": -0.0712,
      "num_tokens": 88156644.0,
      "reward": 0.5055803656578064,
      "reward_std": 0.16517673432826996,
      "rewards/verify_math_reward/mean": 0.5055803656578064,
      "rewards/verify_math_reward/std": 0.5002480745315552,
      "step": 585
    },
    {
      "clip_ratio/high_max": 0.0027591592152020894,
      "clip_ratio/high_mean": 0.0010814143024617806,
      "clip_ratio/low_mean": 0.0007400381391562405,
      "clip_ratio/low_min": 1.3213531019573566e-05,
      "clip_ratio/region_mean": 0.0018214524934592191,
      "epoch": 1.36862058909303,
      "grad_norm": 0.23145802319049835,
      "learning_rate": 1e-06,
      "loss": -0.0714,
      "step": 586
    },
    {
      "clip_ratio/high_max": 0.002570676511822967,
      "clip_ratio/high_mean": 0.0010192344198003411,
      "clip_ratio/low_mean": 0.0008793568467808655,
      "clip_ratio/low_min": 2.402613972662948e-05,
      "clip_ratio/region_mean": 0.0018985912993230158,
      "epoch": 1.3709536307961505,
      "grad_norm": 0.20198185741901398,
      "learning_rate": 1e-06,
      "loss": -0.0716,
      "step": 587
    },
    {
      "clip_ratio/high_max": 0.002756901980319526,
      "clip_ratio/high_mean": 0.0010476693350938149,
      "clip_ratio/low_mean": 0.0010282289204042172,
      "clip_ratio/low_min": 5.233044794294983e-05,
      "clip_ratio/region_mean": 0.0020758982536790427,
      "epoch": 1.373286672499271,
      "grad_norm": 0.19303227961063385,
      "learning_rate": 1e-06,
      "loss": -0.0717,
      "step": 588
    },
    {
      "clip_ratio/high_max": 0.0030537584316334687,
      "clip_ratio/high_mean": 0.0011682229196594562,
      "clip_ratio/low_mean": 0.0005520252943824744,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001720248234050814,
      "completions/clipped_ratio": 0.21875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3188.0,
      "completions/mean_length": 1431.64404296875,
      "completions/mean_terminated_length": 685.624267578125,
      "completions/min_length": 188.0,
      "completions/min_terminated_length": 188.0,
      "epoch": 1.3756197142023914,
      "grad_norm": 0.299399733543396,
      "learning_rate": 1e-06,
      "loss": -0.1017,
      "num_tokens": 88720029.0,
      "reward": 0.53125,
      "reward_std": 0.19854412972927094,
      "rewards/verify_math_reward/mean": 0.53125,
      "rewards/verify_math_reward/std": 0.4993011951446533,
      "step": 589
    },
    {
      "clip_ratio/high_max": 0.0032559526662225835,
      "clip_ratio/high_mean": 0.0012779856951965485,
      "clip_ratio/low_mean": 0.0008215541784011293,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020995398444938473,
      "epoch": 1.3779527559055118,
      "grad_norm": 0.254035621881485,
      "learning_rate": 1e-06,
      "loss": -0.102,
      "step": 590
    },
    {
      "clip_ratio/high_max": 0.003467716960585676,
      "clip_ratio/high_mean": 0.0013532105476770084,
      "clip_ratio/low_mean": 0.001074977066309657,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002428187603072729,
      "epoch": 1.3802857976086322,
      "grad_norm": 0.22269730269908905,
      "learning_rate": 1e-06,
      "loss": -0.1022,
      "step": 591
    },
    {
      "clip_ratio/high_max": 0.0038947790453676134,
      "clip_ratio/high_mean": 0.0014983424262027256,
      "clip_ratio/low_mean": 0.0012840630570281064,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0027824054850498214,
      "epoch": 1.3826188393117527,
      "grad_norm": 0.22764192521572113,
      "learning_rate": 1e-06,
      "loss": -0.1023,
      "step": 592
    },
    {
      "clip_ratio/high_max": 0.0021997439180267975,
      "clip_ratio/high_mean": 0.0010206915576418396,
      "clip_ratio/low_mean": 0.00040406626385447453,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001424757832864998,
      "completions/clipped_ratio": 0.1573660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3147.0,
      "completions/mean_length": 1177.2935791015625,
      "completions/mean_terminated_length": 632.2106323242188,
      "completions/min_length": 151.0,
      "completions/min_terminated_length": 151.0,
      "epoch": 1.384951881014873,
      "grad_norm": 0.3221670985221863,
      "learning_rate": 1e-06,
      "loss": -0.0642,
      "num_tokens": 89281972.0,
      "reward": 0.6116071939468384,
      "reward_std": 0.163072407245636,
      "rewards/verify_math_reward/mean": 0.6116071343421936,
      "rewards/verify_math_reward/std": 0.48765692114830017,
      "step": 593
    },
    {
      "clip_ratio/high_max": 0.003074386069783941,
      "clip_ratio/high_mean": 0.0013346814266697038,
      "clip_ratio/low_mean": 0.0006285099052547594,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001963191360118799,
      "epoch": 1.3872849227179935,
      "grad_norm": 0.26880836486816406,
      "learning_rate": 1e-06,
      "loss": -0.0646,
      "step": 594
    },
    {
      "clip_ratio/high_max": 0.0026226171030430123,
      "clip_ratio/high_mean": 0.0012232904336997308,
      "clip_ratio/low_mean": 0.0008326363977175788,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002055926750472281,
      "epoch": 1.389617964421114,
      "grad_norm": 0.2298322468996048,
      "learning_rate": 1e-06,
      "loss": -0.0647,
      "step": 595
    },
    {
      "clip_ratio/high_max": 0.0031332553407992236,
      "clip_ratio/high_mean": 0.0012834487861255184,
      "clip_ratio/low_mean": 0.0010289407300660969,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023123895152821206,
      "epoch": 1.3919510061242344,
      "grad_norm": 0.17754845321178436,
      "learning_rate": 1e-06,
      "loss": -0.0648,
      "step": 596
    },
    {
      "clip_ratio/high_max": 0.0015846236638026312,
      "clip_ratio/high_mean": 0.0005263043115064647,
      "clip_ratio/low_mean": 0.0002776564401756332,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0008039607509999769,
      "completions/clipped_ratio": 0.1573660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3086.0,
      "completions/mean_length": 1184.8817138671875,
      "completions/mean_terminated_length": 641.2158813476562,
      "completions/min_length": 160.0,
      "completions/min_terminated_length": 160.0,
      "epoch": 1.3942840478273548,
      "grad_norm": 0.17424309253692627,
      "learning_rate": 1e-06,
      "loss": -0.055,
      "num_tokens": 89862346.0,
      "reward": 0.5926339626312256,
      "reward_std": 0.10697808116674423,
      "rewards/verify_math_reward/mean": 0.5926339030265808,
      "rewards/verify_math_reward/std": 0.49161845445632935,
      "step": 597
    },
    {
      "clip_ratio/high_max": 0.0016562970013183076,
      "clip_ratio/high_mean": 0.0005918571487200097,
      "clip_ratio/low_mean": 0.00033543918425493757,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.000927296341615147,
      "epoch": 1.3966170895304755,
      "grad_norm": 0.15606385469436646,
      "learning_rate": 1e-06,
      "loss": -0.0551,
      "step": 598
    },
    {
      "clip_ratio/high_max": 0.0018177629535784945,
      "clip_ratio/high_mean": 0.0006014393620716874,
      "clip_ratio/low_mean": 0.000457512957950712,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0010589522898953874,
      "epoch": 1.3989501312335957,
      "grad_norm": 0.15585005283355713,
      "learning_rate": 1e-06,
      "loss": -0.0551,
      "step": 599
    },
    {
      "clip_ratio/high_max": 0.001933376139277243,
      "clip_ratio/high_mean": 0.0006321645703337708,
      "clip_ratio/low_mean": 0.0005467807463901408,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0011789453419623896,
      "epoch": 1.4012831729367163,
      "grad_norm": 0.14662864804267883,
      "learning_rate": 1e-06,
      "loss": -0.0552,
      "step": 600
    },
    {
      "clip_ratio/high_max": 0.0021389451321738306,
      "clip_ratio/high_mean": 0.0008527110221621115,
      "clip_ratio/low_mean": 0.0005508000540430658,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001403511079843156,
      "completions/clipped_ratio": 0.1852678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3851.0,
      "completions/mean_length": 1268.9832763671875,
      "completions/mean_terminated_length": 626.12744140625,
      "completions/min_length": 136.0,
      "completions/min_terminated_length": 136.0,
      "epoch": 1.4036162146398368,
      "grad_norm": 0.3296069800853729,
      "learning_rate": 1e-06,
      "loss": -0.0579,
      "num_tokens": 90416619.0,
      "reward": 0.5424107313156128,
      "reward_std": 0.15852880477905273,
      "rewards/verify_math_reward/mean": 0.5424107313156128,
      "rewards/verify_math_reward/std": 0.4984763264656067,
      "step": 601
    },
    {
      "clip_ratio/high_max": 0.0026905527847702615,
      "clip_ratio/high_mean": 0.001033009248203598,
      "clip_ratio/low_mean": 0.0008065595188782027,
      "clip_ratio/low_min": 1.9512955987011082e-05,
      "clip_ratio/region_mean": 0.0018395687729935162,
      "epoch": 1.4059492563429572,
      "grad_norm": 0.30061817169189453,
      "learning_rate": 1e-06,
      "loss": -0.0581,
      "step": 602
    },
    {
      "clip_ratio/high_max": 0.002843020934960805,
      "clip_ratio/high_mean": 0.0009820305567700416,
      "clip_ratio/low_mean": 0.0010391084992988908,
      "clip_ratio/low_min": 5.8049536164617166e-05,
      "clip_ratio/region_mean": 0.002021139021962881,
      "epoch": 1.4082822980460776,
      "grad_norm": 0.37664586305618286,
      "learning_rate": 1e-06,
      "loss": -0.0584,
      "step": 603
    },
    {
      "clip_ratio/high_max": 0.002560520537372213,
      "clip_ratio/high_mean": 0.0009429810361325508,
      "clip_ratio/low_mean": 0.001355129446437786,
      "clip_ratio/low_min": 3.9025911974022165e-05,
      "clip_ratio/region_mean": 0.0022981104848440737,
      "epoch": 1.410615339749198,
      "grad_norm": 0.22192707657814026,
      "learning_rate": 1e-06,
      "loss": -0.0584,
      "step": 604
    },
    {
      "clip_ratio/high_max": 0.0021960718695481773,
      "clip_ratio/high_mean": 0.0009144201030721888,
      "clip_ratio/low_mean": 0.0006570868190465262,
      "clip_ratio/low_min": 3.0317729397211224e-05,
      "clip_ratio/region_mean": 0.001571506931213662,
      "completions/clipped_ratio": 0.1328125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3814.0,
      "completions/mean_length": 1080.6138916015625,
      "completions/mean_terminated_length": 618.7979736328125,
      "completions/min_length": 183.0,
      "completions/min_terminated_length": 183.0,
      "epoch": 1.4129483814523185,
      "grad_norm": 0.32067176699638367,
      "learning_rate": 1e-06,
      "loss": -0.0518,
      "num_tokens": 90981905.0,
      "reward": 0.6473214626312256,
      "reward_std": 0.16586002707481384,
      "rewards/verify_math_reward/mean": 0.6473214030265808,
      "rewards/verify_math_reward/std": 0.47807061672210693,
      "step": 605
    },
    {
      "clip_ratio/high_max": 0.0025493590328551363,
      "clip_ratio/high_mean": 0.0010450735744598205,
      "clip_ratio/low_mean": 0.0009328416072094114,
      "clip_ratio/low_min": 6.464625766966492e-05,
      "clip_ratio/region_mean": 0.0019779151625698432,
      "epoch": 1.415281423155439,
      "grad_norm": 0.23323972523212433,
      "learning_rate": 1e-06,
      "loss": -0.0521,
      "step": 606
    },
    {
      "clip_ratio/high_max": 0.002687867498025298,
      "clip_ratio/high_mean": 0.0010876679752982454,
      "clip_ratio/low_mean": 0.0010807789094542386,
      "clip_ratio/low_min": 4.5476594095816836e-05,
      "clip_ratio/region_mean": 0.002168446888390463,
      "epoch": 1.4176144648585594,
      "grad_norm": 0.20053008198738098,
      "learning_rate": 1e-06,
      "loss": -0.0523,
      "step": 607
    },
    {
      "clip_ratio/high_max": 0.0026580548619676847,
      "clip_ratio/high_mean": 0.001002402326776064,
      "clip_ratio/low_mean": 0.0013172924864193192,
      "clip_ratio/low_min": 0.00010611205652821809,
      "clip_ratio/region_mean": 0.0023196948823169805,
      "epoch": 1.4199475065616798,
      "grad_norm": 0.20511440932750702,
      "learning_rate": 1e-06,
      "loss": -0.0523,
      "step": 608
    },
    {
      "clip_ratio/high_max": 0.0022464193098130636,
      "clip_ratio/high_mean": 0.0008555043514206773,
      "clip_ratio/low_mean": 0.0005546764232349233,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014101807100814767,
      "completions/clipped_ratio": 0.1529017857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3897.0,
      "completions/mean_length": 1150.9732666015625,
      "completions/mean_terminated_length": 619.3939208984375,
      "completions/min_length": 152.0,
      "completions/min_terminated_length": 152.0,
      "epoch": 1.4222805482648002,
      "grad_norm": 0.3018054962158203,
      "learning_rate": 1e-06,
      "loss": -0.0708,
      "num_tokens": 91541249.0,
      "reward": 0.5613839626312256,
      "reward_std": 0.16461403667926788,
      "rewards/verify_math_reward/mean": 0.5613839030265808,
      "rewards/verify_math_reward/std": 0.496494859457016,
      "step": 609
    },
    {
      "clip_ratio/high_max": 0.002809393459756393,
      "clip_ratio/high_mean": 0.001057904966728529,
      "clip_ratio/low_mean": 0.0007229856068988738,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017808905977290124,
      "epoch": 1.4246135899679206,
      "grad_norm": 0.2314707636833191,
      "learning_rate": 1e-06,
      "loss": -0.071,
      "step": 610
    },
    {
      "clip_ratio/high_max": 0.0028084185032639652,
      "clip_ratio/high_mean": 0.0010672877251636237,
      "clip_ratio/low_mean": 0.0008877732725522947,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001955060964974109,
      "epoch": 1.426946631671041,
      "grad_norm": 0.4472660720348358,
      "learning_rate": 1e-06,
      "loss": -0.0711,
      "step": 611
    },
    {
      "clip_ratio/high_max": 0.00276724090508651,
      "clip_ratio/high_mean": 0.0010477610740053933,
      "clip_ratio/low_mean": 0.0010973095450026449,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021450706335599534,
      "epoch": 1.4292796733741615,
      "grad_norm": 0.2138577103614807,
      "learning_rate": 1e-06,
      "loss": -0.0712,
      "step": 612
    },
    {
      "clip_ratio/high_max": 0.002117222800734453,
      "clip_ratio/high_mean": 0.0008917890409065876,
      "clip_ratio/low_mean": 0.0007474229569197632,
      "clip_ratio/low_min": 1.2669775060203392e-05,
      "clip_ratio/region_mean": 0.0016392120087402873,
      "completions/clipped_ratio": 0.1462053571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3247.0,
      "completions/mean_length": 1161.48779296875,
      "completions/mean_terminated_length": 658.9765014648438,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 1.431612715077282,
      "grad_norm": 0.2706722617149353,
      "learning_rate": 1e-06,
      "loss": -0.0491,
      "num_tokens": 92127894.0,
      "reward": 0.5803571939468384,
      "reward_std": 0.1737132966518402,
      "rewards/verify_math_reward/mean": 0.5803571343421936,
      "rewards/verify_math_reward/std": 0.4937761425971985,
      "step": 613
    },
    {
      "clip_ratio/high_max": 0.0027357905491953716,
      "clip_ratio/high_mean": 0.001020476996927755,
      "clip_ratio/low_mean": 0.0009552073897793889,
      "clip_ratio/low_min": 1.2669775060203392e-05,
      "clip_ratio/region_mean": 0.001975684383069165,
      "epoch": 1.4339457567804024,
      "grad_norm": 0.2501928210258484,
      "learning_rate": 1e-06,
      "loss": -0.0492,
      "step": 614
    },
    {
      "clip_ratio/high_max": 0.0029026706470176578,
      "clip_ratio/high_mean": 0.0010749668326752726,
      "clip_ratio/low_mean": 0.0011807225237134844,
      "clip_ratio/low_min": 2.5339550120406784e-05,
      "clip_ratio/region_mean": 0.002255689279991202,
      "epoch": 1.436278798483523,
      "grad_norm": 0.2090653032064438,
      "learning_rate": 1e-06,
      "loss": -0.0495,
      "step": 615
    },
    {
      "clip_ratio/high_max": 0.002555416082032025,
      "clip_ratio/high_mean": 0.0009618200056138448,
      "clip_ratio/low_mean": 0.0014350209457916208,
      "clip_ratio/low_min": 2.5572831873432733e-05,
      "clip_ratio/region_mean": 0.002396840944129508,
      "epoch": 1.4386118401866432,
      "grad_norm": 0.2010267674922943,
      "learning_rate": 1e-06,
      "loss": -0.0495,
      "step": 616
    },
    {
      "clip_ratio/high_max": 0.0022100016140029766,
      "clip_ratio/high_mean": 0.0008261815728474176,
      "clip_ratio/low_mean": 0.00046243821452662814,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012886197873740457,
      "completions/clipped_ratio": 0.1238839285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4061.0,
      "completions/mean_length": 1030.7645263671875,
      "completions/mean_terminated_length": 597.3363037109375,
      "completions/min_length": 125.0,
      "completions/min_terminated_length": 125.0,
      "epoch": 1.4409448818897639,
      "grad_norm": 0.30851060152053833,
      "learning_rate": 1e-06,
      "loss": -0.0658,
      "num_tokens": 92681131.0,
      "reward": 0.6428571939468384,
      "reward_std": 0.1633780300617218,
      "rewards/verify_math_reward/mean": 0.6428571343421936,
      "rewards/verify_math_reward/std": 0.4794250428676605,
      "step": 617
    },
    {
      "clip_ratio/high_max": 0.002810432379192207,
      "clip_ratio/high_mean": 0.0011620204331848072,
      "clip_ratio/low_mean": 0.0006768962748537888,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018389166871202178,
      "epoch": 1.443277923592884,
      "grad_norm": 0.2598080039024353,
      "learning_rate": 1e-06,
      "loss": -0.0661,
      "step": 618
    },
    {
      "clip_ratio/high_max": 0.002908035930886399,
      "clip_ratio/high_mean": 0.0011905909195775166,
      "clip_ratio/low_mean": 0.0008262346800620435,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020168255796306767,
      "epoch": 1.4456109652960047,
      "grad_norm": 0.19690977036952972,
      "learning_rate": 1e-06,
      "loss": -0.0663,
      "step": 619
    },
    {
      "clip_ratio/high_max": 0.0025863999107968993,
      "clip_ratio/high_mean": 0.0010566708187980112,
      "clip_ratio/low_mean": 0.0009493513525740127,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002006022143177688,
      "epoch": 1.4479440069991252,
      "grad_norm": 0.23134976625442505,
      "learning_rate": 1e-06,
      "loss": -0.0663,
      "step": 620
    },
    {
      "clip_ratio/high_max": 0.002198130972828949,
      "clip_ratio/high_mean": 0.0007040815144137014,
      "clip_ratio/low_mean": 0.000685114953739685,
      "clip_ratio/low_min": 1.1968594662903342e-05,
      "clip_ratio/region_mean": 0.0013891964590584394,
      "completions/clipped_ratio": 0.1350446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3505.0,
      "completions/mean_length": 1098.3035888671875,
      "completions/mean_terminated_length": 630.276123046875,
      "completions/min_length": 190.0,
      "completions/min_terminated_length": 190.0,
      "epoch": 1.4502770487022456,
      "grad_norm": 0.3217826187610626,
      "learning_rate": 1e-06,
      "loss": -0.0273,
      "num_tokens": 93258899.0,
      "reward": 0.5680803656578064,
      "reward_std": 0.1669451892375946,
      "rewards/verify_math_reward/mean": 0.5680803656578064,
      "rewards/verify_math_reward/std": 0.4956200420856476,
      "step": 621
    },
    {
      "clip_ratio/high_max": 0.002298268344020471,
      "clip_ratio/high_mean": 0.0008164960509020602,
      "clip_ratio/low_mean": 0.0009055617429112317,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017220578229171224,
      "epoch": 1.452610090405366,
      "grad_norm": 0.21622058749198914,
      "learning_rate": 1e-06,
      "loss": -0.0276,
      "step": 622
    },
    {
      "clip_ratio/high_max": 0.0024621934062452056,
      "clip_ratio/high_mean": 0.0007997809279913781,
      "clip_ratio/low_mean": 0.0011010635334969265,
      "clip_ratio/low_min": 1.1968594662903342e-05,
      "clip_ratio/region_mean": 0.001900844436022453,
      "epoch": 1.4549431321084865,
      "grad_norm": 0.20084120333194733,
      "learning_rate": 1e-06,
      "loss": -0.0278,
      "step": 623
    },
    {
      "clip_ratio/high_max": 0.0023159858101280406,
      "clip_ratio/high_mean": 0.0008225301498896442,
      "clip_ratio/low_mean": 0.0012968940864084288,
      "clip_ratio/low_min": 1.1968594662903342e-05,
      "clip_ratio/region_mean": 0.0021194242290221155,
      "epoch": 1.457276173811607,
      "grad_norm": 0.1809324324131012,
      "learning_rate": 1e-06,
      "loss": -0.0279,
      "step": 624
    },
    {
      "clip_ratio/high_max": 0.0020966012234566733,
      "clip_ratio/high_mean": 0.0008217469148803502,
      "clip_ratio/low_mean": 0.0005623789793389733,
      "clip_ratio/low_min": 1.354573032585904e-05,
      "clip_ratio/region_mean": 0.0013841258842148818,
      "completions/clipped_ratio": 0.1685267857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3344.0,
      "completions/mean_length": 1279.501220703125,
      "completions/mean_terminated_length": 708.6403198242188,
      "completions/min_length": 141.0,
      "completions/min_terminated_length": 141.0,
      "epoch": 1.4596092155147273,
      "grad_norm": 0.2877887785434723,
      "learning_rate": 1e-06,
      "loss": -0.0728,
      "num_tokens": 93878340.0,
      "reward": 0.5446428656578064,
      "reward_std": 0.16566601395606995,
      "rewards/verify_math_reward/mean": 0.5446428656578064,
      "rewards/verify_math_reward/std": 0.4982811510562897,
      "step": 625
    },
    {
      "clip_ratio/high_max": 0.00275848429009784,
      "clip_ratio/high_mean": 0.0010338401443732437,
      "clip_ratio/low_mean": 0.0008302076239488088,
      "clip_ratio/low_min": 3.299023592262529e-05,
      "clip_ratio/region_mean": 0.0018640477137523703,
      "epoch": 1.4619422572178478,
      "grad_norm": 0.20693060755729675,
      "learning_rate": 1e-06,
      "loss": -0.0731,
      "step": 626
    },
    {
      "clip_ratio/high_max": 0.0028259688333491795,
      "clip_ratio/high_mean": 0.001058114408806432,
      "clip_ratio/low_mean": 0.0009656888905738015,
      "clip_ratio/low_min": 5.362869706004858e-05,
      "clip_ratio/region_mean": 0.0020238032666384242,
      "epoch": 1.4642752989209682,
      "grad_norm": 0.1813107430934906,
      "learning_rate": 1e-06,
      "loss": -0.0733,
      "step": 627
    },
    {
      "clip_ratio/high_max": 0.0028315739036770537,
      "clip_ratio/high_mean": 0.0009843532243394293,
      "clip_ratio/low_mean": 0.0011155190331919584,
      "clip_ratio/low_min": 9.688676800578833e-05,
      "clip_ratio/region_mean": 0.0020998722466174513,
      "epoch": 1.4666083406240886,
      "grad_norm": 0.17822666466236115,
      "learning_rate": 1e-06,
      "loss": -0.0733,
      "step": 628
    },
    {
      "clip_ratio/high_max": 0.0019807805183518212,
      "clip_ratio/high_mean": 0.0008326768383994931,
      "clip_ratio/low_mean": 0.00034694392343226355,
      "clip_ratio/low_min": 1.3994625987834297e-05,
      "clip_ratio/region_mean": 0.001179620736365905,
      "completions/clipped_ratio": 0.1729910714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3853.0,
      "completions/mean_length": 1252.6484375,
      "completions/mean_terminated_length": 657.88525390625,
      "completions/min_length": 186.0,
      "completions/min_terminated_length": 186.0,
      "epoch": 1.468941382327209,
      "grad_norm": 0.3137721121311188,
      "learning_rate": 1e-06,
      "loss": -0.0701,
      "num_tokens": 94456689.0,
      "reward": 0.5290178656578064,
      "reward_std": 0.15665017068386078,
      "rewards/verify_math_reward/mean": 0.5290178656578064,
      "rewards/verify_math_reward/std": 0.49943602085113525,
      "step": 629
    },
    {
      "clip_ratio/high_max": 0.00227245543646859,
      "clip_ratio/high_mean": 0.0009395831893925788,
      "clip_ratio/low_mean": 0.0005534355395866442,
      "clip_ratio/low_min": 2.0550196495605633e-05,
      "clip_ratio/region_mean": 0.0014930187171557918,
      "epoch": 1.4712744240303295,
      "grad_norm": 0.29060402512550354,
      "learning_rate": 1e-06,
      "loss": -0.0703,
      "step": 630
    },
    {
      "clip_ratio/high_max": 0.00282437232817756,
      "clip_ratio/high_mean": 0.0010493311092432123,
      "clip_ratio/low_mean": 0.0006986428397794953,
      "clip_ratio/low_min": 6.8500658017001115e-06,
      "clip_ratio/region_mean": 0.0017479739581176545,
      "epoch": 1.47360746573345,
      "grad_norm": 0.18267522752285004,
      "learning_rate": 1e-06,
      "loss": -0.0705,
      "step": 631
    },
    {
      "clip_ratio/high_max": 0.0026695308988564648,
      "clip_ratio/high_mean": 0.0010461520869284868,
      "clip_ratio/low_mean": 0.0008833047941152472,
      "clip_ratio/low_min": 4.795045970240608e-05,
      "clip_ratio/region_mean": 0.001929456844663946,
      "epoch": 1.4759405074365703,
      "grad_norm": 0.24536316096782684,
      "learning_rate": 1e-06,
      "loss": -0.0705,
      "step": 632
    },
    {
      "clip_ratio/high_max": 0.0018891414983954746,
      "clip_ratio/high_mean": 0.0008695932247064775,
      "clip_ratio/low_mean": 0.0004981261481589172,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013677193237526808,
      "completions/clipped_ratio": 0.1350446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3659.0,
      "completions/mean_length": 1132.751220703125,
      "completions/mean_terminated_length": 670.1019897460938,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 1.4782735491396908,
      "grad_norm": 0.3137512803077698,
      "learning_rate": 1e-06,
      "loss": -0.0655,
      "num_tokens": 95054562.0,
      "reward": 0.5881696939468384,
      "reward_std": 0.17784713208675385,
      "rewards/verify_math_reward/mean": 0.5881696343421936,
      "rewards/verify_math_reward/std": 0.4924395978450775,
      "step": 633
    },
    {
      "clip_ratio/high_max": 0.0026375512752565555,
      "clip_ratio/high_mean": 0.0011333199290675111,
      "clip_ratio/low_mean": 0.0007825907032383839,
      "clip_ratio/low_min": 2.1226014723652042e-05,
      "clip_ratio/region_mean": 0.0019159107105224393,
      "epoch": 1.4806065908428114,
      "grad_norm": 0.23053784668445587,
      "learning_rate": 1e-06,
      "loss": -0.0658,
      "step": 634
    },
    {
      "clip_ratio/high_max": 0.002540815963584464,
      "clip_ratio/high_mean": 0.0011436544646130642,
      "clip_ratio/low_mean": 0.0009427386703464435,
      "clip_ratio/low_min": 1.0613007361826021e-05,
      "clip_ratio/region_mean": 0.0020863931204075925,
      "epoch": 1.4829396325459316,
      "grad_norm": 0.2093677967786789,
      "learning_rate": 1e-06,
      "loss": -0.0659,
      "step": 635
    },
    {
      "clip_ratio/high_max": 0.002613451928482391,
      "clip_ratio/high_mean": 0.0011095456502516754,
      "clip_ratio/low_mean": 0.0011048078104067827,
      "clip_ratio/low_min": 4.2452029447304085e-05,
      "clip_ratio/region_mean": 0.0022143534224596806,
      "epoch": 1.4852726742490523,
      "grad_norm": 0.1752631515264511,
      "learning_rate": 1e-06,
      "loss": -0.0661,
      "step": 636
    },
    {
      "clip_ratio/high_max": 0.0017787703473004512,
      "clip_ratio/high_mean": 0.0007677474686715868,
      "clip_ratio/low_mean": 0.000416842095546599,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001184589524200419,
      "completions/clipped_ratio": 0.1495535714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3903.0,
      "completions/mean_length": 1191.4888916015625,
      "completions/mean_terminated_length": 680.7218017578125,
      "completions/min_length": 122.0,
      "completions/min_terminated_length": 122.0,
      "epoch": 1.4876057159521727,
      "grad_norm": 0.20887115597724915,
      "learning_rate": 1e-06,
      "loss": -0.0506,
      "num_tokens": 95663376.0,
      "reward": 0.5848214626312256,
      "reward_std": 0.15582671761512756,
      "rewards/verify_math_reward/mean": 0.5848214030265808,
      "rewards/verify_math_reward/std": 0.49302801489830017,
      "step": 637
    },
    {
      "clip_ratio/high_max": 0.0020016869930259418,
      "clip_ratio/high_mean": 0.0008568723087591934,
      "clip_ratio/low_mean": 0.0005421767809821176,
      "clip_ratio/low_min": 1.433815123164095e-05,
      "clip_ratio/region_mean": 0.0013990490406285971,
      "epoch": 1.4899387576552932,
      "grad_norm": 0.24404728412628174,
      "learning_rate": 1e-06,
      "loss": -0.0507,
      "step": 638
    },
    {
      "clip_ratio/high_max": 0.0021760556192020886,
      "clip_ratio/high_mean": 0.0009449577919440344,
      "clip_ratio/low_mean": 0.0007495746449421858,
      "clip_ratio/low_min": 4.012785393570084e-05,
      "clip_ratio/region_mean": 0.0016945324241532944,
      "epoch": 1.4922717993584136,
      "grad_norm": 0.190024733543396,
      "learning_rate": 1e-06,
      "loss": -0.0509,
      "step": 639
    },
    {
      "clip_ratio/high_max": 0.0021634527256537694,
      "clip_ratio/high_mean": 0.0009296696362071089,
      "clip_ratio/low_mean": 0.0008504205998178804,
      "clip_ratio/low_min": 3.849515451292973e-05,
      "clip_ratio/region_mean": 0.0017800902423914522,
      "epoch": 1.494604841061534,
      "grad_norm": 0.17286677658557892,
      "learning_rate": 1e-06,
      "loss": -0.0509,
      "step": 640
    },
    {
      "clip_ratio/high_max": 0.0019069292393396609,
      "clip_ratio/high_mean": 0.000733642582417815,
      "clip_ratio/low_mean": 0.0004784600541825057,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012121026484237518,
      "completions/clipped_ratio": 0.1316964285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2620.0,
      "completions/mean_length": 1143.93310546875,
      "completions/mean_terminated_length": 696.1902465820312,
      "completions/min_length": 131.0,
      "completions/min_terminated_length": 131.0,
      "epoch": 1.4969378827646544,
      "grad_norm": 0.2399642914533615,
      "learning_rate": 1e-06,
      "loss": -0.0393,
      "num_tokens": 96287404.0,
      "reward": 0.5725446939468384,
      "reward_std": 0.15018516778945923,
      "rewards/verify_math_reward/mean": 0.5725446343421936,
      "rewards/verify_math_reward/std": 0.49498558044433594,
      "step": 641
    },
    {
      "clip_ratio/high_max": 0.002361362647206988,
      "clip_ratio/high_mean": 0.0009156956712104147,
      "clip_ratio/low_mean": 0.0007014144121058052,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016171101087820716,
      "epoch": 1.4992709244677749,
      "grad_norm": 0.2206823080778122,
      "learning_rate": 1e-06,
      "loss": -0.0395,
      "step": 642
    },
    {
      "clip_ratio/high_max": 0.0022667571320198476,
      "clip_ratio/high_mean": 0.0009398542388225906,
      "clip_ratio/low_mean": 0.0008125084077619249,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001752362644765526,
      "epoch": 1.5016039661708953,
      "grad_norm": 0.2108108103275299,
      "learning_rate": 1e-06,
      "loss": -0.0396,
      "step": 643
    },
    {
      "clip_ratio/high_max": 0.002244658535346389,
      "clip_ratio/high_mean": 0.0008674152304593008,
      "clip_ratio/low_mean": 0.0009937553195413784,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018611705527291633,
      "epoch": 1.5039370078740157,
      "grad_norm": 0.20998287200927734,
      "learning_rate": 1e-06,
      "loss": -0.0397,
      "step": 644
    },
    {
      "clip_ratio/high_max": 0.002750497020315379,
      "clip_ratio/high_mean": 0.001084525643818779,
      "clip_ratio/low_mean": 0.000655712927255081,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017402385419700295,
      "completions/clipped_ratio": 0.1897321428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3668.0,
      "completions/mean_length": 1344.53125,
      "completions/mean_terminated_length": 700.2479248046875,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 1.5062700495771362,
      "grad_norm": 0.2975298762321472,
      "learning_rate": 1e-06,
      "loss": -0.0639,
      "num_tokens": 96891224.0,
      "reward": 0.4933035969734192,
      "reward_std": 0.2048664689064026,
      "rewards/verify_math_reward/mean": 0.4933035671710968,
      "rewards/verify_math_reward/std": 0.5002344250679016,
      "step": 645
    },
    {
      "clip_ratio/high_max": 0.003178340135491453,
      "clip_ratio/high_mean": 0.0013070478162262589,
      "clip_ratio/low_mean": 0.0008972011510195443,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002204249001806602,
      "epoch": 1.5086030912802566,
      "grad_norm": 0.2677991986274719,
      "learning_rate": 1e-06,
      "loss": -0.0642,
      "step": 646
    },
    {
      "clip_ratio/high_max": 0.003427130046475213,
      "clip_ratio/high_mean": 0.0014177916236803867,
      "clip_ratio/low_mean": 0.0010971521915053017,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0025149437788059004,
      "epoch": 1.510936132983377,
      "grad_norm": 0.21310929954051971,
      "learning_rate": 1e-06,
      "loss": -0.0644,
      "step": 647
    },
    {
      "clip_ratio/high_max": 0.0029649059470102657,
      "clip_ratio/high_mean": 0.001309787250647787,
      "clip_ratio/low_mean": 0.0012619099179573823,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0025716971649671905,
      "epoch": 1.5132691746864975,
      "grad_norm": 0.2274923026561737,
      "learning_rate": 1e-06,
      "loss": -0.0645,
      "step": 648
    },
    {
      "clip_ratio/high_max": 0.002781429029710125,
      "clip_ratio/high_mean": 0.001106526000512531,
      "clip_ratio/low_mean": 0.0006171891363919713,
      "clip_ratio/low_min": 3.19901673719869e-05,
      "clip_ratio/region_mean": 0.00172371514418046,
      "completions/clipped_ratio": 0.1953125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3042.0,
      "completions/mean_length": 1327.8929443359375,
      "completions/mean_terminated_length": 656.022216796875,
      "completions/min_length": 178.0,
      "completions/min_terminated_length": 178.0,
      "epoch": 1.5156022163896181,
      "grad_norm": 0.3981626033782959,
      "learning_rate": 1e-06,
      "loss": -0.0705,
      "num_tokens": 97465832.0,
      "reward": 0.4977678656578064,
      "reward_std": 0.19670122861862183,
      "rewards/verify_math_reward/mean": 0.4977678656578064,
      "rewards/verify_math_reward/std": 0.5002742409706116,
      "step": 649
    },
    {
      "clip_ratio/high_max": 0.0034815168255590834,
      "clip_ratio/high_mean": 0.0014658952386525925,
      "clip_ratio/low_mean": 0.001077957467714441,
      "clip_ratio/low_min": 3.458586161286803e-05,
      "clip_ratio/region_mean": 0.002543852722737938,
      "epoch": 1.5179352580927383,
      "grad_norm": 0.34394311904907227,
      "learning_rate": 1e-06,
      "loss": -0.0708,
      "step": 650
    },
    {
      "clip_ratio/high_max": 0.00356647145235911,
      "clip_ratio/high_mean": 0.0014375105420185719,
      "clip_ratio/low_mean": 0.0013765477087872569,
      "clip_ratio/low_min": 2.7237243102717912e-05,
      "clip_ratio/region_mean": 0.0028140582362539135,
      "epoch": 1.520268299795859,
      "grad_norm": 0.34004664421081543,
      "learning_rate": 1e-06,
      "loss": -0.0711,
      "step": 651
    },
    {
      "clip_ratio/high_max": 0.003366349876159802,
      "clip_ratio/high_mean": 0.0014541119999194052,
      "clip_ratio/low_mean": 0.0016202135921048466,
      "clip_ratio/low_min": 3.19901673719869e-05,
      "clip_ratio/region_mean": 0.0030743255192646757,
      "epoch": 1.5226013414989792,
      "grad_norm": 0.2796070873737335,
      "learning_rate": 1e-06,
      "loss": -0.0712,
      "step": 652
    },
    {
      "clip_ratio/high_max": 0.0021140613534953445,
      "clip_ratio/high_mean": 0.0007855239236960188,
      "clip_ratio/low_mean": 0.0005137370571901556,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012992609772481956,
      "completions/clipped_ratio": 0.1640625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2928.0,
      "completions/mean_length": 1189.734375,
      "completions/mean_terminated_length": 619.3458251953125,
      "completions/min_length": 170.0,
      "completions/min_terminated_length": 170.0,
      "epoch": 1.5249343832020998,
      "grad_norm": 0.25421372056007385,
      "learning_rate": 1e-06,
      "loss": -0.069,
      "num_tokens": 98023210.0,
      "reward": 0.5267857313156128,
      "reward_std": 0.15134702622890472,
      "rewards/verify_math_reward/mean": 0.5267857313156128,
      "rewards/verify_math_reward/std": 0.4995608627796173,
      "step": 653
    },
    {
      "clip_ratio/high_max": 0.002797064465994481,
      "clip_ratio/high_mean": 0.0010685569450288313,
      "clip_ratio/low_mean": 0.0007312407378776697,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001799797682906501,
      "epoch": 1.52726742490522,
      "grad_norm": 0.2175801396369934,
      "learning_rate": 1e-06,
      "loss": -0.0692,
      "step": 654
    },
    {
      "clip_ratio/high_max": 0.0028732655955536757,
      "clip_ratio/high_mean": 0.001066129141690908,
      "clip_ratio/low_mean": 0.0008831094664856209,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019492386272759177,
      "epoch": 1.5296004666083407,
      "grad_norm": 0.20748473703861237,
      "learning_rate": 1e-06,
      "loss": -0.0694,
      "step": 655
    },
    {
      "clip_ratio/high_max": 0.0029239171999506652,
      "clip_ratio/high_mean": 0.0010204350219282787,
      "clip_ratio/low_mean": 0.0011148508874612162,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021352858893806115,
      "epoch": 1.531933508311461,
      "grad_norm": 0.18779869377613068,
      "learning_rate": 1e-06,
      "loss": -0.0695,
      "step": 656
    },
    {
      "clip_ratio/high_max": 0.00239759481701185,
      "clip_ratio/high_mean": 0.0010145321357413195,
      "clip_ratio/low_mean": 0.0006554852961926372,
      "clip_ratio/low_min": 1.1144793461426161e-05,
      "clip_ratio/region_mean": 0.0016700173982826527,
      "completions/clipped_ratio": 0.1819196428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3998.0,
      "completions/mean_length": 1311.765625,
      "completions/mean_terminated_length": 692.6248168945312,
      "completions/min_length": 133.0,
      "completions/min_terminated_length": 133.0,
      "epoch": 1.5342665500145816,
      "grad_norm": 0.283419668674469,
      "learning_rate": 1e-06,
      "loss": -0.1052,
      "num_tokens": 98622792.0,
      "reward": 0.5613839626312256,
      "reward_std": 0.19895784556865692,
      "rewards/verify_math_reward/mean": 0.5613839030265808,
      "rewards/verify_math_reward/std": 0.496494859457016,
      "step": 657
    },
    {
      "clip_ratio/high_max": 0.00306125360657461,
      "clip_ratio/high_mean": 0.0013394121815508697,
      "clip_ratio/low_mean": 0.0008616015857114689,
      "clip_ratio/low_min": 2.059986763924826e-05,
      "clip_ratio/region_mean": 0.00220101373270154,
      "epoch": 1.536599591717702,
      "grad_norm": 0.27444544434547424,
      "learning_rate": 1e-06,
      "loss": -0.1053,
      "step": 658
    },
    {
      "clip_ratio/high_max": 0.0033300267459708266,
      "clip_ratio/high_mean": 0.0012730805719911586,
      "clip_ratio/low_mean": 0.000994399935734691,
      "clip_ratio/low_min": 2.059986763924826e-05,
      "clip_ratio/region_mean": 0.0022674804786220193,
      "epoch": 1.5389326334208224,
      "grad_norm": 0.20978401601314545,
      "learning_rate": 1e-06,
      "loss": -0.1055,
      "step": 659
    },
    {
      "clip_ratio/high_max": 0.0031130103016039357,
      "clip_ratio/high_mean": 0.001303149856539676,
      "clip_ratio/low_mean": 0.0013752163686149288,
      "clip_ratio/low_min": 4.119973527849652e-05,
      "clip_ratio/region_mean": 0.0026783662542584352,
      "epoch": 1.5412656751239429,
      "grad_norm": 0.2526197135448456,
      "learning_rate": 1e-06,
      "loss": -0.1056,
      "step": 660
    },
    {
      "clip_ratio/high_max": 0.0021811685219290666,
      "clip_ratio/high_mean": 0.0008479951702611288,
      "clip_ratio/low_mean": 0.0005780810424766969,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014260761854529846,
      "completions/clipped_ratio": 0.1662946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3428.0,
      "completions/mean_length": 1176.641845703125,
      "completions/mean_terminated_length": 594.3333129882812,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 1.5435987168270633,
      "grad_norm": 0.2901129126548767,
      "learning_rate": 1e-06,
      "loss": -0.0516,
      "num_tokens": 99160319.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.14368948340415955,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 661
    },
    {
      "clip_ratio/high_max": 0.0025631569806137122,
      "clip_ratio/high_mean": 0.0010572301398497075,
      "clip_ratio/low_mean": 0.0007841288665986212,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018413590514683165,
      "epoch": 1.5459317585301837,
      "grad_norm": 0.23473727703094482,
      "learning_rate": 1e-06,
      "loss": -0.0518,
      "step": 662
    },
    {
      "clip_ratio/high_max": 0.0025974552845582366,
      "clip_ratio/high_mean": 0.0010332157580705825,
      "clip_ratio/low_mean": 0.0009663371947681298,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00199955292919185,
      "epoch": 1.5482648002333042,
      "grad_norm": 0.2675100862979889,
      "learning_rate": 1e-06,
      "loss": -0.052,
      "step": 663
    },
    {
      "clip_ratio/high_max": 0.00247337876498932,
      "clip_ratio/high_mean": 0.0009737393938848982,
      "clip_ratio/low_mean": 0.0012372789251458016,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022110183563199826,
      "epoch": 1.5505978419364246,
      "grad_norm": 0.18326008319854736,
      "learning_rate": 1e-06,
      "loss": -0.0521,
      "step": 664
    },
    {
      "clip_ratio/high_max": 0.002578391147835646,
      "clip_ratio/high_mean": 0.001082983690139372,
      "clip_ratio/low_mean": 0.0006028006682754494,
      "clip_ratio/low_min": 7.835025826352648e-06,
      "clip_ratio/region_mean": 0.0016857843802426942,
      "completions/clipped_ratio": 0.1540178571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3464.0,
      "completions/mean_length": 1177.068115234375,
      "completions/mean_terminated_length": 645.653076171875,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 1.552930883639545,
      "grad_norm": 0.31763285398483276,
      "learning_rate": 1e-06,
      "loss": -0.0767,
      "num_tokens": 99732628.0,
      "reward": 0.5881696939468384,
      "reward_std": 0.1853206306695938,
      "rewards/verify_math_reward/mean": 0.5881696343421936,
      "rewards/verify_math_reward/std": 0.4924395978450775,
      "step": 665
    },
    {
      "clip_ratio/high_max": 0.0025883031557896174,
      "clip_ratio/high_mean": 0.0011414089749450795,
      "clip_ratio/low_mean": 0.0008020186633075355,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019434275673120283,
      "epoch": 1.5552639253426657,
      "grad_norm": 0.2709817588329315,
      "learning_rate": 1e-06,
      "loss": -0.0769,
      "step": 666
    },
    {
      "clip_ratio/high_max": 0.00291282403486548,
      "clip_ratio/high_mean": 0.0012065014489053283,
      "clip_ratio/low_mean": 0.0009946562222467037,
      "clip_ratio/low_min": 7.835025826352648e-06,
      "clip_ratio/region_mean": 0.0022011576729710214,
      "epoch": 1.5575969670457859,
      "grad_norm": 0.2394454926252365,
      "learning_rate": 1e-06,
      "loss": -0.0772,
      "step": 667
    },
    {
      "clip_ratio/high_max": 0.002633153955684975,
      "clip_ratio/high_mean": 0.001112860440116492,
      "clip_ratio/low_mean": 0.0012822716889786534,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002395132127276156,
      "epoch": 1.5599300087489065,
      "grad_norm": 0.22501912713050842,
      "learning_rate": 1e-06,
      "loss": -0.0773,
      "step": 668
    },
    {
      "clip_ratio/high_max": 0.0025296201347373426,
      "clip_ratio/high_mean": 0.0008875593648554059,
      "clip_ratio/low_mean": 0.00047986642357500386,
      "clip_ratio/low_min": 1.5363815691671334e-05,
      "clip_ratio/region_mean": 0.0013674257788807154,
      "completions/clipped_ratio": 0.1685267857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3748.0,
      "completions/mean_length": 1283.4810791015625,
      "completions/mean_terminated_length": 713.4268798828125,
      "completions/min_length": 132.0,
      "completions/min_terminated_length": 132.0,
      "epoch": 1.5622630504520267,
      "grad_norm": 0.2584800720214844,
      "learning_rate": 1e-06,
      "loss": -0.0459,
      "num_tokens": 100348475.0,
      "reward": 0.5491071939468384,
      "reward_std": 0.14849409461021423,
      "rewards/verify_math_reward/mean": 0.5491071343421936,
      "rewards/verify_math_reward/std": 0.49786055088043213,
      "step": 669
    },
    {
      "clip_ratio/high_max": 0.0023968944551597815,
      "clip_ratio/high_mean": 0.0009445089151540742,
      "clip_ratio/low_mean": 0.0006333161124985054,
      "clip_ratio/low_min": 1.5363815691671334e-05,
      "clip_ratio/region_mean": 0.001577825049025705,
      "epoch": 1.5645960921551474,
      "grad_norm": 0.23188327252864838,
      "learning_rate": 1e-06,
      "loss": -0.046,
      "step": 670
    },
    {
      "clip_ratio/high_max": 0.00333152122766478,
      "clip_ratio/high_mean": 0.0010798854073073016,
      "clip_ratio/low_mean": 0.0007709232504566899,
      "clip_ratio/low_min": 3.16937112074811e-05,
      "clip_ratio/region_mean": 0.001850808686867822,
      "epoch": 1.5669291338582676,
      "grad_norm": 0.1995295137166977,
      "learning_rate": 1e-06,
      "loss": -0.0462,
      "step": 671
    },
    {
      "clip_ratio/high_max": 0.0030446229757217225,
      "clip_ratio/high_mean": 0.0010142238497792277,
      "clip_ratio/low_mean": 0.0009125107726504211,
      "clip_ratio/low_min": 7.923427801870275e-06,
      "clip_ratio/region_mean": 0.0019267346178821754,
      "epoch": 1.5692621755613883,
      "grad_norm": 0.18869951367378235,
      "learning_rate": 1e-06,
      "loss": -0.0463,
      "step": 672
    },
    {
      "clip_ratio/high_max": 0.0021819696048623882,
      "clip_ratio/high_mean": 0.0008233503558585653,
      "clip_ratio/low_mean": 0.0006891033935971791,
      "clip_ratio/low_min": 8.679349775775336e-06,
      "clip_ratio/region_mean": 0.001512453796749469,
      "completions/clipped_ratio": 0.1964285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3989.0,
      "completions/mean_length": 1307.53466796875,
      "completions/mean_terminated_length": 625.9097290039062,
      "completions/min_length": 148.0,
      "completions/min_terminated_length": 148.0,
      "epoch": 1.5715952172645085,
      "grad_norm": 0.34236404299736023,
      "learning_rate": 1e-06,
      "loss": -0.0669,
      "num_tokens": 100886930.0,
      "reward": 0.5491071939468384,
      "reward_std": 0.18595930933952332,
      "rewards/verify_math_reward/mean": 0.5491071343421936,
      "rewards/verify_math_reward/std": 0.49786055088043213,
      "step": 673
    },
    {
      "clip_ratio/high_max": 0.003019629279151559,
      "clip_ratio/high_mean": 0.0011409016224206425,
      "clip_ratio/low_mean": 0.001045637009156053,
      "clip_ratio/low_min": 4.287306728656404e-05,
      "clip_ratio/region_mean": 0.0021865386806894094,
      "epoch": 1.5739282589676291,
      "grad_norm": 0.2636945843696594,
      "learning_rate": 1e-06,
      "loss": -0.0673,
      "step": 674
    },
    {
      "clip_ratio/high_max": 0.0031311263737734407,
      "clip_ratio/high_mean": 0.001144453988672467,
      "clip_ratio/low_mean": 0.0013594984848168679,
      "clip_ratio/low_min": 2.5514365916023962e-05,
      "clip_ratio/region_mean": 0.0025039524407475255,
      "epoch": 1.5762613006707495,
      "grad_norm": 0.24746200442314148,
      "learning_rate": 1e-06,
      "loss": -0.0675,
      "step": 675
    },
    {
      "clip_ratio/high_max": 0.002691456291358918,
      "clip_ratio/high_mean": 0.0010489637461432721,
      "clip_ratio/low_mean": 0.0015580262552248314,
      "clip_ratio/low_min": 4.2611225580913015e-05,
      "clip_ratio/region_mean": 0.0026069900413858704,
      "epoch": 1.57859434237387,
      "grad_norm": 0.24600592255592346,
      "learning_rate": 1e-06,
      "loss": -0.0676,
      "step": 676
    },
    {
      "clip_ratio/high_max": 0.0024583497288404033,
      "clip_ratio/high_mean": 0.0009814610311877914,
      "clip_ratio/low_mean": 0.000726520012904075,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017079810531868134,
      "completions/clipped_ratio": 0.1350446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4078.0,
      "completions/mean_length": 1136.7723388671875,
      "completions/mean_terminated_length": 674.7509765625,
      "completions/min_length": 141.0,
      "completions/min_terminated_length": 141.0,
      "epoch": 1.5809273840769904,
      "grad_norm": 0.31176334619522095,
      "learning_rate": 1e-06,
      "loss": -0.0515,
      "num_tokens": 101499030.0,
      "reward": 0.546875,
      "reward_std": 0.188248410820961,
      "rewards/verify_math_reward/mean": 0.546875,
      "rewards/verify_math_reward/std": 0.4980759024620056,
      "step": 677
    },
    {
      "clip_ratio/high_max": 0.0029717501529376023,
      "clip_ratio/high_mean": 0.0011934057911275886,
      "clip_ratio/low_mean": 0.0011119712435174733,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002305377049196977,
      "epoch": 1.5832604257801108,
      "grad_norm": 0.27647027373313904,
      "learning_rate": 1e-06,
      "loss": -0.0518,
      "step": 678
    },
    {
      "clip_ratio/high_max": 0.003101067639363464,
      "clip_ratio/high_mean": 0.0012359481625026092,
      "clip_ratio/low_mean": 0.0013322515715117333,
      "clip_ratio/low_min": 2.84349407593254e-05,
      "clip_ratio/region_mean": 0.0025681997431092896,
      "epoch": 1.5855934674832313,
      "grad_norm": 0.25151047110557556,
      "learning_rate": 1e-06,
      "loss": -0.052,
      "step": 679
    },
    {
      "clip_ratio/high_max": 0.0029707483190577477,
      "clip_ratio/high_mean": 0.0011499096181069035,
      "clip_ratio/low_mean": 0.001467522828534129,
      "clip_ratio/low_min": 2.84349407593254e-05,
      "clip_ratio/region_mean": 0.002617432506667683,
      "epoch": 1.5879265091863517,
      "grad_norm": 0.23571977019309998,
      "learning_rate": 1e-06,
      "loss": -0.0521,
      "step": 680
    },
    {
      "clip_ratio/high_max": 0.002559652282798197,
      "clip_ratio/high_mean": 0.0010860990314540686,
      "clip_ratio/low_mean": 0.0007604357206218992,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018465347384335473,
      "completions/clipped_ratio": 0.2008928571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3538.0,
      "completions/mean_length": 1334.6796875,
      "completions/mean_terminated_length": 640.4929809570312,
      "completions/min_length": 162.0,
      "completions/min_terminated_length": 162.0,
      "epoch": 1.5902595508894721,
      "grad_norm": 0.29189226031303406,
      "learning_rate": 1e-06,
      "loss": -0.0747,
      "num_tokens": 102048711.0,
      "reward": 0.5245535969734192,
      "reward_std": 0.1755138486623764,
      "rewards/verify_math_reward/mean": 0.5245535969734192,
      "rewards/verify_math_reward/std": 0.4996756613254547,
      "step": 681
    },
    {
      "clip_ratio/high_max": 0.002816863951011328,
      "clip_ratio/high_mean": 0.0011096646394435083,
      "clip_ratio/low_mean": 0.0011371720538591035,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002246836684207665,
      "epoch": 1.5925925925925926,
      "grad_norm": 0.25846952199935913,
      "learning_rate": 1e-06,
      "loss": -0.0749,
      "step": 682
    },
    {
      "clip_ratio/high_max": 0.003701198067574296,
      "clip_ratio/high_mean": 0.0013540929376176791,
      "clip_ratio/low_mean": 0.001283191236325365,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026372841457487084,
      "epoch": 1.594925634295713,
      "grad_norm": 0.20855620503425598,
      "learning_rate": 1e-06,
      "loss": -0.0752,
      "step": 683
    },
    {
      "clip_ratio/high_max": 0.003204186574293999,
      "clip_ratio/high_mean": 0.001233112501722644,
      "clip_ratio/low_mean": 0.001455334067941294,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026884465478360653,
      "epoch": 1.5972586759988334,
      "grad_norm": 0.21366524696350098,
      "learning_rate": 1e-06,
      "loss": -0.0752,
      "step": 684
    },
    {
      "clip_ratio/high_max": 0.0025050155745702796,
      "clip_ratio/high_mean": 0.0009705617230792996,
      "clip_ratio/low_mean": 0.0006182609540701378,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015888227062532678,
      "completions/clipped_ratio": 0.1875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3316.0,
      "completions/mean_length": 1324.4910888671875,
      "completions/mean_terminated_length": 684.912109375,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 1.599591717701954,
      "grad_norm": 0.2937885522842407,
      "learning_rate": 1e-06,
      "loss": -0.06,
      "num_tokens": 102639079.0,
      "reward": 0.5613839626312256,
      "reward_std": 0.20324954390525818,
      "rewards/verify_math_reward/mean": 0.5613839030265808,
      "rewards/verify_math_reward/std": 0.496494859457016,
      "step": 685
    },
    {
      "clip_ratio/high_max": 0.003114350685791578,
      "clip_ratio/high_mean": 0.001153579549281858,
      "clip_ratio/low_mean": 0.0008343398549186531,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019879194151144475,
      "epoch": 1.6019247594050743,
      "grad_norm": 0.22430022060871124,
      "learning_rate": 1e-06,
      "loss": -0.0603,
      "step": 686
    },
    {
      "clip_ratio/high_max": 0.002859890679246746,
      "clip_ratio/high_mean": 0.0012254288703843486,
      "clip_ratio/low_mean": 0.0010812852524395566,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023067141009960324,
      "epoch": 1.604257801108195,
      "grad_norm": 0.2645739018917084,
      "learning_rate": 1e-06,
      "loss": -0.0604,
      "step": 687
    },
    {
      "clip_ratio/high_max": 0.0028396494963089935,
      "clip_ratio/high_mean": 0.001110363347834209,
      "clip_ratio/low_mean": 0.0012552885236800648,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023656518242205493,
      "epoch": 1.6065908428113151,
      "grad_norm": 0.25388264656066895,
      "learning_rate": 1e-06,
      "loss": -0.0604,
      "step": 688
    },
    {
      "clip_ratio/high_max": 0.002372793183894828,
      "clip_ratio/high_mean": 0.0010088597427966306,
      "clip_ratio/low_mean": 0.0006163730804473744,
      "clip_ratio/low_min": 4.205921868560836e-05,
      "clip_ratio/region_mean": 0.0016252328423433937,
      "completions/clipped_ratio": 0.1551339285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3867.0,
      "completions/mean_length": 1164.2421875,
      "completions/mean_terminated_length": 625.9141845703125,
      "completions/min_length": 183.0,
      "completions/min_terminated_length": 183.0,
      "epoch": 1.6089238845144358,
      "grad_norm": 0.2801320552825928,
      "learning_rate": 1e-06,
      "loss": -0.0958,
      "num_tokens": 103193344.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.19068947434425354,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514803290367126,
      "step": 689
    },
    {
      "clip_ratio/high_max": 0.0032055882111308165,
      "clip_ratio/high_mean": 0.0013251628406578675,
      "clip_ratio/low_mean": 0.0007281958114617737,
      "clip_ratio/low_min": 1.842570782173425e-05,
      "clip_ratio/region_mean": 0.0020533585993689485,
      "epoch": 1.611256926217556,
      "grad_norm": 0.23484808206558228,
      "learning_rate": 1e-06,
      "loss": -0.096,
      "step": 690
    },
    {
      "clip_ratio/high_max": 0.0031799277930986136,
      "clip_ratio/high_mean": 0.0013861646511941217,
      "clip_ratio/low_mean": 0.0009580936939528328,
      "clip_ratio/low_min": 4.6064269554335624e-05,
      "clip_ratio/region_mean": 0.002344258398807142,
      "epoch": 1.6135899679206767,
      "grad_norm": 0.21078988909721375,
      "learning_rate": 1e-06,
      "loss": -0.0962,
      "step": 691
    },
    {
      "clip_ratio/high_max": 0.0031392294986289926,
      "clip_ratio/high_mean": 0.0013050017587374896,
      "clip_ratio/low_mean": 0.0011217644498628943,
      "clip_ratio/low_min": 4.6064269554335624e-05,
      "clip_ratio/region_mean": 0.002426766171993222,
      "epoch": 1.6159230096237969,
      "grad_norm": 0.23015913367271423,
      "learning_rate": 1e-06,
      "loss": -0.0962,
      "step": 692
    },
    {
      "clip_ratio/high_max": 0.0019772380546783097,
      "clip_ratio/high_mean": 0.0007786658425175119,
      "clip_ratio/low_mean": 0.0005568928290813346,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001335558725259034,
      "completions/clipped_ratio": 0.2053571428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3617.0,
      "completions/mean_length": 1319.727783203125,
      "completions/mean_terminated_length": 602.2640380859375,
      "completions/min_length": 167.0,
      "completions/min_terminated_length": 167.0,
      "epoch": 1.6182560513269175,
      "grad_norm": 0.30670201778411865,
      "learning_rate": 1e-06,
      "loss": -0.1051,
      "num_tokens": 103712468.0,
      "reward": 0.5502232313156128,
      "reward_std": 0.18032027781009674,
      "rewards/verify_math_reward/mean": 0.5502232313156128,
      "rewards/verify_math_reward/std": 0.49774909019470215,
      "step": 693
    },
    {
      "clip_ratio/high_max": 0.0029129241302143782,
      "clip_ratio/high_mean": 0.0011786021677835379,
      "clip_ratio/low_mean": 0.0008487720242555952,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002027374204772059,
      "epoch": 1.620589093030038,
      "grad_norm": 0.3664519786834717,
      "learning_rate": 1e-06,
      "loss": -0.1054,
      "step": 694
    },
    {
      "clip_ratio/high_max": 0.002840742585249245,
      "clip_ratio/high_mean": 0.0011297030687273946,
      "clip_ratio/low_mean": 0.0009428865196241532,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020725895592477173,
      "epoch": 1.6229221347331584,
      "grad_norm": 0.25414541363716125,
      "learning_rate": 1e-06,
      "loss": -0.1056,
      "step": 695
    },
    {
      "clip_ratio/high_max": 0.0028106456229579635,
      "clip_ratio/high_mean": 0.0011537669524841476,
      "clip_ratio/low_mean": 0.0011728714871424017,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023266384232556447,
      "epoch": 1.6252551764362788,
      "grad_norm": 0.24437932670116425,
      "learning_rate": 1e-06,
      "loss": -0.1057,
      "step": 696
    },
    {
      "clip_ratio/high_max": 0.002399275341304019,
      "clip_ratio/high_mean": 0.0008898892119759694,
      "clip_ratio/low_mean": 0.000654937247873022,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001544826453027781,
      "completions/clipped_ratio": 0.1975446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3361.0,
      "completions/mean_length": 1324.7366943359375,
      "completions/mean_terminated_length": 642.5202026367188,
      "completions/min_length": 196.0,
      "completions/min_terminated_length": 196.0,
      "epoch": 1.6275882181393992,
      "grad_norm": 0.31290677189826965,
      "learning_rate": 1e-06,
      "loss": -0.0769,
      "num_tokens": 104282808.0,
      "reward": 0.5223214626312256,
      "reward_std": 0.1779131293296814,
      "rewards/verify_math_reward/mean": 0.5223214030265808,
      "rewards/verify_math_reward/std": 0.49978047609329224,
      "step": 697
    },
    {
      "clip_ratio/high_max": 0.0031805944818188436,
      "clip_ratio/high_mean": 0.0011535340945556527,
      "clip_ratio/low_mean": 0.0007945833312987816,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019481174167594872,
      "epoch": 1.6299212598425197,
      "grad_norm": 0.23811165988445282,
      "learning_rate": 1e-06,
      "loss": -0.0771,
      "step": 698
    },
    {
      "clip_ratio/high_max": 0.0027703681626007892,
      "clip_ratio/high_mean": 0.0010411006078356877,
      "clip_ratio/low_mean": 0.0009474635662627406,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019885641668224707,
      "epoch": 1.63225430154564,
      "grad_norm": 0.2243320345878601,
      "learning_rate": 1e-06,
      "loss": -0.0772,
      "step": 699
    },
    {
      "clip_ratio/high_max": 0.0030867442983435467,
      "clip_ratio/high_mean": 0.0010878828761633486,
      "clip_ratio/low_mean": 0.001150815664914262,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002238698521978222,
      "epoch": 1.6345873432487605,
      "grad_norm": 0.18212801218032837,
      "learning_rate": 1e-06,
      "loss": -0.0774,
      "step": 700
    },
    {
      "clip_ratio/high_max": 0.0026252361203660257,
      "clip_ratio/high_mean": 0.0010942014032480074,
      "clip_ratio/low_mean": 0.0005422721651484608,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001636473556573037,
      "completions/clipped_ratio": 0.1875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3663.0,
      "completions/mean_length": 1305.7723388671875,
      "completions/mean_terminated_length": 661.8736572265625,
      "completions/min_length": 178.0,
      "completions/min_terminated_length": 178.0,
      "epoch": 1.636920384951881,
      "grad_norm": 0.33585652709007263,
      "learning_rate": 1e-06,
      "loss": -0.0916,
      "num_tokens": 104854956.0,
      "reward": 0.5725446939468384,
      "reward_std": 0.1841912716627121,
      "rewards/verify_math_reward/mean": 0.5725446343421936,
      "rewards/verify_math_reward/std": 0.49498558044433594,
      "step": 701
    },
    {
      "clip_ratio/high_max": 0.00304038253671024,
      "clip_ratio/high_mean": 0.0013526170332625043,
      "clip_ratio/low_mean": 0.0007281387952389196,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020807558539672755,
      "epoch": 1.6392534266550016,
      "grad_norm": 0.24745501577854156,
      "learning_rate": 1e-06,
      "loss": -0.0918,
      "step": 702
    },
    {
      "clip_ratio/high_max": 0.002915855307946913,
      "clip_ratio/high_mean": 0.0012921984380227514,
      "clip_ratio/low_mean": 0.0009440897629247047,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002236288222775329,
      "epoch": 1.6415864683581218,
      "grad_norm": 0.2093912959098816,
      "learning_rate": 1e-06,
      "loss": -0.0921,
      "step": 703
    },
    {
      "clip_ratio/high_max": 0.0027673358854372054,
      "clip_ratio/high_mean": 0.0013007218585698865,
      "clip_ratio/low_mean": 0.0010847195226233453,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023854413448134437,
      "epoch": 1.6439195100612425,
      "grad_norm": 0.2719499170780182,
      "learning_rate": 1e-06,
      "loss": -0.0922,
      "step": 704
    },
    {
      "clip_ratio/high_max": 0.0023966357875906397,
      "clip_ratio/high_mean": 0.0009143912266154075,
      "clip_ratio/low_mean": 0.0007818282201697002,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016962194422376342,
      "completions/clipped_ratio": 0.2165178571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3769.0,
      "completions/mean_length": 1385.1864013671875,
      "completions/mean_terminated_length": 636.0441284179688,
      "completions/min_length": 153.0,
      "completions/min_terminated_length": 153.0,
      "epoch": 1.6462525517643627,
      "grad_norm": 0.34167617559432983,
      "learning_rate": 1e-06,
      "loss": -0.0604,
      "num_tokens": 105392075.0,
      "reward": 0.4654017984867096,
      "reward_std": 0.19155599176883698,
      "rewards/verify_math_reward/mean": 0.4654017984867096,
      "rewards/verify_math_reward/std": 0.4990801215171814,
      "step": 705
    },
    {
      "clip_ratio/high_max": 0.003016867267433554,
      "clip_ratio/high_mean": 0.0011957475217059255,
      "clip_ratio/low_mean": 0.0011352539168001385,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023310014148592018,
      "epoch": 1.6485855934674833,
      "grad_norm": 0.300221711397171,
      "learning_rate": 1e-06,
      "loss": -0.0606,
      "step": 706
    },
    {
      "clip_ratio/high_max": 0.002963373059174046,
      "clip_ratio/high_mean": 0.0010386884860054124,
      "clip_ratio/low_mean": 0.0012534869038063334,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002292175340699032,
      "epoch": 1.6509186351706036,
      "grad_norm": 0.25997117161750793,
      "learning_rate": 1e-06,
      "loss": -0.0609,
      "step": 707
    },
    {
      "clip_ratio/high_max": 0.002642484017997049,
      "clip_ratio/high_mean": 0.001068166990080499,
      "clip_ratio/low_mean": 0.0016632107199257007,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002731377651798539,
      "epoch": 1.6532516768737242,
      "grad_norm": 0.29943886399269104,
      "learning_rate": 1e-06,
      "loss": -0.0609,
      "step": 708
    },
    {
      "clip_ratio/high_max": 0.0024212583593907766,
      "clip_ratio/high_mean": 0.0010675480953068472,
      "clip_ratio/low_mean": 0.0004458831745068892,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015134312634472735,
      "completions/clipped_ratio": 0.1852678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3461.0,
      "completions/mean_length": 1296.90185546875,
      "completions/mean_terminated_length": 660.39453125,
      "completions/min_length": 172.0,
      "completions/min_terminated_length": 172.0,
      "epoch": 1.6555847185768444,
      "grad_norm": 0.30853670835494995,
      "learning_rate": 1e-06,
      "loss": -0.0938,
      "num_tokens": 105957619.0,
      "reward": 0.5859375,
      "reward_std": 0.19959722459316254,
      "rewards/verify_math_reward/mean": 0.5859375,
      "rewards/verify_math_reward/std": 0.4928344786167145,
      "step": 709
    },
    {
      "clip_ratio/high_max": 0.003249817520554643,
      "clip_ratio/high_mean": 0.0013167445504222997,
      "clip_ratio/low_mean": 0.0007628459834450041,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020795905438717455,
      "epoch": 1.657917760279965,
      "grad_norm": 0.2825876772403717,
      "learning_rate": 1e-06,
      "loss": -0.0941,
      "step": 710
    },
    {
      "clip_ratio/high_max": 0.0033827013758127578,
      "clip_ratio/high_mean": 0.001281385702895932,
      "clip_ratio/low_mean": 0.0009322790992882801,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002213664738519583,
      "epoch": 1.6602508019830855,
      "grad_norm": 0.21970489621162415,
      "learning_rate": 1e-06,
      "loss": -0.0942,
      "step": 711
    },
    {
      "clip_ratio/high_max": 0.003062765797949396,
      "clip_ratio/high_mean": 0.00115851747978013,
      "clip_ratio/low_mean": 0.001118047950512846,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022765654284739867,
      "epoch": 1.662583843686206,
      "grad_norm": 0.22429729998111725,
      "learning_rate": 1e-06,
      "loss": -0.0943,
      "step": 712
    },
    {
      "clip_ratio/high_max": 0.002440858632326126,
      "clip_ratio/high_mean": 0.0009094994184124516,
      "clip_ratio/low_mean": 0.0006458879433921538,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015553873381577432,
      "completions/clipped_ratio": 0.1696428571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3330.0,
      "completions/mean_length": 1200.372802734375,
      "completions/mean_terminated_length": 608.7930297851562,
      "completions/min_length": 127.0,
      "completions/min_terminated_length": 127.0,
      "epoch": 1.6649168853893264,
      "grad_norm": 0.3057929575443268,
      "learning_rate": 1e-06,
      "loss": -0.055,
      "num_tokens": 106492745.0,
      "reward": 0.5446428656578064,
      "reward_std": 0.18821631371974945,
      "rewards/verify_math_reward/mean": 0.5446428656578064,
      "rewards/verify_math_reward/std": 0.4982811510562897,
      "step": 713
    },
    {
      "clip_ratio/high_max": 0.0029230081163404975,
      "clip_ratio/high_mean": 0.0011955939498875523,
      "clip_ratio/low_mean": 0.0008712396156624891,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020668336073867977,
      "epoch": 1.6672499270924468,
      "grad_norm": 0.23520135879516602,
      "learning_rate": 1e-06,
      "loss": -0.0552,
      "step": 714
    },
    {
      "clip_ratio/high_max": 0.0028584764659171924,
      "clip_ratio/high_mean": 0.0011451138452684972,
      "clip_ratio/low_mean": 0.0010565156408119947,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022016294751665555,
      "epoch": 1.6695829687955672,
      "grad_norm": 0.23191626369953156,
      "learning_rate": 1e-06,
      "loss": -0.0554,
      "step": 715
    },
    {
      "clip_ratio/high_max": 0.0028426450662664138,
      "clip_ratio/high_mean": 0.0010478730182512663,
      "clip_ratio/low_mean": 0.001281688870221842,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002329561953956727,
      "epoch": 1.6719160104986877,
      "grad_norm": 0.24943673610687256,
      "learning_rate": 1e-06,
      "loss": -0.0555,
      "step": 716
    },
    {
      "clip_ratio/high_max": 0.0026324251739424653,
      "clip_ratio/high_mean": 0.0009574365285516251,
      "clip_ratio/low_mean": 0.0007995056948857382,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001756942248903215,
      "completions/clipped_ratio": 0.1875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4083.0,
      "completions/mean_length": 1261.8951416015625,
      "completions/mean_terminated_length": 607.8709106445312,
      "completions/min_length": 110.0,
      "completions/min_terminated_length": 110.0,
      "epoch": 1.674249052201808,
      "grad_norm": 0.3139522969722748,
      "learning_rate": 1e-06,
      "loss": -0.0796,
      "num_tokens": 107022179.0,
      "reward": 0.5558035969734192,
      "reward_std": 0.18125459551811218,
      "rewards/verify_math_reward/mean": 0.5558035969734192,
      "rewards/verify_math_reward/std": 0.49715372920036316,
      "step": 717
    },
    {
      "clip_ratio/high_max": 0.003595037505874643,
      "clip_ratio/high_mean": 0.0011868317760672653,
      "clip_ratio/low_mean": 0.0010414816260890802,
      "clip_ratio/low_min": 2.4811433831928298e-05,
      "clip_ratio/region_mean": 0.0022283133985183667,
      "epoch": 1.6765820939049285,
      "grad_norm": 0.28548964858055115,
      "learning_rate": 1e-06,
      "loss": -0.08,
      "step": 718
    },
    {
      "clip_ratio/high_max": 0.0035372318197914865,
      "clip_ratio/high_mean": 0.001203651931064087,
      "clip_ratio/low_mean": 0.0012385282934701536,
      "clip_ratio/low_min": 5.047351623943541e-05,
      "clip_ratio/region_mean": 0.002442180208163336,
      "epoch": 1.678915135608049,
      "grad_norm": 0.24397175014019012,
      "learning_rate": 1e-06,
      "loss": -0.0802,
      "step": 719
    },
    {
      "clip_ratio/high_max": 0.0033241948622162454,
      "clip_ratio/high_mean": 0.0010891852689383086,
      "clip_ratio/low_mean": 0.0014839978211966809,
      "clip_ratio/low_min": 3.90482464354136e-05,
      "clip_ratio/region_mean": 0.0025731830755830742,
      "epoch": 1.6812481773111694,
      "grad_norm": 0.2252047061920166,
      "learning_rate": 1e-06,
      "loss": -0.0802,
      "step": 720
    },
    {
      "clip_ratio/high_max": 0.0020865477781626396,
      "clip_ratio/high_mean": 0.0007889104326750385,
      "clip_ratio/low_mean": 0.0005723328638396197,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013612432612717384,
      "completions/clipped_ratio": 0.1685267857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3685.0,
      "completions/mean_length": 1201.063720703125,
      "completions/mean_terminated_length": 614.3047485351562,
      "completions/min_length": 184.0,
      "completions/min_terminated_length": 184.0,
      "epoch": 1.68358121901429,
      "grad_norm": 0.4459717273712158,
      "learning_rate": 1e-06,
      "loss": -0.072,
      "num_tokens": 107562684.0,
      "reward": 0.5803571939468384,
      "reward_std": 0.17078480124473572,
      "rewards/verify_math_reward/mean": 0.5803571343421936,
      "rewards/verify_math_reward/std": 0.4937761127948761,
      "step": 721
    },
    {
      "clip_ratio/high_max": 0.002799970516207395,
      "clip_ratio/high_mean": 0.0010848278343473794,
      "clip_ratio/low_mean": 0.0008371281196559721,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019219559399061836,
      "epoch": 1.6859142607174102,
      "grad_norm": 0.2520245909690857,
      "learning_rate": 1e-06,
      "loss": -0.0723,
      "step": 722
    },
    {
      "clip_ratio/high_max": 0.002626189547299873,
      "clip_ratio/high_mean": 0.0010460953490110114,
      "clip_ratio/low_mean": 0.000988977945780789,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002035073241131613,
      "epoch": 1.688247302420531,
      "grad_norm": 0.2172216922044754,
      "learning_rate": 1e-06,
      "loss": -0.0725,
      "step": 723
    },
    {
      "clip_ratio/high_max": 0.0027866967429872602,
      "clip_ratio/high_mean": 0.0009972511488740565,
      "clip_ratio/low_mean": 0.0012106410231353948,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022078921720094513,
      "epoch": 1.690580344123651,
      "grad_norm": 0.23183082044124603,
      "learning_rate": 1e-06,
      "loss": -0.0725,
      "step": 724
    },
    {
      "clip_ratio/high_max": 0.0023170647109509446,
      "clip_ratio/high_mean": 0.0008763284749875311,
      "clip_ratio/low_mean": 0.0005670832501891709,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014434117256314494,
      "completions/clipped_ratio": 0.2165178571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3758.0,
      "completions/mean_length": 1413.1998291015625,
      "completions/mean_terminated_length": 671.7991333007812,
      "completions/min_length": 192.0,
      "completions/min_terminated_length": 192.0,
      "epoch": 1.6929133858267718,
      "grad_norm": 0.30320650339126587,
      "learning_rate": 1e-06,
      "loss": -0.0736,
      "num_tokens": 108138791.0,
      "reward": 0.4910714626312256,
      "reward_std": 0.16653545200824738,
      "rewards/verify_math_reward/mean": 0.4910714328289032,
      "rewards/verify_math_reward/std": 0.5001994967460632,
      "step": 725
    },
    {
      "clip_ratio/high_max": 0.003044892131583765,
      "clip_ratio/high_mean": 0.0011829018894786714,
      "clip_ratio/low_mean": 0.0008012801918084733,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019841820576402824,
      "epoch": 1.695246427529892,
      "grad_norm": 0.2436947077512741,
      "learning_rate": 1e-06,
      "loss": -0.0739,
      "step": 726
    },
    {
      "clip_ratio/high_max": 0.0032057481730589643,
      "clip_ratio/high_mean": 0.0011772155739890877,
      "clip_ratio/low_mean": 0.0009244151933671674,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021016307546233293,
      "epoch": 1.6975794692330126,
      "grad_norm": 0.21844151616096497,
      "learning_rate": 1e-06,
      "loss": -0.074,
      "step": 727
    },
    {
      "clip_ratio/high_max": 0.002946708300441969,
      "clip_ratio/high_mean": 0.0011590781032282393,
      "clip_ratio/low_mean": 0.0011810414580395445,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002340119535801932,
      "epoch": 1.6999125109361328,
      "grad_norm": 0.19436538219451904,
      "learning_rate": 1e-06,
      "loss": -0.0741,
      "step": 728
    },
    {
      "clip_ratio/high_max": 0.002840550339897163,
      "clip_ratio/high_mean": 0.0010615747232805006,
      "clip_ratio/low_mean": 0.0006888908933433413,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001750465566146886,
      "completions/clipped_ratio": 0.1372767857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3706.0,
      "completions/mean_length": 1066.88623046875,
      "completions/mean_terminated_length": 584.8926391601562,
      "completions/min_length": 132.0,
      "completions/min_terminated_length": 132.0,
      "epoch": 1.7022455526392535,
      "grad_norm": 0.3603469431400299,
      "learning_rate": 1e-06,
      "loss": -0.0625,
      "num_tokens": 108686689.0,
      "reward": 0.6272321939468384,
      "reward_std": 0.1584189236164093,
      "rewards/verify_math_reward/mean": 0.6272321343421936,
      "rewards/verify_math_reward/std": 0.4838111698627472,
      "step": 729
    },
    {
      "clip_ratio/high_max": 0.00358756499554147,
      "clip_ratio/high_mean": 0.001336411858574138,
      "clip_ratio/low_mean": 0.001047273280164518,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023836851287342142,
      "epoch": 1.704578594342374,
      "grad_norm": 0.3039668798446655,
      "learning_rate": 1e-06,
      "loss": -0.0628,
      "step": 730
    },
    {
      "clip_ratio/high_max": 0.0037036568028270267,
      "clip_ratio/high_mean": 0.001329478091065539,
      "clip_ratio/low_mean": 0.0013035761201081186,
      "clip_ratio/low_min": 1.1415525477787014e-05,
      "clip_ratio/region_mean": 0.00263305424596183,
      "epoch": 1.7069116360454943,
      "grad_norm": 0.27185073494911194,
      "learning_rate": 1e-06,
      "loss": -0.0629,
      "step": 731
    },
    {
      "clip_ratio/high_max": 0.0032777939195511863,
      "clip_ratio/high_mean": 0.001269702181161847,
      "clip_ratio/low_mean": 0.0014182024151523365,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026879046636167914,
      "epoch": 1.7092446777486148,
      "grad_norm": 0.2635992765426636,
      "learning_rate": 1e-06,
      "loss": -0.063,
      "step": 732
    },
    {
      "clip_ratio/high_max": 0.0027937005579588003,
      "clip_ratio/high_mean": 0.0009766038310772274,
      "clip_ratio/low_mean": 0.0004562619997159345,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014328658253361937,
      "completions/clipped_ratio": 0.2087053571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4022.0,
      "completions/mean_length": 1386.69873046875,
      "completions/mean_terminated_length": 672.1156616210938,
      "completions/min_length": 146.0,
      "completions/min_terminated_length": 146.0,
      "epoch": 1.7115777194517352,
      "grad_norm": 0.2532927393913269,
      "learning_rate": 1e-06,
      "loss": -0.1131,
      "num_tokens": 109256067.0,
      "reward": 0.5223214626312256,
      "reward_std": 0.16014528274536133,
      "rewards/verify_math_reward/mean": 0.5223214030265808,
      "rewards/verify_math_reward/std": 0.49978047609329224,
      "step": 733
    },
    {
      "clip_ratio/high_max": 0.003094894185778685,
      "clip_ratio/high_mean": 0.0010996055279974826,
      "clip_ratio/low_mean": 0.0005157897130629863,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016153952456079423,
      "epoch": 1.7139107611548556,
      "grad_norm": 0.27027636766433716,
      "learning_rate": 1e-06,
      "loss": -0.1133,
      "step": 734
    },
    {
      "clip_ratio/high_max": 0.0035217979311710224,
      "clip_ratio/high_mean": 0.0012316461052250816,
      "clip_ratio/low_mean": 0.0007063940056468709,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019380400262889452,
      "epoch": 1.716243802857976,
      "grad_norm": 0.20090660452842712,
      "learning_rate": 1e-06,
      "loss": -0.1135,
      "step": 735
    },
    {
      "clip_ratio/high_max": 0.0032341975238523446,
      "clip_ratio/high_mean": 0.001142957709816983,
      "clip_ratio/low_mean": 0.0009098773225559853,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020528350796666928,
      "epoch": 1.7185768445610965,
      "grad_norm": 0.19518056511878967,
      "learning_rate": 1e-06,
      "loss": -0.1136,
      "step": 736
    },
    {
      "clip_ratio/high_max": 0.0020187238551443443,
      "clip_ratio/high_mean": 0.0007821783510735258,
      "clip_ratio/low_mean": 0.000512348198753898,
      "clip_ratio/low_min": 1.1285662367299665e-05,
      "clip_ratio/region_mean": 0.001294526558922371,
      "completions/clipped_ratio": 0.1852678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3715.0,
      "completions/mean_length": 1289.1820068359375,
      "completions/mean_terminated_length": 650.919189453125,
      "completions/min_length": 152.0,
      "completions/min_terminated_length": 152.0,
      "epoch": 1.720909886264217,
      "grad_norm": 0.27669376134872437,
      "learning_rate": 1e-06,
      "loss": -0.0785,
      "num_tokens": 109817022.0,
      "reward": 0.5111607313156128,
      "reward_std": 0.16732315719127655,
      "rewards/verify_math_reward/mean": 0.5111607313156128,
      "rewards/verify_math_reward/std": 0.5001546144485474,
      "step": 737
    },
    {
      "clip_ratio/high_max": 0.0022142234665807337,
      "clip_ratio/high_mean": 0.000889992157681263,
      "clip_ratio/low_mean": 0.0007480007707272307,
      "clip_ratio/low_min": 2.4206041416618973e-05,
      "clip_ratio/region_mean": 0.00163799295478384,
      "epoch": 1.7232429279673376,
      "grad_norm": 0.26892563700675964,
      "learning_rate": 1e-06,
      "loss": -0.0787,
      "step": 738
    },
    {
      "clip_ratio/high_max": 0.002578092593466863,
      "clip_ratio/high_mean": 0.0010435152689751703,
      "clip_ratio/low_mean": 0.0008472703430015827,
      "clip_ratio/low_min": 2.257132473459933e-05,
      "clip_ratio/region_mean": 0.0018907856065197848,
      "epoch": 1.7255759696704578,
      "grad_norm": 0.1900540441274643,
      "learning_rate": 1e-06,
      "loss": -0.079,
      "step": 739
    },
    {
      "clip_ratio/high_max": 0.0020848819622187875,
      "clip_ratio/high_mean": 0.0008317502379213693,
      "clip_ratio/low_mean": 0.001126940247559105,
      "clip_ratio/low_min": 3.3856988011393696e-05,
      "clip_ratio/region_mean": 0.0019586904818424955,
      "epoch": 1.7279090113735784,
      "grad_norm": 0.1855372041463852,
      "learning_rate": 1e-06,
      "loss": -0.079,
      "step": 740
    },
    {
      "clip_ratio/high_max": 0.002750308034592308,
      "clip_ratio/high_mean": 0.001090041017960175,
      "clip_ratio/low_mean": 0.0005552698330575367,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016453108364657965,
      "completions/clipped_ratio": 0.2209821428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3175.0,
      "completions/mean_length": 1397.0457763671875,
      "completions/mean_terminated_length": 631.4398803710938,
      "completions/min_length": 177.0,
      "completions/min_terminated_length": 177.0,
      "epoch": 1.7302420530766986,
      "grad_norm": 0.37509623169898987,
      "learning_rate": 1e-06,
      "loss": -0.0531,
      "num_tokens": 110349615.0,
      "reward": 0.4687500298023224,
      "reward_std": 0.164472758769989,
      "rewards/verify_math_reward/mean": 0.46875,
      "rewards/verify_math_reward/std": 0.4993011951446533,
      "step": 741
    },
    {
      "clip_ratio/high_max": 0.00314707585494034,
      "clip_ratio/high_mean": 0.001330789462372195,
      "clip_ratio/low_mean": 0.0007947428930492606,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021255323226796463,
      "epoch": 1.7325750947798193,
      "grad_norm": 0.2796333134174347,
      "learning_rate": 1e-06,
      "loss": -0.0534,
      "step": 742
    },
    {
      "clip_ratio/high_max": 0.003252979200624395,
      "clip_ratio/high_mean": 0.0013329988105397206,
      "clip_ratio/low_mean": 0.0010988753110723337,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002431874083413277,
      "epoch": 1.7349081364829395,
      "grad_norm": 0.21326813101768494,
      "learning_rate": 1e-06,
      "loss": -0.0536,
      "step": 743
    },
    {
      "clip_ratio/high_max": 0.003149357784423046,
      "clip_ratio/high_mean": 0.0013002731720916927,
      "clip_ratio/low_mean": 0.0012321688736847136,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0025324420566903427,
      "epoch": 1.7372411781860602,
      "grad_norm": 0.2357005923986435,
      "learning_rate": 1e-06,
      "loss": -0.0537,
      "step": 744
    },
    {
      "clip_ratio/high_max": 0.0026869889843510464,
      "clip_ratio/high_mean": 0.0009339023745269515,
      "clip_ratio/low_mean": 0.00044741597776010167,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013813183722959366,
      "completions/clipped_ratio": 0.1350446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2385.0,
      "completions/mean_length": 1078.5848388671875,
      "completions/mean_terminated_length": 607.478759765625,
      "completions/min_length": 184.0,
      "completions/min_terminated_length": 184.0,
      "epoch": 1.7395742198891804,
      "grad_norm": 0.3794609606266022,
      "learning_rate": 1e-06,
      "loss": -0.0508,
      "num_tokens": 110908155.0,
      "reward": 0.6272321939468384,
      "reward_std": 0.16160815954208374,
      "rewards/verify_math_reward/mean": 0.6272321343421936,
      "rewards/verify_math_reward/std": 0.4838111698627472,
      "step": 745
    },
    {
      "clip_ratio/high_max": 0.0030135468659864273,
      "clip_ratio/high_mean": 0.0011095243444287917,
      "clip_ratio/low_mean": 0.000746576579331304,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018561009055702016,
      "epoch": 1.741907261592301,
      "grad_norm": 0.25194844603538513,
      "learning_rate": 1e-06,
      "loss": -0.0512,
      "step": 746
    },
    {
      "clip_ratio/high_max": 0.003536265438015107,
      "clip_ratio/high_mean": 0.0012160383303125855,
      "clip_ratio/low_mean": 0.0007701759050178225,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019862142580677755,
      "epoch": 1.7442403032954215,
      "grad_norm": 0.2876931428909302,
      "learning_rate": 1e-06,
      "loss": -0.0513,
      "step": 747
    },
    {
      "clip_ratio/high_max": 0.003463274500973057,
      "clip_ratio/high_mean": 0.0012149306567152962,
      "clip_ratio/low_mean": 0.0010488517327758018,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002263782385853119,
      "epoch": 1.7465733449985419,
      "grad_norm": 0.21975436806678772,
      "learning_rate": 1e-06,
      "loss": -0.0515,
      "step": 748
    },
    {
      "clip_ratio/high_max": 0.0023062533073243685,
      "clip_ratio/high_mean": 0.0008066467999015003,
      "clip_ratio/low_mean": 0.0005605802763284373,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013672271052200813,
      "completions/clipped_ratio": 0.2254464285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3654.0,
      "completions/mean_length": 1421.69873046875,
      "completions/mean_terminated_length": 643.2996826171875,
      "completions/min_length": 192.0,
      "completions/min_terminated_length": 192.0,
      "epoch": 1.7489063867016623,
      "grad_norm": 0.3322184681892395,
      "learning_rate": 1e-06,
      "loss": -0.0811,
      "num_tokens": 111445621.0,
      "reward": 0.53125,
      "reward_std": 0.1724012941122055,
      "rewards/verify_math_reward/mean": 0.53125,
      "rewards/verify_math_reward/std": 0.4993011951446533,
      "step": 749
    },
    {
      "clip_ratio/high_max": 0.003089862053457182,
      "clip_ratio/high_mean": 0.0010738891160144703,
      "clip_ratio/low_mean": 0.0006956881975384022,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017695772912702523,
      "epoch": 1.7512394284047827,
      "grad_norm": 0.3031192421913147,
      "learning_rate": 1e-06,
      "loss": -0.0814,
      "step": 750
    },
    {
      "clip_ratio/high_max": 0.0025417940341867507,
      "clip_ratio/high_mean": 0.0009615219769329997,
      "clip_ratio/low_mean": 0.0009298260465584463,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001891348001663573,
      "epoch": 1.7535724701079032,
      "grad_norm": 0.21803531050682068,
      "learning_rate": 1e-06,
      "loss": -0.0816,
      "step": 751
    },
    {
      "clip_ratio/high_max": 0.0026612037072482053,
      "clip_ratio/high_mean": 0.0009760259836184559,
      "clip_ratio/low_mean": 0.001015680973068811,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001991706929402426,
      "epoch": 1.7559055118110236,
      "grad_norm": 0.224185049533844,
      "learning_rate": 1e-06,
      "loss": -0.0817,
      "step": 752
    },
    {
      "clip_ratio/high_max": 0.002623805237817578,
      "clip_ratio/high_mean": 0.0009534142445772886,
      "clip_ratio/low_mean": 0.0004241333190293517,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013775475563306827,
      "completions/clipped_ratio": 0.2053571428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4041.0,
      "completions/mean_length": 1335.7723388671875,
      "completions/mean_terminated_length": 622.455078125,
      "completions/min_length": 134.0,
      "completions/min_terminated_length": 134.0,
      "epoch": 1.758238553514144,
      "grad_norm": 0.30206793546676636,
      "learning_rate": 1e-06,
      "loss": -0.0629,
      "num_tokens": 111985497.0,
      "reward": 0.5569196939468384,
      "reward_std": 0.1587110459804535,
      "rewards/verify_math_reward/mean": 0.5569196343421936,
      "rewards/verify_math_reward/std": 0.49702703952789307,
      "step": 753
    },
    {
      "clip_ratio/high_max": 0.0035839638221659698,
      "clip_ratio/high_mean": 0.0013280071470944677,
      "clip_ratio/low_mean": 0.0007448215110343881,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020728286472149193,
      "epoch": 1.7605715952172645,
      "grad_norm": 0.2561754286289215,
      "learning_rate": 1e-06,
      "loss": -0.0632,
      "step": 754
    },
    {
      "clip_ratio/high_max": 0.003633407446614001,
      "clip_ratio/high_mean": 0.001320830335316714,
      "clip_ratio/low_mean": 0.0008764087033341639,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021972390604787506,
      "epoch": 1.762904636920385,
      "grad_norm": 0.23033028841018677,
      "learning_rate": 1e-06,
      "loss": -0.0634,
      "step": 755
    },
    {
      "clip_ratio/high_max": 0.0034799278510035947,
      "clip_ratio/high_mean": 0.0012372326200420503,
      "clip_ratio/low_mean": 0.0010334521684853826,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022706847521476448,
      "epoch": 1.7652376786235053,
      "grad_norm": 0.221846804022789,
      "learning_rate": 1e-06,
      "loss": -0.0634,
      "step": 756
    },
    {
      "clip_ratio/high_max": 0.002665332518517971,
      "clip_ratio/high_mean": 0.0010277658730046824,
      "clip_ratio/low_mean": 0.0007625566595379496,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017903225743793882,
      "completions/clipped_ratio": 0.1908482142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4020.0,
      "completions/mean_length": 1378.73779296875,
      "completions/mean_terminated_length": 737.838623046875,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 1.767570720326626,
      "grad_norm": 0.3178108036518097,
      "learning_rate": 1e-06,
      "loss": -0.0708,
      "num_tokens": 112605214.0,
      "reward": 0.5334821939468384,
      "reward_std": 0.20185355842113495,
      "rewards/verify_math_reward/mean": 0.5334821343421936,
      "rewards/verify_math_reward/std": 0.49915632605552673,
      "step": 757
    },
    {
      "clip_ratio/high_max": 0.0033235828086617403,
      "clip_ratio/high_mean": 0.0011530254741956014,
      "clip_ratio/low_mean": 0.0010216206592303934,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002174646080675302,
      "epoch": 1.7699037620297462,
      "grad_norm": 0.26831531524658203,
      "learning_rate": 1e-06,
      "loss": -0.071,
      "step": 758
    },
    {
      "clip_ratio/high_max": 0.0035036832850892097,
      "clip_ratio/high_mean": 0.0012240288524481002,
      "clip_ratio/low_mean": 0.0011186093088326743,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023426382322213612,
      "epoch": 1.7722368037328668,
      "grad_norm": 0.2287055402994156,
      "learning_rate": 1e-06,
      "loss": -0.0713,
      "step": 759
    },
    {
      "clip_ratio/high_max": 0.002821169931848999,
      "clip_ratio/high_mean": 0.0010705337044782937,
      "clip_ratio/low_mean": 0.0013855133893230231,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024560470701544546,
      "epoch": 1.774569845435987,
      "grad_norm": 0.2927832305431366,
      "learning_rate": 1e-06,
      "loss": -0.0713,
      "step": 760
    },
    {
      "clip_ratio/high_max": 0.0017815241080825217,
      "clip_ratio/high_mean": 0.0006108197103458224,
      "clip_ratio/low_mean": 0.000504789546084794,
      "clip_ratio/low_min": 1.4180373909766786e-05,
      "clip_ratio/region_mean": 0.0011156092550663743,
      "completions/clipped_ratio": 0.1339285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3427.0,
      "completions/mean_length": 1096.298095703125,
      "completions/mean_terminated_length": 632.426513671875,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 1.7769028871391077,
      "grad_norm": 0.28425121307373047,
      "learning_rate": 1e-06,
      "loss": -0.0622,
      "num_tokens": 113186329.0,
      "reward": 0.5959821939468384,
      "reward_std": 0.14324188232421875,
      "rewards/verify_math_reward/mean": 0.5959821343421936,
      "rewards/verify_math_reward/std": 0.490975022315979,
      "step": 761
    },
    {
      "clip_ratio/high_max": 0.002119680393661838,
      "clip_ratio/high_mean": 0.0007032340054138331,
      "clip_ratio/low_mean": 0.0007346917317363477,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014379257372638676,
      "epoch": 1.779235928842228,
      "grad_norm": 0.18750505149364471,
      "learning_rate": 1e-06,
      "loss": -0.0625,
      "step": 762
    },
    {
      "clip_ratio/high_max": 0.0024135289568221197,
      "clip_ratio/high_mean": 0.0008099280639726203,
      "clip_ratio/low_mean": 0.0007933280356837713,
      "clip_ratio/low_min": 2.127056177414488e-05,
      "clip_ratio/region_mean": 0.00160325612523593,
      "epoch": 1.7815689705453486,
      "grad_norm": 0.24141138792037964,
      "learning_rate": 1e-06,
      "loss": -0.0626,
      "step": 763
    },
    {
      "clip_ratio/high_max": 0.0019172006941516884,
      "clip_ratio/high_mean": 0.0007264084797498072,
      "clip_ratio/low_mean": 0.0009916465546666586,
      "clip_ratio/low_min": 2.127056177414488e-05,
      "clip_ratio/region_mean": 0.001718055016681319,
      "epoch": 1.7839020122484688,
      "grad_norm": 0.17011037468910217,
      "learning_rate": 1e-06,
      "loss": -0.0627,
      "step": 764
    },
    {
      "clip_ratio/high_max": 0.0029642568333656527,
      "clip_ratio/high_mean": 0.0011997792062174994,
      "clip_ratio/low_mean": 0.0006717288924846798,
      "clip_ratio/low_min": 1.4169122550811153e-05,
      "clip_ratio/region_mean": 0.0018715080877882428,
      "completions/clipped_ratio": 0.2087053571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3856.0,
      "completions/mean_length": 1340.204345703125,
      "completions/mean_terminated_length": 613.3582763671875,
      "completions/min_length": 128.0,
      "completions/min_terminated_length": 128.0,
      "epoch": 1.7862350539515894,
      "grad_norm": 0.3163980543613434,
      "learning_rate": 1e-06,
      "loss": -0.124,
      "num_tokens": 113704848.0,
      "reward": 0.5424107313156128,
      "reward_std": 0.20662352442741394,
      "rewards/verify_math_reward/mean": 0.5424107313156128,
      "rewards/verify_math_reward/std": 0.4984763264656067,
      "step": 765
    },
    {
      "clip_ratio/high_max": 0.003829695677268319,
      "clip_ratio/high_mean": 0.00148837465167162,
      "clip_ratio/low_mean": 0.000826685465654009,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002315060126420576,
      "epoch": 1.7885680956547099,
      "grad_norm": 0.3120717406272888,
      "learning_rate": 1e-06,
      "loss": -0.1242,
      "step": 766
    },
    {
      "clip_ratio/high_max": 0.003974874533014372,
      "clip_ratio/high_mean": 0.001538831988000311,
      "clip_ratio/low_mean": 0.0011645123013295233,
      "clip_ratio/low_min": 2.295262493134942e-05,
      "clip_ratio/region_mean": 0.0027033442820538767,
      "epoch": 1.7909011373578303,
      "grad_norm": 0.26164695620536804,
      "learning_rate": 1e-06,
      "loss": -0.1245,
      "step": 767
    },
    {
      "clip_ratio/high_max": 0.003672715203720145,
      "clip_ratio/high_mean": 0.001413390549714677,
      "clip_ratio/low_mean": 0.0012837811300414614,
      "clip_ratio/low_min": 4.620502659236081e-05,
      "clip_ratio/region_mean": 0.0026971717015840113,
      "epoch": 1.7932341790609507,
      "grad_norm": 0.2773039638996124,
      "learning_rate": 1e-06,
      "loss": -0.1245,
      "step": 768
    },
    {
      "clip_ratio/high_max": 0.002676884803804569,
      "clip_ratio/high_mean": 0.0009917986153595848,
      "clip_ratio/low_mean": 0.0005035521880927263,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014953507634345442,
      "completions/clipped_ratio": 0.2142857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3903.0,
      "completions/mean_length": 1422.67529296875,
      "completions/mean_terminated_length": 693.586669921875,
      "completions/min_length": 176.0,
      "completions/min_terminated_length": 176.0,
      "epoch": 1.7955672207640712,
      "grad_norm": 0.3097780644893646,
      "learning_rate": 1e-06,
      "loss": -0.1124,
      "num_tokens": 114285629.0,
      "reward": 0.4765625298023224,
      "reward_std": 0.16796325147151947,
      "rewards/verify_math_reward/mean": 0.4765625,
      "rewards/verify_math_reward/std": 0.49972933530807495,
      "step": 769
    },
    {
      "clip_ratio/high_max": 0.0027745560291805305,
      "clip_ratio/high_mean": 0.0010920167151198257,
      "clip_ratio/low_mean": 0.0006124778379898999,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017044945670932066,
      "epoch": 1.7979002624671916,
      "grad_norm": 0.24015967547893524,
      "learning_rate": 1e-06,
      "loss": -0.1124,
      "step": 770
    },
    {
      "clip_ratio/high_max": 0.0031230877866619267,
      "clip_ratio/high_mean": 0.0010731148413469782,
      "clip_ratio/low_mean": 0.0008133408282446908,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001886455633211881,
      "epoch": 1.800233304170312,
      "grad_norm": 0.20473533868789673,
      "learning_rate": 1e-06,
      "loss": -0.1127,
      "step": 771
    },
    {
      "clip_ratio/high_max": 0.0028612736641662195,
      "clip_ratio/high_mean": 0.0010856566805159673,
      "clip_ratio/low_mean": 0.0009701266862975899,
      "clip_ratio/low_min": 1.3994625987834297e-05,
      "clip_ratio/region_mean": 0.002055783334071748,
      "epoch": 1.8025663458734325,
      "grad_norm": 0.21435610949993134,
      "learning_rate": 1e-06,
      "loss": -0.1127,
      "step": 772
    },
    {
      "clip_ratio/high_max": 0.0022681507507513743,
      "clip_ratio/high_mean": 0.0008565066564187873,
      "clip_ratio/low_mean": 0.0005312314214052094,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013877380770281889,
      "completions/clipped_ratio": 0.1540178571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4023.0,
      "completions/mean_length": 1162.469970703125,
      "completions/mean_terminated_length": 628.3970947265625,
      "completions/min_length": 170.0,
      "completions/min_terminated_length": 170.0,
      "epoch": 1.8048993875765529,
      "grad_norm": 0.3202595114707947,
      "learning_rate": 1e-06,
      "loss": -0.0719,
      "num_tokens": 114847410.0,
      "reward": 0.5758928656578064,
      "reward_std": 0.15744182467460632,
      "rewards/verify_math_reward/mean": 0.5758928656578064,
      "rewards/verify_math_reward/std": 0.49448275566101074,
      "step": 773
    },
    {
      "clip_ratio/high_max": 0.0029461818994604982,
      "clip_ratio/high_mean": 0.0010305059331585653,
      "clip_ratio/low_mean": 0.0007527953118824371,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001783301260729786,
      "epoch": 1.8072324292796735,
      "grad_norm": 0.24247294664382935,
      "learning_rate": 1e-06,
      "loss": -0.0721,
      "step": 774
    },
    {
      "clip_ratio/high_max": 0.0029159980040276423,
      "clip_ratio/high_mean": 0.0010033698526967783,
      "clip_ratio/low_mean": 0.0007841212382118101,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017874911063699983,
      "epoch": 1.8095654709827937,
      "grad_norm": 0.19409358501434326,
      "learning_rate": 1e-06,
      "loss": -0.0722,
      "step": 775
    },
    {
      "clip_ratio/high_max": 0.0026246054694638588,
      "clip_ratio/high_mean": 0.0009365742735099047,
      "clip_ratio/low_mean": 0.0010685592246773012,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020051334613526706,
      "epoch": 1.8118985126859144,
      "grad_norm": 0.21823573112487793,
      "learning_rate": 1e-06,
      "loss": -0.0723,
      "step": 776
    },
    {
      "clip_ratio/high_max": 0.0020401036454131827,
      "clip_ratio/high_mean": 0.0008350547459485824,
      "clip_ratio/low_mean": 0.0006044626568382228,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014395173820958007,
      "completions/clipped_ratio": 0.1651785714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3060.0,
      "completions/mean_length": 1199.04248046875,
      "completions/mean_terminated_length": 625.8475952148438,
      "completions/min_length": 175.0,
      "completions/min_terminated_length": 175.0,
      "epoch": 1.8142315543890346,
      "grad_norm": 0.2843477725982666,
      "learning_rate": 1e-06,
      "loss": -0.0601,
      "num_tokens": 115399928.0,
      "reward": 0.5368303656578064,
      "reward_std": 0.1537223905324936,
      "rewards/verify_math_reward/mean": 0.5368303656578064,
      "rewards/verify_math_reward/std": 0.49892017245292664,
      "step": 777
    },
    {
      "clip_ratio/high_max": 0.002548430689785164,
      "clip_ratio/high_mean": 0.0010934012916550273,
      "clip_ratio/low_mean": 0.0008658578972244868,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019592591634136625,
      "epoch": 1.8165645960921553,
      "grad_norm": 0.24737218022346497,
      "learning_rate": 1e-06,
      "loss": -0.0603,
      "step": 778
    },
    {
      "clip_ratio/high_max": 0.002447107202897314,
      "clip_ratio/high_mean": 0.001076182052202057,
      "clip_ratio/low_mean": 0.0011107966804502212,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002186978621466551,
      "epoch": 1.8188976377952755,
      "grad_norm": 0.21529598534107208,
      "learning_rate": 1e-06,
      "loss": -0.0605,
      "step": 779
    },
    {
      "clip_ratio/high_max": 0.0026493881123315077,
      "clip_ratio/high_mean": 0.001074210747901816,
      "clip_ratio/low_mean": 0.001178176313487711,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022523870793520473,
      "epoch": 1.8212306794983961,
      "grad_norm": 0.22326554358005524,
      "learning_rate": 1e-06,
      "loss": -0.0605,
      "step": 780
    },
    {
      "clip_ratio/high_max": 0.0021476892434293404,
      "clip_ratio/high_mean": 0.0007788121292833239,
      "clip_ratio/low_mean": 0.00047618040298402775,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012549925268103834,
      "completions/clipped_ratio": 0.1584821428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3923.0,
      "completions/mean_length": 1207.443115234375,
      "completions/mean_terminated_length": 663.4442749023438,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 1.8235637212015163,
      "grad_norm": 0.27336016297340393,
      "learning_rate": 1e-06,
      "loss": -0.1015,
      "num_tokens": 115983037.0,
      "reward": 0.5870535969734192,
      "reward_std": 0.16326561570167542,
      "rewards/verify_math_reward/mean": 0.5870535969734192,
      "rewards/verify_math_reward/std": 0.49263837933540344,
      "step": 781
    },
    {
      "clip_ratio/high_max": 0.002120119155733846,
      "clip_ratio/high_mean": 0.0009483648700552294,
      "clip_ratio/low_mean": 0.000684274284139974,
      "clip_ratio/low_min": 1.6382700778194703e-05,
      "clip_ratio/region_mean": 0.0016326391778420657,
      "epoch": 1.825896762904637,
      "grad_norm": 0.19790448248386383,
      "learning_rate": 1e-06,
      "loss": -0.1017,
      "step": 782
    },
    {
      "clip_ratio/high_max": 0.0025395368284080178,
      "clip_ratio/high_mean": 0.0008975396103778621,
      "clip_ratio/low_mean": 0.0007989680589162163,
      "clip_ratio/low_min": 3.2765401556389406e-05,
      "clip_ratio/region_mean": 0.0016965076647466049,
      "epoch": 1.8282298046077574,
      "grad_norm": 0.18685676157474518,
      "learning_rate": 1e-06,
      "loss": -0.1018,
      "step": 783
    },
    {
      "clip_ratio/high_max": 0.002357502049562754,
      "clip_ratio/high_mean": 0.000903857222510851,
      "clip_ratio/low_mean": 0.0008553564457542961,
      "clip_ratio/low_min": 2.734033296292182e-05,
      "clip_ratio/region_mean": 0.001759213650075253,
      "epoch": 1.8305628463108778,
      "grad_norm": 0.22967809438705444,
      "learning_rate": 1e-06,
      "loss": -0.1018,
      "step": 784
    },
    {
      "clip_ratio/high_max": 0.0024036451359279454,
      "clip_ratio/high_mean": 0.000801718235379667,
      "clip_ratio/low_mean": 0.0007686801855015801,
      "clip_ratio/low_min": 1.05645704024937e-05,
      "clip_ratio/region_mean": 0.0015703984281572048,
      "completions/clipped_ratio": 0.2120535714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3896.0,
      "completions/mean_length": 1436.62841796875,
      "completions/mean_terminated_length": 720.9334106445312,
      "completions/min_length": 165.0,
      "completions/min_terminated_length": 165.0,
      "epoch": 1.8328958880139983,
      "grad_norm": 0.35223352909088135,
      "learning_rate": 1e-06,
      "loss": -0.0373,
      "num_tokens": 116592824.0,
      "reward": 0.4598214626312256,
      "reward_std": 0.1575952023267746,
      "rewards/verify_math_reward/mean": 0.4598214328289032,
      "rewards/verify_math_reward/std": 0.4986613988876343,
      "step": 785
    },
    {
      "clip_ratio/high_max": 0.0029588468387373723,
      "clip_ratio/high_mean": 0.0009684900815045694,
      "clip_ratio/low_mean": 0.0009765402992343297,
      "clip_ratio/low_min": 3.16937112074811e-05,
      "clip_ratio/region_mean": 0.0019450303734629415,
      "epoch": 1.8352289297171187,
      "grad_norm": 0.2443126142024994,
      "learning_rate": 1e-06,
      "loss": -0.0374,
      "step": 786
    },
    {
      "clip_ratio/high_max": 0.0035525073762983084,
      "clip_ratio/high_mean": 0.0011505296388349961,
      "clip_ratio/low_mean": 0.0012322973689151695,
      "clip_ratio/low_min": 1.05645704024937e-05,
      "clip_ratio/region_mean": 0.0023828270568628795,
      "epoch": 1.8375619714202391,
      "grad_norm": 0.2040264904499054,
      "learning_rate": 1e-06,
      "loss": -0.0377,
      "step": 787
    },
    {
      "clip_ratio/high_max": 0.002937351026048418,
      "clip_ratio/high_mean": 0.001020425286696991,
      "clip_ratio/low_mean": 0.0013868755704606883,
      "clip_ratio/low_min": 3.16937112074811e-05,
      "clip_ratio/region_mean": 0.002407300882623531,
      "epoch": 1.8398950131233596,
      "grad_norm": 0.22036604583263397,
      "learning_rate": 1e-06,
      "loss": -0.0377,
      "step": 788
    },
    {
      "clip_ratio/high_max": 0.00202498081853264,
      "clip_ratio/high_mean": 0.0006558133309226832,
      "clip_ratio/low_mean": 0.0005988032216919237,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012546165671665221,
      "completions/clipped_ratio": 0.1908482142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2963.0,
      "completions/mean_length": 1309.7935791015625,
      "completions/mean_terminated_length": 652.6331176757812,
      "completions/min_length": 153.0,
      "completions/min_terminated_length": 153.0,
      "epoch": 1.84222805482648,
      "grad_norm": 0.3399260342121124,
      "learning_rate": 1e-06,
      "loss": -0.0444,
      "num_tokens": 117161519.0,
      "reward": 0.4754464626312256,
      "reward_std": 0.13177865743637085,
      "rewards/verify_math_reward/mean": 0.4754464328289032,
      "rewards/verify_math_reward/std": 0.4996756315231323,
      "step": 789
    },
    {
      "clip_ratio/high_max": 0.002574367510533193,
      "clip_ratio/high_mean": 0.0008615906726845424,
      "clip_ratio/low_mean": 0.0006940032471902668,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001555593920784304,
      "epoch": 1.8445610965296004,
      "grad_norm": 0.22516852617263794,
      "learning_rate": 1e-06,
      "loss": -0.0446,
      "step": 790
    },
    {
      "clip_ratio/high_max": 0.0027376520592952147,
      "clip_ratio/high_mean": 0.0008576442141929874,
      "clip_ratio/low_mean": 0.0009341048044007039,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017917489458341151,
      "epoch": 1.8468941382327209,
      "grad_norm": 0.19522975385189056,
      "learning_rate": 1e-06,
      "loss": -0.0448,
      "step": 791
    },
    {
      "clip_ratio/high_max": 0.0026930035164696164,
      "clip_ratio/high_mean": 0.0008588131131546106,
      "clip_ratio/low_mean": 0.0011109570696135052,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00196977020095801,
      "epoch": 1.8492271799358413,
      "grad_norm": 0.2082953304052353,
      "learning_rate": 1e-06,
      "loss": -0.0449,
      "step": 792
    },
    {
      "clip_ratio/high_max": 0.0021975159979774617,
      "clip_ratio/high_mean": 0.0010656295926310122,
      "clip_ratio/low_mean": 0.0006331684171527741,
      "clip_ratio/low_min": 2.886836045945529e-05,
      "clip_ratio/region_mean": 0.0016987979979603551,
      "completions/clipped_ratio": 0.1662946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4013.0,
      "completions/mean_length": 1188.33935546875,
      "completions/mean_terminated_length": 608.3641357421875,
      "completions/min_length": 203.0,
      "completions/min_terminated_length": 203.0,
      "epoch": 1.851560221638962,
      "grad_norm": 0.3365528881549835,
      "learning_rate": 1e-06,
      "loss": -0.0678,
      "num_tokens": 117696535.0,
      "reward": 0.5558035969734192,
      "reward_std": 0.19989821314811707,
      "rewards/verify_math_reward/mean": 0.5558035969734192,
      "rewards/verify_math_reward/std": 0.49715372920036316,
      "step": 793
    },
    {
      "clip_ratio/high_max": 0.003083037110627629,
      "clip_ratio/high_mean": 0.0013537398081098218,
      "clip_ratio/low_mean": 0.0009749114942678716,
      "clip_ratio/low_min": 1.9500779671943747e-05,
      "clip_ratio/region_mean": 0.0023286512951017357,
      "epoch": 1.8538932633420822,
      "grad_norm": 0.32683202624320984,
      "learning_rate": 1e-06,
      "loss": -0.0681,
      "step": 794
    },
    {
      "clip_ratio/high_max": 0.0028669562161667272,
      "clip_ratio/high_mean": 0.0013689995339518646,
      "clip_ratio/low_mean": 0.0012241506410646252,
      "clip_ratio/low_min": 1.9500779671943747e-05,
      "clip_ratio/region_mean": 0.002593150180473458,
      "epoch": 1.8562263050452028,
      "grad_norm": 0.25984010100364685,
      "learning_rate": 1e-06,
      "loss": -0.0683,
      "step": 795
    },
    {
      "clip_ratio/high_max": 0.003130484779831022,
      "clip_ratio/high_mean": 0.0012664634305110667,
      "clip_ratio/low_mean": 0.0014365567985805683,
      "clip_ratio/low_min": 1.9500779671943747e-05,
      "clip_ratio/region_mean": 0.0027030202036257833,
      "epoch": 1.858559346748323,
      "grad_norm": 0.25461524724960327,
      "learning_rate": 1e-06,
      "loss": -0.0684,
      "step": 796
    },
    {
      "clip_ratio/high_max": 0.002819614404870663,
      "clip_ratio/high_mean": 0.0010433404531795532,
      "clip_ratio/low_mean": 0.000562692410312593,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016060328634921461,
      "completions/clipped_ratio": 0.1640625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3896.0,
      "completions/mean_length": 1210.8660888671875,
      "completions/mean_terminated_length": 644.6248779296875,
      "completions/min_length": 143.0,
      "completions/min_terminated_length": 143.0,
      "epoch": 1.8608923884514437,
      "grad_norm": 0.33243247866630554,
      "learning_rate": 1e-06,
      "loss": -0.0822,
      "num_tokens": 118270751.0,
      "reward": 0.5625,
      "reward_std": 0.18307287991046906,
      "rewards/verify_math_reward/mean": 0.5625,
      "rewards/verify_math_reward/std": 0.49635544419288635,
      "step": 797
    },
    {
      "clip_ratio/high_max": 0.003020628049853258,
      "clip_ratio/high_mean": 0.0011996221910521854,
      "clip_ratio/low_mean": 0.0007609488329762826,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001960570953087881,
      "epoch": 1.8632254301545639,
      "grad_norm": 0.2611001431941986,
      "learning_rate": 1e-06,
      "loss": -0.0824,
      "step": 798
    },
    {
      "clip_ratio/high_max": 0.0030056221075938083,
      "clip_ratio/high_mean": 0.0012223275007272605,
      "clip_ratio/low_mean": 0.0008960413233580766,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021183688295423053,
      "epoch": 1.8655584718576845,
      "grad_norm": 0.22762948274612427,
      "learning_rate": 1e-06,
      "loss": -0.0826,
      "step": 799
    },
    {
      "clip_ratio/high_max": 0.0030369122760021128,
      "clip_ratio/high_mean": 0.0011457976415840676,
      "clip_ratio/low_mean": 0.0011843351348943543,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023301327382796444,
      "epoch": 1.8678915135608047,
      "grad_norm": 0.22786091268062592,
      "learning_rate": 1e-06,
      "loss": -0.0827,
      "step": 800
    },
    {
      "clip_ratio/high_max": 0.002827204138156958,
      "clip_ratio/high_mean": 0.0011934108260902576,
      "clip_ratio/low_mean": 0.0006371737217705231,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018305845296708867,
      "completions/clipped_ratio": 0.1618303571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3696.0,
      "completions/mean_length": 1183.9476318359375,
      "completions/mean_terminated_length": 621.7003784179688,
      "completions/min_length": 160.0,
      "completions/min_terminated_length": 160.0,
      "epoch": 1.8702245552639254,
      "grad_norm": 0.3000062108039856,
      "learning_rate": 1e-06,
      "loss": -0.0767,
      "num_tokens": 118818264.0,
      "reward": 0.574776828289032,
      "reward_std": 0.1907682567834854,
      "rewards/verify_math_reward/mean": 0.5747767686843872,
      "rewards/verify_math_reward/std": 0.49465295672416687,
      "step": 801
    },
    {
      "clip_ratio/high_max": 0.0032071580208139494,
      "clip_ratio/high_mean": 0.001341705417871708,
      "clip_ratio/low_mean": 0.0009606630992493592,
      "clip_ratio/low_min": 4.161653941991972e-05,
      "clip_ratio/region_mean": 0.002302368484379258,
      "epoch": 1.8725575969670458,
      "grad_norm": 0.2800460457801819,
      "learning_rate": 1e-06,
      "loss": -0.077,
      "step": 802
    },
    {
      "clip_ratio/high_max": 0.0035844488375005312,
      "clip_ratio/high_mean": 0.001533605121949222,
      "clip_ratio/low_mean": 0.0011396265435905661,
      "clip_ratio/low_min": 1.4282450138125569e-05,
      "clip_ratio/region_mean": 0.002673231625522021,
      "epoch": 1.8748906386701663,
      "grad_norm": 0.25125962495803833,
      "learning_rate": 1e-06,
      "loss": -0.0772,
      "step": 803
    },
    {
      "clip_ratio/high_max": 0.0033888213947648183,
      "clip_ratio/high_mean": 0.0013532129996747244,
      "clip_ratio/low_mean": 0.001306935651882668,
      "clip_ratio/low_min": 4.2847350414376706e-05,
      "clip_ratio/region_mean": 0.0026601486752042547,
      "epoch": 1.8772236803732867,
      "grad_norm": 0.2604830265045166,
      "learning_rate": 1e-06,
      "loss": -0.0773,
      "step": 804
    },
    {
      "clip_ratio/high_max": 0.002367687178775668,
      "clip_ratio/high_mean": 0.000912133504243684,
      "clip_ratio/low_mean": 0.0005989227101963479,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015110562089830637,
      "completions/clipped_ratio": 0.1741071428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3643.0,
      "completions/mean_length": 1275.75341796875,
      "completions/mean_terminated_length": 681.2149047851562,
      "completions/min_length": 152.0,
      "completions/min_terminated_length": 152.0,
      "epoch": 1.8795567220764071,
      "grad_norm": 0.28715670108795166,
      "learning_rate": 1e-06,
      "loss": -0.0577,
      "num_tokens": 119402443.0,
      "reward": 0.5457589626312256,
      "reward_std": 0.16570734977722168,
      "rewards/verify_math_reward/mean": 0.5457589030265808,
      "rewards/verify_math_reward/std": 0.4981798231601715,
      "step": 805
    },
    {
      "clip_ratio/high_max": 0.0029728993104072288,
      "clip_ratio/high_mean": 0.001111834862967953,
      "clip_ratio/low_mean": 0.0007670903178222943,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018789252208080143,
      "epoch": 1.8818897637795275,
      "grad_norm": 0.21860335767269135,
      "learning_rate": 1e-06,
      "loss": -0.058,
      "step": 806
    },
    {
      "clip_ratio/high_max": 0.0027008577017113566,
      "clip_ratio/high_mean": 0.0010888749711739365,
      "clip_ratio/low_mean": 0.0010045799008366885,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002093454902933445,
      "epoch": 1.884222805482648,
      "grad_norm": 0.21858292818069458,
      "learning_rate": 1e-06,
      "loss": -0.058,
      "step": 807
    },
    {
      "clip_ratio/high_max": 0.0027396525474614464,
      "clip_ratio/high_mean": 0.0010619849563227035,
      "clip_ratio/low_mean": 0.0010708690479077632,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002132853987859562,
      "epoch": 1.8865558471857684,
      "grad_norm": 0.20753753185272217,
      "learning_rate": 1e-06,
      "loss": -0.0581,
      "step": 808
    },
    {
      "clip_ratio/high_max": 0.002918562593549723,
      "clip_ratio/high_mean": 0.0010963252407236723,
      "clip_ratio/low_mean": 0.000893263903890329,
      "clip_ratio/low_min": 1.9797276763711125e-05,
      "clip_ratio/region_mean": 0.0019895891564374324,
      "completions/clipped_ratio": 0.2053571428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3374.0,
      "completions/mean_length": 1344.7310791015625,
      "completions/mean_terminated_length": 633.7289428710938,
      "completions/min_length": 138.0,
      "completions/min_terminated_length": 138.0,
      "epoch": 1.8888888888888888,
      "grad_norm": 0.34875476360321045,
      "learning_rate": 1e-06,
      "loss": -0.0851,
      "num_tokens": 119945450.0,
      "reward": 0.4888392984867096,
      "reward_std": 0.19268713891506195,
      "rewards/verify_math_reward/mean": 0.4888392984867096,
      "rewards/verify_math_reward/std": 0.5001546144485474,
      "step": 809
    },
    {
      "clip_ratio/high_max": 0.0038939863807172514,
      "clip_ratio/high_mean": 0.0013524092973966617,
      "clip_ratio/low_mean": 0.0010970616106078523,
      "clip_ratio/low_min": 3.1151354050962254e-05,
      "clip_ratio/region_mean": 0.0024494708195561543,
      "epoch": 1.8912219305920095,
      "grad_norm": 0.3230593204498291,
      "learning_rate": 1e-06,
      "loss": -0.0852,
      "step": 810
    },
    {
      "clip_ratio/high_max": 0.0036844285641564056,
      "clip_ratio/high_mean": 0.0014590889877581503,
      "clip_ratio/low_mean": 0.0015696788232162362,
      "clip_ratio/low_min": 1.9797276763711125e-05,
      "clip_ratio/region_mean": 0.003028767809155397,
      "epoch": 1.8935549722951297,
      "grad_norm": 0.2665470540523529,
      "learning_rate": 1e-06,
      "loss": -0.0857,
      "step": 811
    },
    {
      "clip_ratio/high_max": 0.0036328440000943374,
      "clip_ratio/high_mean": 0.0013039033910899889,
      "clip_ratio/low_mean": 0.0017443379783799173,
      "clip_ratio/low_min": 2.9695913326577283e-05,
      "clip_ratio/region_mean": 0.003048241422220599,
      "epoch": 1.8958880139982504,
      "grad_norm": 0.25983041524887085,
      "learning_rate": 1e-06,
      "loss": -0.0857,
      "step": 812
    },
    {
      "clip_ratio/high_max": 0.0025541995419189334,
      "clip_ratio/high_mean": 0.0010109552877111128,
      "clip_ratio/low_mean": 0.0006645241810474545,
      "clip_ratio/low_min": 1.4460897546086926e-05,
      "clip_ratio/region_mean": 0.0016754794851294719,
      "completions/clipped_ratio": 0.15625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2481.0,
      "completions/mean_length": 1209.9788818359375,
      "completions/mean_terminated_length": 675.5303955078125,
      "completions/min_length": 150.0,
      "completions/min_terminated_length": 150.0,
      "epoch": 1.8982210557013706,
      "grad_norm": 0.2932416796684265,
      "learning_rate": 1e-06,
      "loss": -0.0607,
      "num_tokens": 120553103.0,
      "reward": 0.5502232313156128,
      "reward_std": 0.15782341361045837,
      "rewards/verify_math_reward/mean": 0.5502232313156128,
      "rewards/verify_math_reward/std": 0.49774909019470215,
      "step": 813
    },
    {
      "clip_ratio/high_max": 0.003172433156578336,
      "clip_ratio/high_mean": 0.0012148804053140339,
      "clip_ratio/low_mean": 0.0009071936674445169,
      "clip_ratio/low_min": 2.6595744202495553e-05,
      "clip_ratio/region_mean": 0.002122074060025625,
      "epoch": 1.9005540974044912,
      "grad_norm": 0.23612117767333984,
      "learning_rate": 1e-06,
      "loss": -0.0609,
      "step": 814
    },
    {
      "clip_ratio/high_max": 0.0029380835840129294,
      "clip_ratio/high_mean": 0.001169846629636595,
      "clip_ratio/low_mean": 0.0010059440137411002,
      "clip_ratio/low_min": 2.4418832254013978e-05,
      "clip_ratio/region_mean": 0.0021757906433776952,
      "epoch": 1.9028871391076114,
      "grad_norm": 0.2363535314798355,
      "learning_rate": 1e-06,
      "loss": -0.061,
      "step": 815
    },
    {
      "clip_ratio/high_max": 0.0032878915008041076,
      "clip_ratio/high_mean": 0.0012400051637087017,
      "clip_ratio/low_mean": 0.0012415621458785608,
      "clip_ratio/low_min": 2.8921795092173852e-05,
      "clip_ratio/region_mean": 0.0024815672513796017,
      "epoch": 1.905220180810732,
      "grad_norm": 0.22312773764133453,
      "learning_rate": 1e-06,
      "loss": -0.0612,
      "step": 816
    },
    {
      "clip_ratio/high_max": 0.0027170163157279603,
      "clip_ratio/high_mean": 0.0010297121934854658,
      "clip_ratio/low_mean": 0.0006148753263914841,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016445875189674553,
      "completions/clipped_ratio": 0.1618303571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3367.0,
      "completions/mean_length": 1220.407470703125,
      "completions/mean_terminated_length": 665.19970703125,
      "completions/min_length": 130.0,
      "completions/min_terminated_length": 130.0,
      "epoch": 1.9075532225138523,
      "grad_norm": 0.30903011560440063,
      "learning_rate": 1e-06,
      "loss": -0.078,
      "num_tokens": 121138484.0,
      "reward": 0.5401785969734192,
      "reward_std": 0.17543968558311462,
      "rewards/verify_math_reward/mean": 0.5401785969734192,
      "rewards/verify_math_reward/std": 0.49866142868995667,
      "step": 817
    },
    {
      "clip_ratio/high_max": 0.0035944295013905503,
      "clip_ratio/high_mean": 0.0012570860781124793,
      "clip_ratio/low_mean": 0.0008134864974636002,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002070572561933659,
      "epoch": 1.909886264216973,
      "grad_norm": 0.2809344232082367,
      "learning_rate": 1e-06,
      "loss": -0.0782,
      "step": 818
    },
    {
      "clip_ratio/high_max": 0.0034634525945875794,
      "clip_ratio/high_mean": 0.0012651153083425015,
      "clip_ratio/low_mean": 0.0009896510273392778,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002254766382975504,
      "epoch": 1.9122193059200934,
      "grad_norm": 0.2349781095981598,
      "learning_rate": 1e-06,
      "loss": -0.0785,
      "step": 819
    },
    {
      "clip_ratio/high_max": 0.0031610069345333613,
      "clip_ratio/high_mean": 0.0012248718412593007,
      "clip_ratio/low_mean": 0.0012135470769862877,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002438418916426599,
      "epoch": 1.9145523476232138,
      "grad_norm": 0.24953652918338776,
      "learning_rate": 1e-06,
      "loss": -0.0786,
      "step": 820
    },
    {
      "clip_ratio/high_max": 0.0028672609259956516,
      "clip_ratio/high_mean": 0.0011765829149226192,
      "clip_ratio/low_mean": 0.0007742787074676016,
      "clip_ratio/low_min": 1.4405900401470717e-05,
      "clip_ratio/region_mean": 0.0019508616242092103,
      "completions/clipped_ratio": 0.1674107142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3906.0,
      "completions/mean_length": 1300.1976318359375,
      "completions/mean_terminated_length": 738.0388793945312,
      "completions/min_length": 147.0,
      "completions/min_terminated_length": 147.0,
      "epoch": 1.9168853893263342,
      "grad_norm": 0.30159497261047363,
      "learning_rate": 1e-06,
      "loss": -0.0721,
      "num_tokens": 121768685.0,
      "reward": 0.5089285969734192,
      "reward_std": 0.21560588479042053,
      "rewards/verify_math_reward/mean": 0.5089285969734192,
      "rewards/verify_math_reward/std": 0.5001994967460632,
      "step": 821
    },
    {
      "clip_ratio/high_max": 0.002927249559434131,
      "clip_ratio/high_mean": 0.0013193048471293878,
      "clip_ratio/low_mean": 0.0009674110351625131,
      "clip_ratio/low_min": 9.32696639210917e-06,
      "clip_ratio/region_mean": 0.00228671585500706,
      "epoch": 1.9192184310294547,
      "grad_norm": 0.2845073342323303,
      "learning_rate": 1e-06,
      "loss": -0.0723,
      "step": 822
    },
    {
      "clip_ratio/high_max": 0.0033219540782738477,
      "clip_ratio/high_mean": 0.0013041101992712356,
      "clip_ratio/low_mean": 0.0011976933037658455,
      "clip_ratio/low_min": 2.798089917632751e-05,
      "clip_ratio/region_mean": 0.002501803515770007,
      "epoch": 1.921551472732575,
      "grad_norm": 0.23670266568660736,
      "learning_rate": 1e-06,
      "loss": -0.0726,
      "step": 823
    },
    {
      "clip_ratio/high_max": 0.003028897750482429,
      "clip_ratio/high_mean": 0.0012568753845698666,
      "clip_ratio/low_mean": 0.001436265232769074,
      "clip_ratio/low_min": 2.798089917632751e-05,
      "clip_ratio/region_mean": 0.002693140646442771,
      "epoch": 1.9238845144356955,
      "grad_norm": 0.24771694839000702,
      "learning_rate": 1e-06,
      "loss": -0.0727,
      "step": 824
    },
    {
      "clip_ratio/high_max": 0.002573345904238522,
      "clip_ratio/high_mean": 0.0009996045755542582,
      "clip_ratio/low_mean": 0.0005916766449445277,
      "clip_ratio/low_min": 1.4654161532234866e-05,
      "clip_ratio/region_mean": 0.0015912812450551428,
      "completions/clipped_ratio": 0.1863839285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3848.0,
      "completions/mean_length": 1320.219970703125,
      "completions/mean_terminated_length": 684.341552734375,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 1.926217556138816,
      "grad_norm": 0.3412525951862335,
      "learning_rate": 1e-06,
      "loss": -0.0693,
      "num_tokens": 122361018.0,
      "reward": 0.5703125,
      "reward_std": 0.16457942128181458,
      "rewards/verify_math_reward/mean": 0.5703125,
      "rewards/verify_math_reward/std": 0.49530795216560364,
      "step": 825
    },
    {
      "clip_ratio/high_max": 0.0028175055631436408,
      "clip_ratio/high_mean": 0.0011196003197255777,
      "clip_ratio/low_mean": 0.0007795236651872983,
      "clip_ratio/low_min": 2.9308323064469732e-05,
      "clip_ratio/region_mean": 0.0018991239558090456,
      "epoch": 1.9285505978419364,
      "grad_norm": 0.3357429504394531,
      "learning_rate": 1e-06,
      "loss": -0.0695,
      "step": 826
    },
    {
      "clip_ratio/high_max": 0.002739876370469574,
      "clip_ratio/high_mean": 0.0010583981857053004,
      "clip_ratio/low_mean": 0.0008846381160765304,
      "clip_ratio/low_min": 3.643252784968354e-05,
      "clip_ratio/region_mean": 0.0019430363063293044,
      "epoch": 1.9308836395450568,
      "grad_norm": 0.22380100190639496,
      "learning_rate": 1e-06,
      "loss": -0.0697,
      "step": 827
    },
    {
      "clip_ratio/high_max": 0.002658689318195684,
      "clip_ratio/high_mean": 0.001039948532707058,
      "clip_ratio/low_mean": 0.001099562708986923,
      "clip_ratio/low_min": 7.147962605813518e-05,
      "clip_ratio/region_mean": 0.0021395112489699386,
      "epoch": 1.9332166812481772,
      "grad_norm": 0.22586019337177277,
      "learning_rate": 1e-06,
      "loss": -0.0698,
      "step": 828
    },
    {
      "clip_ratio/high_max": 0.002943986553873401,
      "clip_ratio/high_mean": 0.0011864731186506106,
      "clip_ratio/low_mean": 0.0005792369875052827,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001765710097970441,
      "completions/clipped_ratio": 0.1930803571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2665.0,
      "completions/mean_length": 1342.927490234375,
      "completions/mean_terminated_length": 684.170166015625,
      "completions/min_length": 192.0,
      "completions/min_terminated_length": 192.0,
      "epoch": 1.935549722951298,
      "grad_norm": 0.3448341190814972,
      "learning_rate": 1e-06,
      "loss": -0.1174,
      "num_tokens": 122946665.0,
      "reward": 0.5301339626312256,
      "reward_std": 0.21578949689865112,
      "rewards/verify_math_reward/mean": 0.5301339030265808,
      "rewards/verify_math_reward/std": 0.49936985969543457,
      "step": 829
    },
    {
      "clip_ratio/high_max": 0.003517591430863831,
      "clip_ratio/high_mean": 0.001492794162913924,
      "clip_ratio/low_mean": 0.0008221051957661984,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002314899320481345,
      "epoch": 1.937882764654418,
      "grad_norm": 0.2709249258041382,
      "learning_rate": 1e-06,
      "loss": -0.1177,
      "step": 830
    },
    {
      "clip_ratio/high_max": 0.0034801322108251043,
      "clip_ratio/high_mean": 0.0015445070421264973,
      "clip_ratio/low_mean": 0.0010751401532616,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026196471881121397,
      "epoch": 1.9402158063575388,
      "grad_norm": 0.22620545327663422,
      "learning_rate": 1e-06,
      "loss": -0.118,
      "step": 831
    },
    {
      "clip_ratio/high_max": 0.0033398918312741444,
      "clip_ratio/high_mean": 0.0014138497790554538,
      "clip_ratio/low_mean": 0.0012099658979423111,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026238156788167544,
      "epoch": 1.942548848060659,
      "grad_norm": 0.2483496516942978,
      "learning_rate": 1e-06,
      "loss": -0.118,
      "step": 832
    },
    {
      "clip_ratio/high_max": 0.0026773650461109355,
      "clip_ratio/high_mean": 0.0010678252874640748,
      "clip_ratio/low_mean": 0.00047232585620804457,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015401511409436353,
      "completions/clipped_ratio": 0.125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3823.0,
      "completions/mean_length": 1023.07373046875,
      "completions/mean_terminated_length": 584.0841674804688,
      "completions/min_length": 137.0,
      "completions/min_terminated_length": 137.0,
      "epoch": 1.9448818897637796,
      "grad_norm": 0.35285845398902893,
      "learning_rate": 1e-06,
      "loss": -0.0493,
      "num_tokens": 123489811.0,
      "reward": 0.6495535969734192,
      "reward_std": 0.18645039200782776,
      "rewards/verify_math_reward/mean": 0.6495535969734192,
      "rewards/verify_math_reward/std": 0.477376252412796,
      "step": 833
    },
    {
      "clip_ratio/high_max": 0.0034253023914061487,
      "clip_ratio/high_mean": 0.0013965224497951567,
      "clip_ratio/low_mean": 0.0007195341222541174,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021160565956961364,
      "epoch": 1.9472149314668998,
      "grad_norm": 0.2906281352043152,
      "learning_rate": 1e-06,
      "loss": -0.0498,
      "step": 834
    },
    {
      "clip_ratio/high_max": 0.0034584589520818554,
      "clip_ratio/high_mean": 0.001384539864375256,
      "clip_ratio/low_mean": 0.0009951573447324336,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023796972673153505,
      "epoch": 1.9495479731700205,
      "grad_norm": 0.22655533254146576,
      "learning_rate": 1e-06,
      "loss": -0.05,
      "step": 835
    },
    {
      "clip_ratio/high_max": 0.002966634900076315,
      "clip_ratio/high_mean": 0.0012003569754597265,
      "clip_ratio/low_mean": 0.0012030265734210843,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00240338352159597,
      "epoch": 1.9518810148731407,
      "grad_norm": 0.25467953085899353,
      "learning_rate": 1e-06,
      "loss": -0.05,
      "step": 836
    },
    {
      "clip_ratio/high_max": 0.0023573459111503325,
      "clip_ratio/high_mean": 0.0008272162631328683,
      "clip_ratio/low_mean": 0.0009178584023175063,
      "clip_ratio/low_min": 0.0001526445712443092,
      "clip_ratio/region_mean": 0.0017450746599934064,
      "completions/clipped_ratio": 0.1607142857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3511.0,
      "completions/mean_length": 1196.7410888671875,
      "completions/mean_terminated_length": 641.5637817382812,
      "completions/min_length": 142.0,
      "completions/min_terminated_length": 142.0,
      "epoch": 1.9542140565762613,
      "grad_norm": 0.3911133110523224,
      "learning_rate": 1e-06,
      "loss": -0.0449,
      "num_tokens": 124054891.0,
      "reward": 0.6004464626312256,
      "reward_std": 0.1792774647474289,
      "rewards/verify_math_reward/mean": 0.6004464030265808,
      "rewards/verify_math_reward/std": 0.49008017778396606,
      "step": 837
    },
    {
      "clip_ratio/high_max": 0.0034634961775736883,
      "clip_ratio/high_mean": 0.0011457822802185547,
      "clip_ratio/low_mean": 0.0012875593711214606,
      "clip_ratio/low_min": 9.394089011038886e-05,
      "clip_ratio/region_mean": 0.002433341673167888,
      "epoch": 1.9565470982793818,
      "grad_norm": 0.32581472396850586,
      "learning_rate": 1e-06,
      "loss": -0.0454,
      "step": 838
    },
    {
      "clip_ratio/high_max": 0.0031172664603218436,
      "clip_ratio/high_mean": 0.0010698406240408076,
      "clip_ratio/low_mean": 0.001597761751327198,
      "clip_ratio/low_min": 0.00014222434037947096,
      "clip_ratio/region_mean": 0.0026676024426706135,
      "epoch": 1.9588801399825022,
      "grad_norm": 0.2666594386100769,
      "learning_rate": 1e-06,
      "loss": -0.0456,
      "step": 839
    },
    {
      "clip_ratio/high_max": 0.0027453268921817653,
      "clip_ratio/high_mean": 0.0009260922197427135,
      "clip_ratio/low_mean": 0.002023709843342658,
      "clip_ratio/low_min": 0.00011088754854426952,
      "clip_ratio/region_mean": 0.0029498021976905875,
      "epoch": 1.9612131816856226,
      "grad_norm": 0.2504991292953491,
      "learning_rate": 1e-06,
      "loss": -0.0457,
      "step": 840
    },
    {
      "clip_ratio/high_max": 0.0022448038835136686,
      "clip_ratio/high_mean": 0.0008656120189698413,
      "clip_ratio/low_mean": 0.0005494064107551822,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014150184106256347,
      "completions/clipped_ratio": 0.1383928571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3817.0,
      "completions/mean_length": 1129.015625,
      "completions/mean_terminated_length": 652.453369140625,
      "completions/min_length": 119.0,
      "completions/min_terminated_length": 119.0,
      "epoch": 1.963546223388743,
      "grad_norm": 0.2795056998729706,
      "learning_rate": 1e-06,
      "loss": -0.0576,
      "num_tokens": 124648505.0,
      "reward": 0.578125,
      "reward_std": 0.1662386953830719,
      "rewards/verify_math_reward/mean": 0.578125,
      "rewards/verify_math_reward/std": 0.4941346049308777,
      "step": 841
    },
    {
      "clip_ratio/high_max": 0.0024054357418208383,
      "clip_ratio/high_mean": 0.0010389593844593037,
      "clip_ratio/low_mean": 0.000745671023196337,
      "clip_ratio/low_min": 1.4022884897713084e-05,
      "clip_ratio/region_mean": 0.0017846304181148298,
      "epoch": 1.9658792650918635,
      "grad_norm": 0.22816675901412964,
      "learning_rate": 1e-06,
      "loss": -0.0577,
      "step": 842
    },
    {
      "clip_ratio/high_max": 0.0024598441887064837,
      "clip_ratio/high_mean": 0.001023146014631493,
      "clip_ratio/low_mean": 0.0008784550973359728,
      "clip_ratio/low_min": 1.4022884897713084e-05,
      "clip_ratio/region_mean": 0.0019016010846826248,
      "epoch": 1.968212306794984,
      "grad_norm": 0.21649332344532013,
      "learning_rate": 1e-06,
      "loss": -0.0579,
      "step": 843
    },
    {
      "clip_ratio/high_max": 0.0024888858024496585,
      "clip_ratio/high_mean": 0.0009545433949824655,
      "clip_ratio/low_mean": 0.0010246585416098242,
      "clip_ratio/low_min": 1.4022884897713084e-05,
      "clip_ratio/region_mean": 0.0019792019520536996,
      "epoch": 1.9705453484981044,
      "grad_norm": 0.2124703973531723,
      "learning_rate": 1e-06,
      "loss": -0.058,
      "step": 844
    },
    {
      "clip_ratio/high_max": 0.002497881134331692,
      "clip_ratio/high_mean": 0.0009816796955419704,
      "clip_ratio/low_mean": 0.00041495259642942983,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013966322985652369,
      "completions/clipped_ratio": 0.171875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3616.0,
      "completions/mean_length": 1204.2154541015625,
      "completions/mean_terminated_length": 604.03369140625,
      "completions/min_length": 140.0,
      "completions/min_terminated_length": 140.0,
      "epoch": 1.9728783902012248,
      "grad_norm": 0.33216366171836853,
      "learning_rate": 1e-06,
      "loss": -0.1035,
      "num_tokens": 125181234.0,
      "reward": 0.5948660969734192,
      "reward_std": 0.16175968945026398,
      "rewards/verify_math_reward/mean": 0.5948660969734192,
      "rewards/verify_math_reward/std": 0.49119213223457336,
      "step": 845
    },
    {
      "clip_ratio/high_max": 0.003344422126247082,
      "clip_ratio/high_mean": 0.001384327799314633,
      "clip_ratio/low_mean": 0.0007072462321957573,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020915740460623056,
      "epoch": 1.9752114319043454,
      "grad_norm": 0.2383900135755539,
      "learning_rate": 1e-06,
      "loss": -0.1038,
      "step": 846
    },
    {
      "clip_ratio/high_max": 0.0031971252901712433,
      "clip_ratio/high_mean": 0.0012845905948779546,
      "clip_ratio/low_mean": 0.0008054647610151733,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020900553645333275,
      "epoch": 1.9775444736074657,
      "grad_norm": 0.21354363858699799,
      "learning_rate": 1e-06,
      "loss": -0.104,
      "step": 847
    },
    {
      "clip_ratio/high_max": 0.003370238082425203,
      "clip_ratio/high_mean": 0.0012448482084437273,
      "clip_ratio/low_mean": 0.0009382779990119161,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002183126227464527,
      "epoch": 1.9798775153105863,
      "grad_norm": 0.2377062439918518,
      "learning_rate": 1e-06,
      "loss": -0.104,
      "step": 848
    },
    {
      "clip_ratio/high_max": 0.0023006051560514607,
      "clip_ratio/high_mean": 0.0007807049223629292,
      "clip_ratio/low_mean": 0.0006758223921679019,
      "clip_ratio/low_min": 1.125720427808119e-05,
      "clip_ratio/region_mean": 0.0014565272940672003,
      "completions/clipped_ratio": 0.1540178571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3104.0,
      "completions/mean_length": 1125.243408203125,
      "completions/mean_terminated_length": 584.3931884765625,
      "completions/min_length": 146.0,
      "completions/min_terminated_length": 146.0,
      "epoch": 1.9822105570137065,
      "grad_norm": 0.3098432719707489,
      "learning_rate": 1e-06,
      "loss": -0.0745,
      "num_tokens": 125716956.0,
      "reward": 0.5602678656578064,
      "reward_std": 0.1442854106426239,
      "rewards/verify_math_reward/mean": 0.5602678656578064,
      "rewards/verify_math_reward/std": 0.4966317415237427,
      "step": 849
    },
    {
      "clip_ratio/high_max": 0.002714711728913244,
      "clip_ratio/high_mean": 0.0009240049912477843,
      "clip_ratio/low_mean": 0.0010237854571641947,
      "clip_ratio/low_min": 1.125720427808119e-05,
      "clip_ratio/region_mean": 0.001947790471604094,
      "epoch": 1.9845435987168272,
      "grad_norm": 0.3019099533557892,
      "learning_rate": 1e-06,
      "loss": -0.0748,
      "step": 850
    },
    {
      "clip_ratio/high_max": 0.0033190110116265714,
      "clip_ratio/high_mean": 0.0011018242767022457,
      "clip_ratio/low_mean": 0.0011715921486938896,
      "clip_ratio/low_min": 2.471332481945865e-05,
      "clip_ratio/region_mean": 0.0022734164231223986,
      "epoch": 1.9868766404199474,
      "grad_norm": 0.26513129472732544,
      "learning_rate": 1e-06,
      "loss": -0.075,
      "step": 851
    },
    {
      "clip_ratio/high_max": 0.0026627587794791907,
      "clip_ratio/high_mean": 0.000974896540355985,
      "clip_ratio/low_mean": 0.0013590649268735433,
      "clip_ratio/low_min": 1.125720427808119e-05,
      "clip_ratio/region_mean": 0.0023339615072472952,
      "epoch": 1.989209682123068,
      "grad_norm": 0.2309122234582901,
      "learning_rate": 1e-06,
      "loss": -0.075,
      "step": 852
    },
    {
      "clip_ratio/high_max": 0.0028166246265755035,
      "clip_ratio/high_mean": 0.0010274419109919108,
      "clip_ratio/low_mean": 0.0005627615837511257,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015902034610917326,
      "completions/clipped_ratio": 0.1607142857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3683.0,
      "completions/mean_length": 1177.8560791015625,
      "completions/mean_terminated_length": 619.0625,
      "completions/min_length": 164.0,
      "completions/min_terminated_length": 164.0,
      "epoch": 1.9915427238261882,
      "grad_norm": 0.35362178087234497,
      "learning_rate": 1e-06,
      "loss": -0.0613,
      "num_tokens": 126266563.0,
      "reward": 0.606026828289032,
      "reward_std": 0.14508774876594543,
      "rewards/verify_math_reward/mean": 0.6060267686843872,
      "rewards/verify_math_reward/std": 0.48890194296836853,
      "step": 853
    },
    {
      "clip_ratio/high_max": 0.0031189397268462926,
      "clip_ratio/high_mean": 0.0012358821859379532,
      "clip_ratio/low_mean": 0.0008159829521900974,
      "clip_ratio/low_min": 3.2946758437901735e-05,
      "clip_ratio/region_mean": 0.0020518651654128917,
      "epoch": 1.993875765529309,
      "grad_norm": 0.2607204020023346,
      "learning_rate": 1e-06,
      "loss": -0.0616,
      "step": 854
    },
    {
      "clip_ratio/high_max": 0.003127547497570049,
      "clip_ratio/high_mean": 0.001196386856463505,
      "clip_ratio/low_mean": 0.000987397224889719,
      "clip_ratio/low_min": 4.9498416046844795e-05,
      "clip_ratio/region_mean": 0.0021837840831722133,
      "epoch": 1.9962088072324293,
      "grad_norm": 0.23352815210819244,
      "learning_rate": 1e-06,
      "loss": -0.0618,
      "step": 855
    },
    {
      "clip_ratio/high_max": 0.003170761585352011,
      "clip_ratio/high_mean": 0.0011412914191168966,
      "clip_ratio/low_mean": 0.0012092185934307054,
      "clip_ratio/low_min": 4.9498416046844795e-05,
      "clip_ratio/region_mean": 0.0023505100107286125,
      "epoch": 1.9985418489355498,
      "grad_norm": 0.23547892272472382,
      "learning_rate": 1e-06,
      "loss": -0.0619,
      "step": 856
    },
    {
      "clip_ratio/high_max": 0.003047192654776154,
      "clip_ratio/high_mean": 0.0011949555737373885,
      "clip_ratio/low_mean": 0.0007931789368740283,
      "clip_ratio/low_min": 1.1322464160912205e-05,
      "clip_ratio/region_mean": 0.0019881345360772684,
      "completions/clipped_ratio": 0.1551339285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2954.0,
      "completions/mean_length": 1159.671875,
      "completions/mean_terminated_length": 620.504638671875,
      "completions/min_length": 130.0,
      "completions/min_terminated_length": 130.0,
      "epoch": 2.0023330417031207,
      "grad_norm": 0.4390386939048767,
      "learning_rate": 1e-06,
      "loss": -0.0591,
      "num_tokens": 126824773.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.17874936759471893,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514803290367126,
      "step": 857
    },
    {
      "clip_ratio/high_max": 0.003988200696767308,
      "clip_ratio/high_mean": 0.0014030033053131774,
      "clip_ratio/low_mean": 0.0011837988859042525,
      "clip_ratio/low_min": 5.661231989506632e-05,
      "clip_ratio/region_mean": 0.002586802133009769,
      "epoch": 2.004666083406241,
      "grad_norm": 0.2932032644748688,
      "learning_rate": 1e-06,
      "loss": -0.0593,
      "step": 858
    },
    {
      "clip_ratio/high_max": 0.0027904905364266597,
      "clip_ratio/high_mean": 0.0012420199327607406,
      "clip_ratio/low_mean": 0.0014423892898776103,
      "clip_ratio/low_min": 1.1322464160912205e-05,
      "clip_ratio/region_mean": 0.002684409155335743,
      "epoch": 2.0069991251093615,
      "grad_norm": 0.2708625793457031,
      "learning_rate": 1e-06,
      "loss": -0.0596,
      "step": 859
    },
    {
      "clip_ratio/high_max": 0.0039043327851686627,
      "clip_ratio/high_mean": 0.0014121150343271438,
      "clip_ratio/low_mean": 0.0015946616549626924,
      "clip_ratio/low_min": 4.528985664364882e-05,
      "clip_ratio/region_mean": 0.0030067766856518574,
      "epoch": 2.0093321668124817,
      "grad_norm": 0.2647579312324524,
      "learning_rate": 1e-06,
      "loss": -0.0597,
      "step": 860
    },
    {
      "clip_ratio/high_max": 0.002119186057825573,
      "clip_ratio/high_mean": 0.0007913218196335947,
      "clip_ratio/low_mean": 0.0004452995322026254,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012366213377390523,
      "completions/clipped_ratio": 0.1506696428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2324.0,
      "completions/mean_length": 1149.997802734375,
      "completions/mean_terminated_length": 627.3823852539062,
      "completions/min_length": 143.0,
      "completions/min_terminated_length": 143.0,
      "epoch": 2.0116652085156024,
      "grad_norm": 0.3910108208656311,
      "learning_rate": 1e-06,
      "loss": -0.0548,
      "num_tokens": 127391891.0,
      "reward": 0.590401828289032,
      "reward_std": 0.14556488394737244,
      "rewards/verify_math_reward/mean": 0.5904017686843872,
      "rewards/verify_math_reward/std": 0.49203425645828247,
      "step": 861
    },
    {
      "clip_ratio/high_max": 0.0025455272843828425,
      "clip_ratio/high_mean": 0.0009815586654440267,
      "clip_ratio/low_mean": 0.0006933992754056817,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016749579808674753,
      "epoch": 2.0139982502187226,
      "grad_norm": 0.19863538444042206,
      "learning_rate": 1e-06,
      "loss": -0.0551,
      "step": 862
    },
    {
      "clip_ratio/high_max": 0.0025782149605220184,
      "clip_ratio/high_mean": 0.000941302307182923,
      "clip_ratio/low_mean": 0.000803361215730547,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001744663510180544,
      "epoch": 2.0163312919218432,
      "grad_norm": 0.2755461037158966,
      "learning_rate": 1e-06,
      "loss": -0.0552,
      "step": 863
    },
    {
      "clip_ratio/high_max": 0.002186407222325215,
      "clip_ratio/high_mean": 0.0008494528065057239,
      "clip_ratio/low_mean": 0.000943040229685721,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017924930216395296,
      "epoch": 2.0186643336249634,
      "grad_norm": 0.2356482893228531,
      "learning_rate": 1e-06,
      "loss": -0.0552,
      "step": 864
    },
    {
      "clip_ratio/high_max": 0.0025537680339766666,
      "clip_ratio/high_mean": 0.0010160112324228976,
      "clip_ratio/low_mean": 0.0005593823034359957,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015753935804241337,
      "completions/clipped_ratio": 0.171875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2578.0,
      "completions/mean_length": 1229.4866943359375,
      "completions/mean_terminated_length": 634.5498657226562,
      "completions/min_length": 166.0,
      "completions/min_terminated_length": 166.0,
      "epoch": 2.020997375328084,
      "grad_norm": 0.3333074152469635,
      "learning_rate": 1e-06,
      "loss": -0.0824,
      "num_tokens": 127956423.0,
      "reward": 0.5524553656578064,
      "reward_std": 0.16424313187599182,
      "rewards/verify_math_reward/mean": 0.5524553656578064,
      "rewards/verify_math_reward/std": 0.49751853942871094,
      "step": 865
    },
    {
      "clip_ratio/high_max": 0.002955421070510056,
      "clip_ratio/high_mean": 0.0011224296758882701,
      "clip_ratio/low_mean": 0.0008748107065912336,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001997240397031419,
      "epoch": 2.0233304170312043,
      "grad_norm": 0.25484535098075867,
      "learning_rate": 1e-06,
      "loss": -0.0827,
      "step": 866
    },
    {
      "clip_ratio/high_max": 0.0030243089931900613,
      "clip_ratio/high_mean": 0.0012547181722766254,
      "clip_ratio/low_mean": 0.0010132673451153096,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022679855319438502,
      "epoch": 2.025663458734325,
      "grad_norm": 0.282296746969223,
      "learning_rate": 1e-06,
      "loss": -0.0829,
      "step": 867
    },
    {
      "clip_ratio/high_max": 0.002858953630493488,
      "clip_ratio/high_mean": 0.0011380158939573448,
      "clip_ratio/low_mean": 0.001099641729524592,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022376576234819368,
      "epoch": 2.027996500437445,
      "grad_norm": 0.22792915999889374,
      "learning_rate": 1e-06,
      "loss": -0.083,
      "step": 868
    },
    {
      "clip_ratio/high_max": 0.0023803658405086026,
      "clip_ratio/high_mean": 0.0008531928397133015,
      "clip_ratio/low_mean": 0.0007015458722889889,
      "clip_ratio/low_min": 2.8506270609796047e-05,
      "clip_ratio/region_mean": 0.0015547387047263328,
      "completions/clipped_ratio": 0.1707589285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3985.0,
      "completions/mean_length": 1315.3851318359375,
      "completions/mean_terminated_length": 742.79541015625,
      "completions/min_length": 149.0,
      "completions/min_terminated_length": 149.0,
      "epoch": 2.030329542140566,
      "grad_norm": 0.3597649931907654,
      "learning_rate": 1e-06,
      "loss": -0.066,
      "num_tokens": 128599632.0,
      "reward": 0.5055803656578064,
      "reward_std": 0.1872745305299759,
      "rewards/verify_math_reward/mean": 0.5055803656578064,
      "rewards/verify_math_reward/std": 0.5002480745315552,
      "step": 869
    },
    {
      "clip_ratio/high_max": 0.0026170422934228554,
      "clip_ratio/high_mean": 0.000998484163574176,
      "clip_ratio/low_mean": 0.000972096813711687,
      "clip_ratio/low_min": 3.850102802971378e-05,
      "clip_ratio/region_mean": 0.0019705809900187887,
      "epoch": 2.032662583843686,
      "grad_norm": 0.2861005365848541,
      "learning_rate": 1e-06,
      "loss": -0.0662,
      "step": 870
    },
    {
      "clip_ratio/high_max": 0.003180027582857292,
      "clip_ratio/high_mean": 0.0011111904823337682,
      "clip_ratio/low_mean": 0.0011452123708295403,
      "clip_ratio/low_min": 3.850102802971378e-05,
      "clip_ratio/region_mean": 0.0022564028768101707,
      "epoch": 2.0349956255468067,
      "grad_norm": 0.2486860156059265,
      "learning_rate": 1e-06,
      "loss": -0.0664,
      "step": 871
    },
    {
      "clip_ratio/high_max": 0.00269731388107175,
      "clip_ratio/high_mean": 0.001022971280690399,
      "clip_ratio/low_mean": 0.0013633073940582108,
      "clip_ratio/low_min": 2.5667352019809186e-05,
      "clip_ratio/region_mean": 0.0023862786110839806,
      "epoch": 2.037328667249927,
      "grad_norm": 0.36107203364372253,
      "learning_rate": 1e-06,
      "loss": -0.0665,
      "step": 872
    },
    {
      "clip_ratio/high_max": 0.0027890367164218333,
      "clip_ratio/high_mean": 0.001022377216941095,
      "clip_ratio/low_mean": 0.000612402500337339,
      "clip_ratio/low_min": 1.7193948224303313e-05,
      "clip_ratio/region_mean": 0.001634779735468328,
      "completions/clipped_ratio": 0.1238839285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3322.0,
      "completions/mean_length": 1023.1563110351562,
      "completions/mean_terminated_length": 588.6522216796875,
      "completions/min_length": 133.0,
      "completions/min_terminated_length": 133.0,
      "epoch": 2.0396617089530475,
      "grad_norm": 0.338905930519104,
      "learning_rate": 1e-06,
      "loss": -0.0368,
      "num_tokens": 129149068.0,
      "reward": 0.6517857313156128,
      "reward_std": 0.17029374837875366,
      "rewards/verify_math_reward/mean": 0.6517857313156128,
      "rewards/verify_math_reward/std": 0.47667041420936584,
      "step": 873
    },
    {
      "clip_ratio/high_max": 0.003292863482784014,
      "clip_ratio/high_mean": 0.0011018429668183671,
      "clip_ratio/low_mean": 0.0008577700646128505,
      "clip_ratio/low_min": 1.691932811809238e-05,
      "clip_ratio/region_mean": 0.001959613015060313,
      "epoch": 2.041994750656168,
      "grad_norm": 0.2677460312843323,
      "learning_rate": 1e-06,
      "loss": -0.037,
      "step": 874
    },
    {
      "clip_ratio/high_max": 0.0032929509761743248,
      "clip_ratio/high_mean": 0.001230530640896177,
      "clip_ratio/low_mean": 0.0010246034125884762,
      "clip_ratio/low_min": 5.4959549743216485e-05,
      "clip_ratio/region_mean": 0.0022551340371137485,
      "epoch": 2.0443277923592884,
      "grad_norm": 0.23969261348247528,
      "learning_rate": 1e-06,
      "loss": -0.0372,
      "step": 875
    },
    {
      "clip_ratio/high_max": 0.0036424016871023923,
      "clip_ratio/high_mean": 0.0012846665631514043,
      "clip_ratio/low_mean": 0.0012816219095839188,
      "clip_ratio/low_min": 6.595146260224283e-05,
      "clip_ratio/region_mean": 0.002566288509115111,
      "epoch": 2.046660834062409,
      "grad_norm": 0.22869691252708435,
      "learning_rate": 1e-06,
      "loss": -0.0373,
      "step": 876
    },
    {
      "clip_ratio/high_max": 0.0020634450411307625,
      "clip_ratio/high_mean": 0.0008783011871855706,
      "clip_ratio/low_mean": 0.0007541520153608872,
      "clip_ratio/low_min": 1.429551684850594e-05,
      "clip_ratio/region_mean": 0.001632453189813532,
      "completions/clipped_ratio": 0.2142857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2833.0,
      "completions/mean_length": 1407.37841796875,
      "completions/mean_terminated_length": 674.117919921875,
      "completions/min_length": 206.0,
      "completions/min_terminated_length": 206.0,
      "epoch": 2.0489938757655293,
      "grad_norm": 0.4166792631149292,
      "learning_rate": 1e-06,
      "loss": -0.0569,
      "num_tokens": 129718183.0,
      "reward": 0.5189732313156128,
      "reward_std": 0.18329225480556488,
      "rewards/verify_math_reward/mean": 0.5189732313156128,
      "rewards/verify_math_reward/std": 0.49991893768310547,
      "step": 877
    },
    {
      "clip_ratio/high_max": 0.002453501168929506,
      "clip_ratio/high_mean": 0.0010702426134230336,
      "clip_ratio/low_mean": 0.0010067290804727236,
      "clip_ratio/low_min": 1.429551684850594e-05,
      "clip_ratio/region_mean": 0.0020769717302755453,
      "epoch": 2.05132691746865,
      "grad_norm": 0.24961958825588226,
      "learning_rate": 1e-06,
      "loss": -0.0571,
      "step": 878
    },
    {
      "clip_ratio/high_max": 0.002625891262141522,
      "clip_ratio/high_mean": 0.0011261152212682646,
      "clip_ratio/low_mean": 0.001208353744004853,
      "clip_ratio/low_min": 6.284304254222661e-05,
      "clip_ratio/region_mean": 0.0023344689834630117,
      "epoch": 2.05365995917177,
      "grad_norm": 0.23191159963607788,
      "learning_rate": 1e-06,
      "loss": -0.0573,
      "step": 879
    },
    {
      "clip_ratio/high_max": 0.002479023845808115,
      "clip_ratio/high_mean": 0.0010229359704680974,
      "clip_ratio/low_mean": 0.0014079654647503048,
      "clip_ratio/low_min": 4.4445929233916104e-05,
      "clip_ratio/region_mean": 0.002430901469779201,
      "epoch": 2.055993000874891,
      "grad_norm": 0.21616202592849731,
      "learning_rate": 1e-06,
      "loss": -0.0574,
      "step": 880
    },
    {
      "clip_ratio/high_max": 0.002158216477255337,
      "clip_ratio/high_mean": 0.0007028610289125936,
      "clip_ratio/low_mean": 0.000718300512744463,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014211615598469507,
      "completions/clipped_ratio": 0.1495535714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3887.0,
      "completions/mean_length": 1123.07373046875,
      "completions/mean_terminated_length": 600.2755737304688,
      "completions/min_length": 149.0,
      "completions/min_terminated_length": 149.0,
      "epoch": 2.058326042578011,
      "grad_norm": 0.2821815311908722,
      "learning_rate": 1e-06,
      "loss": -0.0429,
      "num_tokens": 130263505.0,
      "reward": 0.5837053656578064,
      "reward_std": 0.1345210075378418,
      "rewards/verify_math_reward/mean": 0.5837053656578064,
      "rewards/verify_math_reward/std": 0.49321892857551575,
      "step": 881
    },
    {
      "clip_ratio/high_max": 0.0029855417924409267,
      "clip_ratio/high_mean": 0.0009691047216620063,
      "clip_ratio/low_mean": 0.0008622468667454086,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018313515683985315,
      "epoch": 2.0606590842811316,
      "grad_norm": 0.2728791832923889,
      "learning_rate": 1e-06,
      "loss": -0.0432,
      "step": 882
    },
    {
      "clip_ratio/high_max": 0.002525583593524061,
      "clip_ratio/high_mean": 0.0008816239460429642,
      "clip_ratio/low_mean": 0.0010311347541573923,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019127586492686532,
      "epoch": 2.062992125984252,
      "grad_norm": 0.25536054372787476,
      "learning_rate": 1e-06,
      "loss": -0.0433,
      "step": 883
    },
    {
      "clip_ratio/high_max": 0.0026348084647906944,
      "clip_ratio/high_mean": 0.0008546148319510394,
      "clip_ratio/low_mean": 0.001253188143891748,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002107803011313081,
      "epoch": 2.0653251676873725,
      "grad_norm": 0.21684107184410095,
      "learning_rate": 1e-06,
      "loss": -0.0434,
      "step": 884
    },
    {
      "clip_ratio/high_max": 0.002526842705265153,
      "clip_ratio/high_mean": 0.0008572160531912232,
      "clip_ratio/low_mean": 0.00048245869129459606,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013396747344813775,
      "completions/clipped_ratio": 0.1707589285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3284.0,
      "completions/mean_length": 1280.11279296875,
      "completions/mean_terminated_length": 700.259765625,
      "completions/min_length": 172.0,
      "completions/min_terminated_length": 172.0,
      "epoch": 2.0676582093904927,
      "grad_norm": 0.23400527238845825,
      "learning_rate": 1e-06,
      "loss": -0.0832,
      "num_tokens": 130875446.0,
      "reward": 0.5580357313156128,
      "reward_std": 0.16235631704330444,
      "rewards/verify_math_reward/mean": 0.5580357313156128,
      "rewards/verify_math_reward/std": 0.49689778685569763,
      "step": 885
    },
    {
      "clip_ratio/high_max": 0.002830396333592944,
      "clip_ratio/high_mean": 0.001031815798341995,
      "clip_ratio/low_mean": 0.0005952802539468394,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016270960513793398,
      "epoch": 2.0699912510936134,
      "grad_norm": 0.209345743060112,
      "learning_rate": 1e-06,
      "loss": -0.0833,
      "step": 886
    },
    {
      "clip_ratio/high_max": 0.003029807143320795,
      "clip_ratio/high_mean": 0.0010907083978963783,
      "clip_ratio/low_mean": 0.0007710019472142449,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001861710330558708,
      "epoch": 2.0723242927967336,
      "grad_norm": 0.26701340079307556,
      "learning_rate": 1e-06,
      "loss": -0.0834,
      "step": 887
    },
    {
      "clip_ratio/high_max": 0.0031339623310486786,
      "clip_ratio/high_mean": 0.0010413353120384272,
      "clip_ratio/low_mean": 0.0008892381974874297,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001930573518620804,
      "epoch": 2.0746573344998542,
      "grad_norm": 0.18892860412597656,
      "learning_rate": 1e-06,
      "loss": -0.0835,
      "step": 888
    },
    {
      "clip_ratio/high_max": 0.002224198036856251,
      "clip_ratio/high_mean": 0.0007760922162560746,
      "clip_ratio/low_mean": 0.0005144084084349743,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012905006024084287,
      "completions/clipped_ratio": 0.15625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3943.0,
      "completions/mean_length": 1177.1707763671875,
      "completions/mean_terminated_length": 636.6467895507812,
      "completions/min_length": 153.0,
      "completions/min_terminated_length": 153.0,
      "epoch": 2.0769903762029744,
      "grad_norm": 0.3154468238353729,
      "learning_rate": 1e-06,
      "loss": -0.0547,
      "num_tokens": 131457871.0,
      "reward": 0.6071428656578064,
      "reward_std": 0.13139888644218445,
      "rewards/verify_math_reward/mean": 0.6071428656578064,
      "rewards/verify_math_reward/std": 0.48865827918052673,
      "step": 889
    },
    {
      "clip_ratio/high_max": 0.002584983390988782,
      "clip_ratio/high_mean": 0.0009189535740006249,
      "clip_ratio/low_mean": 0.0007904853746367735,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017094390022975858,
      "epoch": 2.079323417906095,
      "grad_norm": 0.23959881067276,
      "learning_rate": 1e-06,
      "loss": -0.0549,
      "step": 890
    },
    {
      "clip_ratio/high_max": 0.00268344135110965,
      "clip_ratio/high_mean": 0.0008809746614133473,
      "clip_ratio/low_mean": 0.0009276802393287653,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018086548880091868,
      "epoch": 2.0816564596092153,
      "grad_norm": 0.20634399354457855,
      "learning_rate": 1e-06,
      "loss": -0.055,
      "step": 891
    },
    {
      "clip_ratio/high_max": 0.002773980326310266,
      "clip_ratio/high_mean": 0.0009158890570688527,
      "clip_ratio/low_mean": 0.0010924457674263977,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002008334849961102,
      "epoch": 2.083989501312336,
      "grad_norm": 0.1906619518995285,
      "learning_rate": 1e-06,
      "loss": -0.0551,
      "step": 892
    },
    {
      "clip_ratio/high_max": 0.0020683441616711207,
      "clip_ratio/high_mean": 0.0007411143869830994,
      "clip_ratio/low_mean": 0.0006184447379382618,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013595591444754973,
      "completions/clipped_ratio": 0.1517857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3906.0,
      "completions/mean_length": 1174.2410888671875,
      "completions/mean_terminated_length": 651.4000244140625,
      "completions/min_length": 184.0,
      "completions/min_terminated_length": 184.0,
      "epoch": 2.0863225430154566,
      "grad_norm": 0.3441523015499115,
      "learning_rate": 1e-06,
      "loss": -0.0756,
      "num_tokens": 132038639.0,
      "reward": 0.6194196939468384,
      "reward_std": 0.13444431126117706,
      "rewards/verify_math_reward/mean": 0.6194196343421936,
      "rewards/verify_math_reward/std": 0.48580074310302734,
      "step": 893
    },
    {
      "clip_ratio/high_max": 0.002379316763835959,
      "clip_ratio/high_mean": 0.0009014450861286605,
      "clip_ratio/low_mean": 0.0008722010829842475,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017736461850290652,
      "epoch": 2.088655584718577,
      "grad_norm": 0.2408902794122696,
      "learning_rate": 1e-06,
      "loss": -0.0758,
      "step": 894
    },
    {
      "clip_ratio/high_max": 0.0024297764684888534,
      "clip_ratio/high_mean": 0.0009308865046477877,
      "clip_ratio/low_mean": 0.0010580024304545077,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001988888980122283,
      "epoch": 2.0909886264216975,
      "grad_norm": 0.2293797731399536,
      "learning_rate": 1e-06,
      "loss": -0.076,
      "step": 895
    },
    {
      "clip_ratio/high_max": 0.002157798364351038,
      "clip_ratio/high_mean": 0.0008320046144945081,
      "clip_ratio/low_mean": 0.0011370203883416252,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001969025001017144,
      "epoch": 2.0933216681248177,
      "grad_norm": 0.1965988725423813,
      "learning_rate": 1e-06,
      "loss": -0.0761,
      "step": 896
    },
    {
      "clip_ratio/high_max": 0.0025361771185998805,
      "clip_ratio/high_mean": 0.000843985199026065,
      "clip_ratio/low_mean": 0.0005831091912114061,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014270944047893863,
      "completions/clipped_ratio": 0.1651785714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4039.0,
      "completions/mean_length": 1246.141845703125,
      "completions/mean_terminated_length": 682.2660522460938,
      "completions/min_length": 164.0,
      "completions/min_terminated_length": 164.0,
      "epoch": 2.0956547098279383,
      "grad_norm": 0.29047125577926636,
      "learning_rate": 1e-06,
      "loss": -0.0672,
      "num_tokens": 132647382.0,
      "reward": 0.5290178656578064,
      "reward_std": 0.15702247619628906,
      "rewards/verify_math_reward/mean": 0.5290178656578064,
      "rewards/verify_math_reward/std": 0.49943605065345764,
      "step": 897
    },
    {
      "clip_ratio/high_max": 0.003130072793283034,
      "clip_ratio/high_mean": 0.0011506956107041333,
      "clip_ratio/low_mean": 0.0007367316311501781,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018874272573157214,
      "epoch": 2.0979877515310585,
      "grad_norm": 0.27660995721817017,
      "learning_rate": 1e-06,
      "loss": -0.0674,
      "step": 898
    },
    {
      "clip_ratio/high_max": 0.0028732167556881905,
      "clip_ratio/high_mean": 0.0011284256470389664,
      "clip_ratio/low_mean": 0.000960379215030116,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020888048675260507,
      "epoch": 2.100320793234179,
      "grad_norm": 0.20674245059490204,
      "learning_rate": 1e-06,
      "loss": -0.0676,
      "step": 899
    },
    {
      "clip_ratio/high_max": 0.0031517503448412754,
      "clip_ratio/high_mean": 0.0010357327191741206,
      "clip_ratio/low_mean": 0.0009695559638203122,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00200528866116656,
      "epoch": 2.1026538349372994,
      "grad_norm": 0.2387486696243286,
      "learning_rate": 1e-06,
      "loss": -0.0676,
      "step": 900
    },
    {
      "clip_ratio/high_max": 0.0027165291612618603,
      "clip_ratio/high_mean": 0.0008888476822903613,
      "clip_ratio/low_mean": 0.0005623961233141017,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014512437846860848,
      "completions/clipped_ratio": 0.1886160714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3551.0,
      "completions/mean_length": 1303.485595703125,
      "completions/mean_terminated_length": 654.3314819335938,
      "completions/min_length": 121.0,
      "completions/min_terminated_length": 121.0,
      "epoch": 2.10498687664042,
      "grad_norm": 0.33062219619750977,
      "learning_rate": 1e-06,
      "loss": -0.0628,
      "num_tokens": 133214929.0,
      "reward": 0.5524553656578064,
      "reward_std": 0.15578144788742065,
      "rewards/verify_math_reward/mean": 0.5524553656578064,
      "rewards/verify_math_reward/std": 0.49751853942871094,
      "step": 901
    },
    {
      "clip_ratio/high_max": 0.00301307208428625,
      "clip_ratio/high_mean": 0.0011444835472502746,
      "clip_ratio/low_mean": 0.0009213571383952512,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020658407011069357,
      "epoch": 2.1073199183435403,
      "grad_norm": 0.2832688093185425,
      "learning_rate": 1e-06,
      "loss": -0.0631,
      "step": 902
    },
    {
      "clip_ratio/high_max": 0.0031409784132847562,
      "clip_ratio/high_mean": 0.0011654591144178994,
      "clip_ratio/low_mean": 0.0010380928288213909,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022035519432392903,
      "epoch": 2.109652960046661,
      "grad_norm": 0.2659349739551544,
      "learning_rate": 1e-06,
      "loss": -0.0633,
      "step": 903
    },
    {
      "clip_ratio/high_max": 0.003178455473971553,
      "clip_ratio/high_mean": 0.001116634884965606,
      "clip_ratio/low_mean": 0.0013046228841631091,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024212577409343794,
      "epoch": 2.111986001749781,
      "grad_norm": 0.23541253805160522,
      "learning_rate": 1e-06,
      "loss": -0.0633,
      "step": 904
    },
    {
      "clip_ratio/high_max": 0.002693645998078864,
      "clip_ratio/high_mean": 0.001170849511254346,
      "clip_ratio/low_mean": 0.000556333008717047,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017271825417992659,
      "completions/clipped_ratio": 0.1819196428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2707.0,
      "completions/mean_length": 1236.67529296875,
      "completions/mean_terminated_length": 600.8363037109375,
      "completions/min_length": 160.0,
      "completions/min_terminated_length": 160.0,
      "epoch": 2.114319043452902,
      "grad_norm": 0.6479451656341553,
      "learning_rate": 1e-06,
      "loss": -0.1152,
      "num_tokens": 133738094.0,
      "reward": 0.590401828289032,
      "reward_std": 0.18558135628700256,
      "rewards/verify_math_reward/mean": 0.5904017686843872,
      "rewards/verify_math_reward/std": 0.49203425645828247,
      "step": 905
    },
    {
      "clip_ratio/high_max": 0.003487138521450106,
      "clip_ratio/high_mean": 0.0013652735142386518,
      "clip_ratio/low_mean": 0.0008751920686336234,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022404655610444024,
      "epoch": 2.116652085156022,
      "grad_norm": 0.3784768581390381,
      "learning_rate": 1e-06,
      "loss": -0.1155,
      "step": 906
    },
    {
      "clip_ratio/high_max": 0.0031059396715136245,
      "clip_ratio/high_mean": 0.0013185599891585298,
      "clip_ratio/low_mean": 0.0009855562639131676,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023041162858135067,
      "epoch": 2.1189851268591426,
      "grad_norm": 0.37493959069252014,
      "learning_rate": 1e-06,
      "loss": -0.1157,
      "step": 907
    },
    {
      "clip_ratio/high_max": 0.0032188539116759785,
      "clip_ratio/high_mean": 0.0013216879997344222,
      "clip_ratio/low_mean": 0.0012801653465430718,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026018533026217483,
      "epoch": 2.121318168562263,
      "grad_norm": 0.23647190630435944,
      "learning_rate": 1e-06,
      "loss": -0.1159,
      "step": 908
    },
    {
      "clip_ratio/high_max": 0.00239591280114837,
      "clip_ratio/high_mean": 0.000833051448353217,
      "clip_ratio/low_mean": 0.0005077994796920393,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001340850900305668,
      "completions/clipped_ratio": 0.1640625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2666.0,
      "completions/mean_length": 1196.2957763671875,
      "completions/mean_terminated_length": 627.1949462890625,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 2.1236512102653835,
      "grad_norm": 0.31013181805610657,
      "learning_rate": 1e-06,
      "loss": -0.0991,
      "num_tokens": 134296271.0,
      "reward": 0.609375,
      "reward_std": 0.15323200821876526,
      "rewards/verify_math_reward/mean": 0.609375,
      "rewards/verify_math_reward/std": 0.48816296458244324,
      "step": 909
    },
    {
      "clip_ratio/high_max": 0.003133546299068257,
      "clip_ratio/high_mean": 0.0010264157244819216,
      "clip_ratio/low_mean": 0.000651560700475784,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001677976415521698,
      "epoch": 2.1259842519685037,
      "grad_norm": 0.2695416808128357,
      "learning_rate": 1e-06,
      "loss": -0.0993,
      "step": 910
    },
    {
      "clip_ratio/high_max": 0.0030468583208858036,
      "clip_ratio/high_mean": 0.001055873934092233,
      "clip_ratio/low_mean": 0.0008514502305843052,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019073241783189587,
      "epoch": 2.1283172936716244,
      "grad_norm": 0.22806468605995178,
      "learning_rate": 1e-06,
      "loss": -0.0995,
      "step": 911
    },
    {
      "clip_ratio/high_max": 0.0028933032081113197,
      "clip_ratio/high_mean": 0.0009494455625826959,
      "clip_ratio/low_mean": 0.001000595917503233,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001950041449163109,
      "epoch": 2.130650335374745,
      "grad_norm": 0.21979136765003204,
      "learning_rate": 1e-06,
      "loss": -0.0996,
      "step": 912
    },
    {
      "clip_ratio/high_max": 0.003457512670138385,
      "clip_ratio/high_mean": 0.0012429355201675207,
      "clip_ratio/low_mean": 0.0006668744454145781,
      "clip_ratio/low_min": 2.9719449230469763e-05,
      "clip_ratio/region_mean": 0.001909809943754226,
      "completions/clipped_ratio": 0.1785714285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3929.0,
      "completions/mean_length": 1314.5301513671875,
      "completions/mean_terminated_length": 709.86279296875,
      "completions/min_length": 189.0,
      "completions/min_terminated_length": 189.0,
      "epoch": 2.1329833770778652,
      "grad_norm": 0.4042540192604065,
      "learning_rate": 1e-06,
      "loss": -0.054,
      "num_tokens": 134913666.0,
      "reward": 0.5290178656578064,
      "reward_std": 0.18719714879989624,
      "rewards/verify_math_reward/mean": 0.5290178656578064,
      "rewards/verify_math_reward/std": 0.49943605065345764,
      "step": 913
    },
    {
      "clip_ratio/high_max": 0.003955599411710864,
      "clip_ratio/high_mean": 0.0015156144054344622,
      "clip_ratio/low_mean": 0.0010538833803366288,
      "clip_ratio/low_min": 7.302063113456825e-05,
      "clip_ratio/region_mean": 0.002569497715739999,
      "epoch": 2.135316418780986,
      "grad_norm": 0.339517742395401,
      "learning_rate": 1e-06,
      "loss": -0.0543,
      "step": 914
    },
    {
      "clip_ratio/high_max": 0.0038658706289425027,
      "clip_ratio/high_mean": 0.001414357790054055,
      "clip_ratio/low_mean": 0.001270181384825264,
      "clip_ratio/low_min": 8.306038125738269e-05,
      "clip_ratio/region_mean": 0.002684539220354054,
      "epoch": 2.137649460484106,
      "grad_norm": 0.300604909658432,
      "learning_rate": 1e-06,
      "loss": -0.0545,
      "step": 915
    },
    {
      "clip_ratio/high_max": 0.003662280912976712,
      "clip_ratio/high_mean": 0.001388209870128776,
      "clip_ratio/low_mean": 0.0015070848130562808,
      "clip_ratio/low_min": 0.00012288177094887942,
      "clip_ratio/region_mean": 0.0028952946086064912,
      "epoch": 2.1399825021872267,
      "grad_norm": 0.24068570137023926,
      "learning_rate": 1e-06,
      "loss": -0.0546,
      "step": 916
    },
    {
      "clip_ratio/high_max": 0.0025139247190963943,
      "clip_ratio/high_mean": 0.0007882582540332805,
      "clip_ratio/low_mean": 0.00047680701118224533,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012650652824959252,
      "completions/clipped_ratio": 0.1350446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3895.0,
      "completions/mean_length": 1088.91748046875,
      "completions/mean_terminated_length": 619.424560546875,
      "completions/min_length": 107.0,
      "completions/min_terminated_length": 107.0,
      "epoch": 2.142315543890347,
      "grad_norm": 0.2798072099685669,
      "learning_rate": 1e-06,
      "loss": -0.0857,
      "num_tokens": 135479312.0,
      "reward": 0.6082589626312256,
      "reward_std": 0.14255832135677338,
      "rewards/verify_math_reward/mean": 0.6082589030265808,
      "rewards/verify_math_reward/std": 0.48841193318367004,
      "step": 917
    },
    {
      "clip_ratio/high_max": 0.0024859030891093425,
      "clip_ratio/high_mean": 0.0008973686544777593,
      "clip_ratio/low_mean": 0.0007205899273685645,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016179585763893556,
      "epoch": 2.1446485855934676,
      "grad_norm": 0.2337394803762436,
      "learning_rate": 1e-06,
      "loss": -0.0859,
      "step": 918
    },
    {
      "clip_ratio/high_max": 0.0025061886408366263,
      "clip_ratio/high_mean": 0.0008488409621350002,
      "clip_ratio/low_mean": 0.0008249815755334566,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016738224949222058,
      "epoch": 2.146981627296588,
      "grad_norm": 0.248878613114357,
      "learning_rate": 1e-06,
      "loss": -0.086,
      "step": 919
    },
    {
      "clip_ratio/high_max": 0.002762799762422219,
      "clip_ratio/high_mean": 0.0009419840571354143,
      "clip_ratio/low_mean": 0.0009382011185152805,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018801851401804015,
      "epoch": 2.1493146689997085,
      "grad_norm": 0.22181767225265503,
      "learning_rate": 1e-06,
      "loss": -0.086,
      "step": 920
    },
    {
      "clip_ratio/high_max": 0.002830226774676703,
      "clip_ratio/high_mean": 0.0008858105793478899,
      "clip_ratio/low_mean": 0.0006732423416906386,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001559052907396108,
      "completions/clipped_ratio": 0.15625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3833.0,
      "completions/mean_length": 1169.8504638671875,
      "completions/mean_terminated_length": 627.9708862304688,
      "completions/min_length": 136.0,
      "completions/min_terminated_length": 136.0,
      "epoch": 2.1516477107028287,
      "grad_norm": 0.36873382329940796,
      "learning_rate": 1e-06,
      "loss": -0.0428,
      "num_tokens": 136043794.0,
      "reward": 0.551339328289032,
      "reward_std": 0.14327649772167206,
      "rewards/verify_math_reward/mean": 0.5513392686843872,
      "rewards/verify_math_reward/std": 0.4976350665092468,
      "step": 921
    },
    {
      "clip_ratio/high_max": 0.0030737649358343333,
      "clip_ratio/high_mean": 0.0010548260834184475,
      "clip_ratio/low_mean": 0.0008977624897852365,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019525885509210639,
      "epoch": 2.1539807524059493,
      "grad_norm": 0.31384843587875366,
      "learning_rate": 1e-06,
      "loss": -0.0432,
      "step": 922
    },
    {
      "clip_ratio/high_max": 0.002901628045947291,
      "clip_ratio/high_mean": 0.0010035861105279764,
      "clip_ratio/low_mean": 0.0010312153040104022,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002034801429545041,
      "epoch": 2.1563137941090695,
      "grad_norm": 0.287413090467453,
      "learning_rate": 1e-06,
      "loss": -0.0433,
      "step": 923
    },
    {
      "clip_ratio/high_max": 0.0030378203809959814,
      "clip_ratio/high_mean": 0.0010518749622860923,
      "clip_ratio/low_mean": 0.001193243349007389,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002245118303108029,
      "epoch": 2.15864683581219,
      "grad_norm": 0.2901656925678253,
      "learning_rate": 1e-06,
      "loss": -0.0434,
      "step": 924
    },
    {
      "clip_ratio/high_max": 0.0021673360315617174,
      "clip_ratio/high_mean": 0.0007874278126109857,
      "clip_ratio/low_mean": 0.00045988218789716484,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012473100287024863,
      "completions/clipped_ratio": 0.1662946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3983.0,
      "completions/mean_length": 1261.01904296875,
      "completions/mean_terminated_length": 695.5408325195312,
      "completions/min_length": 172.0,
      "completions/min_terminated_length": 172.0,
      "epoch": 2.1609798775153104,
      "grad_norm": 0.24593819677829742,
      "learning_rate": 1e-06,
      "loss": -0.0537,
      "num_tokens": 136653739.0,
      "reward": 0.5390625,
      "reward_std": 0.13880637288093567,
      "rewards/verify_math_reward/mean": 0.5390625,
      "rewards/verify_math_reward/std": 0.4987502098083496,
      "step": 925
    },
    {
      "clip_ratio/high_max": 0.0024482507360517047,
      "clip_ratio/high_mean": 0.0008538253787264694,
      "clip_ratio/low_mean": 0.0006132278729182872,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014670532691525295,
      "epoch": 2.163312919218431,
      "grad_norm": 0.2377663403749466,
      "learning_rate": 1e-06,
      "loss": -0.0538,
      "step": 926
    },
    {
      "clip_ratio/high_max": 0.0023770478946971707,
      "clip_ratio/high_mean": 0.0008966462391981622,
      "clip_ratio/low_mean": 0.0008122432655000011,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001708889496512711,
      "epoch": 2.1656459609215517,
      "grad_norm": 0.27022865414619446,
      "learning_rate": 1e-06,
      "loss": -0.054,
      "step": 927
    },
    {
      "clip_ratio/high_max": 0.002444057034153957,
      "clip_ratio/high_mean": 0.0008821404062473448,
      "clip_ratio/low_mean": 0.000945021546613134,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00182716193376109,
      "epoch": 2.167979002624672,
      "grad_norm": 0.25843802094459534,
      "learning_rate": 1e-06,
      "loss": -0.0541,
      "step": 928
    },
    {
      "clip_ratio/high_max": 0.0021551562094828114,
      "clip_ratio/high_mean": 0.0008665754467074294,
      "clip_ratio/low_mean": 0.0007805696077411994,
      "clip_ratio/low_min": 6.278415912674973e-05,
      "clip_ratio/region_mean": 0.0016471450726385228,
      "completions/clipped_ratio": 0.1350446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3538.0,
      "completions/mean_length": 1148.51904296875,
      "completions/mean_terminated_length": 688.3316650390625,
      "completions/min_length": 170.0,
      "completions/min_terminated_length": 170.0,
      "epoch": 2.1703120443277926,
      "grad_norm": 0.3312126696109772,
      "learning_rate": 1e-06,
      "loss": -0.0379,
      "num_tokens": 137278460.0,
      "reward": 0.5167410969734192,
      "reward_std": 0.17908243834972382,
      "rewards/verify_math_reward/mean": 0.5167410969734192,
      "rewards/verify_math_reward/std": 0.4999987483024597,
      "step": 929
    },
    {
      "clip_ratio/high_max": 0.0025835675405687653,
      "clip_ratio/high_mean": 0.001081595979485428,
      "clip_ratio/low_mean": 0.0010114546475961106,
      "clip_ratio/low_min": 0.00011477376574475784,
      "clip_ratio/region_mean": 0.002093050668918295,
      "epoch": 2.1726450860309128,
      "grad_norm": 0.2656216621398926,
      "learning_rate": 1e-06,
      "loss": -0.0381,
      "step": 930
    },
    {
      "clip_ratio/high_max": 0.00256100272963522,
      "clip_ratio/high_mean": 0.0010460949542903109,
      "clip_ratio/low_mean": 0.0011754417610063683,
      "clip_ratio/low_min": 0.00011822393389593344,
      "clip_ratio/region_mean": 0.0022215367353055626,
      "epoch": 2.1749781277340334,
      "grad_norm": 0.2628580927848816,
      "learning_rate": 1e-06,
      "loss": -0.0383,
      "step": 931
    },
    {
      "clip_ratio/high_max": 0.0022051738706068136,
      "clip_ratio/high_mean": 0.000955240908297128,
      "clip_ratio/low_mean": 0.0014340925736178178,
      "clip_ratio/low_min": 0.0001505113068560604,
      "clip_ratio/region_mean": 0.0023893334437161684,
      "epoch": 2.1773111694371536,
      "grad_norm": 0.26694920659065247,
      "learning_rate": 1e-06,
      "loss": -0.0384,
      "step": 932
    },
    {
      "clip_ratio/high_max": 0.002139591691957321,
      "clip_ratio/high_mean": 0.0007941814437799621,
      "clip_ratio/low_mean": 0.0005919145351072075,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013860959879821166,
      "completions/clipped_ratio": 0.1283482142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3572.0,
      "completions/mean_length": 1079.03125,
      "completions/mean_terminated_length": 634.7913208007812,
      "completions/min_length": 155.0,
      "completions/min_terminated_length": 155.0,
      "epoch": 2.1796442111402743,
      "grad_norm": 0.3061717748641968,
      "learning_rate": 1e-06,
      "loss": -0.044,
      "num_tokens": 137881496.0,
      "reward": 0.5089285969734192,
      "reward_std": 0.16006723046302795,
      "rewards/verify_math_reward/mean": 0.5089285969734192,
      "rewards/verify_math_reward/std": 0.5001994967460632,
      "step": 933
    },
    {
      "clip_ratio/high_max": 0.0023438950229319744,
      "clip_ratio/high_mean": 0.000926869637623895,
      "clip_ratio/low_mean": 0.0009254118449462112,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018522815080359578,
      "epoch": 2.1819772528433945,
      "grad_norm": 0.24494388699531555,
      "learning_rate": 1e-06,
      "loss": -0.0442,
      "step": 934
    },
    {
      "clip_ratio/high_max": 0.0024403161805821583,
      "clip_ratio/high_mean": 0.0009846319130701886,
      "clip_ratio/low_mean": 0.0010780941793200327,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002062726103758905,
      "epoch": 2.184310294546515,
      "grad_norm": 0.21183714270591736,
      "learning_rate": 1e-06,
      "loss": -0.0444,
      "step": 935
    },
    {
      "clip_ratio/high_max": 0.0020901129355479497,
      "clip_ratio/high_mean": 0.0008741297460801434,
      "clip_ratio/low_mean": 0.0011600875404838007,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002034217286563944,
      "epoch": 2.1866433362496354,
      "grad_norm": 0.24384094774723053,
      "learning_rate": 1e-06,
      "loss": -0.0443,
      "step": 936
    },
    {
      "clip_ratio/high_max": 0.001768199923390057,
      "clip_ratio/high_mean": 0.000634681118754088,
      "clip_ratio/low_mean": 0.0006225188540156523,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012571999686770141,
      "completions/clipped_ratio": 0.1283482142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3038.0,
      "completions/mean_length": 1036.712158203125,
      "completions/mean_terminated_length": 586.24072265625,
      "completions/min_length": 184.0,
      "completions/min_terminated_length": 184.0,
      "epoch": 2.188976377952756,
      "grad_norm": 0.32948145270347595,
      "learning_rate": 1e-06,
      "loss": -0.0378,
      "num_tokens": 138426646.0,
      "reward": 0.5803571939468384,
      "reward_std": 0.15818998217582703,
      "rewards/verify_math_reward/mean": 0.5803571343421936,
      "rewards/verify_math_reward/std": 0.4937761127948761,
      "step": 937
    },
    {
      "clip_ratio/high_max": 0.002090778416459216,
      "clip_ratio/high_mean": 0.0008678984213474905,
      "clip_ratio/low_mean": 0.0008075540590652963,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001675452458584914,
      "epoch": 2.1913094196558762,
      "grad_norm": 0.43128690123558044,
      "learning_rate": 1e-06,
      "loss": -0.038,
      "step": 938
    },
    {
      "clip_ratio/high_max": 0.0024906518374336883,
      "clip_ratio/high_mean": 0.0009043059508258011,
      "clip_ratio/low_mean": 0.0010591671789370594,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00196347308519762,
      "epoch": 2.193642461358997,
      "grad_norm": 0.21879757940769196,
      "learning_rate": 1e-06,
      "loss": -0.0383,
      "step": 939
    },
    {
      "clip_ratio/high_max": 0.0023015859405859374,
      "clip_ratio/high_mean": 0.0008173199948942056,
      "clip_ratio/low_mean": 0.001166840183941531,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019841601388179697,
      "epoch": 2.195975503062117,
      "grad_norm": 0.22823746502399445,
      "learning_rate": 1e-06,
      "loss": -0.0382,
      "step": 940
    },
    {
      "clip_ratio/high_max": 0.0020483883927227,
      "clip_ratio/high_mean": 0.0007574281025881646,
      "clip_ratio/low_mean": 0.00047466508294746745,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012320932182774413,
      "completions/clipped_ratio": 0.140625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3623.0,
      "completions/mean_length": 1117.8504638671875,
      "completions/mean_terminated_length": 630.516845703125,
      "completions/min_length": 129.0,
      "completions/min_terminated_length": 129.0,
      "epoch": 2.1983085447652377,
      "grad_norm": 0.30982983112335205,
      "learning_rate": 1e-06,
      "loss": -0.0259,
      "num_tokens": 139001576.0,
      "reward": 0.574776828289032,
      "reward_std": 0.15090152621269226,
      "rewards/verify_math_reward/mean": 0.5747767686843872,
      "rewards/verify_math_reward/std": 0.49465295672416687,
      "step": 941
    },
    {
      "clip_ratio/high_max": 0.0023883046815171838,
      "clip_ratio/high_mean": 0.0010080747797474032,
      "clip_ratio/low_mean": 0.0007136938784242375,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017217686836374924,
      "epoch": 2.200641586468358,
      "grad_norm": 0.29436051845550537,
      "learning_rate": 1e-06,
      "loss": -0.0261,
      "step": 942
    },
    {
      "clip_ratio/high_max": 0.0025798613351071253,
      "clip_ratio/high_mean": 0.000991768474705168,
      "clip_ratio/low_mean": 0.0008289616853289772,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001820730212784838,
      "epoch": 2.2029746281714786,
      "grad_norm": 0.22793766856193542,
      "learning_rate": 1e-06,
      "loss": -0.0262,
      "step": 943
    },
    {
      "clip_ratio/high_max": 0.0023700841338722967,
      "clip_ratio/high_mean": 0.00098240349052503,
      "clip_ratio/low_mean": 0.0010814499692060053,
      "clip_ratio/low_min": 1.6587049685767852e-05,
      "clip_ratio/region_mean": 0.0020638534988393076,
      "epoch": 2.205307669874599,
      "grad_norm": 0.26359623670578003,
      "learning_rate": 1e-06,
      "loss": -0.0263,
      "step": 944
    },
    {
      "clip_ratio/high_max": 0.001843147969339043,
      "clip_ratio/high_mean": 0.0006260225245569018,
      "clip_ratio/low_mean": 0.000724763362086378,
      "clip_ratio/low_min": 4.9626807594904676e-05,
      "clip_ratio/region_mean": 0.0013507858820958063,
      "completions/clipped_ratio": 0.15625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2832.0,
      "completions/mean_length": 1200.65185546875,
      "completions/mean_terminated_length": 664.4761962890625,
      "completions/min_length": 166.0,
      "completions/min_terminated_length": 166.0,
      "epoch": 2.2076407115777195,
      "grad_norm": 0.33370280265808105,
      "learning_rate": 1e-06,
      "loss": -0.0472,
      "num_tokens": 139587328.0,
      "reward": 0.5334821939468384,
      "reward_std": 0.15729349851608276,
      "rewards/verify_math_reward/mean": 0.5334821343421936,
      "rewards/verify_math_reward/std": 0.49915632605552673,
      "step": 945
    },
    {
      "clip_ratio/high_max": 0.0019799328547378536,
      "clip_ratio/high_mean": 0.0007350524265348213,
      "clip_ratio/low_mean": 0.0010777700445032679,
      "clip_ratio/low_min": 4.0787468606140465e-05,
      "clip_ratio/region_mean": 0.0018128224764950573,
      "epoch": 2.20997375328084,
      "grad_norm": 0.22046437859535217,
      "learning_rate": 1e-06,
      "loss": -0.0474,
      "step": 946
    },
    {
      "clip_ratio/high_max": 0.002227421813586261,
      "clip_ratio/high_mean": 0.0007772156586725032,
      "clip_ratio/low_mean": 0.0011329547014611308,
      "clip_ratio/low_min": 6.978042074479163e-05,
      "clip_ratio/region_mean": 0.0019101703655906022,
      "epoch": 2.2123067949839603,
      "grad_norm": 0.25371313095092773,
      "learning_rate": 1e-06,
      "loss": -0.0475,
      "step": 947
    },
    {
      "clip_ratio/high_max": 0.002114128517860081,
      "clip_ratio/high_mean": 0.0007653627180843614,
      "clip_ratio/low_mean": 0.001256132894923212,
      "clip_ratio/low_min": 5.815035183331929e-05,
      "clip_ratio/region_mean": 0.002021495543885976,
      "epoch": 2.214639836687081,
      "grad_norm": 0.2694165110588074,
      "learning_rate": 1e-06,
      "loss": -0.0476,
      "step": 948
    },
    {
      "clip_ratio/high_max": 0.0016557084200030658,
      "clip_ratio/high_mean": 0.0005161424414836802,
      "clip_ratio/low_mean": 0.000461753966192191,
      "clip_ratio/low_min": 4.251655445841607e-05,
      "clip_ratio/region_mean": 0.0009778964176803129,
      "completions/clipped_ratio": 0.1417410714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3777.0,
      "completions/mean_length": 1111.8170166015625,
      "completions/mean_terminated_length": 618.98046875,
      "completions/min_length": 148.0,
      "completions/min_terminated_length": 148.0,
      "epoch": 2.216972878390201,
      "grad_norm": 0.276732474565506,
      "learning_rate": 1e-06,
      "loss": -0.041,
      "num_tokens": 140145060.0,
      "reward": 0.5212053656578064,
      "reward_std": 0.13609540462493896,
      "rewards/verify_math_reward/mean": 0.5212053656578064,
      "rewards/verify_math_reward/std": 0.49982914328575134,
      "step": 949
    },
    {
      "clip_ratio/high_max": 0.002127289764757734,
      "clip_ratio/high_mean": 0.0007239664619191899,
      "clip_ratio/low_mean": 0.000746685138437897,
      "clip_ratio/low_min": 4.861864908889402e-05,
      "clip_ratio/region_mean": 0.001470651630370412,
      "epoch": 2.219305920093322,
      "grad_norm": 0.23614975810050964,
      "learning_rate": 1e-06,
      "loss": -0.0412,
      "step": 950
    },
    {
      "clip_ratio/high_max": 0.00227067967352923,
      "clip_ratio/high_mean": 0.0007507983964387677,
      "clip_ratio/low_mean": 0.0008739369732211344,
      "clip_ratio/low_min": 4.017141691292636e-05,
      "clip_ratio/region_mean": 0.0016247353378275875,
      "epoch": 2.221638961796442,
      "grad_norm": 0.18173855543136597,
      "learning_rate": 1e-06,
      "loss": -0.0414,
      "step": 951
    },
    {
      "clip_ratio/high_max": 0.002109660468704533,
      "clip_ratio/high_mean": 0.000666342064505443,
      "clip_ratio/low_mean": 0.001001883705612272,
      "clip_ratio/low_min": 6.294832837738795e-05,
      "clip_ratio/region_mean": 0.0016682257810316514,
      "epoch": 2.2239720034995627,
      "grad_norm": 0.2091158926486969,
      "learning_rate": 1e-06,
      "loss": -0.0414,
      "step": 952
    },
    {
      "clip_ratio/high_max": 0.0026642042430466972,
      "clip_ratio/high_mean": 0.0009831271636357997,
      "clip_ratio/low_mean": 0.0006934364646440372,
      "clip_ratio/low_min": 2.9322072805371135e-05,
      "clip_ratio/region_mean": 0.0016765636173659004,
      "completions/clipped_ratio": 0.1640625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3043.0,
      "completions/mean_length": 1180.571533203125,
      "completions/mean_terminated_length": 608.384521484375,
      "completions/min_length": 139.0,
      "completions/min_terminated_length": 139.0,
      "epoch": 2.226305045202683,
      "grad_norm": 0.3571631610393524,
      "learning_rate": 1e-06,
      "loss": -0.0407,
      "num_tokens": 140691932.0,
      "reward": 0.535714328289032,
      "reward_std": 0.1716071218252182,
      "rewards/verify_math_reward/mean": 0.5357142686843872,
      "rewards/verify_math_reward/std": 0.4990014135837555,
      "step": 953
    },
    {
      "clip_ratio/high_max": 0.00303604166401783,
      "clip_ratio/high_mean": 0.0010800326363096246,
      "clip_ratio/low_mean": 0.0009275444572267588,
      "clip_ratio/low_min": 6.189139821799472e-05,
      "clip_ratio/region_mean": 0.0020075770444236696,
      "epoch": 2.2286380869058036,
      "grad_norm": 0.2970696985721588,
      "learning_rate": 1e-06,
      "loss": -0.0409,
      "step": 954
    },
    {
      "clip_ratio/high_max": 0.0035131540935253724,
      "clip_ratio/high_mean": 0.0012164544277766254,
      "clip_ratio/low_mean": 0.0012272803178348113,
      "clip_ratio/low_min": 0.00010359116276958957,
      "clip_ratio/region_mean": 0.0024437347674393095,
      "epoch": 2.2309711286089238,
      "grad_norm": 0.2447551190853119,
      "learning_rate": 1e-06,
      "loss": -0.0411,
      "step": 955
    },
    {
      "clip_ratio/high_max": 0.0028052897614543326,
      "clip_ratio/high_mean": 0.0010593809638521634,
      "clip_ratio/low_mean": 0.0014988202528911643,
      "clip_ratio/low_min": 0.0001340980379609391,
      "clip_ratio/region_mean": 0.0025582012021914124,
      "epoch": 2.2333041703120444,
      "grad_norm": 0.2844543755054474,
      "learning_rate": 1e-06,
      "loss": -0.0412,
      "step": 956
    },
    {
      "clip_ratio/high_max": 0.0024161854744306765,
      "clip_ratio/high_mean": 0.0010049222473753616,
      "clip_ratio/low_mean": 0.0004824429756808968,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014873652544338256,
      "completions/clipped_ratio": 0.1294642857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3811.0,
      "completions/mean_length": 1056.2723388671875,
      "completions/mean_terminated_length": 604.2102661132812,
      "completions/min_length": 148.0,
      "completions/min_terminated_length": 148.0,
      "epoch": 2.2356372120151646,
      "grad_norm": 0.3601495325565338,
      "learning_rate": 1e-06,
      "loss": -0.044,
      "num_tokens": 141238360.0,
      "reward": 0.6484375,
      "reward_std": 0.16450665891170502,
      "rewards/verify_math_reward/mean": 0.6484375,
      "rewards/verify_math_reward/std": 0.4777248501777649,
      "step": 957
    },
    {
      "clip_ratio/high_max": 0.0032573289645370096,
      "clip_ratio/high_mean": 0.0013352053574635647,
      "clip_ratio/low_mean": 0.0007881219216869795,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021233272418612614,
      "epoch": 2.2379702537182853,
      "grad_norm": 0.2749025523662567,
      "learning_rate": 1e-06,
      "loss": -0.0442,
      "step": 958
    },
    {
      "clip_ratio/high_max": 0.003161405933497008,
      "clip_ratio/high_mean": 0.0013238063256721944,
      "clip_ratio/low_mean": 0.0008990651913336478,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002222871538833715,
      "epoch": 2.2403032954214055,
      "grad_norm": 0.24613353610038757,
      "learning_rate": 1e-06,
      "loss": -0.0444,
      "step": 959
    },
    {
      "clip_ratio/high_max": 0.002812174214341212,
      "clip_ratio/high_mean": 0.0011950203588639852,
      "clip_ratio/low_mean": 0.0010793403234856669,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002274360682349652,
      "epoch": 2.242636337124526,
      "grad_norm": 0.35043787956237793,
      "learning_rate": 1e-06,
      "loss": -0.0445,
      "step": 960
    },
    {
      "clip_ratio/high_max": 0.002419755546725355,
      "clip_ratio/high_mean": 0.0008969339869508985,
      "clip_ratio/low_mean": 0.00043708002590392425,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013340140030777548,
      "completions/clipped_ratio": 0.1540178571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2158.0,
      "completions/mean_length": 1127.540283203125,
      "completions/mean_terminated_length": 587.1082153320312,
      "completions/min_length": 148.0,
      "completions/min_terminated_length": 148.0,
      "epoch": 2.2449693788276464,
      "grad_norm": 0.3806179463863373,
      "learning_rate": 1e-06,
      "loss": -0.0803,
      "num_tokens": 141771868.0,
      "reward": 0.5870535969734192,
      "reward_std": 0.16172580420970917,
      "rewards/verify_math_reward/mean": 0.5870535969734192,
      "rewards/verify_math_reward/std": 0.49263834953308105,
      "step": 961
    },
    {
      "clip_ratio/high_max": 0.0030022057762835175,
      "clip_ratio/high_mean": 0.001246008978341706,
      "clip_ratio/low_mean": 0.000651918781841232,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018979277519974858,
      "epoch": 2.247302420530767,
      "grad_norm": 0.27406615018844604,
      "learning_rate": 1e-06,
      "loss": -0.0807,
      "step": 962
    },
    {
      "clip_ratio/high_max": 0.0028254452408873476,
      "clip_ratio/high_mean": 0.0011592298033065163,
      "clip_ratio/low_mean": 0.0008223621834986261,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019815919440588914,
      "epoch": 2.249635462233887,
      "grad_norm": 0.26929888129234314,
      "learning_rate": 1e-06,
      "loss": -0.0808,
      "step": 963
    },
    {
      "clip_ratio/high_max": 0.0026136957967537455,
      "clip_ratio/high_mean": 0.0010887508469750173,
      "clip_ratio/low_mean": 0.0009752080068210489,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020639588838093914,
      "epoch": 2.251968503937008,
      "grad_norm": 0.27124956250190735,
      "learning_rate": 1e-06,
      "loss": -0.0809,
      "step": 964
    },
    {
      "clip_ratio/high_max": 0.0022181659478519578,
      "clip_ratio/high_mean": 0.000915030407213635,
      "clip_ratio/low_mean": 0.0006308629726845538,
      "clip_ratio/low_min": 2.5375558834639378e-05,
      "clip_ratio/region_mean": 0.0015458933776244521,
      "completions/clipped_ratio": 0.1116071428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2528.0,
      "completions/mean_length": 1011.0848388671875,
      "completions/mean_terminated_length": 623.5326538085938,
      "completions/min_length": 141.0,
      "completions/min_terminated_length": 141.0,
      "epoch": 2.2543015456401285,
      "grad_norm": 0.3349124789237976,
      "learning_rate": 1e-06,
      "loss": -0.0567,
      "num_tokens": 142348168.0,
      "reward": 0.6261160969734192,
      "reward_std": 0.1795709878206253,
      "rewards/verify_math_reward/mean": 0.6261160969734192,
      "rewards/verify_math_reward/std": 0.48410359025001526,
      "step": 965
    },
    {
      "clip_ratio/high_max": 0.002503495335986372,
      "clip_ratio/high_mean": 0.0010731359543569852,
      "clip_ratio/low_mean": 0.000809296878742316,
      "clip_ratio/low_min": 9.833228432398755e-06,
      "clip_ratio/region_mean": 0.0018824328071787022,
      "epoch": 2.2566345873432487,
      "grad_norm": 0.2904164493083954,
      "learning_rate": 1e-06,
      "loss": -0.0569,
      "step": 966
    },
    {
      "clip_ratio/high_max": 0.0024661056631885003,
      "clip_ratio/high_mean": 0.0010312086633348372,
      "clip_ratio/low_mean": 0.0010948240997095127,
      "clip_ratio/low_min": 2.2986392650636844e-05,
      "clip_ratio/region_mean": 0.0021260327084746677,
      "epoch": 2.2589676290463694,
      "grad_norm": 0.2515934109687805,
      "learning_rate": 1e-06,
      "loss": -0.0571,
      "step": 967
    },
    {
      "clip_ratio/high_max": 0.0027019068111258093,
      "clip_ratio/high_mean": 0.00110522161321569,
      "clip_ratio/low_mean": 0.0012917765252495883,
      "clip_ratio/low_min": 6.47084725642344e-05,
      "clip_ratio/region_mean": 0.0023969980975380167,
      "epoch": 2.2613006707494896,
      "grad_norm": 0.2741507887840271,
      "learning_rate": 1e-06,
      "loss": -0.0572,
      "step": 968
    },
    {
      "clip_ratio/high_max": 0.002199630605900893,
      "clip_ratio/high_mean": 0.0008182023684639717,
      "clip_ratio/low_mean": 0.0006519424650832661,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014701447835250292,
      "completions/clipped_ratio": 0.1082589285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3819.0,
      "completions/mean_length": 1000.7511596679688,
      "completions/mean_terminated_length": 624.9824829101562,
      "completions/min_length": 191.0,
      "completions/min_terminated_length": 191.0,
      "epoch": 2.2636337124526102,
      "grad_norm": 0.3289185166358948,
      "learning_rate": 1e-06,
      "loss": -0.0485,
      "num_tokens": 142933185.0,
      "reward": 0.598214328289032,
      "reward_std": 0.16198793053627014,
      "rewards/verify_math_reward/mean": 0.5982142686843872,
      "rewards/verify_math_reward/std": 0.49053287506103516,
      "step": 969
    },
    {
      "clip_ratio/high_max": 0.0025676776858745143,
      "clip_ratio/high_mean": 0.0009590076369931921,
      "clip_ratio/low_mean": 0.0008741687543079024,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018331763858441263,
      "epoch": 2.2659667541557305,
      "grad_norm": 0.2584741711616516,
      "learning_rate": 1e-06,
      "loss": -0.0487,
      "step": 970
    },
    {
      "clip_ratio/high_max": 0.002799642810714431,
      "clip_ratio/high_mean": 0.0010030778466898482,
      "clip_ratio/low_mean": 0.0010420067264931276,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020450845404411666,
      "epoch": 2.268299795858851,
      "grad_norm": 0.23060286045074463,
      "learning_rate": 1e-06,
      "loss": -0.0489,
      "step": 971
    },
    {
      "clip_ratio/high_max": 0.002635231299791485,
      "clip_ratio/high_mean": 0.0009523928201815579,
      "clip_ratio/low_mean": 0.001234156496138894,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021865493181394413,
      "epoch": 2.2706328375619713,
      "grad_norm": 0.23002904653549194,
      "learning_rate": 1e-06,
      "loss": -0.049,
      "step": 972
    },
    {
      "clip_ratio/high_max": 0.0028391991509124637,
      "clip_ratio/high_mean": 0.001106439893192146,
      "clip_ratio/low_mean": 0.0006090573533583665,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017154972338175867,
      "completions/clipped_ratio": 0.1439732142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2396.0,
      "completions/mean_length": 1066.6551513671875,
      "completions/mean_terminated_length": 557.1564331054688,
      "completions/min_length": 128.0,
      "completions/min_terminated_length": 128.0,
      "epoch": 2.272965879265092,
      "grad_norm": 0.3089137673377991,
      "learning_rate": 1e-06,
      "loss": -0.0821,
      "num_tokens": 143437588.0,
      "reward": 0.645089328289032,
      "reward_std": 0.15808121860027313,
      "rewards/verify_math_reward/mean": 0.6450892686843872,
      "rewards/verify_math_reward/std": 0.4787535071372986,
      "step": 973
    },
    {
      "clip_ratio/high_max": 0.0032532387995161116,
      "clip_ratio/high_mean": 0.0012893593338958453,
      "clip_ratio/low_mean": 0.0007900426817286643,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020794020492758136,
      "epoch": 2.275298920968212,
      "grad_norm": 0.2660689949989319,
      "learning_rate": 1e-06,
      "loss": -0.0822,
      "step": 974
    },
    {
      "clip_ratio/high_max": 0.0037110813573235646,
      "clip_ratio/high_mean": 0.0013476625608745962,
      "clip_ratio/low_mean": 0.0009197242743539391,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00226738685159944,
      "epoch": 2.277631962671333,
      "grad_norm": 0.2527943253517151,
      "learning_rate": 1e-06,
      "loss": -0.0825,
      "step": 975
    },
    {
      "clip_ratio/high_max": 0.0033753456664271653,
      "clip_ratio/high_mean": 0.001257763993635308,
      "clip_ratio/low_mean": 0.0011084291199949803,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002366193075431511,
      "epoch": 2.279965004374453,
      "grad_norm": 0.2624468505382538,
      "learning_rate": 1e-06,
      "loss": -0.0825,
      "step": 976
    },
    {
      "clip_ratio/high_max": 0.0023874520557001233,
      "clip_ratio/high_mean": 0.0010449714463902637,
      "clip_ratio/low_mean": 0.0005359301885619061,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015809016913408414,
      "completions/clipped_ratio": 0.1674107142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3252.0,
      "completions/mean_length": 1222.703125,
      "completions/mean_terminated_length": 644.9624633789062,
      "completions/min_length": 174.0,
      "completions/min_terminated_length": 174.0,
      "epoch": 2.2822980460775737,
      "grad_norm": 0.36057326197624207,
      "learning_rate": 1e-06,
      "loss": -0.0536,
      "num_tokens": 144001986.0,
      "reward": 0.578125,
      "reward_std": 0.17299975454807281,
      "rewards/verify_math_reward/mean": 0.578125,
      "rewards/verify_math_reward/std": 0.4941346049308777,
      "step": 977
    },
    {
      "clip_ratio/high_max": 0.0027978644429822452,
      "clip_ratio/high_mean": 0.0011361790857336018,
      "clip_ratio/low_mean": 0.0007619837706442922,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018981628454639576,
      "epoch": 2.284631087780694,
      "grad_norm": 0.2731843888759613,
      "learning_rate": 1e-06,
      "loss": -0.0537,
      "step": 978
    },
    {
      "clip_ratio/high_max": 0.002700283053854946,
      "clip_ratio/high_mean": 0.0011772612706408836,
      "clip_ratio/low_mean": 0.000963448102993425,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002140709380910266,
      "epoch": 2.2869641294838146,
      "grad_norm": 0.2100088894367218,
      "learning_rate": 1e-06,
      "loss": -0.0539,
      "step": 979
    },
    {
      "clip_ratio/high_max": 0.00309593380370643,
      "clip_ratio/high_mean": 0.0012225247410242446,
      "clip_ratio/low_mean": 0.001080663063476095,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002303187829966191,
      "epoch": 2.289297171186935,
      "grad_norm": 0.21596242487430573,
      "learning_rate": 1e-06,
      "loss": -0.0539,
      "step": 980
    },
    {
      "clip_ratio/high_max": 0.002694579445233103,
      "clip_ratio/high_mean": 0.0012348649215709884,
      "clip_ratio/low_mean": 0.0006861911151645472,
      "clip_ratio/low_min": 4.292476296541281e-05,
      "clip_ratio/region_mean": 0.0019210560130886734,
      "completions/clipped_ratio": 0.1595982142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4012.0,
      "completions/mean_length": 1203.9710693359375,
      "completions/mean_terminated_length": 654.7543334960938,
      "completions/min_length": 151.0,
      "completions/min_terminated_length": 151.0,
      "epoch": 2.2916302128900554,
      "grad_norm": 0.28912052512168884,
      "learning_rate": 1e-06,
      "loss": -0.0814,
      "num_tokens": 144588008.0,
      "reward": 0.5167410969734192,
      "reward_std": 0.20376215875148773,
      "rewards/verify_math_reward/mean": 0.5167410969734192,
      "rewards/verify_math_reward/std": 0.4999987483024597,
      "step": 981
    },
    {
      "clip_ratio/high_max": 0.002998442665557377,
      "clip_ratio/high_mean": 0.0013558516875491478,
      "clip_ratio/low_mean": 0.0009762507397681475,
      "clip_ratio/low_min": 6.391544957295991e-05,
      "clip_ratio/region_mean": 0.0023321025000768714,
      "epoch": 2.2939632545931756,
      "grad_norm": 0.2508056163787842,
      "learning_rate": 1e-06,
      "loss": -0.0816,
      "step": 982
    },
    {
      "clip_ratio/high_max": 0.0032239833162748255,
      "clip_ratio/high_mean": 0.001394302129483549,
      "clip_ratio/low_mean": 0.0011156422642670805,
      "clip_ratio/low_min": 7.90116610005498e-05,
      "clip_ratio/region_mean": 0.002509944373741746,
      "epoch": 2.2962962962962963,
      "grad_norm": 0.2622165083885193,
      "learning_rate": 1e-06,
      "loss": -0.0818,
      "step": 983
    },
    {
      "clip_ratio/high_max": 0.0029617280670208856,
      "clip_ratio/high_mean": 0.0013805006765323924,
      "clip_ratio/low_mean": 0.0013676582530024461,
      "clip_ratio/low_min": 9.943220175046008e-05,
      "clip_ratio/region_mean": 0.0027481589568196796,
      "epoch": 2.298629337999417,
      "grad_norm": 0.2608332335948944,
      "learning_rate": 1e-06,
      "loss": -0.0819,
      "step": 984
    },
    {
      "clip_ratio/high_max": 0.002652029063028749,
      "clip_ratio/high_mean": 0.0007925968584459042,
      "clip_ratio/low_mean": 0.0007248047786561074,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015174016079981811,
      "completions/clipped_ratio": 0.140625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2870.0,
      "completions/mean_length": 1135.5882568359375,
      "completions/mean_terminated_length": 651.1571044921875,
      "completions/min_length": 140.0,
      "completions/min_terminated_length": 140.0,
      "epoch": 2.300962379702537,
      "grad_norm": 0.3000766336917877,
      "learning_rate": 1e-06,
      "loss": -0.0537,
      "num_tokens": 145185087.0,
      "reward": 0.559151828289032,
      "reward_std": 0.1624336987733841,
      "rewards/verify_math_reward/mean": 0.5591517686843872,
      "rewards/verify_math_reward/std": 0.496766060590744,
      "step": 985
    },
    {
      "clip_ratio/high_max": 0.0030103262324701063,
      "clip_ratio/high_mean": 0.0010665687896107556,
      "clip_ratio/low_mean": 0.0009453985421714606,
      "clip_ratio/low_min": 4.060199353261851e-05,
      "clip_ratio/region_mean": 0.002011967313592322,
      "epoch": 2.303295421405658,
      "grad_norm": 0.2218482494354248,
      "learning_rate": 1e-06,
      "loss": -0.054,
      "step": 986
    },
    {
      "clip_ratio/high_max": 0.002894998098781798,
      "clip_ratio/high_mean": 0.0009617551313567674,
      "clip_ratio/low_mean": 0.0010639281244948506,
      "clip_ratio/low_min": 1.2250097825017292e-05,
      "clip_ratio/region_mean": 0.002025683337706141,
      "epoch": 2.305628463108778,
      "grad_norm": 0.22861462831497192,
      "learning_rate": 1e-06,
      "loss": -0.0541,
      "step": 987
    },
    {
      "clip_ratio/high_max": 0.0030679903575219214,
      "clip_ratio/high_mean": 0.0010229348026769003,
      "clip_ratio/low_mean": 0.0012137892845203169,
      "clip_ratio/low_min": 1.2250097825017292e-05,
      "clip_ratio/region_mean": 0.002236724059912376,
      "epoch": 2.3079615048118987,
      "grad_norm": 0.2555590569972992,
      "learning_rate": 1e-06,
      "loss": -0.0542,
      "step": 988
    },
    {
      "clip_ratio/high_max": 0.0031955152153386734,
      "clip_ratio/high_mean": 0.0011687080059346044,
      "clip_ratio/low_mean": 0.0006758173572052328,
      "clip_ratio/low_min": 2.572965604485944e-05,
      "clip_ratio/region_mean": 0.0018445253372192383,
      "completions/clipped_ratio": 0.1662946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2483.0,
      "completions/mean_length": 1158.669677734375,
      "completions/mean_terminated_length": 572.7764282226562,
      "completions/min_length": 175.0,
      "completions/min_terminated_length": 175.0,
      "epoch": 2.310294546515019,
      "grad_norm": 0.3772180378437042,
      "learning_rate": 1e-06,
      "loss": -0.066,
      "num_tokens": 145703951.0,
      "reward": 0.5558035969734192,
      "reward_std": 0.1875448375940323,
      "rewards/verify_math_reward/mean": 0.5558035969734192,
      "rewards/verify_math_reward/std": 0.49715372920036316,
      "step": 989
    },
    {
      "clip_ratio/high_max": 0.0033459850965300575,
      "clip_ratio/high_mean": 0.001259680666407803,
      "clip_ratio/low_mean": 0.0010045460803667083,
      "clip_ratio/low_min": 3.2691279557184316e-05,
      "clip_ratio/region_mean": 0.0022642267504124902,
      "epoch": 2.3126275882181395,
      "grad_norm": 0.29148250818252563,
      "learning_rate": 1e-06,
      "loss": -0.0664,
      "step": 990
    },
    {
      "clip_ratio/high_max": 0.003272478556027636,
      "clip_ratio/high_mean": 0.0013310912727320101,
      "clip_ratio/low_mean": 0.001179783198494988,
      "clip_ratio/low_min": 2.8312570066191256e-05,
      "clip_ratio/region_mean": 0.0025108745030593127,
      "epoch": 2.3149606299212597,
      "grad_norm": 0.3249438405036926,
      "learning_rate": 1e-06,
      "loss": -0.0665,
      "step": 991
    },
    {
      "clip_ratio/high_max": 0.0033494883828097954,
      "clip_ratio/high_mean": 0.0013296841680130456,
      "clip_ratio/low_mean": 0.0014279158840508899,
      "clip_ratio/low_min": 8.391755181946792e-05,
      "clip_ratio/region_mean": 0.0027576000647968613,
      "epoch": 2.3172936716243804,
      "grad_norm": 0.2720489203929901,
      "learning_rate": 1e-06,
      "loss": -0.0667,
      "step": 992
    },
    {
      "clip_ratio/high_max": 0.0027341580935171805,
      "clip_ratio/high_mean": 0.0010409573660581373,
      "clip_ratio/low_mean": 0.000581146584408998,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001622103933186736,
      "completions/clipped_ratio": 0.1529017857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3695.0,
      "completions/mean_length": 1190.938720703125,
      "completions/mean_terminated_length": 666.5731201171875,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 2.3196267133275006,
      "grad_norm": 0.39617666602134705,
      "learning_rate": 1e-06,
      "loss": -0.0477,
      "num_tokens": 146300536.0,
      "reward": 0.5390625,
      "reward_std": 0.16044628620147705,
      "rewards/verify_math_reward/mean": 0.5390625,
      "rewards/verify_math_reward/std": 0.4987502098083496,
      "step": 993
    },
    {
      "clip_ratio/high_max": 0.0033796231291489676,
      "clip_ratio/high_mean": 0.001176894274976803,
      "clip_ratio/low_mean": 0.0008453242789983051,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020222185557940975,
      "epoch": 2.3219597550306212,
      "grad_norm": 0.31302952766418457,
      "learning_rate": 1e-06,
      "loss": -0.0478,
      "step": 994
    },
    {
      "clip_ratio/high_max": 0.0033353502658428624,
      "clip_ratio/high_mean": 0.0012012331644655205,
      "clip_ratio/low_mean": 0.0009508791226835456,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021521122835110873,
      "epoch": 2.3242927967337415,
      "grad_norm": 0.2497008889913559,
      "learning_rate": 1e-06,
      "loss": -0.0481,
      "step": 995
    },
    {
      "clip_ratio/high_max": 0.0029132952477084473,
      "clip_ratio/high_mean": 0.0010657403581717517,
      "clip_ratio/low_mean": 0.0011637679217528785,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002229508289019577,
      "epoch": 2.326625838436862,
      "grad_norm": 0.224105566740036,
      "learning_rate": 1e-06,
      "loss": -0.0482,
      "step": 996
    },
    {
      "clip_ratio/high_max": 0.0020946715885656886,
      "clip_ratio/high_mean": 0.0008001404403330525,
      "clip_ratio/low_mean": 0.00042710497018561,
      "clip_ratio/low_min": 1.6129031791933812e-05,
      "clip_ratio/region_mean": 0.0012272453859623056,
      "completions/clipped_ratio": 0.125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3185.0,
      "completions/mean_length": 1075.9140625,
      "completions/mean_terminated_length": 644.4732055664062,
      "completions/min_length": 179.0,
      "completions/min_terminated_length": 179.0,
      "epoch": 2.3289588801399823,
      "grad_norm": 0.2584588825702667,
      "learning_rate": 1e-06,
      "loss": -0.0593,
      "num_tokens": 146888867.0,
      "reward": 0.6071428656578064,
      "reward_std": 0.15706203877925873,
      "rewards/verify_math_reward/mean": 0.6071428656578064,
      "rewards/verify_math_reward/std": 0.48865827918052673,
      "step": 997
    },
    {
      "clip_ratio/high_max": 0.002415469105471857,
      "clip_ratio/high_mean": 0.0009024233272612037,
      "clip_ratio/low_mean": 0.0006431267738662427,
      "clip_ratio/low_min": 5.172770397621207e-05,
      "clip_ratio/region_mean": 0.0015455501197720878,
      "epoch": 2.331291921843103,
      "grad_norm": 0.29324856400489807,
      "learning_rate": 1e-06,
      "loss": -0.0595,
      "step": 998
    },
    {
      "clip_ratio/high_max": 0.0024191764314309694,
      "clip_ratio/high_mean": 0.0009480474327574484,
      "clip_ratio/low_mean": 0.0008097914069367107,
      "clip_ratio/low_min": 1.6129031791933812e-05,
      "clip_ratio/region_mean": 0.0017578388178662863,
      "epoch": 2.3336249635462236,
      "grad_norm": 0.20104141533374786,
      "learning_rate": 1e-06,
      "loss": -0.0596,
      "step": 999
    },
    {
      "clip_ratio/high_max": 0.0022805830703873653,
      "clip_ratio/high_mean": 0.0008394501201109961,
      "clip_ratio/low_mean": 0.0008450898640148807,
      "clip_ratio/low_min": 6.451612716773525e-05,
      "clip_ratio/region_mean": 0.0016845399732119404,
      "epoch": 2.335958005249344,
      "grad_norm": 0.20965102314949036,
      "learning_rate": 1e-06,
      "loss": -0.0597,
      "step": 1000
    },
    {
      "clip_ratio/high_max": 0.002165381643862929,
      "clip_ratio/high_mean": 0.000781822514909436,
      "clip_ratio/low_mean": 0.0005456448689074023,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013274673838168383,
      "completions/clipped_ratio": 0.1573660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3491.0,
      "completions/mean_length": 1153.9107666015625,
      "completions/mean_terminated_length": 604.4609375,
      "completions/min_length": 130.0,
      "completions/min_terminated_length": 130.0,
      "epoch": 2.338291046952464,
      "grad_norm": 0.3027136027812958,
      "learning_rate": 1e-06,
      "loss": -0.0542,
      "num_tokens": 147437075.0,
      "reward": 0.5334821939468384,
      "reward_std": 0.13500885665416718,
      "rewards/verify_math_reward/mean": 0.5334821343421936,
      "rewards/verify_math_reward/std": 0.49915632605552673,
      "step": 1001
    },
    {
      "clip_ratio/high_max": 0.0026408979974803515,
      "clip_ratio/high_mean": 0.0009627417293813778,
      "clip_ratio/low_mean": 0.0007521888182964176,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017149305313068908,
      "epoch": 2.3406240886555847,
      "grad_norm": 0.25148019194602966,
      "learning_rate": 1e-06,
      "loss": -0.0543,
      "step": 1002
    },
    {
      "clip_ratio/high_max": 0.0027020912239095196,
      "clip_ratio/high_mean": 0.0009959432627510978,
      "clip_ratio/low_mean": 0.0009145451731455978,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019104883976979181,
      "epoch": 2.3429571303587053,
      "grad_norm": 0.20745614171028137,
      "learning_rate": 1e-06,
      "loss": -0.0545,
      "step": 1003
    },
    {
      "clip_ratio/high_max": 0.002835997562215198,
      "clip_ratio/high_mean": 0.000995350506855175,
      "clip_ratio/low_mean": 0.0011339418560964987,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002129292370227631,
      "epoch": 2.3452901720618256,
      "grad_norm": 0.23285433650016785,
      "learning_rate": 1e-06,
      "loss": -0.0546,
      "step": 1004
    },
    {
      "clip_ratio/high_max": 0.0025864409108180553,
      "clip_ratio/high_mean": 0.0009867948083410738,
      "clip_ratio/low_mean": 0.0006698022280033911,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016565970581723377,
      "completions/clipped_ratio": 0.15625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2938.0,
      "completions/mean_length": 1144.094970703125,
      "completions/mean_terminated_length": 597.4457397460938,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 2.347623213764946,
      "grad_norm": 0.32017406821250916,
      "learning_rate": 1e-06,
      "loss": -0.0705,
      "num_tokens": 147974072.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.17277081310749054,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 1005
    },
    {
      "clip_ratio/high_max": 0.003541181984473951,
      "clip_ratio/high_mean": 0.0012657882925850572,
      "clip_ratio/low_mean": 0.0010236859016004018,
      "clip_ratio/low_min": 2.3710166715318337e-05,
      "clip_ratio/region_mean": 0.0022894742214703,
      "epoch": 2.3499562554680664,
      "grad_norm": 0.2364875078201294,
      "learning_rate": 1e-06,
      "loss": -0.0709,
      "step": 1006
    },
    {
      "clip_ratio/high_max": 0.0031734051372040994,
      "clip_ratio/high_mean": 0.0012652313234866597,
      "clip_ratio/low_mean": 0.0011457300988695351,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002410961387795396,
      "epoch": 2.352289297171187,
      "grad_norm": 0.245115265250206,
      "learning_rate": 1e-06,
      "loss": -0.0709,
      "step": 1007
    },
    {
      "clip_ratio/high_max": 0.0032782024645712227,
      "clip_ratio/high_mean": 0.0012094823105144314,
      "clip_ratio/low_mean": 0.0013976999325677752,
      "clip_ratio/low_min": 2.3710166715318337e-05,
      "clip_ratio/region_mean": 0.002607182163046673,
      "epoch": 2.3546223388743073,
      "grad_norm": 0.21835707128047943,
      "learning_rate": 1e-06,
      "loss": -0.0711,
      "step": 1008
    },
    {
      "clip_ratio/high_max": 0.0016837205039337277,
      "clip_ratio/high_mean": 0.0006111365655669942,
      "clip_ratio/low_mean": 0.00039568928423250327,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0010068258379760664,
      "completions/clipped_ratio": 0.1450892857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2747.0,
      "completions/mean_length": 1125.196533203125,
      "completions/mean_terminated_length": 621.0130615234375,
      "completions/min_length": 155.0,
      "completions/min_terminated_length": 155.0,
      "epoch": 2.356955380577428,
      "grad_norm": 0.2512587904930115,
      "learning_rate": 1e-06,
      "loss": -0.0546,
      "num_tokens": 148534904.0,
      "reward": 0.629464328289032,
      "reward_std": 0.13534656167030334,
      "rewards/verify_math_reward/mean": 0.6294642686843872,
      "rewards/verify_math_reward/std": 0.4832179844379425,
      "step": 1009
    },
    {
      "clip_ratio/high_max": 0.002148231546016177,
      "clip_ratio/high_mean": 0.0008400733331654919,
      "clip_ratio/low_mean": 0.0005824126833431365,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014224859842215665,
      "epoch": 2.359288422280548,
      "grad_norm": 0.18119673430919647,
      "learning_rate": 1e-06,
      "loss": -0.0548,
      "step": 1010
    },
    {
      "clip_ratio/high_max": 0.0024765461348579265,
      "clip_ratio/high_mean": 0.0009033664991875412,
      "clip_ratio/low_mean": 0.0006529972429234476,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015563637716695666,
      "epoch": 2.361621463983669,
      "grad_norm": 0.18178009986877441,
      "learning_rate": 1e-06,
      "loss": -0.0549,
      "step": 1011
    },
    {
      "clip_ratio/high_max": 0.0024374165113840718,
      "clip_ratio/high_mean": 0.000785118581916322,
      "clip_ratio/low_mean": 0.0007713841982877057,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015565027970296796,
      "epoch": 2.363954505686789,
      "grad_norm": 0.16140949726104736,
      "learning_rate": 1e-06,
      "loss": -0.055,
      "step": 1012
    },
    {
      "clip_ratio/high_max": 0.001982963345653843,
      "clip_ratio/high_mean": 0.0006908731138537405,
      "clip_ratio/low_mean": 0.0006206093303262605,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013114824396325275,
      "completions/clipped_ratio": 0.1450892857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2940.0,
      "completions/mean_length": 1175.548095703125,
      "completions/mean_terminated_length": 679.909912109375,
      "completions/min_length": 170.0,
      "completions/min_terminated_length": 170.0,
      "epoch": 2.3662875473899097,
      "grad_norm": 0.28620150685310364,
      "learning_rate": 1e-06,
      "loss": -0.0503,
      "num_tokens": 149138171.0,
      "reward": 0.5613839626312256,
      "reward_std": 0.13654935359954834,
      "rewards/verify_math_reward/mean": 0.5613839030265808,
      "rewards/verify_math_reward/std": 0.496494859457016,
      "step": 1013
    },
    {
      "clip_ratio/high_max": 0.0024065280231297947,
      "clip_ratio/high_mean": 0.0007314348376894486,
      "clip_ratio/low_mean": 0.000729183364455821,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014606182121497113,
      "epoch": 2.36862058909303,
      "grad_norm": 0.20217272639274597,
      "learning_rate": 1e-06,
      "loss": -0.0505,
      "step": 1014
    },
    {
      "clip_ratio/high_max": 0.002407199404842686,
      "clip_ratio/high_mean": 0.0007553441428171936,
      "clip_ratio/low_mean": 0.0008421047932642978,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015974489360814914,
      "epoch": 2.3709536307961505,
      "grad_norm": 0.22331973910331726,
      "learning_rate": 1e-06,
      "loss": -0.0505,
      "step": 1015
    },
    {
      "clip_ratio/high_max": 0.0020915623936161865,
      "clip_ratio/high_mean": 0.0007299595235963352,
      "clip_ratio/low_mean": 0.0010612890218908433,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017912485091073904,
      "epoch": 2.3732866724992707,
      "grad_norm": 0.20704081654548645,
      "learning_rate": 1e-06,
      "loss": -0.0506,
      "step": 1016
    },
    {
      "clip_ratio/high_max": 0.0034728839236777276,
      "clip_ratio/high_mean": 0.0011303817245789105,
      "clip_ratio/low_mean": 0.0005705017920263344,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017008835238812026,
      "completions/clipped_ratio": 0.1238839285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3766.0,
      "completions/mean_length": 1060.7645263671875,
      "completions/mean_terminated_length": 631.578369140625,
      "completions/min_length": 157.0,
      "completions/min_terminated_length": 157.0,
      "epoch": 2.3756197142023914,
      "grad_norm": 0.36774662137031555,
      "learning_rate": 1e-06,
      "loss": -0.057,
      "num_tokens": 149718992.0,
      "reward": 0.6171875,
      "reward_std": 0.1595052033662796,
      "rewards/verify_math_reward/mean": 0.6171875,
      "rewards/verify_math_reward/std": 0.4863446056842804,
      "step": 1017
    },
    {
      "clip_ratio/high_max": 0.0044531360326800495,
      "clip_ratio/high_mean": 0.0013801455934299156,
      "clip_ratio/low_mean": 0.000763638318858284,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002143783975043334,
      "epoch": 2.377952755905512,
      "grad_norm": 0.28642094135284424,
      "learning_rate": 1e-06,
      "loss": -0.0574,
      "step": 1018
    },
    {
      "clip_ratio/high_max": 0.005310886037477758,
      "clip_ratio/high_mean": 0.0015749102640256751,
      "clip_ratio/low_mean": 0.0009763622192622279,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0025512724605505355,
      "epoch": 2.3802857976086322,
      "grad_norm": 0.30652979016304016,
      "learning_rate": 1e-06,
      "loss": -0.0576,
      "step": 1019
    },
    {
      "clip_ratio/high_max": 0.005440009721496608,
      "clip_ratio/high_mean": 0.001479281030697166,
      "clip_ratio/low_mean": 0.0011621802077570464,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026414612657390535,
      "epoch": 2.382618839311753,
      "grad_norm": 0.2900792062282562,
      "learning_rate": 1e-06,
      "loss": -0.0576,
      "step": 1020
    },
    {
      "clip_ratio/high_max": 0.0016350403457181528,
      "clip_ratio/high_mean": 0.0006149514538265066,
      "clip_ratio/low_mean": 0.00037639642846443166,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.000991347873423365,
      "completions/clipped_ratio": 0.1227678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3706.0,
      "completions/mean_length": 1115.727783203125,
      "completions/mean_terminated_length": 698.6412353515625,
      "completions/min_length": 171.0,
      "completions/min_terminated_length": 171.0,
      "epoch": 2.384951881014873,
      "grad_norm": 0.22175031900405884,
      "learning_rate": 1e-06,
      "loss": -0.0308,
      "num_tokens": 150361820.0,
      "reward": 0.5491071939468384,
      "reward_std": 0.13354600965976715,
      "rewards/verify_math_reward/mean": 0.5491071343421936,
      "rewards/verify_math_reward/std": 0.49786055088043213,
      "step": 1021
    },
    {
      "clip_ratio/high_max": 0.001988006792089436,
      "clip_ratio/high_mean": 0.000717979422006465,
      "clip_ratio/low_mean": 0.0005121810263517546,
      "clip_ratio/low_min": 1.3724198652198538e-05,
      "clip_ratio/region_mean": 0.0012301604547246825,
      "epoch": 2.3872849227179938,
      "grad_norm": 0.21414272487163544,
      "learning_rate": 1e-06,
      "loss": -0.0309,
      "step": 1022
    },
    {
      "clip_ratio/high_max": 0.0021304474903445225,
      "clip_ratio/high_mean": 0.0008804438002698589,
      "clip_ratio/low_mean": 0.000570450026771141,
      "clip_ratio/low_min": 1.941295158758294e-05,
      "clip_ratio/region_mean": 0.001450893840228673,
      "epoch": 2.389617964421114,
      "grad_norm": 0.22635546326637268,
      "learning_rate": 1e-06,
      "loss": -0.0311,
      "step": 1023
    },
    {
      "clip_ratio/high_max": 0.0018546131504990626,
      "clip_ratio/high_mean": 0.000691392498993082,
      "clip_ratio/low_mean": 0.000757499012252083,
      "clip_ratio/low_min": 2.911942829086911e-05,
      "clip_ratio/region_mean": 0.0014488915076071862,
      "epoch": 2.3919510061242346,
      "grad_norm": 0.17617785930633545,
      "learning_rate": 1e-06,
      "loss": -0.0311,
      "step": 1024
    },
    {
      "clip_ratio/high_max": 0.002508030127501115,
      "clip_ratio/high_mean": 0.0010209910560661228,
      "clip_ratio/low_mean": 0.0005794368807983119,
      "clip_ratio/low_min": 1.5337423974415287e-05,
      "clip_ratio/region_mean": 0.0016004279714252334,
      "completions/clipped_ratio": 0.1339285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4045.0,
      "completions/mean_length": 1116.724365234375,
      "completions/mean_terminated_length": 656.0115966796875,
      "completions/min_length": 144.0,
      "completions/min_terminated_length": 144.0,
      "epoch": 2.394284047827355,
      "grad_norm": 0.2999490201473236,
      "learning_rate": 1e-06,
      "loss": -0.055,
      "num_tokens": 150959397.0,
      "reward": 0.578125,
      "reward_std": 0.17739249765872955,
      "rewards/verify_math_reward/mean": 0.578125,
      "rewards/verify_math_reward/std": 0.4941346049308777,
      "step": 1025
    },
    {
      "clip_ratio/high_max": 0.002997482515638694,
      "clip_ratio/high_mean": 0.0012299517911742441,
      "clip_ratio/low_mean": 0.0007308456479222514,
      "clip_ratio/low_min": 1.5337423974415287e-05,
      "clip_ratio/region_mean": 0.001960797468200326,
      "epoch": 2.3966170895304755,
      "grad_norm": 0.2786708176136017,
      "learning_rate": 1e-06,
      "loss": -0.0553,
      "step": 1026
    },
    {
      "clip_ratio/high_max": 0.003061028866795823,
      "clip_ratio/high_mean": 0.0012740095735352952,
      "clip_ratio/low_mean": 0.0009666674104664708,
      "clip_ratio/low_min": 1.5337423974415287e-05,
      "clip_ratio/region_mean": 0.002240676956716925,
      "epoch": 2.3989501312335957,
      "grad_norm": 0.30744946002960205,
      "learning_rate": 1e-06,
      "loss": -0.0554,
      "step": 1027
    },
    {
      "clip_ratio/high_max": 0.0029087264701956883,
      "clip_ratio/high_mean": 0.0011492716112115886,
      "clip_ratio/low_mean": 0.0011242707532801433,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022735423772246577,
      "epoch": 2.4012831729367163,
      "grad_norm": 0.19937247037887573,
      "learning_rate": 1e-06,
      "loss": -0.0555,
      "step": 1028
    },
    {
      "clip_ratio/high_max": 0.0026239179860567674,
      "clip_ratio/high_mean": 0.0011437333923822735,
      "clip_ratio/low_mean": 0.0006050125793990446,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017487459990661591,
      "completions/clipped_ratio": 0.1205357142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3868.0,
      "completions/mean_length": 1022.3281860351562,
      "completions/mean_terminated_length": 601.0634155273438,
      "completions/min_length": 164.0,
      "completions/min_terminated_length": 164.0,
      "epoch": 2.4036162146398365,
      "grad_norm": 0.3392671048641205,
      "learning_rate": 1e-06,
      "loss": -0.0527,
      "num_tokens": 151517939.0,
      "reward": 0.625,
      "reward_std": 0.1799846738576889,
      "rewards/verify_math_reward/mean": 0.625,
      "rewards/verify_math_reward/std": 0.48439329862594604,
      "step": 1029
    },
    {
      "clip_ratio/high_max": 0.002875018493796233,
      "clip_ratio/high_mean": 0.0012252770939085167,
      "clip_ratio/low_mean": 0.0008593679367550067,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002084644991555251,
      "epoch": 2.405949256342957,
      "grad_norm": 0.3104827404022217,
      "learning_rate": 1e-06,
      "loss": -0.053,
      "step": 1030
    },
    {
      "clip_ratio/high_max": 0.0031917464584694244,
      "clip_ratio/high_mean": 0.001287612994929077,
      "clip_ratio/low_mean": 0.0010495392143639037,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023371522111119702,
      "epoch": 2.4082822980460774,
      "grad_norm": 0.29952681064605713,
      "learning_rate": 1e-06,
      "loss": -0.0532,
      "step": 1031
    },
    {
      "clip_ratio/high_max": 0.0029809527841280214,
      "clip_ratio/high_mean": 0.0011874449846800417,
      "clip_ratio/low_mean": 0.0012573567855724832,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024448017647955567,
      "epoch": 2.410615339749198,
      "grad_norm": 0.24343644082546234,
      "learning_rate": 1e-06,
      "loss": -0.0533,
      "step": 1032
    },
    {
      "clip_ratio/high_max": 0.002099279656249564,
      "clip_ratio/high_mean": 0.0008029598102439195,
      "clip_ratio/low_mean": 0.000499091003803187,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013020508013141807,
      "completions/clipped_ratio": 0.1149553571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3196.0,
      "completions/mean_length": 1026.536865234375,
      "completions/mean_terminated_length": 627.85498046875,
      "completions/min_length": 184.0,
      "completions/min_terminated_length": 184.0,
      "epoch": 2.4129483814523183,
      "grad_norm": 0.28373295068740845,
      "learning_rate": 1e-06,
      "loss": -0.0644,
      "num_tokens": 152107364.0,
      "reward": 0.6160714626312256,
      "reward_std": 0.16743192076683044,
      "rewards/verify_math_reward/mean": 0.6160714030265808,
      "rewards/verify_math_reward/std": 0.486612468957901,
      "step": 1033
    },
    {
      "clip_ratio/high_max": 0.0028957660470041446,
      "clip_ratio/high_mean": 0.0010994240874424577,
      "clip_ratio/low_mean": 0.000717490160241141,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018169142858823761,
      "epoch": 2.415281423155439,
      "grad_norm": 0.22463740408420563,
      "learning_rate": 1e-06,
      "loss": -0.0646,
      "step": 1034
    },
    {
      "clip_ratio/high_max": 0.0024197215352614876,
      "clip_ratio/high_mean": 0.0009790176372916903,
      "clip_ratio/low_mean": 0.0008822561085253255,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00186127373308409,
      "epoch": 2.417614464858559,
      "grad_norm": 0.24838964641094208,
      "learning_rate": 1e-06,
      "loss": -0.0648,
      "step": 1035
    },
    {
      "clip_ratio/high_max": 0.002677409182069823,
      "clip_ratio/high_mean": 0.0010236194138997234,
      "clip_ratio/low_mean": 0.0010017153799708467,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002025334812060464,
      "epoch": 2.41994750656168,
      "grad_norm": 0.21755583584308624,
      "learning_rate": 1e-06,
      "loss": -0.0648,
      "step": 1036
    },
    {
      "clip_ratio/high_max": 0.0024117432403727435,
      "clip_ratio/high_mean": 0.0010259137743560132,
      "clip_ratio/low_mean": 0.0008399565867875936,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018658704066183418,
      "completions/clipped_ratio": 0.1685267857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3989.0,
      "completions/mean_length": 1220.51904296875,
      "completions/mean_terminated_length": 637.703369140625,
      "completions/min_length": 146.0,
      "completions/min_terminated_length": 146.0,
      "epoch": 2.4222805482648004,
      "grad_norm": 0.39747315645217896,
      "learning_rate": 1e-06,
      "loss": -0.0534,
      "num_tokens": 152674085.0,
      "reward": 0.5100446939468384,
      "reward_std": 0.2037724405527115,
      "rewards/verify_math_reward/mean": 0.5100446343421936,
      "rewards/verify_math_reward/std": 0.5001782774925232,
      "step": 1037
    },
    {
      "clip_ratio/high_max": 0.0029730666137766093,
      "clip_ratio/high_mean": 0.00125672075591865,
      "clip_ratio/low_mean": 0.0010882800452236552,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00234500078659039,
      "epoch": 2.4246135899679206,
      "grad_norm": 0.45811083912849426,
      "learning_rate": 1e-06,
      "loss": -0.0538,
      "step": 1038
    },
    {
      "clip_ratio/high_max": 0.0027848283207276836,
      "clip_ratio/high_mean": 0.0011850921364384703,
      "clip_ratio/low_mean": 0.0014201014419086277,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026051936220028438,
      "epoch": 2.4269466316710413,
      "grad_norm": 0.25683873891830444,
      "learning_rate": 1e-06,
      "loss": -0.0539,
      "step": 1039
    },
    {
      "clip_ratio/high_max": 0.0030117457718006335,
      "clip_ratio/high_mean": 0.001205594224302331,
      "clip_ratio/low_mean": 0.001628739686566405,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0028343339363345876,
      "epoch": 2.4292796733741615,
      "grad_norm": 0.2934616208076477,
      "learning_rate": 1e-06,
      "loss": -0.0541,
      "step": 1040
    },
    {
      "clip_ratio/high_max": 0.0023620007559657097,
      "clip_ratio/high_mean": 0.0009517699872958474,
      "clip_ratio/low_mean": 0.0006037206144355878,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015554906567558646,
      "completions/clipped_ratio": 0.1395089285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2973.0,
      "completions/mean_length": 1104.765625,
      "completions/mean_terminated_length": 619.805419921875,
      "completions/min_length": 148.0,
      "completions/min_terminated_length": 148.0,
      "epoch": 2.431612715077282,
      "grad_norm": 0.32303157448768616,
      "learning_rate": 1e-06,
      "loss": -0.0482,
      "num_tokens": 153247235.0,
      "reward": 0.5524553656578064,
      "reward_std": 0.17922443151474,
      "rewards/verify_math_reward/mean": 0.5524553656578064,
      "rewards/verify_math_reward/std": 0.49751853942871094,
      "step": 1041
    },
    {
      "clip_ratio/high_max": 0.0029223363890196197,
      "clip_ratio/high_mean": 0.0011394780394766713,
      "clip_ratio/low_mean": 0.0009088044116651872,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020482824838836677,
      "epoch": 2.4339457567804024,
      "grad_norm": 0.22415462136268616,
      "learning_rate": 1e-06,
      "loss": -0.0485,
      "step": 1042
    },
    {
      "clip_ratio/high_max": 0.002557893552875612,
      "clip_ratio/high_mean": 0.0010014909148594597,
      "clip_ratio/low_mean": 0.0011282869018032216,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002129777829395607,
      "epoch": 2.436278798483523,
      "grad_norm": 0.24331408739089966,
      "learning_rate": 1e-06,
      "loss": -0.0486,
      "step": 1043
    },
    {
      "clip_ratio/high_max": 0.0026559678371995687,
      "clip_ratio/high_mean": 0.0010543445932853501,
      "clip_ratio/low_mean": 0.0012289675414649537,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002283312140207272,
      "epoch": 2.4386118401866432,
      "grad_norm": 0.24428428709506989,
      "learning_rate": 1e-06,
      "loss": -0.0487,
      "step": 1044
    },
    {
      "clip_ratio/high_max": 0.0021358276244427543,
      "clip_ratio/high_mean": 0.000681489653288736,
      "clip_ratio/low_mean": 0.0005759707082688692,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001257460353372153,
      "completions/clipped_ratio": 0.1584821428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2308.0,
      "completions/mean_length": 1167.421875,
      "completions/mean_terminated_length": 615.8859252929688,
      "completions/min_length": 187.0,
      "completions/min_terminated_length": 187.0,
      "epoch": 2.440944881889764,
      "grad_norm": 0.34599801898002625,
      "learning_rate": 1e-06,
      "loss": -0.0598,
      "num_tokens": 153800453.0,
      "reward": 0.5569196939468384,
      "reward_std": 0.1539819985628128,
      "rewards/verify_math_reward/mean": 0.5569196343421936,
      "rewards/verify_math_reward/std": 0.4970270097255707,
      "step": 1045
    },
    {
      "clip_ratio/high_max": 0.0024830914499034407,
      "clip_ratio/high_mean": 0.0008351252122338337,
      "clip_ratio/low_mean": 0.0009217281767632812,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017568534167367034,
      "epoch": 2.443277923592884,
      "grad_norm": 0.23306939005851746,
      "learning_rate": 1e-06,
      "loss": -0.0601,
      "step": 1046
    },
    {
      "clip_ratio/high_max": 0.002836348914570408,
      "clip_ratio/high_mean": 0.0009229661163772107,
      "clip_ratio/low_mean": 0.0010264891225233441,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019494552543619648,
      "epoch": 2.4456109652960047,
      "grad_norm": 0.2903509736061096,
      "learning_rate": 1e-06,
      "loss": -0.0603,
      "step": 1047
    },
    {
      "clip_ratio/high_max": 0.0025230854234905564,
      "clip_ratio/high_mean": 0.0007946755479224521,
      "clip_ratio/low_mean": 0.0012072050776623655,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020018806608277373,
      "epoch": 2.447944006999125,
      "grad_norm": 0.23268888890743256,
      "learning_rate": 1e-06,
      "loss": -0.0604,
      "step": 1048
    },
    {
      "clip_ratio/high_max": 0.0020995121230953373,
      "clip_ratio/high_mean": 0.0006992659818934044,
      "clip_ratio/low_mean": 0.0008514704632034409,
      "clip_ratio/low_min": 1.7846945411292836e-05,
      "clip_ratio/region_mean": 0.001550736455101287,
      "completions/clipped_ratio": 0.1316964285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3066.0,
      "completions/mean_length": 1065.997802734375,
      "completions/mean_terminated_length": 606.4344482421875,
      "completions/min_length": 119.0,
      "completions/min_terminated_length": 119.0,
      "epoch": 2.4502770487022456,
      "grad_norm": 0.3484068810939789,
      "learning_rate": 1e-06,
      "loss": -0.0566,
      "num_tokens": 154357067.0,
      "reward": 0.5647321939468384,
      "reward_std": 0.14943771064281464,
      "rewards/verify_math_reward/mean": 0.5647321343421936,
      "rewards/verify_math_reward/std": 0.49606895446777344,
      "step": 1049
    },
    {
      "clip_ratio/high_max": 0.0028557084588101134,
      "clip_ratio/high_mean": 0.0009366729300381849,
      "clip_ratio/low_mean": 0.0011126888975923066,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002049361872195732,
      "epoch": 2.452610090405366,
      "grad_norm": 0.37618282437324524,
      "learning_rate": 1e-06,
      "loss": -0.0569,
      "step": 1050
    },
    {
      "clip_ratio/high_max": 0.0027996127391816117,
      "clip_ratio/high_mean": 0.0009656961301516276,
      "clip_ratio/low_mean": 0.001358412255285657,
      "clip_ratio/low_min": 1.5879064449109137e-05,
      "clip_ratio/region_mean": 0.002324108405446168,
      "epoch": 2.4549431321084865,
      "grad_norm": 0.2839452922344208,
      "learning_rate": 1e-06,
      "loss": -0.057,
      "step": 1051
    },
    {
      "clip_ratio/high_max": 0.0028057046947651543,
      "clip_ratio/high_mean": 0.0009376704547321424,
      "clip_ratio/low_mean": 0.001566952392749954,
      "clip_ratio/low_min": 1.984126902243588e-05,
      "clip_ratio/region_mean": 0.0025046228474820964,
      "epoch": 2.457276173811607,
      "grad_norm": 0.2504903972148895,
      "learning_rate": 1e-06,
      "loss": -0.0571,
      "step": 1052
    },
    {
      "clip_ratio/high_max": 0.0025813447864493355,
      "clip_ratio/high_mean": 0.0010606997839204269,
      "clip_ratio/low_mean": 0.0006247893743420718,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016854891619004775,
      "completions/clipped_ratio": 0.1417410714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3553.0,
      "completions/mean_length": 1123.673095703125,
      "completions/mean_terminated_length": 632.7945556640625,
      "completions/min_length": 163.0,
      "completions/min_terminated_length": 163.0,
      "epoch": 2.4596092155147273,
      "grad_norm": 0.30488279461860657,
      "learning_rate": 1e-06,
      "loss": -0.0629,
      "num_tokens": 154937918.0,
      "reward": 0.543526828289032,
      "reward_std": 0.1814917027950287,
      "rewards/verify_math_reward/mean": 0.5435267686843872,
      "rewards/verify_math_reward/std": 0.49838000535964966,
      "step": 1053
    },
    {
      "clip_ratio/high_max": 0.0030096502196101937,
      "clip_ratio/high_mean": 0.0012082152516086353,
      "clip_ratio/low_mean": 0.0008828598911350127,
      "clip_ratio/low_min": 1.0511267646506894e-05,
      "clip_ratio/region_mean": 0.002091075155476574,
      "epoch": 2.4619422572178475,
      "grad_norm": 0.2740449011325836,
      "learning_rate": 1e-06,
      "loss": -0.0631,
      "step": 1054
    },
    {
      "clip_ratio/high_max": 0.0032181386195588857,
      "clip_ratio/high_mean": 0.0013170854817872168,
      "clip_ratio/low_mean": 0.0009968225913326023,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023139080876717344,
      "epoch": 2.464275298920968,
      "grad_norm": 0.2314031422138214,
      "learning_rate": 1e-06,
      "loss": -0.0633,
      "step": 1055
    },
    {
      "clip_ratio/high_max": 0.002719180021813372,
      "clip_ratio/high_mean": 0.001063241807059967,
      "clip_ratio/low_mean": 0.0011379034131095978,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022011452019796707,
      "epoch": 2.466608340624089,
      "grad_norm": 0.2776223123073578,
      "learning_rate": 1e-06,
      "loss": -0.0633,
      "step": 1056
    },
    {
      "clip_ratio/high_max": 0.00232488347319304,
      "clip_ratio/high_mean": 0.0007725332707195776,
      "clip_ratio/low_mean": 0.0004357351265298348,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001208268353366293,
      "completions/clipped_ratio": 0.171875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3643.0,
      "completions/mean_length": 1210.6551513671875,
      "completions/mean_terminated_length": 611.8099365234375,
      "completions/min_length": 106.0,
      "completions/min_terminated_length": 106.0,
      "epoch": 2.468941382327209,
      "grad_norm": 0.33703455328941345,
      "learning_rate": 1e-06,
      "loss": -0.0685,
      "num_tokens": 155480625.0,
      "reward": 0.5625,
      "reward_std": 0.13467255234718323,
      "rewards/verify_math_reward/mean": 0.5625,
      "rewards/verify_math_reward/std": 0.49635544419288635,
      "step": 1057
    },
    {
      "clip_ratio/high_max": 0.002764700235275086,
      "clip_ratio/high_mean": 0.0009506015940132784,
      "clip_ratio/low_mean": 0.0006605244466300064,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016111260156321805,
      "epoch": 2.4712744240303297,
      "grad_norm": 0.3677258789539337,
      "learning_rate": 1e-06,
      "loss": -0.0688,
      "step": 1058
    },
    {
      "clip_ratio/high_max": 0.0029668719434994273,
      "clip_ratio/high_mean": 0.0009761436376720667,
      "clip_ratio/low_mean": 0.0008182404535546084,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017943840939551592,
      "epoch": 2.47360746573345,
      "grad_norm": 0.29488086700439453,
      "learning_rate": 1e-06,
      "loss": -0.0689,
      "step": 1059
    },
    {
      "clip_ratio/high_max": 0.0024077127018244937,
      "clip_ratio/high_mean": 0.000827830983325839,
      "clip_ratio/low_mean": 0.001010290815429471,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018381218214926776,
      "epoch": 2.4759405074365706,
      "grad_norm": 0.22862303256988525,
      "learning_rate": 1e-06,
      "loss": -0.0689,
      "step": 1060
    },
    {
      "clip_ratio/high_max": 0.0029381492131506093,
      "clip_ratio/high_mean": 0.0012637385625566822,
      "clip_ratio/low_mean": 0.0007906910850579152,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020544295985018834,
      "completions/clipped_ratio": 0.1629464285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3153.0,
      "completions/mean_length": 1159.727783203125,
      "completions/mean_terminated_length": 588.13330078125,
      "completions/min_length": 168.0,
      "completions/min_terminated_length": 168.0,
      "epoch": 2.478273549139691,
      "grad_norm": 0.3715926706790924,
      "learning_rate": 1e-06,
      "loss": -0.0747,
      "num_tokens": 156005069.0,
      "reward": 0.5993303656578064,
      "reward_std": 0.1802881807088852,
      "rewards/verify_math_reward/mean": 0.5993303656578064,
      "rewards/verify_math_reward/std": 0.49030786752700806,
      "step": 1061
    },
    {
      "clip_ratio/high_max": 0.0034059446115861647,
      "clip_ratio/high_mean": 0.0013395270325418096,
      "clip_ratio/low_mean": 0.0010574301250017015,
      "clip_ratio/low_min": 1.3001872503082268e-05,
      "clip_ratio/region_mean": 0.002396957170276437,
      "epoch": 2.4806065908428114,
      "grad_norm": 0.29294928908348083,
      "learning_rate": 1e-06,
      "loss": -0.075,
      "step": 1062
    },
    {
      "clip_ratio/high_max": 0.003073756306548603,
      "clip_ratio/high_mean": 0.001338158421276603,
      "clip_ratio/low_mean": 0.0012771279207299813,
      "clip_ratio/low_min": 2.6003745006164536e-05,
      "clip_ratio/region_mean": 0.0026152863647439517,
      "epoch": 2.4829396325459316,
      "grad_norm": 0.29861995577812195,
      "learning_rate": 1e-06,
      "loss": -0.0751,
      "step": 1063
    },
    {
      "clip_ratio/high_max": 0.0038295917911455035,
      "clip_ratio/high_mean": 0.0014337743923533708,
      "clip_ratio/low_mean": 0.0014355370567500358,
      "clip_ratio/low_min": 1.3001872503082268e-05,
      "clip_ratio/region_mean": 0.0028693113999906927,
      "epoch": 2.4852726742490523,
      "grad_norm": 0.27737295627593994,
      "learning_rate": 1e-06,
      "loss": -0.0753,
      "step": 1064
    },
    {
      "clip_ratio/high_max": 0.002714105990889948,
      "clip_ratio/high_mean": 0.0009218911745847436,
      "clip_ratio/low_mean": 0.0005902984266867861,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015121895994525403,
      "completions/clipped_ratio": 0.1428571428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2535.0,
      "completions/mean_length": 1139.15625,
      "completions/mean_terminated_length": 646.3489990234375,
      "completions/min_length": 174.0,
      "completions/min_terminated_length": 174.0,
      "epoch": 2.4876057159521725,
      "grad_norm": 0.2811812162399292,
      "learning_rate": 1e-06,
      "loss": -0.0534,
      "num_tokens": 156587673.0,
      "reward": 0.5948660969734192,
      "reward_std": 0.15857158601284027,
      "rewards/verify_math_reward/mean": 0.5948660969734192,
      "rewards/verify_math_reward/std": 0.49119213223457336,
      "step": 1065
    },
    {
      "clip_ratio/high_max": 0.0030347594947670586,
      "clip_ratio/high_mean": 0.001111539651901694,
      "clip_ratio/low_mean": 0.0007866604082664708,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018982000910909846,
      "epoch": 2.489938757655293,
      "grad_norm": 0.260781854391098,
      "learning_rate": 1e-06,
      "loss": -0.0537,
      "step": 1066
    },
    {
      "clip_ratio/high_max": 0.0033222405400010757,
      "clip_ratio/high_mean": 0.0011177840005984763,
      "clip_ratio/low_mean": 0.0009584187682776246,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020762028070748784,
      "epoch": 2.4922717993584134,
      "grad_norm": 0.2341073453426361,
      "learning_rate": 1e-06,
      "loss": -0.0539,
      "step": 1067
    },
    {
      "clip_ratio/high_max": 0.002542661844927352,
      "clip_ratio/high_mean": 0.0010024788025475573,
      "clip_ratio/low_mean": 0.0011134628221043386,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002115941599186044,
      "epoch": 2.494604841061534,
      "grad_norm": 0.237702876329422,
      "learning_rate": 1e-06,
      "loss": -0.0539,
      "step": 1068
    },
    {
      "clip_ratio/high_max": 0.0029128591049811803,
      "clip_ratio/high_mean": 0.001122817240684526,
      "clip_ratio/low_mean": 0.0003881173327044962,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015109345877135638,
      "completions/clipped_ratio": 0.1964285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3621.0,
      "completions/mean_length": 1375.216552734375,
      "completions/mean_terminated_length": 710.1361083984375,
      "completions/min_length": 180.0,
      "completions/min_terminated_length": 180.0,
      "epoch": 2.4969378827646542,
      "grad_norm": 0.30831193923950195,
      "learning_rate": 1e-06,
      "loss": -0.0844,
      "num_tokens": 157197947.0,
      "reward": 0.5401785969734192,
      "reward_std": 0.16841016709804535,
      "rewards/verify_math_reward/mean": 0.5401785969734192,
      "rewards/verify_math_reward/std": 0.49866142868995667,
      "step": 1069
    },
    {
      "clip_ratio/high_max": 0.003005316306371242,
      "clip_ratio/high_mean": 0.0012329331184446346,
      "clip_ratio/low_mean": 0.0005590850987573504,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017920181853696704,
      "epoch": 2.499270924467775,
      "grad_norm": 0.29480475187301636,
      "learning_rate": 1e-06,
      "loss": -0.0846,
      "step": 1070
    },
    {
      "clip_ratio/high_max": 0.003337475980515592,
      "clip_ratio/high_mean": 0.0013035204883635743,
      "clip_ratio/low_mean": 0.0007201540975074749,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020236746131558903,
      "epoch": 2.5016039661708955,
      "grad_norm": 0.21608605980873108,
      "learning_rate": 1e-06,
      "loss": -0.0848,
      "step": 1071
    },
    {
      "clip_ratio/high_max": 0.0032416925678262487,
      "clip_ratio/high_mean": 0.001307557919062674,
      "clip_ratio/low_mean": 0.0008353501461897395,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021429080952657387,
      "epoch": 2.5039370078740157,
      "grad_norm": 0.2243119776248932,
      "learning_rate": 1e-06,
      "loss": -0.0849,
      "step": 1072
    },
    {
      "clip_ratio/high_max": 0.0021912348056503106,
      "clip_ratio/high_mean": 0.0009612147186999209,
      "clip_ratio/low_mean": 0.0006734274684276897,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00163464221986942,
      "completions/clipped_ratio": 0.1517857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3212.0,
      "completions/mean_length": 1191.6373291015625,
      "completions/mean_terminated_length": 671.9092407226562,
      "completions/min_length": 142.0,
      "completions/min_terminated_length": 142.0,
      "epoch": 2.506270049577136,
      "grad_norm": 0.32219722867012024,
      "learning_rate": 1e-06,
      "loss": -0.0749,
      "num_tokens": 157793438.0,
      "reward": 0.590401828289032,
      "reward_std": 0.17484305799007416,
      "rewards/verify_math_reward/mean": 0.5904017686843872,
      "rewards/verify_math_reward/std": 0.49203425645828247,
      "step": 1073
    },
    {
      "clip_ratio/high_max": 0.0026715986750787124,
      "clip_ratio/high_mean": 0.0010693446074583335,
      "clip_ratio/low_mean": 0.0008631762666482246,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001932520914124325,
      "epoch": 2.5086030912802566,
      "grad_norm": 0.254000723361969,
      "learning_rate": 1e-06,
      "loss": -0.0752,
      "step": 1074
    },
    {
      "clip_ratio/high_max": 0.0026933696353808045,
      "clip_ratio/high_mean": 0.00115871095113107,
      "clip_ratio/low_mean": 0.001100592824514024,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022593038374907337,
      "epoch": 2.5109361329833773,
      "grad_norm": 0.2328997105360031,
      "learning_rate": 1e-06,
      "loss": -0.0753,
      "step": 1075
    },
    {
      "clip_ratio/high_max": 0.0028204722548252903,
      "clip_ratio/high_mean": 0.0010656727863533888,
      "clip_ratio/low_mean": 0.001168592214526143,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022342649899655953,
      "epoch": 2.5132691746864975,
      "grad_norm": 0.2636139392852783,
      "learning_rate": 1e-06,
      "loss": -0.0753,
      "step": 1076
    },
    {
      "clip_ratio/high_max": 0.0027340222732163966,
      "clip_ratio/high_mean": 0.0009299906196247321,
      "clip_ratio/low_mean": 0.0007416015505441464,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001671592181082815,
      "completions/clipped_ratio": 0.1674107142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3079.0,
      "completions/mean_length": 1231.919677734375,
      "completions/mean_terminated_length": 656.0321655273438,
      "completions/min_length": 140.0,
      "completions/min_terminated_length": 140.0,
      "epoch": 2.515602216389618,
      "grad_norm": 0.34658893942832947,
      "learning_rate": 1e-06,
      "loss": -0.0651,
      "num_tokens": 158376614.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.15890717506408691,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514803290367126,
      "step": 1077
    },
    {
      "clip_ratio/high_max": 0.002843072659743484,
      "clip_ratio/high_mean": 0.0009990428261517081,
      "clip_ratio/low_mean": 0.0010295792199030984,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002028622016950976,
      "epoch": 2.5179352580927383,
      "grad_norm": 0.2607556879520416,
      "learning_rate": 1e-06,
      "loss": -0.0652,
      "step": 1078
    },
    {
      "clip_ratio/high_max": 0.0030230559350457042,
      "clip_ratio/high_mean": 0.0010666556881915312,
      "clip_ratio/low_mean": 0.0012042350135743618,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002270890661748126,
      "epoch": 2.520268299795859,
      "grad_norm": 0.23274657130241394,
      "learning_rate": 1e-06,
      "loss": -0.0654,
      "step": 1079
    },
    {
      "clip_ratio/high_max": 0.0026979696049238555,
      "clip_ratio/high_mean": 0.0009742539659782778,
      "clip_ratio/low_mean": 0.0013522894550987985,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023265434792847373,
      "epoch": 2.522601341498979,
      "grad_norm": 0.2244875580072403,
      "learning_rate": 1e-06,
      "loss": -0.0655,
      "step": 1080
    },
    {
      "clip_ratio/high_max": 0.0018652241342351772,
      "clip_ratio/high_mean": 0.0006941004485270241,
      "clip_ratio/low_mean": 0.0004151139025907469,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0011092143586211023,
      "completions/clipped_ratio": 0.1629464285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3628.0,
      "completions/mean_length": 1252.193115234375,
      "completions/mean_terminated_length": 698.5986328125,
      "completions/min_length": 178.0,
      "completions/min_terminated_length": 178.0,
      "epoch": 2.5249343832021,
      "grad_norm": 0.3375288248062134,
      "learning_rate": 1e-06,
      "loss": -0.0258,
      "num_tokens": 159011563.0,
      "reward": 0.520089328289032,
      "reward_std": 0.1151694729924202,
      "rewards/verify_math_reward/mean": 0.5200892686843872,
      "rewards/verify_math_reward/std": 0.4998753070831299,
      "step": 1081
    },
    {
      "clip_ratio/high_max": 0.002928855452410062,
      "clip_ratio/high_mean": 0.0008989085581561085,
      "clip_ratio/low_mean": 0.0005837733792759536,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014826819660811452,
      "epoch": 2.52726742490522,
      "grad_norm": 0.2433030605316162,
      "learning_rate": 1e-06,
      "loss": -0.0261,
      "step": 1082
    },
    {
      "clip_ratio/high_max": 0.0029415451353997923,
      "clip_ratio/high_mean": 0.0009354234507554793,
      "clip_ratio/low_mean": 0.0006812501569584128,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001616673598618945,
      "epoch": 2.5296004666083407,
      "grad_norm": 0.20506641268730164,
      "learning_rate": 1e-06,
      "loss": -0.0263,
      "step": 1083
    },
    {
      "clip_ratio/high_max": 0.0024541875172872096,
      "clip_ratio/high_mean": 0.0008309054683195427,
      "clip_ratio/low_mean": 0.0008008740928744373,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001631779559829738,
      "epoch": 2.531933508311461,
      "grad_norm": 0.21931932866573334,
      "learning_rate": 1e-06,
      "loss": -0.0262,
      "step": 1084
    },
    {
      "clip_ratio/high_max": 0.0022768886665289756,
      "clip_ratio/high_mean": 0.000846750735945534,
      "clip_ratio/low_mean": 0.0005720474728150293,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014187982269504573,
      "completions/clipped_ratio": 0.1339285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3729.0,
      "completions/mean_length": 1088.0703125,
      "completions/mean_terminated_length": 622.926513671875,
      "completions/min_length": 151.0,
      "completions/min_terminated_length": 151.0,
      "epoch": 2.5342665500145816,
      "grad_norm": 0.370697945356369,
      "learning_rate": 1e-06,
      "loss": -0.0638,
      "num_tokens": 159590042.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.1486460566520691,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514803290367126,
      "step": 1085
    },
    {
      "clip_ratio/high_max": 0.002662609243998304,
      "clip_ratio/high_mean": 0.001032141619361937,
      "clip_ratio/low_mean": 0.0008312277450386318,
      "clip_ratio/low_min": 4.931276271236129e-05,
      "clip_ratio/region_mean": 0.0018633693834999576,
      "epoch": 2.536599591717702,
      "grad_norm": 0.29990503191947937,
      "learning_rate": 1e-06,
      "loss": -0.064,
      "step": 1086
    },
    {
      "clip_ratio/high_max": 0.0032535009559069294,
      "clip_ratio/high_mean": 0.0010791864497150527,
      "clip_ratio/low_mean": 0.0009813344277063152,
      "clip_ratio/low_min": 4.732724482892081e-05,
      "clip_ratio/region_mean": 0.0020605208810593467,
      "epoch": 2.5389326334208224,
      "grad_norm": 0.23016409575939178,
      "learning_rate": 1e-06,
      "loss": -0.0642,
      "step": 1087
    },
    {
      "clip_ratio/high_max": 0.003006082901265472,
      "clip_ratio/high_mean": 0.0011064850477850996,
      "clip_ratio/low_mean": 0.001108902984924498,
      "clip_ratio/low_min": 1.1548410839168355e-05,
      "clip_ratio/region_mean": 0.0022153881000122055,
      "epoch": 2.5412656751239426,
      "grad_norm": 0.2070586383342743,
      "learning_rate": 1e-06,
      "loss": -0.0643,
      "step": 1088
    },
    {
      "clip_ratio/high_max": 0.0025503474244032986,
      "clip_ratio/high_mean": 0.0011560129387362394,
      "clip_ratio/low_mean": 0.0007998784167284612,
      "clip_ratio/low_min": 2.5105442546191625e-05,
      "clip_ratio/region_mean": 0.0019558913918444887,
      "completions/clipped_ratio": 0.1372767857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2836.0,
      "completions/mean_length": 1123.438720703125,
      "completions/mean_terminated_length": 650.4437255859375,
      "completions/min_length": 183.0,
      "completions/min_terminated_length": 183.0,
      "epoch": 2.5435987168270633,
      "grad_norm": 0.3656218349933624,
      "learning_rate": 1e-06,
      "loss": -0.0502,
      "num_tokens": 160174971.0,
      "reward": 0.5535714626312256,
      "reward_std": 0.197794571518898,
      "rewards/verify_math_reward/mean": 0.5535714030265808,
      "rewards/verify_math_reward/std": 0.4973995089530945,
      "step": 1089
    },
    {
      "clip_ratio/high_max": 0.002961598554975353,
      "clip_ratio/high_mean": 0.001348602381767705,
      "clip_ratio/low_mean": 0.0010555532207945362,
      "clip_ratio/low_min": 2.5105442546191625e-05,
      "clip_ratio/region_mean": 0.0024041556171141565,
      "epoch": 2.545931758530184,
      "grad_norm": 0.3478671908378601,
      "learning_rate": 1e-06,
      "loss": -0.0506,
      "step": 1090
    },
    {
      "clip_ratio/high_max": 0.0028374877947499044,
      "clip_ratio/high_mean": 0.0013154700263839914,
      "clip_ratio/low_mean": 0.0012582149520312669,
      "clip_ratio/low_min": 2.5105442546191625e-05,
      "clip_ratio/region_mean": 0.0025736849784152582,
      "epoch": 2.548264800233304,
      "grad_norm": 0.2995145618915558,
      "learning_rate": 1e-06,
      "loss": -0.0508,
      "step": 1091
    },
    {
      "clip_ratio/high_max": 0.0030797913859714754,
      "clip_ratio/high_mean": 0.0012829393544961931,
      "clip_ratio/low_mean": 0.0015215690691547934,
      "clip_ratio/low_min": 1.4302059753390495e-05,
      "clip_ratio/region_mean": 0.0028045084109180607,
      "epoch": 2.5505978419364244,
      "grad_norm": 0.33942118287086487,
      "learning_rate": 1e-06,
      "loss": -0.051,
      "step": 1092
    },
    {
      "clip_ratio/high_max": 0.0027154575218446553,
      "clip_ratio/high_mean": 0.0009403135609318269,
      "clip_ratio/low_mean": 0.0004383714103823877,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013786849558528047,
      "completions/clipped_ratio": 0.1774553571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3796.0,
      "completions/mean_length": 1330.404052734375,
      "completions/mean_terminated_length": 733.7557983398438,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 2.552930883639545,
      "grad_norm": 0.3024427592754364,
      "learning_rate": 1e-06,
      "loss": -0.0898,
      "num_tokens": 160800557.0,
      "reward": 0.53125,
      "reward_std": 0.1597309112548828,
      "rewards/verify_math_reward/mean": 0.53125,
      "rewards/verify_math_reward/std": 0.4993011951446533,
      "step": 1093
    },
    {
      "clip_ratio/high_max": 0.003726006718352437,
      "clip_ratio/high_mean": 0.0012743875086016487,
      "clip_ratio/low_mean": 0.0005434095000964589,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018177969686803408,
      "epoch": 2.5552639253426657,
      "grad_norm": 0.25026634335517883,
      "learning_rate": 1e-06,
      "loss": -0.0899,
      "step": 1094
    },
    {
      "clip_ratio/high_max": 0.0036933976589352824,
      "clip_ratio/high_mean": 0.0012457342709240038,
      "clip_ratio/low_mean": 0.0006731112216584734,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019188454753020778,
      "epoch": 2.557596967045786,
      "grad_norm": 0.23310990631580353,
      "learning_rate": 1e-06,
      "loss": -0.0901,
      "step": 1095
    },
    {
      "clip_ratio/high_max": 0.003475955832982436,
      "clip_ratio/high_mean": 0.0011531089658092242,
      "clip_ratio/low_mean": 0.0007925101235741749,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019456190675555263,
      "epoch": 2.5599300087489065,
      "grad_norm": 0.24037988483905792,
      "learning_rate": 1e-06,
      "loss": -0.0902,
      "step": 1096
    },
    {
      "clip_ratio/high_max": 0.0025270912956329994,
      "clip_ratio/high_mean": 0.0011240604162594536,
      "clip_ratio/low_mean": 0.0009068060808203882,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020308665043557994,
      "completions/clipped_ratio": 0.1774553571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3769.0,
      "completions/mean_length": 1283.2578125,
      "completions/mean_terminated_length": 676.4382934570312,
      "completions/min_length": 160.0,
      "completions/min_terminated_length": 160.0,
      "epoch": 2.5622630504520267,
      "grad_norm": 0.35510119795799255,
      "learning_rate": 1e-06,
      "loss": -0.0802,
      "num_tokens": 161386172.0,
      "reward": 0.5178571939468384,
      "reward_std": 0.18468938767910004,
      "rewards/verify_math_reward/mean": 0.5178571343421936,
      "rewards/verify_math_reward/std": 0.4999600946903229,
      "step": 1097
    },
    {
      "clip_ratio/high_max": 0.003006246748554986,
      "clip_ratio/high_mean": 0.0013421027542790398,
      "clip_ratio/low_mean": 0.0010785836548166117,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002420686447294429,
      "epoch": 2.5645960921551474,
      "grad_norm": 0.37266677618026733,
      "learning_rate": 1e-06,
      "loss": -0.0804,
      "step": 1098
    },
    {
      "clip_ratio/high_max": 0.002982464611704927,
      "clip_ratio/high_mean": 0.0012667034025071189,
      "clip_ratio/low_mean": 0.0013548635179176927,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002621566949528642,
      "epoch": 2.5669291338582676,
      "grad_norm": 0.23731666803359985,
      "learning_rate": 1e-06,
      "loss": -0.0806,
      "step": 1099
    },
    {
      "clip_ratio/high_max": 0.00288599579653237,
      "clip_ratio/high_mean": 0.0012670163796428824,
      "clip_ratio/low_mean": 0.001627523721253965,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002894540004490409,
      "epoch": 2.5692621755613883,
      "grad_norm": 0.2168024182319641,
      "learning_rate": 1e-06,
      "loss": -0.0808,
      "step": 1100
    },
    {
      "clip_ratio/high_max": 0.002326938461919781,
      "clip_ratio/high_mean": 0.0009346261258542654,
      "clip_ratio/low_mean": 0.0005173175809431996,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014519437027047388,
      "completions/clipped_ratio": 0.1272321428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2913.0,
      "completions/mean_length": 1109.4710693359375,
      "completions/mean_terminated_length": 674.0946044921875,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 2.5715952172645085,
      "grad_norm": 0.3829411268234253,
      "learning_rate": 1e-06,
      "loss": -0.0594,
      "num_tokens": 162010586.0,
      "reward": 0.6383928656578064,
      "reward_std": 0.1537972390651703,
      "rewards/verify_math_reward/mean": 0.6383928656578064,
      "rewards/verify_math_reward/std": 0.4807341694831848,
      "step": 1101
    },
    {
      "clip_ratio/high_max": 0.0024993760089273565,
      "clip_ratio/high_mean": 0.0010787901337607764,
      "clip_ratio/low_mean": 0.0007505401317757787,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018293302418896928,
      "epoch": 2.573928258967629,
      "grad_norm": 0.2564522325992584,
      "learning_rate": 1e-06,
      "loss": -0.0596,
      "step": 1102
    },
    {
      "clip_ratio/high_max": 0.0030906805586710107,
      "clip_ratio/high_mean": 0.0011437192715675337,
      "clip_ratio/low_mean": 0.0009046731920534512,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020483924672589637,
      "epoch": 2.5762613006707493,
      "grad_norm": 0.20611624419689178,
      "learning_rate": 1e-06,
      "loss": -0.0598,
      "step": 1103
    },
    {
      "clip_ratio/high_max": 0.0025507345417281613,
      "clip_ratio/high_mean": 0.0010343282538087806,
      "clip_ratio/low_mean": 0.0010219057076028548,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020562340068863705,
      "epoch": 2.57859434237387,
      "grad_norm": 0.30897054076194763,
      "learning_rate": 1e-06,
      "loss": -0.0599,
      "step": 1104
    },
    {
      "clip_ratio/high_max": 0.0027421125705586746,
      "clip_ratio/high_mean": 0.0010197161354881246,
      "clip_ratio/low_mean": 0.0004768837093251932,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001496599848906044,
      "completions/clipped_ratio": 0.1305803571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3928.0,
      "completions/mean_length": 1054.1317138671875,
      "completions/mean_terminated_length": 597.2657470703125,
      "completions/min_length": 146.0,
      "completions/min_terminated_length": 146.0,
      "epoch": 2.5809273840769906,
      "grad_norm": 0.32677143812179565,
      "learning_rate": 1e-06,
      "loss": -0.0551,
      "num_tokens": 162562792.0,
      "reward": 0.5959821939468384,
      "reward_std": 0.15357083082199097,
      "rewards/verify_math_reward/mean": 0.5959821343421936,
      "rewards/verify_math_reward/std": 0.490975022315979,
      "step": 1105
    },
    {
      "clip_ratio/high_max": 0.0025771559012355283,
      "clip_ratio/high_mean": 0.0010178010234085377,
      "clip_ratio/low_mean": 0.000736704959308554,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017545059599797241,
      "epoch": 2.583260425780111,
      "grad_norm": 0.27515554428100586,
      "learning_rate": 1e-06,
      "loss": -0.0553,
      "step": 1106
    },
    {
      "clip_ratio/high_max": 0.0029067202340229414,
      "clip_ratio/high_mean": 0.0011057470001105685,
      "clip_ratio/low_mean": 0.000855755053635221,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001961502101039514,
      "epoch": 2.585593467483231,
      "grad_norm": 0.22235415875911713,
      "learning_rate": 1e-06,
      "loss": -0.0555,
      "step": 1107
    },
    {
      "clip_ratio/high_max": 0.0027928195049753413,
      "clip_ratio/high_mean": 0.0010616704385029152,
      "clip_ratio/low_mean": 0.0010224524312434369,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002084122876112815,
      "epoch": 2.5879265091863517,
      "grad_norm": 0.27045929431915283,
      "learning_rate": 1e-06,
      "loss": -0.0555,
      "step": 1108
    },
    {
      "clip_ratio/high_max": 0.002150590888049919,
      "clip_ratio/high_mean": 0.0010427422366774408,
      "clip_ratio/low_mean": 0.000490098445879994,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015328407025663182,
      "completions/clipped_ratio": 0.1261160714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3760.0,
      "completions/mean_length": 1091.25,
      "completions/mean_terminated_length": 657.6142578125,
      "completions/min_length": 178.0,
      "completions/min_terminated_length": 178.0,
      "epoch": 2.5902595508894724,
      "grad_norm": 0.4157736897468567,
      "learning_rate": 1e-06,
      "loss": -0.06,
      "num_tokens": 163170296.0,
      "reward": 0.5647321939468384,
      "reward_std": 0.16578474640846252,
      "rewards/verify_math_reward/mean": 0.5647321343421936,
      "rewards/verify_math_reward/std": 0.49606895446777344,
      "step": 1109
    },
    {
      "clip_ratio/high_max": 0.0026978680325555615,
      "clip_ratio/high_mean": 0.0011257234727963805,
      "clip_ratio/low_mean": 0.0007382088842859957,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018639323971001431,
      "epoch": 2.5925925925925926,
      "grad_norm": 0.25867852568626404,
      "learning_rate": 1e-06,
      "loss": -0.0603,
      "step": 1110
    },
    {
      "clip_ratio/high_max": 0.0028489320684457198,
      "clip_ratio/high_mean": 0.0012609830337169115,
      "clip_ratio/low_mean": 0.0008413531631958904,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002102336278767325,
      "epoch": 2.5949256342957128,
      "grad_norm": 0.25433149933815,
      "learning_rate": 1e-06,
      "loss": -0.0605,
      "step": 1111
    },
    {
      "clip_ratio/high_max": 0.0024162857080227695,
      "clip_ratio/high_mean": 0.0010507349834369961,
      "clip_ratio/low_mean": 0.0009205314836435718,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019712664870894514,
      "epoch": 2.5972586759988334,
      "grad_norm": 0.2389124184846878,
      "learning_rate": 1e-06,
      "loss": -0.0605,
      "step": 1112
    },
    {
      "clip_ratio/high_max": 0.00232755062461365,
      "clip_ratio/high_mean": 0.0009195824804919539,
      "clip_ratio/low_mean": 0.0005153276651981287,
      "clip_ratio/low_min": 1.2817883543903008e-05,
      "clip_ratio/region_mean": 0.001434910129319178,
      "completions/clipped_ratio": 0.1495535714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3426.0,
      "completions/mean_length": 1138.8426513671875,
      "completions/mean_terminated_length": 618.8175659179688,
      "completions/min_length": 157.0,
      "completions/min_terminated_length": 157.0,
      "epoch": 2.599591717701954,
      "grad_norm": 0.33814549446105957,
      "learning_rate": 1e-06,
      "loss": -0.0936,
      "num_tokens": 163728859.0,
      "reward": 0.6272321939468384,
      "reward_std": 0.1795371025800705,
      "rewards/verify_math_reward/mean": 0.6272321343421936,
      "rewards/verify_math_reward/std": 0.4838111698627472,
      "step": 1113
    },
    {
      "clip_ratio/high_max": 0.0033839424868347123,
      "clip_ratio/high_mean": 0.0012554665518109687,
      "clip_ratio/low_mean": 0.0007289757140824804,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00198444229317829,
      "epoch": 2.6019247594050743,
      "grad_norm": 0.25683408975601196,
      "learning_rate": 1e-06,
      "loss": -0.0939,
      "step": 1114
    },
    {
      "clip_ratio/high_max": 0.002822003727487754,
      "clip_ratio/high_mean": 0.001149767565948423,
      "clip_ratio/low_mean": 0.000779302723458386,
      "clip_ratio/low_min": 1.2332280675764196e-05,
      "clip_ratio/region_mean": 0.0019290702839498408,
      "epoch": 2.604257801108195,
      "grad_norm": 0.26113370060920715,
      "learning_rate": 1e-06,
      "loss": -0.094,
      "step": 1115
    },
    {
      "clip_ratio/high_max": 0.002664569779881276,
      "clip_ratio/high_mean": 0.0010907177729677642,
      "clip_ratio/low_mean": 0.0010436815664434107,
      "clip_ratio/low_min": 2.466456135152839e-05,
      "clip_ratio/region_mean": 0.002134399299393408,
      "epoch": 2.606590842811315,
      "grad_norm": 0.21102070808410645,
      "learning_rate": 1e-06,
      "loss": -0.0941,
      "step": 1116
    },
    {
      "clip_ratio/high_max": 0.0020429506421351107,
      "clip_ratio/high_mean": 0.0006444957043640898,
      "clip_ratio/low_mean": 0.000556617698066475,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012011134040221805,
      "completions/clipped_ratio": 0.1216517857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3814.0,
      "completions/mean_length": 1052.1273193359375,
      "completions/mean_terminated_length": 630.5488891601562,
      "completions/min_length": 135.0,
      "completions/min_terminated_length": 135.0,
      "epoch": 2.608923884514436,
      "grad_norm": 0.3030131161212921,
      "learning_rate": 1e-06,
      "loss": -0.0359,
      "num_tokens": 164309485.0,
      "reward": 0.5524553656578064,
      "reward_std": 0.13876289129257202,
      "rewards/verify_math_reward/mean": 0.5524553656578064,
      "rewards/verify_math_reward/std": 0.49751853942871094,
      "step": 1117
    },
    {
      "clip_ratio/high_max": 0.0025315484544989886,
      "clip_ratio/high_mean": 0.0008370420373466914,
      "clip_ratio/low_mean": 0.0007493260559385817,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001586368085554568,
      "epoch": 2.611256926217556,
      "grad_norm": 0.25326409935951233,
      "learning_rate": 1e-06,
      "loss": -0.036,
      "step": 1118
    },
    {
      "clip_ratio/high_max": 0.0025373918942932505,
      "clip_ratio/high_mean": 0.00083686521384152,
      "clip_ratio/low_mean": 0.001015362488487881,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018522277350712102,
      "epoch": 2.6135899679206767,
      "grad_norm": 0.2460647076368332,
      "learning_rate": 1e-06,
      "loss": -0.0362,
      "step": 1119
    },
    {
      "clip_ratio/high_max": 0.002477307054505218,
      "clip_ratio/high_mean": 0.0008151182400979451,
      "clip_ratio/low_mean": 0.0010352167614655627,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018503350001992658,
      "epoch": 2.615923009623797,
      "grad_norm": 0.24024319648742676,
      "learning_rate": 1e-06,
      "loss": -0.0363,
      "step": 1120
    },
    {
      "clip_ratio/high_max": 0.0021482690281118266,
      "clip_ratio/high_mean": 0.0008689304322615499,
      "clip_ratio/low_mean": 0.0005232768307905644,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013922072866989765,
      "completions/clipped_ratio": 0.1127232142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3888.0,
      "completions/mean_length": 996.0848388671875,
      "completions/mean_terminated_length": 602.2590942382812,
      "completions/min_length": 175.0,
      "completions/min_terminated_length": 175.0,
      "epoch": 2.6182560513269175,
      "grad_norm": 0.2719813585281372,
      "learning_rate": 1e-06,
      "loss": -0.0492,
      "num_tokens": 164869201.0,
      "reward": 0.660714328289032,
      "reward_std": 0.15390713512897491,
      "rewards/verify_math_reward/mean": 0.6607142686843872,
      "rewards/verify_math_reward/std": 0.4737313687801361,
      "step": 1121
    },
    {
      "clip_ratio/high_max": 0.002664324529177975,
      "clip_ratio/high_mean": 0.0010516172642383026,
      "clip_ratio/low_mean": 0.0006718581489622011,
      "clip_ratio/low_min": 1.4859724615234882e-05,
      "clip_ratio/region_mean": 0.001723475430480903,
      "epoch": 2.6205890930300377,
      "grad_norm": 0.3058817982673645,
      "learning_rate": 1e-06,
      "loss": -0.0493,
      "step": 1122
    },
    {
      "clip_ratio/high_max": 0.0028086251113563776,
      "clip_ratio/high_mean": 0.001154651603428647,
      "clip_ratio/low_mean": 0.0008613234749645926,
      "clip_ratio/low_min": 2.2080905182519928e-05,
      "clip_ratio/region_mean": 0.002015975085669197,
      "epoch": 2.6229221347331584,
      "grad_norm": 0.19979745149612427,
      "learning_rate": 1e-06,
      "loss": -0.0496,
      "step": 1123
    },
    {
      "clip_ratio/high_max": 0.0024760094711382408,
      "clip_ratio/high_mean": 0.0009517693533780403,
      "clip_ratio/low_mean": 0.00111147116876964,
      "clip_ratio/low_min": 1.1040452591259964e-05,
      "clip_ratio/region_mean": 0.0020632405357901007,
      "epoch": 2.625255176436279,
      "grad_norm": 0.2403215914964676,
      "learning_rate": 1e-06,
      "loss": -0.0496,
      "step": 1124
    },
    {
      "clip_ratio/high_max": 0.0027720848738681525,
      "clip_ratio/high_mean": 0.0009584093750163447,
      "clip_ratio/low_mean": 0.000537255859853758,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014956652375985868,
      "completions/clipped_ratio": 0.1160714285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3061.0,
      "completions/mean_length": 1032.552490234375,
      "completions/mean_terminated_length": 630.2815551757812,
      "completions/min_length": 143.0,
      "completions/min_terminated_length": 143.0,
      "epoch": 2.6275882181393992,
      "grad_norm": 0.2966957688331604,
      "learning_rate": 1e-06,
      "loss": -0.061,
      "num_tokens": 165453048.0,
      "reward": 0.6517857313156128,
      "reward_std": 0.17476709187030792,
      "rewards/verify_math_reward/mean": 0.6517857313156128,
      "rewards/verify_math_reward/std": 0.47667041420936584,
      "step": 1125
    },
    {
      "clip_ratio/high_max": 0.00300396719831042,
      "clip_ratio/high_mean": 0.001154687659436604,
      "clip_ratio/low_mean": 0.0008199599869840313,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001974647617316805,
      "epoch": 2.6299212598425195,
      "grad_norm": 0.25790533423423767,
      "learning_rate": 1e-06,
      "loss": -0.0612,
      "step": 1126
    },
    {
      "clip_ratio/high_max": 0.0032114461428136565,
      "clip_ratio/high_mean": 0.0012655668942898046,
      "clip_ratio/low_mean": 0.0009432689948880579,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022088358527980745,
      "epoch": 2.63225430154564,
      "grad_norm": 0.21397745609283447,
      "learning_rate": 1e-06,
      "loss": -0.0614,
      "step": 1127
    },
    {
      "clip_ratio/high_max": 0.0031452557304874063,
      "clip_ratio/high_mean": 0.001191447696328396,
      "clip_ratio/low_mean": 0.001080630059732357,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022720778069924563,
      "epoch": 2.6345873432487608,
      "grad_norm": 0.25344526767730713,
      "learning_rate": 1e-06,
      "loss": -0.0614,
      "step": 1128
    },
    {
      "clip_ratio/high_max": 0.0022820700614829548,
      "clip_ratio/high_mean": 0.0009935147118085297,
      "clip_ratio/low_mean": 0.0007081885323714232,
      "clip_ratio/low_min": 4.7674659072072245e-05,
      "clip_ratio/region_mean": 0.0017017032078001648,
      "completions/clipped_ratio": 0.1506696428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3312.0,
      "completions/mean_length": 1139.97216796875,
      "completions/mean_terminated_length": 615.5781860351562,
      "completions/min_length": 169.0,
      "completions/min_terminated_length": 169.0,
      "epoch": 2.636920384951881,
      "grad_norm": 0.32561057806015015,
      "learning_rate": 1e-06,
      "loss": -0.0586,
      "num_tokens": 166002015.0,
      "reward": 0.5837053656578064,
      "reward_std": 0.1879894882440567,
      "rewards/verify_math_reward/mean": 0.5837053656578064,
      "rewards/verify_math_reward/std": 0.49321895837783813,
      "step": 1129
    },
    {
      "clip_ratio/high_max": 0.002762396055913996,
      "clip_ratio/high_mean": 0.001226898548338795,
      "clip_ratio/low_mean": 0.000930823360249633,
      "clip_ratio/low_min": 7.519316386606079e-05,
      "clip_ratio/region_mean": 0.0021577219049504492,
      "epoch": 2.6392534266550016,
      "grad_norm": 0.25686073303222656,
      "learning_rate": 1e-06,
      "loss": -0.0589,
      "step": 1130
    },
    {
      "clip_ratio/high_max": 0.0028286995220696554,
      "clip_ratio/high_mean": 0.0012117004789615748,
      "clip_ratio/low_mean": 0.0010972727650369052,
      "clip_ratio/low_min": 3.666993507067673e-05,
      "clip_ratio/region_mean": 0.002308973234903533,
      "epoch": 2.641586468358122,
      "grad_norm": 0.2529727518558502,
      "learning_rate": 1e-06,
      "loss": -0.0591,
      "step": 1131
    },
    {
      "clip_ratio/high_max": 0.0026503606131882407,
      "clip_ratio/high_mean": 0.0011134245469293091,
      "clip_ratio/low_mean": 0.0013780929548374843,
      "clip_ratio/low_min": 9.493100515101105e-05,
      "clip_ratio/region_mean": 0.0024915175017667934,
      "epoch": 2.6439195100612425,
      "grad_norm": 0.25941675901412964,
      "learning_rate": 1e-06,
      "loss": -0.0592,
      "step": 1132
    },
    {
      "clip_ratio/high_max": 0.0023725707796984352,
      "clip_ratio/high_mean": 0.0009655496851337375,
      "clip_ratio/low_mean": 0.0005347466503735632,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015002963627921417,
      "completions/clipped_ratio": 0.109375,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3983.0,
      "completions/mean_length": 982.0357666015625,
      "completions/mean_terminated_length": 599.6190185546875,
      "completions/min_length": 183.0,
      "completions/min_terminated_length": 183.0,
      "epoch": 2.6462525517643627,
      "grad_norm": 0.2986929714679718,
      "learning_rate": 1e-06,
      "loss": -0.0672,
      "num_tokens": 166575767.0,
      "reward": 0.6194196939468384,
      "reward_std": 0.189485564827919,
      "rewards/verify_math_reward/mean": 0.6194196343421936,
      "rewards/verify_math_reward/std": 0.48580074310302734,
      "step": 1133
    },
    {
      "clip_ratio/high_max": 0.0027116593992104754,
      "clip_ratio/high_mean": 0.0011672088367049582,
      "clip_ratio/low_mean": 0.0007560647736681858,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001923273601278197,
      "epoch": 2.6485855934674833,
      "grad_norm": 0.272161066532135,
      "learning_rate": 1e-06,
      "loss": -0.0675,
      "step": 1134
    },
    {
      "clip_ratio/high_max": 0.002445779689878691,
      "clip_ratio/high_mean": 0.0011025048988813069,
      "clip_ratio/low_mean": 0.0009129698701144662,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002015474768995773,
      "epoch": 2.6509186351706036,
      "grad_norm": 0.204036146402359,
      "learning_rate": 1e-06,
      "loss": -0.0676,
      "step": 1135
    },
    {
      "clip_ratio/high_max": 0.0024223813525168225,
      "clip_ratio/high_mean": 0.0011369093772373162,
      "clip_ratio/low_mean": 0.0010873437604459468,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022242531340452842,
      "epoch": 2.653251676873724,
      "grad_norm": 0.2310946136713028,
      "learning_rate": 1e-06,
      "loss": -0.0677,
      "step": 1136
    },
    {
      "clip_ratio/high_max": 0.0019093765258730855,
      "clip_ratio/high_mean": 0.0007528779151471099,
      "clip_ratio/low_mean": 0.0005131302204972599,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00126600810472155,
      "completions/clipped_ratio": 0.1573660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4006.0,
      "completions/mean_length": 1196.7523193359375,
      "completions/mean_terminated_length": 655.3033447265625,
      "completions/min_length": 224.0,
      "completions/min_terminated_length": 224.0,
      "epoch": 2.6555847185768444,
      "grad_norm": 0.2802563011646271,
      "learning_rate": 1e-06,
      "loss": -0.056,
      "num_tokens": 167160089.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.15916569530963898,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 1137
    },
    {
      "clip_ratio/high_max": 0.0025058464889298193,
      "clip_ratio/high_mean": 0.0009987526918848744,
      "clip_ratio/low_mean": 0.0006417501290343353,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016405027927248739,
      "epoch": 2.657917760279965,
      "grad_norm": 0.23167243599891663,
      "learning_rate": 1e-06,
      "loss": -0.0562,
      "step": 1138
    },
    {
      "clip_ratio/high_max": 0.002632463161717169,
      "clip_ratio/high_mean": 0.000962874670221936,
      "clip_ratio/low_mean": 0.0008900784996512812,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018529531444073655,
      "epoch": 2.6602508019830857,
      "grad_norm": 0.32529324293136597,
      "learning_rate": 1e-06,
      "loss": -0.0564,
      "step": 1139
    },
    {
      "clip_ratio/high_max": 0.0023656452540308237,
      "clip_ratio/high_mean": 0.0009692981020634761,
      "clip_ratio/low_mean": 0.0009637194434617413,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019330175855429843,
      "epoch": 2.662583843686206,
      "grad_norm": 0.27710026502609253,
      "learning_rate": 1e-06,
      "loss": -0.0564,
      "step": 1140
    },
    {
      "clip_ratio/high_max": 0.0022393335384549573,
      "clip_ratio/high_mean": 0.0007632964770891704,
      "clip_ratio/low_mean": 0.0005693107414117549,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013326072366908193,
      "completions/clipped_ratio": 0.1841517857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3621.0,
      "completions/mean_length": 1360.2723388671875,
      "completions/mean_terminated_length": 742.7688598632812,
      "completions/min_length": 169.0,
      "completions/min_terminated_length": 169.0,
      "epoch": 2.664916885389326,
      "grad_norm": 0.2509084641933441,
      "learning_rate": 1e-06,
      "loss": -0.0584,
      "num_tokens": 167792013.0,
      "reward": 0.463169664144516,
      "reward_std": 0.16214200854301453,
      "rewards/verify_math_reward/mean": 0.4631696343421936,
      "rewards/verify_math_reward/std": 0.49892017245292664,
      "step": 1141
    },
    {
      "clip_ratio/high_max": 0.0027459394914330915,
      "clip_ratio/high_mean": 0.0010035582890850492,
      "clip_ratio/low_mean": 0.0008170855089701945,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018206437562184874,
      "epoch": 2.667249927092447,
      "grad_norm": 0.22688522934913635,
      "learning_rate": 1e-06,
      "loss": -0.0586,
      "step": 1142
    },
    {
      "clip_ratio/high_max": 0.0028888552769785747,
      "clip_ratio/high_mean": 0.0009991747792810202,
      "clip_ratio/low_mean": 0.0009259738062610268,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019251485900895204,
      "epoch": 2.6695829687955674,
      "grad_norm": 0.22150474786758423,
      "learning_rate": 1e-06,
      "loss": -0.0587,
      "step": 1143
    },
    {
      "clip_ratio/high_max": 0.002439021052850876,
      "clip_ratio/high_mean": 0.0009087577436730498,
      "clip_ratio/low_mean": 0.0010164745017391397,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019252322454121895,
      "epoch": 2.6719160104986877,
      "grad_norm": 0.20555339753627777,
      "learning_rate": 1e-06,
      "loss": -0.0588,
      "step": 1144
    },
    {
      "clip_ratio/high_max": 0.0018441054489812814,
      "clip_ratio/high_mean": 0.0007074962049955502,
      "clip_ratio/low_mean": 0.0005324415769791813,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001239937777427258,
      "completions/clipped_ratio": 0.1428571428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4079.0,
      "completions/mean_length": 1170.54248046875,
      "completions/mean_terminated_length": 682.9661865234375,
      "completions/min_length": 173.0,
      "completions/min_terminated_length": 173.0,
      "epoch": 2.674249052201808,
      "grad_norm": 0.2705370783805847,
      "learning_rate": 1e-06,
      "loss": -0.0643,
      "num_tokens": 168411147.0,
      "reward": 0.5089285969734192,
      "reward_std": 0.14042216539382935,
      "rewards/verify_math_reward/mean": 0.5089285969734192,
      "rewards/verify_math_reward/std": 0.5001994967460632,
      "step": 1145
    },
    {
      "clip_ratio/high_max": 0.002201586488808971,
      "clip_ratio/high_mean": 0.0008728033153602155,
      "clip_ratio/low_mean": 0.0007337162696785526,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016065196141425986,
      "epoch": 2.6765820939049285,
      "grad_norm": 0.24536322057247162,
      "learning_rate": 1e-06,
      "loss": -0.0645,
      "step": 1146
    },
    {
      "clip_ratio/high_max": 0.002139316660759505,
      "clip_ratio/high_mean": 0.0007986662731127581,
      "clip_ratio/low_mean": 0.0007943360560602741,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015930023946566507,
      "epoch": 2.678915135608049,
      "grad_norm": 0.24811631441116333,
      "learning_rate": 1e-06,
      "loss": -0.0646,
      "step": 1147
    },
    {
      "clip_ratio/high_max": 0.0023903483706817497,
      "clip_ratio/high_mean": 0.0008260114354925463,
      "clip_ratio/low_mean": 0.0009675408273324138,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017935522046172991,
      "epoch": 2.6812481773111694,
      "grad_norm": 0.2024756819009781,
      "learning_rate": 1e-06,
      "loss": -0.0647,
      "step": 1148
    },
    {
      "clip_ratio/high_max": 0.0022558805030712392,
      "clip_ratio/high_mean": 0.0008015481953407289,
      "clip_ratio/low_mean": 0.0005789182469015941,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001380466444970807,
      "completions/clipped_ratio": 0.1439732142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3994.0,
      "completions/mean_length": 1187.8560791015625,
      "completions/mean_terminated_length": 698.7418212890625,
      "completions/min_length": 227.0,
      "completions/min_terminated_length": 227.0,
      "epoch": 2.68358121901429,
      "grad_norm": 0.28669261932373047,
      "learning_rate": 1e-06,
      "loss": -0.0455,
      "num_tokens": 169031930.0,
      "reward": 0.5457589626312256,
      "reward_std": 0.15522870421409607,
      "rewards/verify_math_reward/mean": 0.5457589030265808,
      "rewards/verify_math_reward/std": 0.4981797933578491,
      "step": 1149
    },
    {
      "clip_ratio/high_max": 0.0021396095071395393,
      "clip_ratio/high_mean": 0.0008520158498868113,
      "clip_ratio/low_mean": 0.0006523220808958285,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015043379098642617,
      "epoch": 2.6859142607174102,
      "grad_norm": 0.2215685099363327,
      "learning_rate": 1e-06,
      "loss": -0.0455,
      "step": 1150
    },
    {
      "clip_ratio/high_max": 0.0026519415332586505,
      "clip_ratio/high_mean": 0.0009802410786505789,
      "clip_ratio/low_mean": 0.000823186988782254,
      "clip_ratio/low_min": 1.5277439160854556e-05,
      "clip_ratio/region_mean": 0.001803428036510013,
      "epoch": 2.688247302420531,
      "grad_norm": 0.24253995716571808,
      "learning_rate": 1e-06,
      "loss": -0.0457,
      "step": 1151
    },
    {
      "clip_ratio/high_max": 0.002195755841967184,
      "clip_ratio/high_mean": 0.0008404661348322406,
      "clip_ratio/low_mean": 0.0009513325512671145,
      "clip_ratio/low_min": 4.391512266010977e-05,
      "clip_ratio/region_mean": 0.001791798677004408,
      "epoch": 2.690580344123651,
      "grad_norm": 0.20786729454994202,
      "learning_rate": 1e-06,
      "loss": -0.0458,
      "step": 1152
    },
    {
      "clip_ratio/high_max": 0.002137276365829166,
      "clip_ratio/high_mean": 0.0008418489796895301,
      "clip_ratio/low_mean": 0.00039431780010090733,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012361667722871061,
      "completions/clipped_ratio": 0.1462053571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3274.0,
      "completions/mean_length": 1187.6429443359375,
      "completions/mean_terminated_length": 689.6104736328125,
      "completions/min_length": 155.0,
      "completions/min_terminated_length": 155.0,
      "epoch": 2.6929133858267718,
      "grad_norm": 0.26343563199043274,
      "learning_rate": 1e-06,
      "loss": -0.0643,
      "num_tokens": 169643770.0,
      "reward": 0.5703125,
      "reward_std": 0.1406080424785614,
      "rewards/verify_math_reward/mean": 0.5703125,
      "rewards/verify_math_reward/std": 0.49530795216560364,
      "step": 1153
    },
    {
      "clip_ratio/high_max": 0.0022794629840063863,
      "clip_ratio/high_mean": 0.0009593052564014215,
      "clip_ratio/low_mean": 0.0005149428202457784,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014742480670975056,
      "epoch": 2.695246427529892,
      "grad_norm": 0.21706290543079376,
      "learning_rate": 1e-06,
      "loss": -0.0644,
      "step": 1154
    },
    {
      "clip_ratio/high_max": 0.0026719845554907806,
      "clip_ratio/high_mean": 0.00097825049306266,
      "clip_ratio/low_mean": 0.0005811857872686232,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015594363103446085,
      "epoch": 2.6975794692330126,
      "grad_norm": 0.45889315009117126,
      "learning_rate": 1e-06,
      "loss": -0.0645,
      "step": 1155
    },
    {
      "clip_ratio/high_max": 0.0024308937645400874,
      "clip_ratio/high_mean": 0.0009150312143901829,
      "clip_ratio/low_mean": 0.0007390931596091832,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016541243676329032,
      "epoch": 2.699912510936133,
      "grad_norm": 0.20172478258609772,
      "learning_rate": 1e-06,
      "loss": -0.0646,
      "step": 1156
    },
    {
      "clip_ratio/high_max": 0.001911978415591875,
      "clip_ratio/high_mean": 0.000815377288745367,
      "clip_ratio/low_mean": 0.0006175725966386381,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014329498771985527,
      "completions/clipped_ratio": 0.1361607142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3168.0,
      "completions/mean_length": 1103.0257568359375,
      "completions/mean_terminated_length": 631.264892578125,
      "completions/min_length": 179.0,
      "completions/min_terminated_length": 179.0,
      "epoch": 2.7022455526392535,
      "grad_norm": 0.3399980962276459,
      "learning_rate": 1e-06,
      "loss": -0.0641,
      "num_tokens": 170219817.0,
      "reward": 0.6183035969734192,
      "reward_std": 0.15819111466407776,
      "rewards/verify_math_reward/mean": 0.6183035969734192,
      "rewards/verify_math_reward/std": 0.4860740303993225,
      "step": 1157
    },
    {
      "clip_ratio/high_max": 0.002753663487965241,
      "clip_ratio/high_mean": 0.0010733545605035033,
      "clip_ratio/low_mean": 0.000898408230568748,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019717628456419334,
      "epoch": 2.704578594342374,
      "grad_norm": 0.285305917263031,
      "learning_rate": 1e-06,
      "loss": -0.0643,
      "step": 1158
    },
    {
      "clip_ratio/high_max": 0.002390883913903963,
      "clip_ratio/high_mean": 0.0009104770833801012,
      "clip_ratio/low_mean": 0.0010506581602385268,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001961135247256607,
      "epoch": 2.7069116360454943,
      "grad_norm": 0.2516597807407379,
      "learning_rate": 1e-06,
      "loss": -0.0644,
      "step": 1159
    },
    {
      "clip_ratio/high_max": 0.0023764417856000364,
      "clip_ratio/high_mean": 0.0008862508093443466,
      "clip_ratio/low_mean": 0.0012949681567988591,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002181219002522994,
      "epoch": 2.7092446777486145,
      "grad_norm": 0.2154473513364792,
      "learning_rate": 1e-06,
      "loss": -0.0646,
      "step": 1160
    },
    {
      "clip_ratio/high_max": 0.0021184427678235807,
      "clip_ratio/high_mean": 0.0007785532925481675,
      "clip_ratio/low_mean": 0.0004812925199075835,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012598458342836238,
      "completions/clipped_ratio": 0.0970982142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2231.0,
      "completions/mean_length": 987.177490234375,
      "completions/mean_terminated_length": 652.8541259765625,
      "completions/min_length": 204.0,
      "completions/min_terminated_length": 204.0,
      "epoch": 2.711577719451735,
      "grad_norm": 0.2807345390319824,
      "learning_rate": 1e-06,
      "loss": -0.0352,
      "num_tokens": 170838272.0,
      "reward": 0.5803571939468384,
      "reward_std": 0.14004239439964294,
      "rewards/verify_math_reward/mean": 0.5803571343421936,
      "rewards/verify_math_reward/std": 0.4937761425971985,
      "step": 1161
    },
    {
      "clip_ratio/high_max": 0.0022629649320151657,
      "clip_ratio/high_mean": 0.0008516832112945849,
      "clip_ratio/low_mean": 0.0006314157672022702,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014830990257905796,
      "epoch": 2.713910761154856,
      "grad_norm": 0.2561696469783783,
      "learning_rate": 1e-06,
      "loss": -0.0353,
      "step": 1162
    },
    {
      "clip_ratio/high_max": 0.002565336508268956,
      "clip_ratio/high_mean": 0.0009492197441431927,
      "clip_ratio/low_mean": 0.0007205124074971536,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016697321334504522,
      "epoch": 2.716243802857976,
      "grad_norm": 0.202647864818573,
      "learning_rate": 1e-06,
      "loss": -0.0354,
      "step": 1163
    },
    {
      "clip_ratio/high_max": 0.002372305709286593,
      "clip_ratio/high_mean": 0.0009050377866515191,
      "clip_ratio/low_mean": 0.0009203308145515621,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018253686212119646,
      "epoch": 2.7185768445610963,
      "grad_norm": 0.1613180786371231,
      "learning_rate": 1e-06,
      "loss": -0.0355,
      "step": 1164
    },
    {
      "clip_ratio/high_max": 0.002923695297795348,
      "clip_ratio/high_mean": 0.0010887382250075461,
      "clip_ratio/low_mean": 0.0005822753610118525,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001671013578743441,
      "completions/clipped_ratio": 0.1875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3938.0,
      "completions/mean_length": 1305.7154541015625,
      "completions/mean_terminated_length": 661.8035888671875,
      "completions/min_length": 180.0,
      "completions/min_terminated_length": 180.0,
      "epoch": 2.720909886264217,
      "grad_norm": 0.34382542967796326,
      "learning_rate": 1e-06,
      "loss": -0.0553,
      "num_tokens": 171422433.0,
      "reward": 0.5245535969734192,
      "reward_std": 0.1658179610967636,
      "rewards/verify_math_reward/mean": 0.5245535969734192,
      "rewards/verify_math_reward/std": 0.4996756613254547,
      "step": 1165
    },
    {
      "clip_ratio/high_max": 0.00350352305395063,
      "clip_ratio/high_mean": 0.0011558006335690152,
      "clip_ratio/low_mean": 0.00077688737110293,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019326879591972101,
      "epoch": 2.7232429279673376,
      "grad_norm": 0.29401642084121704,
      "learning_rate": 1e-06,
      "loss": -0.0555,
      "step": 1166
    },
    {
      "clip_ratio/high_max": 0.0030248135080910288,
      "clip_ratio/high_mean": 0.0012315093026700197,
      "clip_ratio/low_mean": 0.0010095744673890295,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002241083755507134,
      "epoch": 2.725575969670458,
      "grad_norm": 0.2268916368484497,
      "learning_rate": 1e-06,
      "loss": -0.0557,
      "step": 1167
    },
    {
      "clip_ratio/high_max": 0.003271043417043984,
      "clip_ratio/high_mean": 0.0011021570935554337,
      "clip_ratio/low_mean": 0.0012381917695165612,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023403488739859313,
      "epoch": 2.7279090113735784,
      "grad_norm": 0.2651258111000061,
      "learning_rate": 1e-06,
      "loss": -0.0558,
      "step": 1168
    },
    {
      "clip_ratio/high_max": 0.0018323434414924122,
      "clip_ratio/high_mean": 0.0007325293281610357,
      "clip_ratio/low_mean": 0.0006358508017001441,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001368380144413095,
      "completions/clipped_ratio": 0.1484375,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3717.0,
      "completions/mean_length": 1216.068115234375,
      "completions/mean_terminated_length": 714.0616455078125,
      "completions/min_length": 162.0,
      "completions/min_terminated_length": 162.0,
      "epoch": 2.7302420530766986,
      "grad_norm": 0.2982531189918518,
      "learning_rate": 1e-06,
      "loss": -0.0591,
      "num_tokens": 172058222.0,
      "reward": 0.5111607313156128,
      "reward_std": 0.18160229921340942,
      "rewards/verify_math_reward/mean": 0.5111607313156128,
      "rewards/verify_math_reward/std": 0.5001546144485474,
      "step": 1169
    },
    {
      "clip_ratio/high_max": 0.0024237072066171095,
      "clip_ratio/high_mean": 0.0009196715491270879,
      "clip_ratio/low_mean": 0.0009615821636543842,
      "clip_ratio/low_min": 6.521401428472018e-05,
      "clip_ratio/region_mean": 0.001881253694591578,
      "epoch": 2.7325750947798193,
      "grad_norm": 0.2623019516468048,
      "learning_rate": 1e-06,
      "loss": -0.0595,
      "step": 1170
    },
    {
      "clip_ratio/high_max": 0.0023440176228177734,
      "clip_ratio/high_mean": 0.0008444089853583137,
      "clip_ratio/low_mean": 0.0010867597684409702,
      "clip_ratio/low_min": 5.1829283620463684e-05,
      "clip_ratio/region_mean": 0.001931168750161305,
      "epoch": 2.7349081364829395,
      "grad_norm": 0.26250067353248596,
      "learning_rate": 1e-06,
      "loss": -0.0595,
      "step": 1171
    },
    {
      "clip_ratio/high_max": 0.002036090547335334,
      "clip_ratio/high_mean": 0.0008630632728454657,
      "clip_ratio/low_mean": 0.0012405992401909316,
      "clip_ratio/low_min": 8.95601797310519e-05,
      "clip_ratio/region_mean": 0.002103662511217408,
      "epoch": 2.73724117818606,
      "grad_norm": 0.21721115708351135,
      "learning_rate": 1e-06,
      "loss": -0.0596,
      "step": 1172
    },
    {
      "clip_ratio/high_max": 0.0025588344724383205,
      "clip_ratio/high_mean": 0.0009905137867463054,
      "clip_ratio/low_mean": 0.00043740399178204825,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014279178067226894,
      "completions/clipped_ratio": 0.1607142857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3473.0,
      "completions/mean_length": 1183.5904541015625,
      "completions/mean_terminated_length": 625.8948974609375,
      "completions/min_length": 191.0,
      "completions/min_terminated_length": 191.0,
      "epoch": 2.7395742198891804,
      "grad_norm": 0.3492778539657593,
      "learning_rate": 1e-06,
      "loss": -0.0756,
      "num_tokens": 172629727.0,
      "reward": 0.5535714626312256,
      "reward_std": 0.14661546051502228,
      "rewards/verify_math_reward/mean": 0.5535714030265808,
      "rewards/verify_math_reward/std": 0.4973994791507721,
      "step": 1173
    },
    {
      "clip_ratio/high_max": 0.003237783705117181,
      "clip_ratio/high_mean": 0.001139278729169746,
      "clip_ratio/low_mean": 0.0006850259969723993,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018243047306896187,
      "epoch": 2.741907261592301,
      "grad_norm": 0.33572831749916077,
      "learning_rate": 1e-06,
      "loss": -0.0758,
      "step": 1174
    },
    {
      "clip_ratio/high_max": 0.002890553376346361,
      "clip_ratio/high_mean": 0.0011176306452398421,
      "clip_ratio/low_mean": 0.0008286535630759317,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019462841810309328,
      "epoch": 2.7442403032954212,
      "grad_norm": 0.21871332824230194,
      "learning_rate": 1e-06,
      "loss": -0.076,
      "step": 1175
    },
    {
      "clip_ratio/high_max": 0.0031910955658531748,
      "clip_ratio/high_mean": 0.001188287607874372,
      "clip_ratio/low_mean": 0.0009394897433594451,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021277773412293755,
      "epoch": 2.746573344998542,
      "grad_norm": 0.2713559865951538,
      "learning_rate": 1e-06,
      "loss": -0.0762,
      "step": 1176
    },
    {
      "clip_ratio/high_max": 0.002106075917254202,
      "clip_ratio/high_mean": 0.000794560561189428,
      "clip_ratio/low_mean": 0.0007055580717860721,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001500118654803373,
      "completions/clipped_ratio": 0.1305803571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3589.0,
      "completions/mean_length": 1118.3046875,
      "completions/mean_terminated_length": 671.0770263671875,
      "completions/min_length": 183.0,
      "completions/min_terminated_length": 183.0,
      "epoch": 2.7489063867016625,
      "grad_norm": 0.356035053730011,
      "learning_rate": 1e-06,
      "loss": -0.0457,
      "num_tokens": 173238416.0,
      "reward": 0.6071428656578064,
      "reward_std": 0.16281278431415558,
      "rewards/verify_math_reward/mean": 0.6071428656578064,
      "rewards/verify_math_reward/std": 0.48865827918052673,
      "step": 1177
    },
    {
      "clip_ratio/high_max": 0.0025467468294664286,
      "clip_ratio/high_mean": 0.0009412623439857271,
      "clip_ratio/low_mean": 0.0008732760998100275,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018145384601666592,
      "epoch": 2.7512394284047827,
      "grad_norm": 0.24822822213172913,
      "learning_rate": 1e-06,
      "loss": -0.046,
      "step": 1178
    },
    {
      "clip_ratio/high_max": 0.002678291952179279,
      "clip_ratio/high_mean": 0.0009726418738864595,
      "clip_ratio/low_mean": 0.0009707488916319562,
      "clip_ratio/low_min": 2.565154863987118e-05,
      "clip_ratio/region_mean": 0.001943390758242458,
      "epoch": 2.753572470107903,
      "grad_norm": 0.2327621877193451,
      "learning_rate": 1e-06,
      "loss": -0.0461,
      "step": 1179
    },
    {
      "clip_ratio/high_max": 0.0026214629324385896,
      "clip_ratio/high_mean": 0.0008822716972645139,
      "clip_ratio/low_mean": 0.0013206969961174764,
      "clip_ratio/low_min": 2.565154863987118e-05,
      "clip_ratio/region_mean": 0.002202968746132683,
      "epoch": 2.7559055118110236,
      "grad_norm": 0.23097383975982666,
      "learning_rate": 1e-06,
      "loss": -0.0462,
      "step": 1180
    },
    {
      "clip_ratio/high_max": 0.0026383646836620755,
      "clip_ratio/high_mean": 0.000980668108240934,
      "clip_ratio/low_mean": 0.0007353683377004927,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017160364659503102,
      "completions/clipped_ratio": 0.1696428571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4095.0,
      "completions/mean_length": 1247.09716796875,
      "completions/mean_terminated_length": 665.0631713867188,
      "completions/min_length": 155.0,
      "completions/min_terminated_length": 155.0,
      "epoch": 2.7582385535141443,
      "grad_norm": 0.32525336742401123,
      "learning_rate": 1e-06,
      "loss": -0.0709,
      "num_tokens": 173821887.0,
      "reward": 0.5345982313156128,
      "reward_std": 0.16930988430976868,
      "rewards/verify_math_reward/mean": 0.5345982313156128,
      "rewards/verify_math_reward/std": 0.4990801215171814,
      "step": 1181
    },
    {
      "clip_ratio/high_max": 0.002790924299915787,
      "clip_ratio/high_mean": 0.0010920978220383404,
      "clip_ratio/low_mean": 0.0009414568048669025,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020335546214482747,
      "epoch": 2.7605715952172645,
      "grad_norm": 0.25325608253479004,
      "learning_rate": 1e-06,
      "loss": -0.0711,
      "step": 1182
    },
    {
      "clip_ratio/high_max": 0.002858604333596304,
      "clip_ratio/high_mean": 0.0010866996581171406,
      "clip_ratio/low_mean": 0.0011052198096876964,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021919194914516993,
      "epoch": 2.7629046369203847,
      "grad_norm": 0.22945712506771088,
      "learning_rate": 1e-06,
      "loss": -0.0713,
      "step": 1183
    },
    {
      "clip_ratio/high_max": 0.002792059247440193,
      "clip_ratio/high_mean": 0.0011309509209240787,
      "clip_ratio/low_mean": 0.0013075094866508152,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002438460389385,
      "epoch": 2.7652376786235053,
      "grad_norm": 0.23172591626644135,
      "learning_rate": 1e-06,
      "loss": -0.0714,
      "step": 1184
    },
    {
      "clip_ratio/high_max": 0.003013772060512565,
      "clip_ratio/high_mean": 0.0011258381564402953,
      "clip_ratio/low_mean": 0.0008641459953651065,
      "clip_ratio/low_min": 1.304801662627142e-05,
      "clip_ratio/region_mean": 0.0019899841499864124,
      "completions/clipped_ratio": 0.1227678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3706.0,
      "completions/mean_length": 1096.640625,
      "completions/mean_terminated_length": 676.8829345703125,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 2.767570720326626,
      "grad_norm": 0.4517523944377899,
      "learning_rate": 1e-06,
      "loss": -0.0366,
      "num_tokens": 174442205.0,
      "reward": 0.5424107313156128,
      "reward_std": 0.18754372000694275,
      "rewards/verify_math_reward/mean": 0.5424107313156128,
      "rewards/verify_math_reward/std": 0.4984763562679291,
      "step": 1185
    },
    {
      "clip_ratio/high_max": 0.003333640306664165,
      "clip_ratio/high_mean": 0.001336344128503697,
      "clip_ratio/low_mean": 0.0012563255713757826,
      "clip_ratio/low_min": 6.52400849503465e-05,
      "clip_ratio/region_mean": 0.002592669588921126,
      "epoch": 2.769903762029746,
      "grad_norm": 0.3139788508415222,
      "learning_rate": 1e-06,
      "loss": -0.037,
      "step": 1186
    },
    {
      "clip_ratio/high_max": 0.0034307411697227508,
      "clip_ratio/high_mean": 0.0012566710538521875,
      "clip_ratio/low_mean": 0.0014524489160976373,
      "clip_ratio/low_min": 5.219206650508568e-05,
      "clip_ratio/region_mean": 0.002709119929932058,
      "epoch": 2.772236803732867,
      "grad_norm": 0.26903483271598816,
      "learning_rate": 1e-06,
      "loss": -0.0371,
      "step": 1187
    },
    {
      "clip_ratio/high_max": 0.0035063183604506776,
      "clip_ratio/high_mean": 0.00135813991073519,
      "clip_ratio/low_mean": 0.0017541509369038977,
      "clip_ratio/low_min": 7.303107850020751e-05,
      "clip_ratio/region_mean": 0.0031122907894314267,
      "epoch": 2.774569845435987,
      "grad_norm": 0.2814794182777405,
      "learning_rate": 1e-06,
      "loss": -0.0373,
      "step": 1188
    },
    {
      "clip_ratio/high_max": 0.0020838035343331285,
      "clip_ratio/high_mean": 0.0008119172434817301,
      "clip_ratio/low_mean": 0.0005080953824290191,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001320012645010138,
      "completions/clipped_ratio": 0.1729910714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4013.0,
      "completions/mean_length": 1269.9576416015625,
      "completions/mean_terminated_length": 678.8150634765625,
      "completions/min_length": 129.0,
      "completions/min_terminated_length": 129.0,
      "epoch": 2.7769028871391077,
      "grad_norm": 0.2438543736934662,
      "learning_rate": 1e-06,
      "loss": -0.0851,
      "num_tokens": 175037479.0,
      "reward": 0.5178571939468384,
      "reward_std": 0.1632988154888153,
      "rewards/verify_math_reward/mean": 0.5178571343421936,
      "rewards/verify_math_reward/std": 0.4999600946903229,
      "step": 1189
    },
    {
      "clip_ratio/high_max": 0.002365332911722362,
      "clip_ratio/high_mean": 0.0009482004315941595,
      "clip_ratio/low_mean": 0.000811347272247076,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017595476601854898,
      "epoch": 2.779235928842228,
      "grad_norm": 0.21394813060760498,
      "learning_rate": 1e-06,
      "loss": -0.0853,
      "step": 1190
    },
    {
      "clip_ratio/high_max": 0.0028085903213650454,
      "clip_ratio/high_mean": 0.001010134254102013,
      "clip_ratio/low_mean": 0.0008846170476317639,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018947512508020736,
      "epoch": 2.7815689705453486,
      "grad_norm": 0.20263056457042694,
      "learning_rate": 1e-06,
      "loss": -0.0854,
      "step": 1191
    },
    {
      "clip_ratio/high_max": 0.00252992239620653,
      "clip_ratio/high_mean": 0.0009914755883073667,
      "clip_ratio/low_mean": 0.0010134298354387283,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020049054437549785,
      "epoch": 2.783902012248469,
      "grad_norm": 0.20704755187034607,
      "learning_rate": 1e-06,
      "loss": -0.0854,
      "step": 1192
    },
    {
      "clip_ratio/high_max": 0.002115512397722341,
      "clip_ratio/high_mean": 0.0009101857149289572,
      "clip_ratio/low_mean": 0.0007816509087206214,
      "clip_ratio/low_min": 1.2021542715956457e-05,
      "clip_ratio/region_mean": 0.0016918366600293666,
      "completions/clipped_ratio": 0.1629464285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3332.0,
      "completions/mean_length": 1241.036865234375,
      "completions/mean_terminated_length": 685.2706298828125,
      "completions/min_length": 190.0,
      "completions/min_terminated_length": 190.0,
      "epoch": 2.7862350539515894,
      "grad_norm": 0.339778333902359,
      "learning_rate": 1e-06,
      "loss": -0.0496,
      "num_tokens": 175639264.0,
      "reward": 0.5212053656578064,
      "reward_std": 0.19377093017101288,
      "rewards/verify_math_reward/mean": 0.5212053656578064,
      "rewards/verify_math_reward/std": 0.49982914328575134,
      "step": 1193
    },
    {
      "clip_ratio/high_max": 0.002788091187539976,
      "clip_ratio/high_mean": 0.0011203964841115521,
      "clip_ratio/low_mean": 0.0010449898163642501,
      "clip_ratio/low_min": 1.3845812645740807e-05,
      "clip_ratio/region_mean": 0.0021653863150277175,
      "epoch": 2.7885680956547096,
      "grad_norm": 0.28704142570495605,
      "learning_rate": 1e-06,
      "loss": -0.0499,
      "step": 1194
    },
    {
      "clip_ratio/high_max": 0.003262251972046215,
      "clip_ratio/high_mean": 0.0012426850935298717,
      "clip_ratio/low_mean": 0.0013735083193751052,
      "clip_ratio/low_min": 2.6974536012858152e-05,
      "clip_ratio/region_mean": 0.002616193421999924,
      "epoch": 2.7909011373578303,
      "grad_norm": 0.24440589547157288,
      "learning_rate": 1e-06,
      "loss": -0.0501,
      "step": 1195
    },
    {
      "clip_ratio/high_max": 0.002734982001129538,
      "clip_ratio/high_mean": 0.0011236375848966418,
      "clip_ratio/low_mean": 0.0015607085042574909,
      "clip_ratio/low_min": 4.940060534863733e-05,
      "clip_ratio/region_mean": 0.002684346094611101,
      "epoch": 2.793234179060951,
      "grad_norm": 0.2630174458026886,
      "learning_rate": 1e-06,
      "loss": -0.0502,
      "step": 1196
    },
    {
      "clip_ratio/high_max": 0.003300103453511838,
      "clip_ratio/high_mean": 0.001191360746815917,
      "clip_ratio/low_mean": 0.0007367967900790973,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019281575951026753,
      "completions/clipped_ratio": 0.1573660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3343.0,
      "completions/mean_length": 1237.2467041015625,
      "completions/mean_terminated_length": 703.3602905273438,
      "completions/min_length": 137.0,
      "completions/min_terminated_length": 137.0,
      "epoch": 2.795567220764071,
      "grad_norm": 0.30969828367233276,
      "learning_rate": 1e-06,
      "loss": -0.0785,
      "num_tokens": 176254261.0,
      "reward": 0.5792410969734192,
      "reward_std": 0.16995178163051605,
      "rewards/verify_math_reward/mean": 0.5792410969734192,
      "rewards/verify_math_reward/std": 0.49395665526390076,
      "step": 1197
    },
    {
      "clip_ratio/high_max": 0.003714725600730162,
      "clip_ratio/high_mean": 0.0013485967392625753,
      "clip_ratio/low_mean": 0.0009606522280591889,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002309248950041365,
      "epoch": 2.7979002624671914,
      "grad_norm": 0.29674312472343445,
      "learning_rate": 1e-06,
      "loss": -0.0786,
      "step": 1198
    },
    {
      "clip_ratio/high_max": 0.0038086366112111136,
      "clip_ratio/high_mean": 0.0014416351423278684,
      "clip_ratio/low_mean": 0.0009979936794479727,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00243962890090188,
      "epoch": 2.800233304170312,
      "grad_norm": 0.3184277415275574,
      "learning_rate": 1e-06,
      "loss": -0.0788,
      "step": 1199
    },
    {
      "clip_ratio/high_max": 0.003577950566977961,
      "clip_ratio/high_mean": 0.0012790588716597995,
      "clip_ratio/low_mean": 0.001276157284337387,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0025552161896484904,
      "epoch": 2.8025663458734327,
      "grad_norm": 0.22036480903625488,
      "learning_rate": 1e-06,
      "loss": -0.079,
      "step": 1200
    },
    {
      "clip_ratio/high_max": 0.0021206787823757622,
      "clip_ratio/high_mean": 0.0008117430406855419,
      "clip_ratio/low_mean": 0.0005327217286321684,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013444648102449719,
      "completions/clipped_ratio": 0.1529017857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3625.0,
      "completions/mean_length": 1157.3828125,
      "completions/mean_terminated_length": 626.9605102539062,
      "completions/min_length": 183.0,
      "completions/min_terminated_length": 183.0,
      "epoch": 2.804899387576553,
      "grad_norm": 0.26239198446273804,
      "learning_rate": 1e-06,
      "loss": -0.0505,
      "num_tokens": 176810772.0,
      "reward": 0.5580357313156128,
      "reward_std": 0.14676883816719055,
      "rewards/verify_math_reward/mean": 0.5580357313156128,
      "rewards/verify_math_reward/std": 0.49689778685569763,
      "step": 1201
    },
    {
      "clip_ratio/high_max": 0.0025539233974996023,
      "clip_ratio/high_mean": 0.0009800572079257108,
      "clip_ratio/low_mean": 0.0007396926212095423,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017197498382302001,
      "epoch": 2.8072324292796735,
      "grad_norm": 0.19713591039180756,
      "learning_rate": 1e-06,
      "loss": -0.0507,
      "step": 1202
    },
    {
      "clip_ratio/high_max": 0.002599708161142189,
      "clip_ratio/high_mean": 0.0009262524527002824,
      "clip_ratio/low_mean": 0.0008012518883333541,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017275043101108167,
      "epoch": 2.8095654709827937,
      "grad_norm": 0.24003306031227112,
      "learning_rate": 1e-06,
      "loss": -0.0507,
      "step": 1203
    },
    {
      "clip_ratio/high_max": 0.002897015765483957,
      "clip_ratio/high_mean": 0.0010553792744758539,
      "clip_ratio/low_mean": 0.000998868641545414,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002054247903288342,
      "epoch": 2.8118985126859144,
      "grad_norm": 0.20766811072826385,
      "learning_rate": 1e-06,
      "loss": -0.0508,
      "step": 1204
    },
    {
      "clip_ratio/high_max": 0.0023108789464458823,
      "clip_ratio/high_mean": 0.0009054496049429872,
      "clip_ratio/low_mean": 0.0006721101754010306,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015775597821630072,
      "completions/clipped_ratio": 0.1852678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2730.0,
      "completions/mean_length": 1284.1239013671875,
      "completions/mean_terminated_length": 644.7109985351562,
      "completions/min_length": 164.0,
      "completions/min_terminated_length": 164.0,
      "epoch": 2.8142315543890346,
      "grad_norm": 0.33877506852149963,
      "learning_rate": 1e-06,
      "loss": -0.0541,
      "num_tokens": 177365099.0,
      "reward": 0.5245535969734192,
      "reward_std": 0.17570818960666656,
      "rewards/verify_math_reward/mean": 0.5245535969734192,
      "rewards/verify_math_reward/std": 0.4996756613254547,
      "step": 1205
    },
    {
      "clip_ratio/high_max": 0.002984140723128803,
      "clip_ratio/high_mean": 0.0012275164845050313,
      "clip_ratio/low_mean": 0.0009287758030041005,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021562922847806476,
      "epoch": 2.8165645960921553,
      "grad_norm": 0.3308020532131195,
      "learning_rate": 1e-06,
      "loss": -0.0544,
      "step": 1206
    },
    {
      "clip_ratio/high_max": 0.003067373127123574,
      "clip_ratio/high_mean": 0.0012576047138281865,
      "clip_ratio/low_mean": 0.0011606596617639298,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024182643537642434,
      "epoch": 2.8188976377952755,
      "grad_norm": 0.2738145589828491,
      "learning_rate": 1e-06,
      "loss": -0.0547,
      "step": 1207
    },
    {
      "clip_ratio/high_max": 0.0028996927321713883,
      "clip_ratio/high_mean": 0.0010932534914900316,
      "clip_ratio/low_mean": 0.0014001720101077808,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024934255197877064,
      "epoch": 2.821230679498396,
      "grad_norm": 0.35904988646507263,
      "learning_rate": 1e-06,
      "loss": -0.0547,
      "step": 1208
    },
    {
      "clip_ratio/high_max": 0.00207980446430156,
      "clip_ratio/high_mean": 0.0007250223825394642,
      "clip_ratio/low_mean": 0.0006736360783179407,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013986584708618466,
      "completions/clipped_ratio": 0.1729910714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4080.0,
      "completions/mean_length": 1330.8460693359375,
      "completions/mean_terminated_length": 752.43994140625,
      "completions/min_length": 109.0,
      "completions/min_terminated_length": 109.0,
      "epoch": 2.8235637212015163,
      "grad_norm": 0.2817864418029785,
      "learning_rate": 1e-06,
      "loss": -0.0639,
      "num_tokens": 178012521.0,
      "reward": 0.53125,
      "reward_std": 0.16198793053627014,
      "rewards/verify_math_reward/mean": 0.53125,
      "rewards/verify_math_reward/std": 0.4993011951446533,
      "step": 1209
    },
    {
      "clip_ratio/high_max": 0.0026954747299896553,
      "clip_ratio/high_mean": 0.0008574896291975165,
      "clip_ratio/low_mean": 0.0009417580295121297,
      "clip_ratio/low_min": 1.9860184693243355e-05,
      "clip_ratio/region_mean": 0.001799247671442572,
      "epoch": 2.825896762904637,
      "grad_norm": 0.2551283538341522,
      "learning_rate": 1e-06,
      "loss": -0.0641,
      "step": 1210
    },
    {
      "clip_ratio/high_max": 0.0026105109573109075,
      "clip_ratio/high_mean": 0.0008668896643939661,
      "clip_ratio/low_mean": 0.0011566333250812022,
      "clip_ratio/low_min": 2.131287328666076e-05,
      "clip_ratio/region_mean": 0.002023522974923253,
      "epoch": 2.8282298046077576,
      "grad_norm": 0.22527474164962769,
      "learning_rate": 1e-06,
      "loss": -0.0642,
      "step": 1211
    },
    {
      "clip_ratio/high_max": 0.002564192283898592,
      "clip_ratio/high_mean": 0.0008459131749987137,
      "clip_ratio/low_mean": 0.0012112642543797847,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002057177436654456,
      "epoch": 2.830562846310878,
      "grad_norm": 0.22328247129917145,
      "learning_rate": 1e-06,
      "loss": -0.0643,
      "step": 1212
    },
    {
      "clip_ratio/high_max": 0.002531326266762335,
      "clip_ratio/high_mean": 0.0010001528316934127,
      "clip_ratio/low_mean": 0.0007164640178416448,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017166168399853632,
      "completions/clipped_ratio": 0.1473214285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2916.0,
      "completions/mean_length": 1123.37060546875,
      "completions/mean_terminated_length": 609.77490234375,
      "completions/min_length": 175.0,
      "completions/min_terminated_length": 175.0,
      "epoch": 2.832895888013998,
      "grad_norm": 0.3418889343738556,
      "learning_rate": 1e-06,
      "loss": -0.0552,
      "num_tokens": 178569413.0,
      "reward": 0.6004464626312256,
      "reward_std": 0.17634011805057526,
      "rewards/verify_math_reward/mean": 0.6004464030265808,
      "rewards/verify_math_reward/std": 0.49008017778396606,
      "step": 1213
    },
    {
      "clip_ratio/high_max": 0.002907997964939568,
      "clip_ratio/high_mean": 0.0012003184456261806,
      "clip_ratio/low_mean": 0.000967797907833301,
      "clip_ratio/low_min": 1.5166221601248253e-05,
      "clip_ratio/region_mean": 0.0021681163270841353,
      "epoch": 2.8352289297171187,
      "grad_norm": 0.2897513210773468,
      "learning_rate": 1e-06,
      "loss": -0.0554,
      "step": 1214
    },
    {
      "clip_ratio/high_max": 0.002957124510430731,
      "clip_ratio/high_mean": 0.0010831455474544782,
      "clip_ratio/low_mean": 0.0012507559713412775,
      "clip_ratio/low_min": 1.1959433322772384e-05,
      "clip_ratio/region_mean": 0.002333901538804639,
      "epoch": 2.8375619714202394,
      "grad_norm": 0.2763117849826813,
      "learning_rate": 1e-06,
      "loss": -0.0556,
      "step": 1215
    },
    {
      "clip_ratio/high_max": 0.0029035662228125148,
      "clip_ratio/high_mean": 0.001143075820436934,
      "clip_ratio/low_mean": 0.001538982873171335,
      "clip_ratio/low_min": 3.0332443202496506e-05,
      "clip_ratio/region_mean": 0.0026820587008842267,
      "epoch": 2.8398950131233596,
      "grad_norm": 0.25847601890563965,
      "learning_rate": 1e-06,
      "loss": -0.0557,
      "step": 1216
    },
    {
      "clip_ratio/high_max": 0.002246396463306155,
      "clip_ratio/high_mean": 0.0008512069762218744,
      "clip_ratio/low_mean": 0.0006472689628935768,
      "clip_ratio/low_min": 1.4785900020797271e-05,
      "clip_ratio/region_mean": 0.0014984759218350518,
      "completions/clipped_ratio": 0.1696428571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4058.0,
      "completions/mean_length": 1240.4609375,
      "completions/mean_terminated_length": 657.0712280273438,
      "completions/min_length": 153.0,
      "completions/min_terminated_length": 153.0,
      "epoch": 2.8422280548264798,
      "grad_norm": 0.3724900484085083,
      "learning_rate": 1e-06,
      "loss": -0.0451,
      "num_tokens": 179157538.0,
      "reward": 0.559151828289032,
      "reward_std": 0.15642336010932922,
      "rewards/verify_math_reward/mean": 0.5591517686843872,
      "rewards/verify_math_reward/std": 0.496766060590744,
      "step": 1217
    },
    {
      "clip_ratio/high_max": 0.002594152738311095,
      "clip_ratio/high_mean": 0.0010427189536130754,
      "clip_ratio/low_mean": 0.0009653071083448594,
      "clip_ratio/low_min": 1.4785900020797271e-05,
      "clip_ratio/region_mean": 0.0020080260692338925,
      "epoch": 2.8445610965296004,
      "grad_norm": 0.27912214398384094,
      "learning_rate": 1e-06,
      "loss": -0.0455,
      "step": 1218
    },
    {
      "clip_ratio/high_max": 0.002910590839746874,
      "clip_ratio/high_mean": 0.001090924324671505,
      "clip_ratio/low_mean": 0.0010437631863169372,
      "clip_ratio/low_min": 2.9571800041594543e-05,
      "clip_ratio/region_mean": 0.0021346874891605694,
      "epoch": 2.846894138232721,
      "grad_norm": 0.25943613052368164,
      "learning_rate": 1e-06,
      "loss": -0.0456,
      "step": 1219
    },
    {
      "clip_ratio/high_max": 0.00258619509259006,
      "clip_ratio/high_mean": 0.0009054034726432292,
      "clip_ratio/low_mean": 0.0012875327192887198,
      "clip_ratio/low_min": 4.4357700971886516e-05,
      "clip_ratio/region_mean": 0.0021929361901129596,
      "epoch": 2.8492271799358413,
      "grad_norm": 0.23242810368537903,
      "learning_rate": 1e-06,
      "loss": -0.0457,
      "step": 1220
    },
    {
      "clip_ratio/high_max": 0.002463534357957542,
      "clip_ratio/high_mean": 0.0009862915576377418,
      "clip_ratio/low_mean": 0.000740970635888516,
      "clip_ratio/low_min": 3.55821248376742e-05,
      "clip_ratio/region_mean": 0.0017272621698793955,
      "completions/clipped_ratio": 0.171875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3266.0,
      "completions/mean_length": 1230.34716796875,
      "completions/mean_terminated_length": 635.5889282226562,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 2.851560221638962,
      "grad_norm": 0.33978813886642456,
      "learning_rate": 1e-06,
      "loss": -0.0675,
      "num_tokens": 179712281.0,
      "reward": 0.5446428656578064,
      "reward_std": 0.1726941168308258,
      "rewards/verify_math_reward/mean": 0.5446428656578064,
      "rewards/verify_math_reward/std": 0.49828118085861206,
      "step": 1221
    },
    {
      "clip_ratio/high_max": 0.0031510016779066063,
      "clip_ratio/high_mean": 0.0011794015936175128,
      "clip_ratio/low_mean": 0.0008715644653420895,
      "clip_ratio/low_min": 4.572310263029067e-05,
      "clip_ratio/region_mean": 0.002050966075330507,
      "epoch": 2.853893263342082,
      "grad_norm": 0.3055119216442108,
      "learning_rate": 1e-06,
      "loss": -0.0677,
      "step": 1222
    },
    {
      "clip_ratio/high_max": 0.0033840225733001716,
      "clip_ratio/high_mean": 0.0012274939244889538,
      "clip_ratio/low_mean": 0.0011448321802163264,
      "clip_ratio/low_min": 1.3261192179925274e-05,
      "clip_ratio/region_mean": 0.0023723261037957855,
      "epoch": 2.856226305045203,
      "grad_norm": 0.2623078227043152,
      "learning_rate": 1e-06,
      "loss": -0.068,
      "step": 1223
    },
    {
      "clip_ratio/high_max": 0.0027565543532546144,
      "clip_ratio/high_mean": 0.0010786218917928636,
      "clip_ratio/low_mean": 0.0014120109372015577,
      "clip_ratio/low_min": 9.161041270999704e-05,
      "clip_ratio/region_mean": 0.0024906328108045273,
      "epoch": 2.858559346748323,
      "grad_norm": 0.2556232213973999,
      "learning_rate": 1e-06,
      "loss": -0.068,
      "step": 1224
    },
    {
      "clip_ratio/high_max": 0.002132123066985514,
      "clip_ratio/high_mean": 0.000865826683366322,
      "clip_ratio/low_mean": 0.0004899367630741835,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013557634192693513,
      "completions/clipped_ratio": 0.1841517857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3233.0,
      "completions/mean_length": 1271.3717041015625,
      "completions/mean_terminated_length": 633.8016967773438,
      "completions/min_length": 157.0,
      "completions/min_terminated_length": 157.0,
      "epoch": 2.8608923884514437,
      "grad_norm": 0.3052181601524353,
      "learning_rate": 1e-06,
      "loss": -0.0542,
      "num_tokens": 180259150.0,
      "reward": 0.5345982313156128,
      "reward_std": 0.1515420526266098,
      "rewards/verify_math_reward/mean": 0.5345982313156128,
      "rewards/verify_math_reward/std": 0.4990801215171814,
      "step": 1225
    },
    {
      "clip_ratio/high_max": 0.00272468279581517,
      "clip_ratio/high_mean": 0.0010511810796742793,
      "clip_ratio/low_mean": 0.0006787887932659942,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017299698738497682,
      "epoch": 2.863225430154564,
      "grad_norm": 0.3244656026363373,
      "learning_rate": 1e-06,
      "loss": -0.0545,
      "step": 1226
    },
    {
      "clip_ratio/high_max": 0.0026379339178674854,
      "clip_ratio/high_mean": 0.001092517632059753,
      "clip_ratio/low_mean": 0.0008542970224425517,
      "clip_ratio/low_min": 2.0647505152737722e-05,
      "clip_ratio/region_mean": 0.0019468145837890916,
      "epoch": 2.8655584718576845,
      "grad_norm": 0.2317638099193573,
      "learning_rate": 1e-06,
      "loss": -0.0546,
      "step": 1227
    },
    {
      "clip_ratio/high_max": 0.002808631288644392,
      "clip_ratio/high_mean": 0.0010549850030656671,
      "clip_ratio/low_mean": 0.0010281999293511035,
      "clip_ratio/low_min": 1.575497844896745e-05,
      "clip_ratio/region_mean": 0.002083184925140813,
      "epoch": 2.8678915135608047,
      "grad_norm": 0.2297726422548294,
      "learning_rate": 1e-06,
      "loss": -0.0546,
      "step": 1228
    },
    {
      "clip_ratio/high_max": 0.0020344149161246605,
      "clip_ratio/high_mean": 0.0008280365382233867,
      "clip_ratio/low_mean": 0.0006267763910727808,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001454812940210104,
      "completions/clipped_ratio": 0.1339285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2822.0,
      "completions/mean_length": 1098.884033203125,
      "completions/mean_terminated_length": 635.412353515625,
      "completions/min_length": 168.0,
      "completions/min_terminated_length": 168.0,
      "epoch": 2.8702245552639254,
      "grad_norm": 0.37846052646636963,
      "learning_rate": 1e-06,
      "loss": -0.0355,
      "num_tokens": 180839038.0,
      "reward": 0.6049107313156128,
      "reward_std": 0.1555236279964447,
      "rewards/verify_math_reward/mean": 0.6049107313156128,
      "rewards/verify_math_reward/std": 0.48914292454719543,
      "step": 1229
    },
    {
      "clip_ratio/high_max": 0.002511028382286895,
      "clip_ratio/high_mean": 0.0010764880844362779,
      "clip_ratio/low_mean": 0.0007981017679412616,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018745899287750944,
      "epoch": 2.872557596967046,
      "grad_norm": 0.2716333568096161,
      "learning_rate": 1e-06,
      "loss": -0.0358,
      "step": 1230
    },
    {
      "clip_ratio/high_max": 0.0027851527556777,
      "clip_ratio/high_mean": 0.0010962301967083476,
      "clip_ratio/low_mean": 0.0010780406737467274,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00217427085590316,
      "epoch": 2.8748906386701663,
      "grad_norm": 0.24592313170433044,
      "learning_rate": 1e-06,
      "loss": -0.0361,
      "step": 1231
    },
    {
      "clip_ratio/high_max": 0.0025209656268998515,
      "clip_ratio/high_mean": 0.000988474640507775,
      "clip_ratio/low_mean": 0.0013377151772147045,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023261897658812813,
      "epoch": 2.8772236803732865,
      "grad_norm": 0.2778870463371277,
      "learning_rate": 1e-06,
      "loss": -0.0361,
      "step": 1232
    },
    {
      "clip_ratio/high_max": 0.0020293524794396944,
      "clip_ratio/high_mean": 0.0008347573566425126,
      "clip_ratio/low_mean": 0.0005649715913023101,
      "clip_ratio/low_min": 1.323311425949214e-05,
      "clip_ratio/region_mean": 0.0013997289497638121,
      "completions/clipped_ratio": 0.1707589285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3990.0,
      "completions/mean_length": 1249.10498046875,
      "completions/mean_terminated_length": 662.8667602539062,
      "completions/min_length": 162.0,
      "completions/min_terminated_length": 162.0,
      "epoch": 2.879556722076407,
      "grad_norm": 0.2888152599334717,
      "learning_rate": 1e-06,
      "loss": -0.0342,
      "num_tokens": 181418684.0,
      "reward": 0.520089328289032,
      "reward_std": 0.1504133939743042,
      "rewards/verify_math_reward/mean": 0.5200892686843872,
      "rewards/verify_math_reward/std": 0.4998753070831299,
      "step": 1233
    },
    {
      "clip_ratio/high_max": 0.0025165256338368636,
      "clip_ratio/high_mean": 0.0009329646563855931,
      "clip_ratio/low_mean": 0.0008046613766055088,
      "clip_ratio/low_min": 1.8296252164873295e-05,
      "clip_ratio/region_mean": 0.0017376260075252503,
      "epoch": 2.8818897637795278,
      "grad_norm": 0.22874966263771057,
      "learning_rate": 1e-06,
      "loss": -0.0344,
      "step": 1234
    },
    {
      "clip_ratio/high_max": 0.00254748542465677,
      "clip_ratio/high_mean": 0.0010051646777355927,
      "clip_ratio/low_mean": 0.0009650562169554178,
      "clip_ratio/low_min": 1.3403388038568664e-05,
      "clip_ratio/region_mean": 0.001970220921066357,
      "epoch": 2.884222805482648,
      "grad_norm": 0.21959078311920166,
      "learning_rate": 1e-06,
      "loss": -0.0345,
      "step": 1235
    },
    {
      "clip_ratio/high_max": 0.0026146117743337527,
      "clip_ratio/high_mean": 0.000941455909924116,
      "clip_ratio/low_mean": 0.0010380630519648548,
      "clip_ratio/low_min": 5.293245703796856e-05,
      "clip_ratio/region_mean": 0.00197951905647642,
      "epoch": 2.886555847185768,
      "grad_norm": 0.28595098853111267,
      "learning_rate": 1e-06,
      "loss": -0.0346,
      "step": 1236
    },
    {
      "clip_ratio/high_max": 0.002767169673461467,
      "clip_ratio/high_mean": 0.0011425521242927061,
      "clip_ratio/low_mean": 0.0009676702266006032,
      "clip_ratio/low_min": 4.705656374426326e-05,
      "clip_ratio/region_mean": 0.0021102223327034153,
      "completions/clipped_ratio": 0.1662946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2665.0,
      "completions/mean_length": 1280.4320068359375,
      "completions/mean_terminated_length": 718.8259887695312,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 2.888888888888889,
      "grad_norm": 0.3602418303489685,
      "learning_rate": 1e-06,
      "loss": -0.0748,
      "num_tokens": 182040983.0,
      "reward": 0.4988839626312256,
      "reward_std": 0.20282992720603943,
      "rewards/verify_math_reward/mean": 0.4988839328289032,
      "rewards/verify_math_reward/std": 0.5002779960632324,
      "step": 1237
    },
    {
      "clip_ratio/high_max": 0.003788474656175822,
      "clip_ratio/high_mean": 0.0013509797827282455,
      "clip_ratio/low_mean": 0.0012202237157907803,
      "clip_ratio/low_min": 2.8419009140634444e-05,
      "clip_ratio/region_mean": 0.0025712034985190257,
      "epoch": 2.8912219305920095,
      "grad_norm": 0.2973235547542572,
      "learning_rate": 1e-06,
      "loss": -0.0751,
      "step": 1238
    },
    {
      "clip_ratio/high_max": 0.0034605427645146847,
      "clip_ratio/high_mean": 0.0013756530061073136,
      "clip_ratio/low_mean": 0.0014561256757588126,
      "clip_ratio/low_min": 9.341506392956944e-05,
      "clip_ratio/region_mean": 0.002831778627296444,
      "epoch": 2.8935549722951297,
      "grad_norm": 0.2692912220954895,
      "learning_rate": 1e-06,
      "loss": -0.0753,
      "step": 1239
    },
    {
      "clip_ratio/high_max": 0.0036129713262198493,
      "clip_ratio/high_mean": 0.0013319130812305957,
      "clip_ratio/low_mean": 0.001683001533820061,
      "clip_ratio/low_min": 7.718307097093202e-05,
      "clip_ratio/region_mean": 0.0030149146332405508,
      "epoch": 2.8958880139982504,
      "grad_norm": 0.27325916290283203,
      "learning_rate": 1e-06,
      "loss": -0.0754,
      "step": 1240
    },
    {
      "clip_ratio/high_max": 0.0032859176135389134,
      "clip_ratio/high_mean": 0.0013003923741052859,
      "clip_ratio/low_mean": 0.0005826023161716876,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018829947512131184,
      "completions/clipped_ratio": 0.1584821428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3380.0,
      "completions/mean_length": 1168.388427734375,
      "completions/mean_terminated_length": 617.0344848632812,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 2.8982210557013706,
      "grad_norm": 0.4320470690727234,
      "learning_rate": 1e-06,
      "loss": -0.0908,
      "num_tokens": 182605411.0,
      "reward": 0.6127232313156128,
      "reward_std": 0.17979852855205536,
      "rewards/verify_math_reward/mean": 0.6127232313156128,
      "rewards/verify_math_reward/std": 0.4873998463153839,
      "step": 1241
    },
    {
      "clip_ratio/high_max": 0.003133914178761188,
      "clip_ratio/high_mean": 0.0012624690498341806,
      "clip_ratio/low_mean": 0.0008868542624895781,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021493233216460794,
      "epoch": 2.900554097404491,
      "grad_norm": 0.33980104327201843,
      "learning_rate": 1e-06,
      "loss": -0.091,
      "step": 1242
    },
    {
      "clip_ratio/high_max": 0.003558458687621169,
      "clip_ratio/high_mean": 0.0014487796688626986,
      "clip_ratio/low_mean": 0.0011010279954462021,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002549807686591521,
      "epoch": 2.9028871391076114,
      "grad_norm": 0.2559306025505066,
      "learning_rate": 1e-06,
      "loss": -0.0913,
      "step": 1243
    },
    {
      "clip_ratio/high_max": 0.0033493489027023315,
      "clip_ratio/high_mean": 0.0013537378908949904,
      "clip_ratio/low_mean": 0.0012544961227831664,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002608234019135125,
      "epoch": 2.905220180810732,
      "grad_norm": 0.26972684264183044,
      "learning_rate": 1e-06,
      "loss": -0.0913,
      "step": 1244
    },
    {
      "clip_ratio/high_max": 0.0020876192866126075,
      "clip_ratio/high_mean": 0.0006878449212308624,
      "clip_ratio/low_mean": 0.0005604099387710448,
      "clip_ratio/low_min": 1.269551103177946e-05,
      "clip_ratio/region_mean": 0.0012482548845582642,
      "completions/clipped_ratio": 0.1808035714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3991.0,
      "completions/mean_length": 1270.688720703125,
      "completions/mean_terminated_length": 647.1185302734375,
      "completions/min_length": 105.0,
      "completions/min_terminated_length": 105.0,
      "epoch": 2.9075532225138523,
      "grad_norm": 0.23925110697746277,
      "learning_rate": 1e-06,
      "loss": -0.0772,
      "num_tokens": 183170740.0,
      "reward": 0.5680803656578064,
      "reward_std": 0.14078985154628754,
      "rewards/verify_math_reward/mean": 0.5680803656578064,
      "rewards/verify_math_reward/std": 0.4956200420856476,
      "step": 1245
    },
    {
      "clip_ratio/high_max": 0.002451334377838066,
      "clip_ratio/high_mean": 0.0008019774531931034,
      "clip_ratio/low_mean": 0.0007614677442688844,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001563445184729062,
      "epoch": 2.909886264216973,
      "grad_norm": 0.24351325631141663,
      "learning_rate": 1e-06,
      "loss": -0.0773,
      "step": 1246
    },
    {
      "clip_ratio/high_max": 0.002328195438167313,
      "clip_ratio/high_mean": 0.0008338033185282256,
      "clip_ratio/low_mean": 0.0008368225990125211,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001670625941187609,
      "epoch": 2.912219305920093,
      "grad_norm": 0.23359844088554382,
      "learning_rate": 1e-06,
      "loss": -0.0774,
      "step": 1247
    },
    {
      "clip_ratio/high_max": 0.0023826419565011747,
      "clip_ratio/high_mean": 0.0008263971176347695,
      "clip_ratio/low_mean": 0.0010366100559622282,
      "clip_ratio/low_min": 2.891510484914761e-05,
      "clip_ratio/region_mean": 0.0018630072263476904,
      "epoch": 2.914552347623214,
      "grad_norm": 0.1954878568649292,
      "learning_rate": 1e-06,
      "loss": -0.0775,
      "step": 1248
    },
    {
      "clip_ratio/high_max": 0.002909617527620867,
      "clip_ratio/high_mean": 0.001141853274020832,
      "clip_ratio/low_mean": 0.0006739021519024391,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018157554331992287,
      "completions/clipped_ratio": 0.1997767857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3484.0,
      "completions/mean_length": 1377.587158203125,
      "completions/mean_terminated_length": 698.931640625,
      "completions/min_length": 191.0,
      "completions/min_terminated_length": 191.0,
      "epoch": 2.9168853893263345,
      "grad_norm": 0.3355540931224823,
      "learning_rate": 1e-06,
      "loss": -0.1229,
      "num_tokens": 183765530.0,
      "reward": 0.5133928656578064,
      "reward_std": 0.2020464688539505,
      "rewards/verify_math_reward/mean": 0.5133928656578064,
      "rewards/verify_math_reward/std": 0.500099778175354,
      "step": 1249
    },
    {
      "clip_ratio/high_max": 0.003443351000896655,
      "clip_ratio/high_mean": 0.0013787734824290965,
      "clip_ratio/low_mean": 0.0009193578698614147,
      "clip_ratio/low_min": 1.67358411999885e-05,
      "clip_ratio/region_mean": 0.002298131315910723,
      "epoch": 2.9192184310294547,
      "grad_norm": 0.3063414394855499,
      "learning_rate": 1e-06,
      "loss": -0.1232,
      "step": 1250
    },
    {
      "clip_ratio/high_max": 0.003658572015410755,
      "clip_ratio/high_mean": 0.0014169354108162224,
      "clip_ratio/low_mean": 0.0011746121272153687,
      "clip_ratio/low_min": 8.36792059999425e-06,
      "clip_ratio/region_mean": 0.0025915475343936123,
      "epoch": 2.921551472732575,
      "grad_norm": 0.24443690478801727,
      "learning_rate": 1e-06,
      "loss": -0.1235,
      "step": 1251
    },
    {
      "clip_ratio/high_max": 0.0033247855681111105,
      "clip_ratio/high_mean": 0.0012643429545278195,
      "clip_ratio/low_mean": 0.0012776335424860008,
      "clip_ratio/low_min": 1.67358411999885e-05,
      "clip_ratio/region_mean": 0.002541976544307545,
      "epoch": 2.9238845144356955,
      "grad_norm": 0.2677989602088928,
      "learning_rate": 1e-06,
      "loss": -0.1235,
      "step": 1252
    },
    {
      "clip_ratio/high_max": 0.0031514803486061282,
      "clip_ratio/high_mean": 0.0010818075570568908,
      "clip_ratio/low_mean": 0.0005988120160509425,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016806195926619694,
      "completions/clipped_ratio": 0.2265625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3857.0,
      "completions/mean_length": 1470.782470703125,
      "completions/mean_terminated_length": 701.7792358398438,
      "completions/min_length": 168.0,
      "completions/min_terminated_length": 168.0,
      "epoch": 2.926217556138816,
      "grad_norm": 0.36490005254745483,
      "learning_rate": 1e-06,
      "loss": -0.0912,
      "num_tokens": 184341335.0,
      "reward": 0.5033482313156128,
      "reward_std": 0.17340727150440216,
      "rewards/verify_math_reward/mean": 0.5033482313156128,
      "rewards/verify_math_reward/std": 0.5002680420875549,
      "step": 1253
    },
    {
      "clip_ratio/high_max": 0.003511617695039604,
      "clip_ratio/high_mean": 0.0012043408587487647,
      "clip_ratio/low_mean": 0.0009400233684573323,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00214436421811115,
      "epoch": 2.9285505978419364,
      "grad_norm": 0.3625325560569763,
      "learning_rate": 1e-06,
      "loss": -0.0915,
      "step": 1254
    },
    {
      "clip_ratio/high_max": 0.0036964921237085946,
      "clip_ratio/high_mean": 0.0013299772872414906,
      "clip_ratio/low_mean": 0.0011411833074816968,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024711605874472298,
      "epoch": 2.9308836395450566,
      "grad_norm": 0.2712169885635376,
      "learning_rate": 1e-06,
      "loss": -0.0917,
      "step": 1255
    },
    {
      "clip_ratio/high_max": 0.0035515071940608323,
      "clip_ratio/high_mean": 0.0011886032934853574,
      "clip_ratio/low_mean": 0.001462831896787975,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026514351993682794,
      "epoch": 2.9332166812481772,
      "grad_norm": 0.28189143538475037,
      "learning_rate": 1e-06,
      "loss": -0.0918,
      "step": 1256
    },
    {
      "clip_ratio/high_max": 0.0029168042092351243,
      "clip_ratio/high_mean": 0.0011313008290017024,
      "clip_ratio/low_mean": 0.0006421525376936188,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017734533903421834,
      "completions/clipped_ratio": 0.1729910714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3267.0,
      "completions/mean_length": 1248.1785888671875,
      "completions/mean_terminated_length": 652.4804077148438,
      "completions/min_length": 142.0,
      "completions/min_terminated_length": 142.0,
      "epoch": 2.935549722951298,
      "grad_norm": 0.32276833057403564,
      "learning_rate": 1e-06,
      "loss": -0.0691,
      "num_tokens": 184919535.0,
      "reward": 0.5446428656578064,
      "reward_std": 0.17130404710769653,
      "rewards/verify_math_reward/mean": 0.5446428656578064,
      "rewards/verify_math_reward/std": 0.4982811510562897,
      "step": 1257
    },
    {
      "clip_ratio/high_max": 0.003362276627740357,
      "clip_ratio/high_mean": 0.0013551082702178974,
      "clip_ratio/low_mean": 0.0008055710422922857,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021606793234241195,
      "epoch": 2.937882764654418,
      "grad_norm": 0.2880367338657379,
      "learning_rate": 1e-06,
      "loss": -0.0692,
      "step": 1258
    },
    {
      "clip_ratio/high_max": 0.0032669836000422947,
      "clip_ratio/high_mean": 0.0013611893409688491,
      "clip_ratio/low_mean": 0.0010897846113948617,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024509740105713718,
      "epoch": 2.9402158063575388,
      "grad_norm": 0.2908535301685333,
      "learning_rate": 1e-06,
      "loss": -0.0694,
      "step": 1259
    },
    {
      "clip_ratio/high_max": 0.003219017293304205,
      "clip_ratio/high_mean": 0.0012895310610474553,
      "clip_ratio/low_mean": 0.0012497504794737324,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002539281500503421,
      "epoch": 2.942548848060659,
      "grad_norm": 0.25888124108314514,
      "learning_rate": 1e-06,
      "loss": -0.0695,
      "step": 1260
    },
    {
      "clip_ratio/high_max": 0.0031690335308667272,
      "clip_ratio/high_mean": 0.0010736659132817294,
      "clip_ratio/low_mean": 0.000851191491165082,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019248574171797372,
      "completions/clipped_ratio": 0.1316964285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4078.0,
      "completions/mean_length": 1107.841552734375,
      "completions/mean_terminated_length": 654.6246948242188,
      "completions/min_length": 181.0,
      "completions/min_terminated_length": 181.0,
      "epoch": 2.9448818897637796,
      "grad_norm": 0.41687285900115967,
      "learning_rate": 1e-06,
      "loss": -0.0384,
      "num_tokens": 185523905.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.17070811986923218,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514803290367126,
      "step": 1261
    },
    {
      "clip_ratio/high_max": 0.003776175442908425,
      "clip_ratio/high_mean": 0.0013129156450304436,
      "clip_ratio/low_mean": 0.0011358087704138597,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024487245027557947,
      "epoch": 2.9472149314669,
      "grad_norm": 0.31890761852264404,
      "learning_rate": 1e-06,
      "loss": -0.0386,
      "step": 1262
    },
    {
      "clip_ratio/high_max": 0.0034182021554443054,
      "clip_ratio/high_mean": 0.0012003357005596627,
      "clip_ratio/low_mean": 0.0012306724784139078,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024310082371812314,
      "epoch": 2.9495479731700205,
      "grad_norm": 0.31810462474823,
      "learning_rate": 1e-06,
      "loss": -0.0388,
      "step": 1263
    },
    {
      "clip_ratio/high_max": 0.0037536010349867865,
      "clip_ratio/high_mean": 0.0012760615500155836,
      "clip_ratio/low_mean": 0.0015386828272312414,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0028147443663328886,
      "epoch": 2.9518810148731407,
      "grad_norm": 0.330525279045105,
      "learning_rate": 1e-06,
      "loss": -0.0389,
      "step": 1264
    },
    {
      "clip_ratio/high_max": 0.003201582541805692,
      "clip_ratio/high_mean": 0.0011561989849724341,
      "clip_ratio/low_mean": 0.0007509991464758059,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019071980932494625,
      "completions/clipped_ratio": 0.2131696428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3647.0,
      "completions/mean_length": 1442.813720703125,
      "completions/mean_terminated_length": 724.007080078125,
      "completions/min_length": 162.0,
      "completions/min_terminated_length": 162.0,
      "epoch": 2.9542140565762613,
      "grad_norm": 0.3820577561855316,
      "learning_rate": 1e-06,
      "loss": -0.1146,
      "num_tokens": 186121642.0,
      "reward": 0.520089328289032,
      "reward_std": 0.19857734441757202,
      "rewards/verify_math_reward/mean": 0.5200892686843872,
      "rewards/verify_math_reward/std": 0.4998753070831299,
      "step": 1265
    },
    {
      "clip_ratio/high_max": 0.0037811147049069405,
      "clip_ratio/high_mean": 0.001389671135257231,
      "clip_ratio/low_mean": 0.0010202501507592387,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002409921267826576,
      "epoch": 2.9565470982793816,
      "grad_norm": 0.27282702922821045,
      "learning_rate": 1e-06,
      "loss": -0.1149,
      "step": 1266
    },
    {
      "clip_ratio/high_max": 0.0034939453471451998,
      "clip_ratio/high_mean": 0.0013106262995279394,
      "clip_ratio/low_mean": 0.001251272336958209,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002561898661952,
      "epoch": 2.958880139982502,
      "grad_norm": 0.27828308939933777,
      "learning_rate": 1e-06,
      "loss": -0.1151,
      "step": 1267
    },
    {
      "clip_ratio/high_max": 0.0037431169039336964,
      "clip_ratio/high_mean": 0.0013020098085689824,
      "clip_ratio/low_mean": 0.001392197300447151,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0026942070908262394,
      "epoch": 2.961213181685623,
      "grad_norm": 0.3036958575248718,
      "learning_rate": 1e-06,
      "loss": -0.1152,
      "step": 1268
    },
    {
      "clip_ratio/high_max": 0.0024892619549063966,
      "clip_ratio/high_mean": 0.0008576363125030184,
      "clip_ratio/low_mean": 0.0006019526936142938,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001459589009755291,
      "completions/clipped_ratio": 0.1674107142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2883.0,
      "completions/mean_length": 1226.3046875,
      "completions/mean_terminated_length": 649.2882080078125,
      "completions/min_length": 175.0,
      "completions/min_terminated_length": 175.0,
      "epoch": 2.963546223388743,
      "grad_norm": 0.30294013023376465,
      "learning_rate": 1e-06,
      "loss": -0.088,
      "num_tokens": 186695163.0,
      "reward": 0.5636160969734192,
      "reward_std": 0.1510867178440094,
      "rewards/verify_math_reward/mean": 0.5636160969734192,
      "rewards/verify_math_reward/std": 0.49621346592903137,
      "step": 1269
    },
    {
      "clip_ratio/high_max": 0.0030007847453816794,
      "clip_ratio/high_mean": 0.0010451485723024234,
      "clip_ratio/low_mean": 0.0007863998307584552,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018315484121558256,
      "epoch": 2.9658792650918633,
      "grad_norm": 0.2306789755821228,
      "learning_rate": 1e-06,
      "loss": -0.0882,
      "step": 1270
    },
    {
      "clip_ratio/high_max": 0.002864803740521893,
      "clip_ratio/high_mean": 0.0010004250652855262,
      "clip_ratio/low_mean": 0.0008507482689310564,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018511733578634448,
      "epoch": 2.968212306794984,
      "grad_norm": 0.23171015083789825,
      "learning_rate": 1e-06,
      "loss": -0.0883,
      "step": 1271
    },
    {
      "clip_ratio/high_max": 0.0022417276850319467,
      "clip_ratio/high_mean": 0.0008938817300077062,
      "clip_ratio/low_mean": 0.0010590481469989754,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019529298806446604,
      "epoch": 2.9705453484981046,
      "grad_norm": 0.21633273363113403,
      "learning_rate": 1e-06,
      "loss": -0.0883,
      "step": 1272
    },
    {
      "clip_ratio/high_max": 0.002702313297049841,
      "clip_ratio/high_mean": 0.000888308519279235,
      "clip_ratio/low_mean": 0.0006352464279189007,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015235549326462205,
      "completions/clipped_ratio": 0.203125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4057.0,
      "completions/mean_length": 1356.8382568359375,
      "completions/mean_terminated_length": 658.6204833984375,
      "completions/min_length": 160.0,
      "completions/min_terminated_length": 160.0,
      "epoch": 2.972878390201225,
      "grad_norm": 0.3829348683357239,
      "learning_rate": 1e-06,
      "loss": -0.0772,
      "num_tokens": 187255202.0,
      "reward": 0.5223214626312256,
      "reward_std": 0.15774601697921753,
      "rewards/verify_math_reward/mean": 0.5223214030265808,
      "rewards/verify_math_reward/std": 0.49978047609329224,
      "step": 1273
    },
    {
      "clip_ratio/high_max": 0.003664480900624767,
      "clip_ratio/high_mean": 0.001145835303759668,
      "clip_ratio/low_mean": 0.0008248937992902938,
      "clip_ratio/low_min": 1.3519359526981134e-05,
      "clip_ratio/region_mean": 0.001970729099411983,
      "epoch": 2.9752114319043454,
      "grad_norm": 0.2908918261528015,
      "learning_rate": 1e-06,
      "loss": -0.0773,
      "step": 1274
    },
    {
      "clip_ratio/high_max": 0.0035258611969766207,
      "clip_ratio/high_mean": 0.0011402282234485028,
      "clip_ratio/low_mean": 0.0010692770429159282,
      "clip_ratio/low_min": 1.970987068489194e-05,
      "clip_ratio/region_mean": 0.0022095053136581555,
      "epoch": 2.9775444736074657,
      "grad_norm": 0.2749641239643097,
      "learning_rate": 1e-06,
      "loss": -0.0776,
      "step": 1275
    },
    {
      "clip_ratio/high_max": 0.0031806512270122766,
      "clip_ratio/high_mean": 0.001110293977035326,
      "clip_ratio/low_mean": 0.0012063731919624843,
      "clip_ratio/low_min": 1.970987068489194e-05,
      "clip_ratio/region_mean": 0.0023166672108345665,
      "epoch": 2.9798775153105863,
      "grad_norm": 0.2742725610733032,
      "learning_rate": 1e-06,
      "loss": -0.0776,
      "step": 1276
    },
    {
      "clip_ratio/high_max": 0.0028186667332192883,
      "clip_ratio/high_mean": 0.0011238940751354676,
      "clip_ratio/low_mean": 0.0006910904971846321,
      "clip_ratio/low_min": 3.5132095945300534e-05,
      "clip_ratio/region_mean": 0.0018149845491279848,
      "completions/clipped_ratio": 0.1495535714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4083.0,
      "completions/mean_length": 1191.9520263671875,
      "completions/mean_terminated_length": 681.2664184570312,
      "completions/min_length": 176.0,
      "completions/min_terminated_length": 176.0,
      "epoch": 2.9822105570137065,
      "grad_norm": 0.3575637936592102,
      "learning_rate": 1e-06,
      "loss": -0.0632,
      "num_tokens": 187871031.0,
      "reward": 0.6194196939468384,
      "reward_std": 0.17720773816108704,
      "rewards/verify_math_reward/mean": 0.6194196343421936,
      "rewards/verify_math_reward/std": 0.48580074310302734,
      "step": 1277
    },
    {
      "clip_ratio/high_max": 0.003162377870467026,
      "clip_ratio/high_mean": 0.0012078278268745635,
      "clip_ratio/low_mean": 0.0010228057308268035,
      "clip_ratio/low_min": 3.4270047763129696e-05,
      "clip_ratio/region_mean": 0.002230633508588653,
      "epoch": 2.984543598716827,
      "grad_norm": 0.2947334051132202,
      "learning_rate": 1e-06,
      "loss": -0.0634,
      "step": 1278
    },
    {
      "clip_ratio/high_max": 0.003265526691393461,
      "clip_ratio/high_mean": 0.0011708471392921638,
      "clip_ratio/low_mean": 0.0011104116874776082,
      "clip_ratio/low_min": 1.7566047972650267e-05,
      "clip_ratio/region_mean": 0.0022812587922089733,
      "epoch": 2.9868766404199474,
      "grad_norm": 0.24825675785541534,
      "learning_rate": 1e-06,
      "loss": -0.0636,
      "step": 1279
    },
    {
      "clip_ratio/high_max": 0.0033663997164694592,
      "clip_ratio/high_mean": 0.0013209114513301756,
      "clip_ratio/low_mean": 0.0014029893063707277,
      "clip_ratio/low_min": 5.3065035899635404e-05,
      "clip_ratio/region_mean": 0.002723900804994628,
      "epoch": 2.989209682123068,
      "grad_norm": 0.2688109576702118,
      "learning_rate": 1e-06,
      "loss": -0.0637,
      "step": 1280
    },
    {
      "clip_ratio/high_max": 0.0024496471014572307,
      "clip_ratio/high_mean": 0.0010025371338997502,
      "clip_ratio/low_mean": 0.0006610128039028496,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016635499450785574,
      "completions/clipped_ratio": 0.2020089285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3437.0,
      "completions/mean_length": 1351.2913818359375,
      "completions/mean_terminated_length": 656.4769287109375,
      "completions/min_length": 132.0,
      "completions/min_terminated_length": 132.0,
      "epoch": 2.9915427238261882,
      "grad_norm": 0.4109784662723541,
      "learning_rate": 1e-06,
      "loss": -0.0831,
      "num_tokens": 188444508.0,
      "reward": 0.5725446939468384,
      "reward_std": 0.16671767830848694,
      "rewards/verify_math_reward/mean": 0.5725446343421936,
      "rewards/verify_math_reward/std": 0.49498558044433594,
      "step": 1281
    },
    {
      "clip_ratio/high_max": 0.0033538812494953163,
      "clip_ratio/high_mean": 0.0012586799457494635,
      "clip_ratio/low_mean": 0.0009316196228610352,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002190299579524435,
      "epoch": 2.993875765529309,
      "grad_norm": 0.3125744163990021,
      "learning_rate": 1e-06,
      "loss": -0.0833,
      "step": 1282
    },
    {
      "clip_ratio/high_max": 0.0035722587344935164,
      "clip_ratio/high_mean": 0.0013289053640619386,
      "clip_ratio/low_mean": 0.0011280564067419618,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002456961723510176,
      "epoch": 2.9962088072324295,
      "grad_norm": 0.2602030038833618,
      "learning_rate": 1e-06,
      "loss": -0.0835,
      "step": 1283
    },
    {
      "clip_ratio/high_max": 0.0030436663510045037,
      "clip_ratio/high_mean": 0.0011226604074181523,
      "clip_ratio/low_mean": 0.0013129749859217554,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024356354406336322,
      "epoch": 2.9985418489355498,
      "grad_norm": 0.2609206438064575,
      "learning_rate": 1e-06,
      "loss": -0.0836,
      "step": 1284
    },
    {
      "clip_ratio/high_max": 0.0025205490383086726,
      "clip_ratio/high_mean": 0.0008495939018757781,
      "clip_ratio/low_mean": 0.000519059556609136,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013686535021406598,
      "completions/clipped_ratio": 0.2243303571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3577.0,
      "completions/mean_length": 1500.4029541015625,
      "completions/mean_terminated_length": 749.7338256835938,
      "completions/min_length": 135.0,
      "completions/min_terminated_length": 135.0,
      "epoch": 3.0023330417031207,
      "grad_norm": 0.32256942987442017,
      "learning_rate": 1e-06,
      "loss": -0.0761,
      "num_tokens": 189055861.0,
      "reward": 0.4720982313156128,
      "reward_std": 0.15582603216171265,
      "rewards/verify_math_reward/mean": 0.4720982015132904,
      "rewards/verify_math_reward/std": 0.49949970841407776,
      "step": 1285
    },
    {
      "clip_ratio/high_max": 0.002837576321326196,
      "clip_ratio/high_mean": 0.0010215287147730123,
      "clip_ratio/low_mean": 0.0007602082423545653,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001781736958946567,
      "epoch": 3.004666083406241,
      "grad_norm": 0.23747889697551727,
      "learning_rate": 1e-06,
      "loss": -0.0764,
      "step": 1286
    },
    {
      "clip_ratio/high_max": 0.0030404913741222117,
      "clip_ratio/high_mean": 0.0010446386586409062,
      "clip_ratio/low_mean": 0.0009466475057706703,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001991286117117852,
      "epoch": 3.0069991251093615,
      "grad_norm": 0.23243752121925354,
      "learning_rate": 1e-06,
      "loss": -0.0765,
      "step": 1287
    },
    {
      "clip_ratio/high_max": 0.0028408026046236046,
      "clip_ratio/high_mean": 0.0010033117487182608,
      "clip_ratio/low_mean": 0.001017110549582867,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002020422303758096,
      "epoch": 3.0093321668124817,
      "grad_norm": 0.2571777403354645,
      "learning_rate": 1e-06,
      "loss": -0.0766,
      "step": 1288
    },
    {
      "clip_ratio/high_max": 0.0020180494066153187,
      "clip_ratio/high_mean": 0.0007340943420786061,
      "clip_ratio/low_mean": 0.0005057745229350985,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012398688468238106,
      "completions/clipped_ratio": 0.1830357142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3731.0,
      "completions/mean_length": 1326.7890625,
      "completions/mean_terminated_length": 706.36474609375,
      "completions/min_length": 101.0,
      "completions/min_terminated_length": 101.0,
      "epoch": 3.0116652085156024,
      "grad_norm": 0.350341796875,
      "learning_rate": 1e-06,
      "loss": -0.0606,
      "num_tokens": 189655416.0,
      "reward": 0.5703125,
      "reward_std": 0.13527169823646545,
      "rewards/verify_math_reward/mean": 0.5703125,
      "rewards/verify_math_reward/std": 0.49530795216560364,
      "step": 1289
    },
    {
      "clip_ratio/high_max": 0.0026527890586294234,
      "clip_ratio/high_mean": 0.0009219521225531935,
      "clip_ratio/low_mean": 0.0006422939604817657,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015642460857634433,
      "epoch": 3.0139982502187226,
      "grad_norm": 0.23119445145130157,
      "learning_rate": 1e-06,
      "loss": -0.0608,
      "step": 1290
    },
    {
      "clip_ratio/high_max": 0.0021421724777610507,
      "clip_ratio/high_mean": 0.0008223191680372111,
      "clip_ratio/low_mean": 0.0007526751051045721,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015749942758702673,
      "epoch": 3.0163312919218432,
      "grad_norm": 0.21958255767822266,
      "learning_rate": 1e-06,
      "loss": -0.061,
      "step": 1291
    },
    {
      "clip_ratio/high_max": 0.002493561652954668,
      "clip_ratio/high_mean": 0.0009063939414772904,
      "clip_ratio/low_mean": 0.0010313017191947438,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019376956952328328,
      "epoch": 3.0186643336249634,
      "grad_norm": 0.20196911692619324,
      "learning_rate": 1e-06,
      "loss": -0.0611,
      "step": 1292
    },
    {
      "clip_ratio/high_max": 0.002537009750085417,
      "clip_ratio/high_mean": 0.000978603291514446,
      "clip_ratio/low_mean": 0.0005903242995373148,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015689275933254976,
      "completions/clipped_ratio": 0.2075892857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3190.0,
      "completions/mean_length": 1396.829345703125,
      "completions/mean_terminated_length": 689.7225341796875,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 3.020997375328084,
      "grad_norm": 0.375221848487854,
      "learning_rate": 1e-06,
      "loss": -0.087,
      "num_tokens": 190234783.0,
      "reward": 0.5256696939468384,
      "reward_std": 0.19271855056285858,
      "rewards/verify_math_reward/mean": 0.5256696343421936,
      "rewards/verify_math_reward/std": 0.4996195435523987,
      "step": 1293
    },
    {
      "clip_ratio/high_max": 0.0033749637150322087,
      "clip_ratio/high_mean": 0.001275835860724328,
      "clip_ratio/low_mean": 0.0007964365959196584,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002072272458462976,
      "epoch": 3.0233304170312043,
      "grad_norm": 0.26271942257881165,
      "learning_rate": 1e-06,
      "loss": -0.0873,
      "step": 1294
    },
    {
      "clip_ratio/high_max": 0.003290289307187777,
      "clip_ratio/high_mean": 0.0012362351626507007,
      "clip_ratio/low_mean": 0.0009485341597610386,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002184769276937004,
      "epoch": 3.025663458734325,
      "grad_norm": 0.2825155258178711,
      "learning_rate": 1e-06,
      "loss": -0.0875,
      "step": 1295
    },
    {
      "clip_ratio/high_max": 0.0031547732069157064,
      "clip_ratio/high_mean": 0.0012714094336843118,
      "clip_ratio/low_mean": 0.0011926287315873196,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024640381743665785,
      "epoch": 3.027996500437445,
      "grad_norm": 0.2432880997657776,
      "learning_rate": 1e-06,
      "loss": -0.0875,
      "step": 1296
    },
    {
      "clip_ratio/high_max": 0.0024913674205890857,
      "clip_ratio/high_mean": 0.0008098549060377991,
      "clip_ratio/low_mean": 0.0005295911487337435,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013394460802373942,
      "completions/clipped_ratio": 0.1863839285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4066.0,
      "completions/mean_length": 1315.8248291015625,
      "completions/mean_terminated_length": 678.9396362304688,
      "completions/min_length": 147.0,
      "completions/min_terminated_length": 147.0,
      "epoch": 3.030329542140566,
      "grad_norm": 0.2805899977684021,
      "learning_rate": 1e-06,
      "loss": -0.0622,
      "num_tokens": 190820322.0,
      "reward": 0.5625,
      "reward_std": 0.1513993740081787,
      "rewards/verify_math_reward/mean": 0.5625,
      "rewards/verify_math_reward/std": 0.49635544419288635,
      "step": 1297
    },
    {
      "clip_ratio/high_max": 0.002850104974640999,
      "clip_ratio/high_mean": 0.0009773864730959758,
      "clip_ratio/low_mean": 0.0007491319574910449,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017265184615098406,
      "epoch": 3.032662583843686,
      "grad_norm": 0.238377183675766,
      "learning_rate": 1e-06,
      "loss": -0.0624,
      "step": 1298
    },
    {
      "clip_ratio/high_max": 0.003161500957503449,
      "clip_ratio/high_mean": 0.0009961924697563518,
      "clip_ratio/low_mean": 0.0009126104705501348,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019088029439444654,
      "epoch": 3.0349956255468067,
      "grad_norm": 0.228261336684227,
      "learning_rate": 1e-06,
      "loss": -0.0626,
      "step": 1299
    },
    {
      "clip_ratio/high_max": 0.002946636486740317,
      "clip_ratio/high_mean": 0.0009655360481701791,
      "clip_ratio/low_mean": 0.0010455251995153958,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020110612640564796,
      "epoch": 3.037328667249927,
      "grad_norm": 0.22765207290649414,
      "learning_rate": 1e-06,
      "loss": -0.0626,
      "step": 1300
    },
    {
      "clip_ratio/high_max": 0.002572289333329536,
      "clip_ratio/high_mean": 0.0010175790775974747,
      "clip_ratio/low_mean": 0.0005612465474769124,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015788256023370195,
      "completions/clipped_ratio": 0.1629464285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3429.0,
      "completions/mean_length": 1212.0491943359375,
      "completions/mean_terminated_length": 650.6400146484375,
      "completions/min_length": 103.0,
      "completions/min_terminated_length": 103.0,
      "epoch": 3.0396617089530475,
      "grad_norm": 0.3184731900691986,
      "learning_rate": 1e-06,
      "loss": -0.0827,
      "num_tokens": 191388566.0,
      "reward": 0.609375,
      "reward_std": 0.1598842889070511,
      "rewards/verify_math_reward/mean": 0.609375,
      "rewards/verify_math_reward/std": 0.48816296458244324,
      "step": 1301
    },
    {
      "clip_ratio/high_max": 0.002737827966484474,
      "clip_ratio/high_mean": 0.001159976873168489,
      "clip_ratio/low_mean": 0.0006855983087916684,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018455751924193464,
      "epoch": 3.041994750656168,
      "grad_norm": 0.28312474489212036,
      "learning_rate": 1e-06,
      "loss": -0.083,
      "step": 1302
    },
    {
      "clip_ratio/high_max": 0.003003618709044531,
      "clip_ratio/high_mean": 0.001171711934148334,
      "clip_ratio/low_mean": 0.0009604678125469945,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002132179753971286,
      "epoch": 3.0443277923592884,
      "grad_norm": 0.24512238800525665,
      "learning_rate": 1e-06,
      "loss": -0.0831,
      "step": 1303
    },
    {
      "clip_ratio/high_max": 0.0028580743819475174,
      "clip_ratio/high_mean": 0.0011533901815710124,
      "clip_ratio/low_mean": 0.0010000799777571956,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002153470122721046,
      "epoch": 3.046660834062409,
      "grad_norm": 0.3053385615348816,
      "learning_rate": 1e-06,
      "loss": -0.0831,
      "step": 1304
    },
    {
      "clip_ratio/high_max": 0.0020921457835356705,
      "clip_ratio/high_mean": 0.0007462162448064191,
      "clip_ratio/low_mean": 0.0004180340970378893,
      "clip_ratio/low_min": 1.9913972209906206e-05,
      "clip_ratio/region_mean": 0.0011642503486655187,
      "completions/clipped_ratio": 0.1785714285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3833.0,
      "completions/mean_length": 1298.3717041015625,
      "completions/mean_terminated_length": 690.1915893554688,
      "completions/min_length": 188.0,
      "completions/min_terminated_length": 188.0,
      "epoch": 3.0489938757655293,
      "grad_norm": 0.2806278169155121,
      "learning_rate": 1e-06,
      "loss": -0.058,
      "num_tokens": 191988699.0,
      "reward": 0.5792410969734192,
      "reward_std": 0.1340659260749817,
      "rewards/verify_math_reward/mean": 0.5792410969734192,
      "rewards/verify_math_reward/std": 0.49395665526390076,
      "step": 1305
    },
    {
      "clip_ratio/high_max": 0.002359187856200151,
      "clip_ratio/high_mean": 0.0009155851730611175,
      "clip_ratio/low_mean": 0.0005765044097643113,
      "clip_ratio/low_min": 2.3773298380547203e-05,
      "clip_ratio/region_mean": 0.0014920895664545242,
      "epoch": 3.05132691746865,
      "grad_norm": 0.28606978058815,
      "learning_rate": 1e-06,
      "loss": -0.0582,
      "step": 1306
    },
    {
      "clip_ratio/high_max": 0.0025636116333771497,
      "clip_ratio/high_mean": 0.0008925399761210429,
      "clip_ratio/low_mean": 0.0007514707058362546,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001644010713789612,
      "epoch": 3.05365995917177,
      "grad_norm": 0.1945660561323166,
      "learning_rate": 1e-06,
      "loss": -0.0584,
      "step": 1307
    },
    {
      "clip_ratio/high_max": 0.0023899032203189563,
      "clip_ratio/high_mean": 0.0008637774208182236,
      "clip_ratio/low_mean": 0.0008972271461971104,
      "clip_ratio/low_min": 1.9913972209906206e-05,
      "clip_ratio/region_mean": 0.001761004503350705,
      "epoch": 3.055993000874891,
      "grad_norm": 0.20888541638851166,
      "learning_rate": 1e-06,
      "loss": -0.0585,
      "step": 1308
    },
    {
      "clip_ratio/high_max": 0.0028958607435924932,
      "clip_ratio/high_mean": 0.0011245208734180778,
      "clip_ratio/low_mean": 0.0005285132112931024,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016530340944882482,
      "completions/clipped_ratio": 0.2321428571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4034.0,
      "completions/mean_length": 1454.505615234375,
      "completions/mean_terminated_length": 655.9142456054688,
      "completions/min_length": 151.0,
      "completions/min_terminated_length": 151.0,
      "epoch": 3.058326042578011,
      "grad_norm": 0.30097147822380066,
      "learning_rate": 1e-06,
      "loss": -0.0859,
      "num_tokens": 192532720.0,
      "reward": 0.5558035969734192,
      "reward_std": 0.1601438969373703,
      "rewards/verify_math_reward/mean": 0.5558035969734192,
      "rewards/verify_math_reward/std": 0.49715372920036316,
      "step": 1309
    },
    {
      "clip_ratio/high_max": 0.003858191121253185,
      "clip_ratio/high_mean": 0.0013613364571938291,
      "clip_ratio/low_mean": 0.0007244455564432428,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020857820054516196,
      "epoch": 3.0606590842811316,
      "grad_norm": 0.29711002111434937,
      "learning_rate": 1e-06,
      "loss": -0.0862,
      "step": 1310
    },
    {
      "clip_ratio/high_max": 0.0034845851332647726,
      "clip_ratio/high_mean": 0.0013190685895096976,
      "clip_ratio/low_mean": 0.0008171836466317473,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021362522566050757,
      "epoch": 3.062992125984252,
      "grad_norm": 0.2635003328323364,
      "learning_rate": 1e-06,
      "loss": -0.0863,
      "step": 1311
    },
    {
      "clip_ratio/high_max": 0.0035168484901078045,
      "clip_ratio/high_mean": 0.001348177953332197,
      "clip_ratio/low_mean": 0.0010513546640140703,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002399532611889299,
      "epoch": 3.0653251676873725,
      "grad_norm": 0.24283172190189362,
      "learning_rate": 1e-06,
      "loss": -0.0864,
      "step": 1312
    },
    {
      "clip_ratio/high_max": 0.0025021589008247247,
      "clip_ratio/high_mean": 0.0010183860122197075,
      "clip_ratio/low_mean": 0.0004949986587234889,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001513384671852691,
      "completions/clipped_ratio": 0.1863839285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2602.0,
      "completions/mean_length": 1248.216552734375,
      "completions/mean_terminated_length": 595.8436279296875,
      "completions/min_length": 168.0,
      "completions/min_terminated_length": 168.0,
      "epoch": 3.0676582093904927,
      "grad_norm": 0.37111696600914,
      "learning_rate": 1e-06,
      "loss": -0.0602,
      "num_tokens": 193057898.0,
      "reward": 0.6037946939468384,
      "reward_std": 0.1543617695569992,
      "rewards/verify_math_reward/mean": 0.6037946343421936,
      "rewards/verify_math_reward/std": 0.48938122391700745,
      "step": 1313
    },
    {
      "clip_ratio/high_max": 0.0034964400474564172,
      "clip_ratio/high_mean": 0.001275072372663999,
      "clip_ratio/low_mean": 0.0007646513663530641,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020397236621647608,
      "epoch": 3.0699912510936134,
      "grad_norm": 0.2912351191043854,
      "learning_rate": 1e-06,
      "loss": -0.0605,
      "step": 1314
    },
    {
      "clip_ratio/high_max": 0.003168079365423182,
      "clip_ratio/high_mean": 0.0012253887607585057,
      "clip_ratio/low_mean": 0.0009435755418962799,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021689642599085346,
      "epoch": 3.0723242927967336,
      "grad_norm": 0.2485773265361786,
      "learning_rate": 1e-06,
      "loss": -0.0607,
      "step": 1315
    },
    {
      "clip_ratio/high_max": 0.0031388264724228065,
      "clip_ratio/high_mean": 0.0012513939327618573,
      "clip_ratio/low_mean": 0.0010505451255085063,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023019390864646994,
      "epoch": 3.0746573344998542,
      "grad_norm": 0.26045605540275574,
      "learning_rate": 1e-06,
      "loss": -0.0607,
      "step": 1316
    },
    {
      "clip_ratio/high_max": 0.0022871091787237674,
      "clip_ratio/high_mean": 0.000889468803507043,
      "clip_ratio/low_mean": 0.0005702042794837325,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014596730943594594,
      "completions/clipped_ratio": 0.1819196428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3486.0,
      "completions/mean_length": 1329.7109375,
      "completions/mean_terminated_length": 714.5607299804688,
      "completions/min_length": 175.0,
      "completions/min_terminated_length": 175.0,
      "epoch": 3.0769903762029744,
      "grad_norm": 0.35755103826522827,
      "learning_rate": 1e-06,
      "loss": -0.075,
      "num_tokens": 193677407.0,
      "reward": 0.5011160969734192,
      "reward_std": 0.15090125799179077,
      "rewards/verify_math_reward/mean": 0.5011160969734192,
      "rewards/verify_math_reward/std": 0.5002779960632324,
      "step": 1317
    },
    {
      "clip_ratio/high_max": 0.00302785503299674,
      "clip_ratio/high_mean": 0.0011055794529966079,
      "clip_ratio/low_mean": 0.0008301125717480318,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019356920092832297,
      "epoch": 3.079323417906095,
      "grad_norm": 0.23217631876468658,
      "learning_rate": 1e-06,
      "loss": -0.0752,
      "step": 1318
    },
    {
      "clip_ratio/high_max": 0.0029139655307517387,
      "clip_ratio/high_mean": 0.0010585647978587076,
      "clip_ratio/low_mean": 0.0008228557717302465,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018814205541275442,
      "epoch": 3.0816564596092153,
      "grad_norm": 0.24049969017505646,
      "learning_rate": 1e-06,
      "loss": -0.0753,
      "step": 1319
    },
    {
      "clip_ratio/high_max": 0.002689016386284493,
      "clip_ratio/high_mean": 0.0010127902824024204,
      "clip_ratio/low_mean": 0.0010295547544956207,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00204234504053602,
      "epoch": 3.083989501312336,
      "grad_norm": 0.22247229516506195,
      "learning_rate": 1e-06,
      "loss": -0.0754,
      "step": 1320
    },
    {
      "clip_ratio/high_max": 0.0022223807318368927,
      "clip_ratio/high_mean": 0.000926491069549229,
      "clip_ratio/low_mean": 0.0005344799919839716,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014609710924560204,
      "completions/clipped_ratio": 0.15625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3970.0,
      "completions/mean_length": 1175.2913818359375,
      "completions/mean_terminated_length": 634.4193115234375,
      "completions/min_length": 141.0,
      "completions/min_terminated_length": 141.0,
      "epoch": 3.0863225430154566,
      "grad_norm": 0.3505702614784241,
      "learning_rate": 1e-06,
      "loss": -0.0763,
      "num_tokens": 194243804.0,
      "reward": 0.637276828289032,
      "reward_std": 0.16040420532226562,
      "rewards/verify_math_reward/mean": 0.6372767686843872,
      "rewards/verify_math_reward/std": 0.481054425239563,
      "step": 1321
    },
    {
      "clip_ratio/high_max": 0.002552797508542426,
      "clip_ratio/high_mean": 0.00113710213554441,
      "clip_ratio/low_mean": 0.0007302283574972535,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00186733047303278,
      "epoch": 3.088655584718577,
      "grad_norm": 0.27257534861564636,
      "learning_rate": 1e-06,
      "loss": -0.0766,
      "step": 1322
    },
    {
      "clip_ratio/high_max": 0.0024929726532718632,
      "clip_ratio/high_mean": 0.0010018376469815848,
      "clip_ratio/low_mean": 0.0008190456974261906,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00182088326982921,
      "epoch": 3.0909886264216975,
      "grad_norm": 0.2139657884836197,
      "learning_rate": 1e-06,
      "loss": -0.0767,
      "step": 1323
    },
    {
      "clip_ratio/high_max": 0.0023928607697598636,
      "clip_ratio/high_mean": 0.001074929530659574,
      "clip_ratio/low_mean": 0.0009412316030648071,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020161611028015614,
      "epoch": 3.0933216681248177,
      "grad_norm": 0.21198685467243195,
      "learning_rate": 1e-06,
      "loss": -0.0768,
      "step": 1324
    },
    {
      "clip_ratio/high_max": 0.0020417028936208226,
      "clip_ratio/high_mean": 0.0007766804992570542,
      "clip_ratio/low_mean": 0.0005703795477529638,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013470600388245657,
      "completions/clipped_ratio": 0.1517857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3445.0,
      "completions/mean_length": 1186.0101318359375,
      "completions/mean_terminated_length": 665.2750244140625,
      "completions/min_length": 132.0,
      "completions/min_terminated_length": 132.0,
      "epoch": 3.0956547098279383,
      "grad_norm": 0.2741696238517761,
      "learning_rate": 1e-06,
      "loss": -0.058,
      "num_tokens": 194836629.0,
      "reward": 0.5915178656578064,
      "reward_std": 0.15567587316036224,
      "rewards/verify_math_reward/mean": 0.5915178656578064,
      "rewards/verify_math_reward/std": 0.49182769656181335,
      "step": 1325
    },
    {
      "clip_ratio/high_max": 0.002530857287638355,
      "clip_ratio/high_mean": 0.0010029010190919507,
      "clip_ratio/low_mean": 0.0007089033119882515,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017118043251684867,
      "epoch": 3.0979877515310585,
      "grad_norm": 0.24755199253559113,
      "learning_rate": 1e-06,
      "loss": -0.0582,
      "step": 1326
    },
    {
      "clip_ratio/high_max": 0.0025047990056918934,
      "clip_ratio/high_mean": 0.0009640709140512627,
      "clip_ratio/low_mean": 0.0007944907483761199,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001758561615133658,
      "epoch": 3.100320793234179,
      "grad_norm": 0.32930606603622437,
      "learning_rate": 1e-06,
      "loss": -0.0583,
      "step": 1327
    },
    {
      "clip_ratio/high_max": 0.0024294221111631487,
      "clip_ratio/high_mean": 0.0008861384922056459,
      "clip_ratio/low_mean": 0.0010210815198661294,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019072200229857117,
      "epoch": 3.1026538349372994,
      "grad_norm": 0.20770609378814697,
      "learning_rate": 1e-06,
      "loss": -0.0584,
      "step": 1328
    },
    {
      "clip_ratio/high_max": 0.0025223410448234063,
      "clip_ratio/high_mean": 0.0008677193072799128,
      "clip_ratio/low_mean": 0.0005741742888858425,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014418935825233348,
      "completions/clipped_ratio": 0.1573660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2366.0,
      "completions/mean_length": 1134.923095703125,
      "completions/mean_terminated_length": 581.9271850585938,
      "completions/min_length": 146.0,
      "completions/min_terminated_length": 146.0,
      "epoch": 3.10498687664042,
      "grad_norm": 0.3837740421295166,
      "learning_rate": 1e-06,
      "loss": -0.0606,
      "num_tokens": 195371312.0,
      "reward": 0.613839328289032,
      "reward_std": 0.1515427529811859,
      "rewards/verify_math_reward/mean": 0.6138392686843872,
      "rewards/verify_math_reward/std": 0.48714008927345276,
      "step": 1329
    },
    {
      "clip_ratio/high_max": 0.0028084589139325544,
      "clip_ratio/high_mean": 0.0010732359787652967,
      "clip_ratio/low_mean": 0.0008945150748331798,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001967751064512413,
      "epoch": 3.1073199183435403,
      "grad_norm": 0.3349197208881378,
      "learning_rate": 1e-06,
      "loss": -0.0609,
      "step": 1330
    },
    {
      "clip_ratio/high_max": 0.0028585824693436734,
      "clip_ratio/high_mean": 0.0010309245662938338,
      "clip_ratio/low_mean": 0.0009570273650751915,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001987951938644983,
      "epoch": 3.109652960046661,
      "grad_norm": 0.2798576354980469,
      "learning_rate": 1e-06,
      "loss": -0.0611,
      "step": 1331
    },
    {
      "clip_ratio/high_max": 0.0028027815860696137,
      "clip_ratio/high_mean": 0.0010115866825799458,
      "clip_ratio/low_mean": 0.0010797360355354613,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002091322712658439,
      "epoch": 3.111986001749781,
      "grad_norm": 0.2401420623064041,
      "learning_rate": 1e-06,
      "loss": -0.0612,
      "step": 1332
    },
    {
      "clip_ratio/high_max": 0.002997489478730131,
      "clip_ratio/high_mean": 0.0010483501500857528,
      "clip_ratio/low_mean": 0.0006875655117255519,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017359156845486723,
      "completions/clipped_ratio": 0.2075892857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4033.0,
      "completions/mean_length": 1388.8985595703125,
      "completions/mean_terminated_length": 679.7140502929688,
      "completions/min_length": 205.0,
      "completions/min_terminated_length": 205.0,
      "epoch": 3.114319043452902,
      "grad_norm": 0.2810341417789459,
      "learning_rate": 1e-06,
      "loss": -0.0808,
      "num_tokens": 195954741.0,
      "reward": 0.5390625,
      "reward_std": 0.148578941822052,
      "rewards/verify_math_reward/mean": 0.5390625,
      "rewards/verify_math_reward/std": 0.4987502098083496,
      "step": 1333
    },
    {
      "clip_ratio/high_max": 0.0032104247584356926,
      "clip_ratio/high_mean": 0.0011600665020523593,
      "clip_ratio/low_mean": 0.000777182036472368,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019372485403437167,
      "epoch": 3.116652085156022,
      "grad_norm": 0.26112329959869385,
      "learning_rate": 1e-06,
      "loss": -0.081,
      "step": 1334
    },
    {
      "clip_ratio/high_max": 0.0038014510937500745,
      "clip_ratio/high_mean": 0.0013046984822722152,
      "clip_ratio/low_mean": 0.0011017159140465083,
      "clip_ratio/low_min": 1.90200844372157e-05,
      "clip_ratio/region_mean": 0.0024064143872237764,
      "epoch": 3.1189851268591426,
      "grad_norm": 0.22285813093185425,
      "learning_rate": 1e-06,
      "loss": -0.0812,
      "step": 1335
    },
    {
      "clip_ratio/high_max": 0.0032983868804876693,
      "clip_ratio/high_mean": 0.0011769670236390084,
      "clip_ratio/low_mean": 0.0011258692848059582,
      "clip_ratio/low_min": 9.577076525602024e-06,
      "clip_ratio/region_mean": 0.002302836270246189,
      "epoch": 3.121318168562263,
      "grad_norm": 0.22714033722877502,
      "learning_rate": 1e-06,
      "loss": -0.0812,
      "step": 1336
    },
    {
      "clip_ratio/high_max": 0.002241811485873768,
      "clip_ratio/high_mean": 0.000784858597398852,
      "clip_ratio/low_mean": 0.0004849584456678713,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012698170139628928,
      "completions/clipped_ratio": 0.1930803571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3163.0,
      "completions/mean_length": 1253.9598388671875,
      "completions/mean_terminated_length": 573.9142456054688,
      "completions/min_length": 188.0,
      "completions/min_terminated_length": 188.0,
      "epoch": 3.1236512102653835,
      "grad_norm": 0.361855685710907,
      "learning_rate": 1e-06,
      "loss": -0.0635,
      "num_tokens": 196451225.0,
      "reward": 0.6305803656578064,
      "reward_std": 0.13080225884914398,
      "rewards/verify_math_reward/mean": 0.6305803656578064,
      "rewards/verify_math_reward/std": 0.4829172194004059,
      "step": 1337
    },
    {
      "clip_ratio/high_max": 0.002304529039975023,
      "clip_ratio/high_mean": 0.0008729561104701133,
      "clip_ratio/low_mean": 0.0006413904757209821,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015143465643632226,
      "epoch": 3.1259842519685037,
      "grad_norm": 0.26827290654182434,
      "learning_rate": 1e-06,
      "loss": -0.0636,
      "step": 1338
    },
    {
      "clip_ratio/high_max": 0.0027912609184568282,
      "clip_ratio/high_mean": 0.0009762080953805707,
      "clip_ratio/low_mean": 0.0008614428024884546,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018376508523942903,
      "epoch": 3.1283172936716244,
      "grad_norm": 0.21968616545200348,
      "learning_rate": 1e-06,
      "loss": -0.0637,
      "step": 1339
    },
    {
      "clip_ratio/high_max": 0.0025011100333358627,
      "clip_ratio/high_mean": 0.0008304880302603124,
      "clip_ratio/low_mean": 0.0009597744519851403,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001790262478607474,
      "epoch": 3.130650335374745,
      "grad_norm": 0.22235487401485443,
      "learning_rate": 1e-06,
      "loss": -0.0639,
      "step": 1340
    },
    {
      "clip_ratio/high_max": 0.002429226202366408,
      "clip_ratio/high_mean": 0.0009180686429317575,
      "clip_ratio/low_mean": 0.0007322976243813173,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016503662773175165,
      "completions/clipped_ratio": 0.1640625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3774.0,
      "completions/mean_length": 1257.1138916015625,
      "completions/mean_terminated_length": 699.9492797851562,
      "completions/min_length": 217.0,
      "completions/min_terminated_length": 217.0,
      "epoch": 3.1329833770778652,
      "grad_norm": 0.33593234419822693,
      "learning_rate": 1e-06,
      "loss": -0.0803,
      "num_tokens": 197065823.0,
      "reward": 0.5133928656578064,
      "reward_std": 0.1994110643863678,
      "rewards/verify_math_reward/mean": 0.5133928656578064,
      "rewards/verify_math_reward/std": 0.500099778175354,
      "step": 1341
    },
    {
      "clip_ratio/high_max": 0.0027345400958438404,
      "clip_ratio/high_mean": 0.001094697949156398,
      "clip_ratio/low_mean": 0.0009638746232667472,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002058572594251018,
      "epoch": 3.135316418780986,
      "grad_norm": 0.3292315602302551,
      "learning_rate": 1e-06,
      "loss": -0.0805,
      "step": 1342
    },
    {
      "clip_ratio/high_max": 0.003092270919296425,
      "clip_ratio/high_mean": 0.0011939085452468134,
      "clip_ratio/low_mean": 0.0012318274784774985,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002425736100121867,
      "epoch": 3.137649460484106,
      "grad_norm": 0.24277986586093903,
      "learning_rate": 1e-06,
      "loss": -0.0807,
      "step": 1343
    },
    {
      "clip_ratio/high_max": 0.003047489677555859,
      "clip_ratio/high_mean": 0.00112897347935359,
      "clip_ratio/low_mean": 0.0013810614182148129,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0025100349012063816,
      "epoch": 3.1399825021872267,
      "grad_norm": 0.22549711167812347,
      "learning_rate": 1e-06,
      "loss": -0.0808,
      "step": 1344
    },
    {
      "clip_ratio/high_max": 0.002825524723448325,
      "clip_ratio/high_mean": 0.0009551564926368883,
      "clip_ratio/low_mean": 0.0005220908387855161,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014772473223274574,
      "completions/clipped_ratio": 0.1618303571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2880.0,
      "completions/mean_length": 1180.544677734375,
      "completions/mean_terminated_length": 617.6404418945312,
      "completions/min_length": 151.0,
      "completions/min_terminated_length": 151.0,
      "epoch": 3.142315543890347,
      "grad_norm": 0.3498491644859314,
      "learning_rate": 1e-06,
      "loss": -0.0639,
      "num_tokens": 197616975.0,
      "reward": 0.6238839626312256,
      "reward_std": 0.13414262235164642,
      "rewards/verify_math_reward/mean": 0.6238839030265808,
      "rewards/verify_math_reward/std": 0.4846802353858948,
      "step": 1345
    },
    {
      "clip_ratio/high_max": 0.002978554868604988,
      "clip_ratio/high_mean": 0.0010518930212128907,
      "clip_ratio/low_mean": 0.0006826523404015461,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001734545348881511,
      "epoch": 3.1446485855934676,
      "grad_norm": 0.31992146372795105,
      "learning_rate": 1e-06,
      "loss": -0.0641,
      "step": 1346
    },
    {
      "clip_ratio/high_max": 0.0032743963893153705,
      "clip_ratio/high_mean": 0.0010800384643516736,
      "clip_ratio/low_mean": 0.0009262666280847043,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020063050906173885,
      "epoch": 3.146981627296588,
      "grad_norm": 0.28409144282341003,
      "learning_rate": 1e-06,
      "loss": -0.0643,
      "step": 1347
    },
    {
      "clip_ratio/high_max": 0.00318326481647091,
      "clip_ratio/high_mean": 0.0010985936642100569,
      "clip_ratio/low_mean": 0.0011267528734606458,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002225346521299798,
      "epoch": 3.1493146689997085,
      "grad_norm": 0.233980193734169,
      "learning_rate": 1e-06,
      "loss": -0.0644,
      "step": 1348
    },
    {
      "clip_ratio/high_max": 0.002139932978025172,
      "clip_ratio/high_mean": 0.0009107685109484009,
      "clip_ratio/low_mean": 0.000569330952657765,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014800994686083868,
      "completions/clipped_ratio": 0.2287946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3305.0,
      "completions/mean_length": 1480.96435546875,
      "completions/mean_terminated_length": 705.15771484375,
      "completions/min_length": 141.0,
      "completions/min_terminated_length": 141.0,
      "epoch": 3.1516477107028287,
      "grad_norm": 0.3200514018535614,
      "learning_rate": 1e-06,
      "loss": -0.0999,
      "num_tokens": 198218823.0,
      "reward": 0.4810267984867096,
      "reward_std": 0.15747570991516113,
      "rewards/verify_math_reward/mean": 0.4810267984867096,
      "rewards/verify_math_reward/std": 0.49991899728775024,
      "step": 1349
    },
    {
      "clip_ratio/high_max": 0.003589366293454077,
      "clip_ratio/high_mean": 0.0013790113334835041,
      "clip_ratio/low_mean": 0.0007359550145338289,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002114966333465418,
      "epoch": 3.1539807524059493,
      "grad_norm": 0.2880544364452362,
      "learning_rate": 1e-06,
      "loss": -0.1002,
      "step": 1350
    },
    {
      "clip_ratio/high_max": 0.0030352926878549624,
      "clip_ratio/high_mean": 0.0012039461980748456,
      "clip_ratio/low_mean": 0.0009654550049162935,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021694011993531603,
      "epoch": 3.1563137941090695,
      "grad_norm": 0.22166696190834045,
      "learning_rate": 1e-06,
      "loss": -0.1003,
      "step": 1351
    },
    {
      "clip_ratio/high_max": 0.0034358482298557647,
      "clip_ratio/high_mean": 0.0012475168659875635,
      "clip_ratio/low_mean": 0.0010006393131334335,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022481561463791877,
      "epoch": 3.15864683581219,
      "grad_norm": 0.3134857714176178,
      "learning_rate": 1e-06,
      "loss": -0.1004,
      "step": 1352
    },
    {
      "clip_ratio/high_max": 0.0018772271469060797,
      "clip_ratio/high_mean": 0.0006525577919092029,
      "clip_ratio/low_mean": 0.0005370141607272672,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001189571961731417,
      "completions/clipped_ratio": 0.1540178571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3762.0,
      "completions/mean_length": 1193.2578125,
      "completions/mean_terminated_length": 664.790283203125,
      "completions/min_length": 182.0,
      "completions/min_terminated_length": 182.0,
      "epoch": 3.1609798775153104,
      "grad_norm": 0.2849535048007965,
      "learning_rate": 1e-06,
      "loss": -0.0671,
      "num_tokens": 198815262.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.14327509701251984,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514806270599365,
      "step": 1353
    },
    {
      "clip_ratio/high_max": 0.0023393456867779605,
      "clip_ratio/high_mean": 0.0008391679275518982,
      "clip_ratio/low_mean": 0.0008357585793419275,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001674926512350794,
      "epoch": 3.163312919218431,
      "grad_norm": 0.20166900753974915,
      "learning_rate": 1e-06,
      "loss": -0.0674,
      "step": 1354
    },
    {
      "clip_ratio/high_max": 0.002438125266053248,
      "clip_ratio/high_mean": 0.0008892816931620473,
      "clip_ratio/low_mean": 0.0008286072243208764,
      "clip_ratio/low_min": 7.640586773050018e-06,
      "clip_ratio/region_mean": 0.0017178889029310085,
      "epoch": 3.1656459609215517,
      "grad_norm": 0.22612257301807404,
      "learning_rate": 1e-06,
      "loss": -0.0674,
      "step": 1355
    },
    {
      "clip_ratio/high_max": 0.0021777985348307993,
      "clip_ratio/high_mean": 0.0007701308604737278,
      "clip_ratio/low_mean": 0.001180693635433272,
      "clip_ratio/low_min": 2.2921760319150053e-05,
      "clip_ratio/region_mean": 0.0019508245386532508,
      "epoch": 3.167979002624672,
      "grad_norm": 0.19750025868415833,
      "learning_rate": 1e-06,
      "loss": -0.0676,
      "step": 1356
    },
    {
      "clip_ratio/high_max": 0.0021138735828571953,
      "clip_ratio/high_mean": 0.0008293652481370373,
      "clip_ratio/low_mean": 0.0005782775979241705,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014076428415137343,
      "completions/clipped_ratio": 0.1852678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3315.0,
      "completions/mean_length": 1272.3973388671875,
      "completions/mean_terminated_length": 630.3178100585938,
      "completions/min_length": 175.0,
      "completions/min_terminated_length": 175.0,
      "epoch": 3.1703120443277926,
      "grad_norm": 0.3227970600128174,
      "learning_rate": 1e-06,
      "loss": -0.0737,
      "num_tokens": 199360290.0,
      "reward": 0.5736607313156128,
      "reward_std": 0.148612841963768,
      "rewards/verify_math_reward/mean": 0.5736607313156128,
      "rewards/verify_math_reward/std": 0.4948205351829529,
      "step": 1357
    },
    {
      "clip_ratio/high_max": 0.0030250952040660195,
      "clip_ratio/high_mean": 0.0011142691564600682,
      "clip_ratio/low_mean": 0.0007781955441714672,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018924646901723463,
      "epoch": 3.1726450860309128,
      "grad_norm": 0.246768981218338,
      "learning_rate": 1e-06,
      "loss": -0.0739,
      "step": 1358
    },
    {
      "clip_ratio/high_max": 0.0028133710147812963,
      "clip_ratio/high_mean": 0.001045112068823073,
      "clip_ratio/low_mean": 0.000986219386504672,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002031331448961282,
      "epoch": 3.1749781277340334,
      "grad_norm": 0.2945663034915924,
      "learning_rate": 1e-06,
      "loss": -0.0742,
      "step": 1359
    },
    {
      "clip_ratio/high_max": 0.0024650287450640462,
      "clip_ratio/high_mean": 0.0009208102055708878,
      "clip_ratio/low_mean": 0.0010797965260280762,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00200060671340907,
      "epoch": 3.1773111694371536,
      "grad_norm": 0.22011803090572357,
      "learning_rate": 1e-06,
      "loss": -0.0742,
      "step": 1360
    },
    {
      "clip_ratio/high_max": 0.0028081150812795386,
      "clip_ratio/high_mean": 0.0010802435426739976,
      "clip_ratio/low_mean": 0.0005825233292853227,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016627668555884156,
      "completions/clipped_ratio": 0.1908482142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3346.0,
      "completions/mean_length": 1339.078125,
      "completions/mean_terminated_length": 688.8248291015625,
      "completions/min_length": 170.0,
      "completions/min_terminated_length": 170.0,
      "epoch": 3.1796442111402743,
      "grad_norm": 0.3312619924545288,
      "learning_rate": 1e-06,
      "loss": -0.1132,
      "num_tokens": 199942360.0,
      "reward": 0.520089328289032,
      "reward_std": 0.1837480217218399,
      "rewards/verify_math_reward/mean": 0.5200892686843872,
      "rewards/verify_math_reward/std": 0.4998753070831299,
      "step": 1361
    },
    {
      "clip_ratio/high_max": 0.0037408599819173105,
      "clip_ratio/high_mean": 0.001404201702825958,
      "clip_ratio/low_mean": 0.0008379365717701148,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022421382745960727,
      "epoch": 3.1819772528433945,
      "grad_norm": 0.29121676087379456,
      "learning_rate": 1e-06,
      "loss": -0.1135,
      "step": 1362
    },
    {
      "clip_ratio/high_max": 0.0035552595218177885,
      "clip_ratio/high_mean": 0.001331019233475672,
      "clip_ratio/low_mean": 0.0009413602620043093,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002272379490023013,
      "epoch": 3.184310294546515,
      "grad_norm": 0.24848061800003052,
      "learning_rate": 1e-06,
      "loss": -0.1137,
      "step": 1363
    },
    {
      "clip_ratio/high_max": 0.003499184087559115,
      "clip_ratio/high_mean": 0.0012648745578189846,
      "clip_ratio/low_mean": 0.001018966988340253,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022838415825390257,
      "epoch": 3.1866433362496354,
      "grad_norm": 0.2734673321247101,
      "learning_rate": 1e-06,
      "loss": -0.1138,
      "step": 1364
    },
    {
      "clip_ratio/high_max": 0.0033895653759827837,
      "clip_ratio/high_mean": 0.001157745202363003,
      "clip_ratio/low_mean": 0.0006268017434649664,
      "clip_ratio/low_min": 9.797773600439541e-06,
      "clip_ratio/region_mean": 0.001784546908311313,
      "completions/clipped_ratio": 0.2142857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3118.0,
      "completions/mean_length": 1455.985595703125,
      "completions/mean_terminated_length": 735.9815673828125,
      "completions/min_length": 164.0,
      "completions/min_terminated_length": 164.0,
      "epoch": 3.188976377952756,
      "grad_norm": 0.3164910078048706,
      "learning_rate": 1e-06,
      "loss": -0.1014,
      "num_tokens": 200553963.0,
      "reward": 0.504464328289032,
      "reward_std": 0.18614476919174194,
      "rewards/verify_math_reward/mean": 0.5044642686843872,
      "rewards/verify_math_reward/std": 0.5002593398094177,
      "step": 1365
    },
    {
      "clip_ratio/high_max": 0.0032233367674052715,
      "clip_ratio/high_mean": 0.0011809509014710784,
      "clip_ratio/low_mean": 0.0009057749102794332,
      "clip_ratio/low_min": 1.9595547200879082e-05,
      "clip_ratio/region_mean": 0.00208672575536184,
      "epoch": 3.1913094196558762,
      "grad_norm": 0.28135693073272705,
      "learning_rate": 1e-06,
      "loss": -0.1016,
      "step": 1366
    },
    {
      "clip_ratio/high_max": 0.0033793382317526266,
      "clip_ratio/high_mean": 0.0012781320911017247,
      "clip_ratio/low_mean": 0.0011014879482900142,
      "clip_ratio/low_min": 3.9191094401758164e-05,
      "clip_ratio/region_mean": 0.002379620047577191,
      "epoch": 3.193642461358997,
      "grad_norm": 0.23779508471488953,
      "learning_rate": 1e-06,
      "loss": -0.1018,
      "step": 1367
    },
    {
      "clip_ratio/high_max": 0.0032954135313048027,
      "clip_ratio/high_mean": 0.0012228189552843105,
      "clip_ratio/low_mean": 0.00129534852476354,
      "clip_ratio/low_min": 1.9595547200879082e-05,
      "clip_ratio/region_mean": 0.002518167566449847,
      "epoch": 3.195975503062117,
      "grad_norm": 0.24185238778591156,
      "learning_rate": 1e-06,
      "loss": -0.1019,
      "step": 1368
    },
    {
      "clip_ratio/high_max": 0.0022182962638908066,
      "clip_ratio/high_mean": 0.0008027746625884902,
      "clip_ratio/low_mean": 0.0006631616515733185,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001465936285967473,
      "completions/clipped_ratio": 0.2008928571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3785.0,
      "completions/mean_length": 1357.7545166015625,
      "completions/mean_terminated_length": 669.3687133789062,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 3.1983085447652377,
      "grad_norm": 0.30065393447875977,
      "learning_rate": 1e-06,
      "loss": -0.0765,
      "num_tokens": 201119335.0,
      "reward": 0.5546875,
      "reward_std": 0.14902400970458984,
      "rewards/verify_math_reward/mean": 0.5546875,
      "rewards/verify_math_reward/std": 0.4972778558731079,
      "step": 1369
    },
    {
      "clip_ratio/high_max": 0.002445177990011871,
      "clip_ratio/high_mean": 0.0009415408603672404,
      "clip_ratio/low_mean": 0.0008758692583796801,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00181741012056591,
      "epoch": 3.200641586468358,
      "grad_norm": 0.32187435030937195,
      "learning_rate": 1e-06,
      "loss": -0.0765,
      "step": 1370
    },
    {
      "clip_ratio/high_max": 0.0026672029853216372,
      "clip_ratio/high_mean": 0.0010266938543281867,
      "clip_ratio/low_mean": 0.001034640639772988,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020613344968296587,
      "epoch": 3.2029746281714786,
      "grad_norm": 0.26315632462501526,
      "learning_rate": 1e-06,
      "loss": -0.0768,
      "step": 1371
    },
    {
      "clip_ratio/high_max": 0.0024598382879048586,
      "clip_ratio/high_mean": 0.001008868899589288,
      "clip_ratio/low_mean": 0.0014022009199834429,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024110698213917203,
      "epoch": 3.205307669874599,
      "grad_norm": 0.23852814733982086,
      "learning_rate": 1e-06,
      "loss": -0.0769,
      "step": 1372
    },
    {
      "clip_ratio/high_max": 0.0026599104603519663,
      "clip_ratio/high_mean": 0.001088276407244848,
      "clip_ratio/low_mean": 0.00048437131954415236,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015726477358839475,
      "completions/clipped_ratio": 0.2120535714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3145.0,
      "completions/mean_length": 1406.2344970703125,
      "completions/mean_terminated_length": 682.3597412109375,
      "completions/min_length": 173.0,
      "completions/min_terminated_length": 173.0,
      "epoch": 3.2076407115777195,
      "grad_norm": 0.3115938603878021,
      "learning_rate": 1e-06,
      "loss": -0.0721,
      "num_tokens": 201689641.0,
      "reward": 0.5022321939468384,
      "reward_std": 0.16660960018634796,
      "rewards/verify_math_reward/mean": 0.5022321343421936,
      "rewards/verify_math_reward/std": 0.5002743005752563,
      "step": 1373
    },
    {
      "clip_ratio/high_max": 0.002917425306804944,
      "clip_ratio/high_mean": 0.0012557908485177904,
      "clip_ratio/low_mean": 0.0006526048437081045,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019083957158727571,
      "epoch": 3.20997375328084,
      "grad_norm": 0.33406734466552734,
      "learning_rate": 1e-06,
      "loss": -0.0724,
      "step": 1374
    },
    {
      "clip_ratio/high_max": 0.00291656961053377,
      "clip_ratio/high_mean": 0.001195302145788446,
      "clip_ratio/low_mean": 0.0007796578788656916,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019749599960050546,
      "epoch": 3.2123067949839603,
      "grad_norm": 0.26957711577415466,
      "learning_rate": 1e-06,
      "loss": -0.0725,
      "step": 1375
    },
    {
      "clip_ratio/high_max": 0.0030238851468311623,
      "clip_ratio/high_mean": 0.001208049950946588,
      "clip_ratio/low_mean": 0.0008903388452381478,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002098388780723326,
      "epoch": 3.214639836687081,
      "grad_norm": 0.25391343235969543,
      "learning_rate": 1e-06,
      "loss": -0.0725,
      "step": 1376
    },
    {
      "clip_ratio/high_max": 0.0023889686999609694,
      "clip_ratio/high_mean": 0.000987817853456363,
      "clip_ratio/low_mean": 0.0006864479155410663,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016742657790018711,
      "completions/clipped_ratio": 0.2209821428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3866.0,
      "completions/mean_length": 1485.3751220703125,
      "completions/mean_terminated_length": 744.8252563476562,
      "completions/min_length": 206.0,
      "completions/min_terminated_length": 206.0,
      "epoch": 3.216972878390201,
      "grad_norm": 0.2967849373817444,
      "learning_rate": 1e-06,
      "loss": -0.098,
      "num_tokens": 202304313.0,
      "reward": 0.4765625298023224,
      "reward_std": 0.1740477830171585,
      "rewards/verify_math_reward/mean": 0.4765625,
      "rewards/verify_math_reward/std": 0.49972933530807495,
      "step": 1377
    },
    {
      "clip_ratio/high_max": 0.0024445352246402763,
      "clip_ratio/high_mean": 0.0010376444915891625,
      "clip_ratio/low_mean": 0.0009159694309346378,
      "clip_ratio/low_min": 1.9647908629849553e-05,
      "clip_ratio/region_mean": 0.001953613871592097,
      "epoch": 3.219305920093322,
      "grad_norm": 0.30563458800315857,
      "learning_rate": 1e-06,
      "loss": -0.0981,
      "step": 1378
    },
    {
      "clip_ratio/high_max": 0.0023477618451579474,
      "clip_ratio/high_mean": 0.0010650950243871193,
      "clip_ratio/low_mean": 0.0010782867429952603,
      "clip_ratio/low_min": 1.60503332153894e-05,
      "clip_ratio/region_mean": 0.0021433818037621677,
      "epoch": 3.221638961796442,
      "grad_norm": 0.263799786567688,
      "learning_rate": 1e-06,
      "loss": -0.0984,
      "step": 1379
    },
    {
      "clip_ratio/high_max": 0.002721808406931814,
      "clip_ratio/high_mean": 0.0011329717963235453,
      "clip_ratio/low_mean": 0.0011568615827854956,
      "clip_ratio/low_min": 5.0730519433273e-05,
      "clip_ratio/region_mean": 0.0022898333336343057,
      "epoch": 3.2239720034995627,
      "grad_norm": 0.23858009278774261,
      "learning_rate": 1e-06,
      "loss": -0.0985,
      "step": 1380
    },
    {
      "clip_ratio/high_max": 0.0021201352865318768,
      "clip_ratio/high_mean": 0.000750652090573567,
      "clip_ratio/low_mean": 0.0005253635126791778,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001276015624171123,
      "completions/clipped_ratio": 0.1696428571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3517.0,
      "completions/mean_length": 1212.1685791015625,
      "completions/mean_terminated_length": 622.9986572265625,
      "completions/min_length": 181.0,
      "completions/min_terminated_length": 181.0,
      "epoch": 3.226305045202683,
      "grad_norm": 0.316988080739975,
      "learning_rate": 1e-06,
      "loss": -0.0555,
      "num_tokens": 202851776.0,
      "reward": 0.621651828289032,
      "reward_std": 0.1516508162021637,
      "rewards/verify_math_reward/mean": 0.6216517686843872,
      "rewards/verify_math_reward/std": 0.4852459728717804,
      "step": 1381
    },
    {
      "clip_ratio/high_max": 0.0027541725648916326,
      "clip_ratio/high_mean": 0.0009396336899953894,
      "clip_ratio/low_mean": 0.0006619510313612409,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016015847795642912,
      "epoch": 3.2286380869058036,
      "grad_norm": 0.25705021619796753,
      "learning_rate": 1e-06,
      "loss": -0.0557,
      "step": 1382
    },
    {
      "clip_ratio/high_max": 0.002651364127814304,
      "clip_ratio/high_mean": 0.000985535503787105,
      "clip_ratio/low_mean": 0.0008871033442119369,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018726388880168088,
      "epoch": 3.2309711286089238,
      "grad_norm": 0.31974393129348755,
      "learning_rate": 1e-06,
      "loss": -0.0558,
      "step": 1383
    },
    {
      "clip_ratio/high_max": 0.0023458923678845167,
      "clip_ratio/high_mean": 0.0008305495575768873,
      "clip_ratio/low_mean": 0.0010812032996909693,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019117528572678566,
      "epoch": 3.2333041703120444,
      "grad_norm": 0.2726139724254608,
      "learning_rate": 1e-06,
      "loss": -0.0558,
      "step": 1384
    },
    {
      "clip_ratio/high_max": 0.0024199914041673765,
      "clip_ratio/high_mean": 0.0009865360079857055,
      "clip_ratio/low_mean": 0.0006013246538714156,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015878607009653933,
      "completions/clipped_ratio": 0.1875,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3653.0,
      "completions/mean_length": 1295.06591796875,
      "completions/mean_terminated_length": 648.6964721679688,
      "completions/min_length": 156.0,
      "completions/min_terminated_length": 156.0,
      "epoch": 3.2356372120151646,
      "grad_norm": 0.31969305872917175,
      "learning_rate": 1e-06,
      "loss": -0.0838,
      "num_tokens": 203415259.0,
      "reward": 0.5569196939468384,
      "reward_std": 0.15097934007644653,
      "rewards/verify_math_reward/mean": 0.5569196343421936,
      "rewards/verify_math_reward/std": 0.49702703952789307,
      "step": 1385
    },
    {
      "clip_ratio/high_max": 0.0032785937219159678,
      "clip_ratio/high_mean": 0.0013131868727214169,
      "clip_ratio/low_mean": 0.000833667117603909,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021468539998750202,
      "epoch": 3.2379702537182853,
      "grad_norm": 0.25600868463516235,
      "learning_rate": 1e-06,
      "loss": -0.084,
      "step": 1386
    },
    {
      "clip_ratio/high_max": 0.002963113114674343,
      "clip_ratio/high_mean": 0.0012603523737197975,
      "clip_ratio/low_mean": 0.0008865321888151811,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002146884515241254,
      "epoch": 3.2403032954214055,
      "grad_norm": 0.22946487367153168,
      "learning_rate": 1e-06,
      "loss": -0.0842,
      "step": 1387
    },
    {
      "clip_ratio/high_max": 0.002728587744059041,
      "clip_ratio/high_mean": 0.001143316458183108,
      "clip_ratio/low_mean": 0.0010875751013372792,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002230891514045652,
      "epoch": 3.242636337124526,
      "grad_norm": 0.20421616733074188,
      "learning_rate": 1e-06,
      "loss": -0.0842,
      "step": 1388
    },
    {
      "clip_ratio/high_max": 0.0031228039442794397,
      "clip_ratio/high_mean": 0.0011625596416706685,
      "clip_ratio/low_mean": 0.000449032223514223,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016115919133881107,
      "completions/clipped_ratio": 0.1361607142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3356.0,
      "completions/mean_length": 1118.03466796875,
      "completions/mean_terminated_length": 648.6395263671875,
      "completions/min_length": 165.0,
      "completions/min_terminated_length": 165.0,
      "epoch": 3.2449693788276464,
      "grad_norm": 0.3249204456806183,
      "learning_rate": 1e-06,
      "loss": -0.0685,
      "num_tokens": 204008506.0,
      "reward": 0.5390625,
      "reward_std": 0.18876947462558746,
      "rewards/verify_math_reward/mean": 0.5390625,
      "rewards/verify_math_reward/std": 0.4987502098083496,
      "step": 1389
    },
    {
      "clip_ratio/high_max": 0.0030883293802617118,
      "clip_ratio/high_mean": 0.0012925277842441574,
      "clip_ratio/low_mean": 0.0007287934840860544,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002021321233769413,
      "epoch": 3.247302420530767,
      "grad_norm": 0.27527910470962524,
      "learning_rate": 1e-06,
      "loss": -0.0688,
      "step": 1390
    },
    {
      "clip_ratio/high_max": 0.0035688294083229266,
      "clip_ratio/high_mean": 0.0013115837173245382,
      "clip_ratio/low_mean": 0.000894385588253499,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002205969278293196,
      "epoch": 3.249635462233887,
      "grad_norm": 0.22996072471141815,
      "learning_rate": 1e-06,
      "loss": -0.0689,
      "step": 1391
    },
    {
      "clip_ratio/high_max": 0.0032063006656244397,
      "clip_ratio/high_mean": 0.0012513813744590152,
      "clip_ratio/low_mean": 0.0010004925225075567,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022518738987855613,
      "epoch": 3.251968503937008,
      "grad_norm": 0.2599335014820099,
      "learning_rate": 1e-06,
      "loss": -0.069,
      "step": 1392
    },
    {
      "clip_ratio/high_max": 0.0017107321291405242,
      "clip_ratio/high_mean": 0.0005945113161942572,
      "clip_ratio/low_mean": 0.0004417756699695019,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0010362869943492115,
      "completions/clipped_ratio": 0.2522321428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3981.0,
      "completions/mean_length": 1552.5357666015625,
      "completions/mean_terminated_length": 694.591064453125,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 3.2543015456401285,
      "grad_norm": 0.2847426235675812,
      "learning_rate": 1e-06,
      "loss": -0.0712,
      "num_tokens": 204568730.0,
      "reward": 0.4676339626312256,
      "reward_std": 0.12531113624572754,
      "rewards/verify_math_reward/mean": 0.4676339328289032,
      "rewards/verify_math_reward/std": 0.4992299973964691,
      "step": 1393
    },
    {
      "clip_ratio/high_max": 0.0022250008332775906,
      "clip_ratio/high_mean": 0.0007966230305100908,
      "clip_ratio/low_mean": 0.0006261913163143618,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014228143299988005,
      "epoch": 3.2566345873432487,
      "grad_norm": 0.25943008065223694,
      "learning_rate": 1e-06,
      "loss": -0.0715,
      "step": 1394
    },
    {
      "clip_ratio/high_max": 0.0021694751958420966,
      "clip_ratio/high_mean": 0.0007335646751016611,
      "clip_ratio/low_mean": 0.0007502906291847466,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014838552997389343,
      "epoch": 3.2589676290463694,
      "grad_norm": 0.22387363016605377,
      "learning_rate": 1e-06,
      "loss": -0.0715,
      "step": 1395
    },
    {
      "clip_ratio/high_max": 0.00225412361578492,
      "clip_ratio/high_mean": 0.0006936196227798064,
      "clip_ratio/low_mean": 0.0008708087239028828,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015644283539586468,
      "epoch": 3.2613006707494896,
      "grad_norm": 0.22405408322811127,
      "learning_rate": 1e-06,
      "loss": -0.0716,
      "step": 1396
    },
    {
      "clip_ratio/high_max": 0.002606538255349733,
      "clip_ratio/high_mean": 0.0008888039737939835,
      "clip_ratio/low_mean": 0.0007069700986903626,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015957740324665792,
      "completions/clipped_ratio": 0.1618303571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3575.0,
      "completions/mean_length": 1223.63623046875,
      "completions/mean_terminated_length": 669.0518798828125,
      "completions/min_length": 152.0,
      "completions/min_terminated_length": 152.0,
      "epoch": 3.2636337124526102,
      "grad_norm": 0.3812796473503113,
      "learning_rate": 1e-06,
      "loss": -0.0572,
      "num_tokens": 205163652.0,
      "reward": 0.515625,
      "reward_std": 0.16153399646282196,
      "rewards/verify_math_reward/mean": 0.515625,
      "rewards/verify_math_reward/std": 0.5000349283218384,
      "step": 1397
    },
    {
      "clip_ratio/high_max": 0.0026634673340595327,
      "clip_ratio/high_mean": 0.001081415710359579,
      "clip_ratio/low_mean": 0.0009601121455489192,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002041527899564244,
      "epoch": 3.2659667541557305,
      "grad_norm": 0.31780505180358887,
      "learning_rate": 1e-06,
      "loss": -0.0574,
      "step": 1398
    },
    {
      "clip_ratio/high_max": 0.0027923797897528857,
      "clip_ratio/high_mean": 0.0010067439689009916,
      "clip_ratio/low_mean": 0.001261196823179489,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002267940857564099,
      "epoch": 3.268299795858851,
      "grad_norm": 0.28611859679222107,
      "learning_rate": 1e-06,
      "loss": -0.0576,
      "step": 1399
    },
    {
      "clip_ratio/high_max": 0.002342934902117122,
      "clip_ratio/high_mean": 0.0009197864255838795,
      "clip_ratio/low_mean": 0.0013291147406562231,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002248901131679304,
      "epoch": 3.2706328375619713,
      "grad_norm": 0.232257679104805,
      "learning_rate": 1e-06,
      "loss": -0.0577,
      "step": 1400
    },
    {
      "clip_ratio/high_max": 0.0023585003764310386,
      "clip_ratio/high_mean": 0.000794408446381567,
      "clip_ratio/low_mean": 0.0005302667259456939,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013246751987026073,
      "completions/clipped_ratio": 0.1852678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2986.0,
      "completions/mean_length": 1287.828125,
      "completions/mean_terminated_length": 649.257568359375,
      "completions/min_length": 153.0,
      "completions/min_terminated_length": 153.0,
      "epoch": 3.272965879265092,
      "grad_norm": 0.3569932281970978,
      "learning_rate": 1e-06,
      "loss": -0.0867,
      "num_tokens": 205724266.0,
      "reward": 0.4988839626312256,
      "reward_std": 0.13985875248908997,
      "rewards/verify_math_reward/mean": 0.4988839328289032,
      "rewards/verify_math_reward/std": 0.5002779960632324,
      "step": 1401
    },
    {
      "clip_ratio/high_max": 0.003075787186389789,
      "clip_ratio/high_mean": 0.0009645216541684931,
      "clip_ratio/low_mean": 0.0008519003768014954,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018164220455219038,
      "epoch": 3.275298920968212,
      "grad_norm": 0.29467490315437317,
      "learning_rate": 1e-06,
      "loss": -0.087,
      "step": 1402
    },
    {
      "clip_ratio/high_max": 0.003196276338712778,
      "clip_ratio/high_mean": 0.0009206524755427381,
      "clip_ratio/low_mean": 0.0010368306466261856,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019574831567297224,
      "epoch": 3.277631962671333,
      "grad_norm": 0.2431877851486206,
      "learning_rate": 1e-06,
      "loss": -0.0872,
      "step": 1403
    },
    {
      "clip_ratio/high_max": 0.0028092034262954257,
      "clip_ratio/high_mean": 0.0009348743496957468,
      "clip_ratio/low_mean": 0.0012063762533216504,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021412506757769734,
      "epoch": 3.279965004374453,
      "grad_norm": 0.22555813193321228,
      "learning_rate": 1e-06,
      "loss": -0.0873,
      "step": 1404
    },
    {
      "clip_ratio/high_max": 0.002152534943888895,
      "clip_ratio/high_mean": 0.0008175220791599713,
      "clip_ratio/low_mean": 0.0006683234432784957,
      "clip_ratio/low_min": 7.524680768256076e-06,
      "clip_ratio/region_mean": 0.0014858455178909935,
      "completions/clipped_ratio": 0.2198660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4090.0,
      "completions/mean_length": 1495.5726318359375,
      "completions/mean_terminated_length": 762.6909790039062,
      "completions/min_length": 133.0,
      "completions/min_terminated_length": 133.0,
      "epoch": 3.2822980460775737,
      "grad_norm": 0.29808303713798523,
      "learning_rate": 1e-06,
      "loss": -0.1224,
      "num_tokens": 206352035.0,
      "reward": 0.4966517984867096,
      "reward_std": 0.18396486341953278,
      "rewards/verify_math_reward/mean": 0.4966517984867096,
      "rewards/verify_math_reward/std": 0.5002680420875549,
      "step": 1405
    },
    {
      "clip_ratio/high_max": 0.002606508096505422,
      "clip_ratio/high_mean": 0.0010587483557173982,
      "clip_ratio/low_mean": 0.0008699759018782061,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019287242466816679,
      "epoch": 3.284631087780694,
      "grad_norm": 0.2458135187625885,
      "learning_rate": 1e-06,
      "loss": -0.1226,
      "step": 1406
    },
    {
      "clip_ratio/high_max": 0.00274065002304269,
      "clip_ratio/high_mean": 0.0010915798920905218,
      "clip_ratio/low_mean": 0.0010510491247259779,
      "clip_ratio/low_min": 1.3536929145629983e-05,
      "clip_ratio/region_mean": 0.002142629018635489,
      "epoch": 3.2869641294838146,
      "grad_norm": 0.2649960219860077,
      "learning_rate": 1e-06,
      "loss": -0.1227,
      "step": 1407
    },
    {
      "clip_ratio/high_max": 0.00254956914432114,
      "clip_ratio/high_mean": 0.0010358562685723882,
      "clip_ratio/low_mean": 0.0011377867303963285,
      "clip_ratio/low_min": 1.3536929145629983e-05,
      "clip_ratio/region_mean": 0.0021736430426244624,
      "epoch": 3.289297171186935,
      "grad_norm": 0.22067445516586304,
      "learning_rate": 1e-06,
      "loss": -0.1228,
      "step": 1408
    },
    {
      "clip_ratio/high_max": 0.002485864573827712,
      "clip_ratio/high_mean": 0.0009715916239656508,
      "clip_ratio/low_mean": 0.0006173867241159314,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015889783389866352,
      "completions/clipped_ratio": 0.1975446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3935.0,
      "completions/mean_length": 1307.6663818359375,
      "completions/mean_terminated_length": 621.24755859375,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 3.2916302128900554,
      "grad_norm": 0.4216117858886719,
      "learning_rate": 1e-06,
      "loss": -0.0581,
      "num_tokens": 206892232.0,
      "reward": 0.5803571939468384,
      "reward_std": 0.14166000485420227,
      "rewards/verify_math_reward/mean": 0.5803571343421936,
      "rewards/verify_math_reward/std": 0.4937761127948761,
      "step": 1409
    },
    {
      "clip_ratio/high_max": 0.0030081197182880715,
      "clip_ratio/high_mean": 0.0011857901899929857,
      "clip_ratio/low_mean": 0.0008584561774114263,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020442463573999703,
      "epoch": 3.2939632545931756,
      "grad_norm": 0.2833123505115509,
      "learning_rate": 1e-06,
      "loss": -0.0584,
      "step": 1410
    },
    {
      "clip_ratio/high_max": 0.0028013972550979815,
      "clip_ratio/high_mean": 0.001152902543253731,
      "clip_ratio/low_mean": 0.0011308564498904161,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022837589713162743,
      "epoch": 3.2962962962962963,
      "grad_norm": 0.28059399127960205,
      "learning_rate": 1e-06,
      "loss": -0.0587,
      "step": 1411
    },
    {
      "clip_ratio/high_max": 0.0031393519675475545,
      "clip_ratio/high_mean": 0.0010532768737903098,
      "clip_ratio/low_mean": 0.0012799748437828384,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002333251715754159,
      "epoch": 3.298629337999417,
      "grad_norm": 0.3258727490901947,
      "learning_rate": 1e-06,
      "loss": -0.0587,
      "step": 1412
    },
    {
      "clip_ratio/high_max": 0.0029762344929622486,
      "clip_ratio/high_mean": 0.0011315756164549384,
      "clip_ratio/low_mean": 0.000596388447775098,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017279640451306477,
      "completions/clipped_ratio": 0.1986607142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3880.0,
      "completions/mean_length": 1383.8360595703125,
      "completions/mean_terminated_length": 711.4609985351562,
      "completions/min_length": 129.0,
      "completions/min_terminated_length": 129.0,
      "epoch": 3.300962379702537,
      "grad_norm": 0.3155386447906494,
      "learning_rate": 1e-06,
      "loss": -0.0868,
      "num_tokens": 207493981.0,
      "reward": 0.4899553656578064,
      "reward_std": 0.1817513257265091,
      "rewards/verify_math_reward/mean": 0.4899553656578064,
      "rewards/verify_math_reward/std": 0.5001782774925232,
      "step": 1413
    },
    {
      "clip_ratio/high_max": 0.003695404317113571,
      "clip_ratio/high_mean": 0.0013442114504869096,
      "clip_ratio/low_mean": 0.0008084241362666944,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021526355121750385,
      "epoch": 3.303295421405658,
      "grad_norm": 0.3200008273124695,
      "learning_rate": 1e-06,
      "loss": -0.087,
      "step": 1414
    },
    {
      "clip_ratio/high_max": 0.00346468755742535,
      "clip_ratio/high_mean": 0.0013513403246179223,
      "clip_ratio/low_mean": 0.000933639259528718,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022849796223454177,
      "epoch": 3.305628463108778,
      "grad_norm": 0.24863432347774506,
      "learning_rate": 1e-06,
      "loss": -0.0872,
      "step": 1415
    },
    {
      "clip_ratio/high_max": 0.0036600856838049367,
      "clip_ratio/high_mean": 0.0013390149251790717,
      "clip_ratio/low_mean": 0.0011202372443221975,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0024592521294835024,
      "epoch": 3.3079615048118987,
      "grad_norm": 0.23993004858493805,
      "learning_rate": 1e-06,
      "loss": -0.0872,
      "step": 1416
    },
    {
      "clip_ratio/high_max": 0.0019119550051982515,
      "clip_ratio/high_mean": 0.0006607695031561889,
      "clip_ratio/low_mean": 0.0004249778171470098,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00108574733894784,
      "completions/clipped_ratio": 0.1417410714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3274.0,
      "completions/mean_length": 1082.376220703125,
      "completions/mean_terminated_length": 584.677490234375,
      "completions/min_length": 191.0,
      "completions/min_terminated_length": 191.0,
      "epoch": 3.310294546515019,
      "grad_norm": 0.30247485637664795,
      "learning_rate": 1e-06,
      "loss": -0.0402,
      "num_tokens": 208020966.0,
      "reward": 0.6361607313156128,
      "reward_std": 0.12267616391181946,
      "rewards/verify_math_reward/mean": 0.6361607313156128,
      "rewards/verify_math_reward/std": 0.4813718795776367,
      "step": 1417
    },
    {
      "clip_ratio/high_max": 0.002196392902988009,
      "clip_ratio/high_mean": 0.0008674347182022757,
      "clip_ratio/low_mean": 0.0006404634668797371,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00150789816325414,
      "epoch": 3.3126275882181395,
      "grad_norm": 0.27442285418510437,
      "learning_rate": 1e-06,
      "loss": -0.0405,
      "step": 1418
    },
    {
      "clip_ratio/high_max": 0.0025767409315449186,
      "clip_ratio/high_mean": 0.0009292753056797665,
      "clip_ratio/low_mean": 0.0008339601104125904,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017632354065426625,
      "epoch": 3.3149606299212597,
      "grad_norm": 0.2677547037601471,
      "learning_rate": 1e-06,
      "loss": -0.0406,
      "step": 1419
    },
    {
      "clip_ratio/high_max": 0.0021459552117448766,
      "clip_ratio/high_mean": 0.0007863909440857242,
      "clip_ratio/low_mean": 0.0009221609288943,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017085518957173917,
      "epoch": 3.3172936716243804,
      "grad_norm": 0.28091907501220703,
      "learning_rate": 1e-06,
      "loss": -0.0406,
      "step": 1420
    },
    {
      "clip_ratio/high_max": 0.0026617609764798544,
      "clip_ratio/high_mean": 0.001080154237570241,
      "clip_ratio/low_mean": 0.0006827125680501922,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017628667847020552,
      "completions/clipped_ratio": 0.1573660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2586.0,
      "completions/mean_length": 1220.9085693359375,
      "completions/mean_terminated_length": 683.9708862304688,
      "completions/min_length": 151.0,
      "completions/min_terminated_length": 151.0,
      "epoch": 3.3196267133275006,
      "grad_norm": 0.31796377897262573,
      "learning_rate": 1e-06,
      "loss": -0.0813,
      "num_tokens": 208628124.0,
      "reward": 0.5859375,
      "reward_std": 0.1989564299583435,
      "rewards/verify_math_reward/mean": 0.5859375,
      "rewards/verify_math_reward/std": 0.4928344786167145,
      "step": 1421
    },
    {
      "clip_ratio/high_max": 0.003174099554598797,
      "clip_ratio/high_mean": 0.001284394380490994,
      "clip_ratio/low_mean": 0.000976650461780082,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022610448650084436,
      "epoch": 3.3219597550306212,
      "grad_norm": 0.2940112054347992,
      "learning_rate": 1e-06,
      "loss": -0.0816,
      "step": 1422
    },
    {
      "clip_ratio/high_max": 0.003210231261618901,
      "clip_ratio/high_mean": 0.0013108856364851817,
      "clip_ratio/low_mean": 0.001243566661287332,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002554452257754747,
      "epoch": 3.3242927967337415,
      "grad_norm": 0.36122965812683105,
      "learning_rate": 1e-06,
      "loss": -0.0819,
      "step": 1423
    },
    {
      "clip_ratio/high_max": 0.0031171913651633076,
      "clip_ratio/high_mean": 0.00118163075967459,
      "clip_ratio/low_mean": 0.0014060313278605463,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002587662158475723,
      "epoch": 3.326625838436862,
      "grad_norm": 0.2434595227241516,
      "learning_rate": 1e-06,
      "loss": -0.0819,
      "step": 1424
    },
    {
      "clip_ratio/high_max": 0.0022868446685606614,
      "clip_ratio/high_mean": 0.0009595526171324309,
      "clip_ratio/low_mean": 0.0006868742293590913,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016464268555864692,
      "completions/clipped_ratio": 0.2287946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3040.0,
      "completions/mean_length": 1486.864990234375,
      "completions/mean_terminated_length": 712.8089599609375,
      "completions/min_length": 179.0,
      "completions/min_terminated_length": 179.0,
      "epoch": 3.3289588801399823,
      "grad_norm": 0.33154115080833435,
      "learning_rate": 1e-06,
      "loss": -0.1223,
      "num_tokens": 209217779.0,
      "reward": 0.5345982313156128,
      "reward_std": 0.17682796716690063,
      "rewards/verify_math_reward/mean": 0.5345982313156128,
      "rewards/verify_math_reward/std": 0.4990801215171814,
      "step": 1425
    },
    {
      "clip_ratio/high_max": 0.0027643133798846975,
      "clip_ratio/high_mean": 0.0012076245911885053,
      "clip_ratio/low_mean": 0.0008671995437907754,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020748240931425244,
      "epoch": 3.331291921843103,
      "grad_norm": 0.2755907475948334,
      "learning_rate": 1e-06,
      "loss": -0.1226,
      "step": 1426
    },
    {
      "clip_ratio/high_max": 0.0028445334028219804,
      "clip_ratio/high_mean": 0.001203192128741648,
      "clip_ratio/low_mean": 0.000988325117759814,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021915172765147872,
      "epoch": 3.3336249635462236,
      "grad_norm": 0.2327805757522583,
      "learning_rate": 1e-06,
      "loss": -0.1227,
      "step": 1427
    },
    {
      "clip_ratio/high_max": 0.0022473930803244,
      "clip_ratio/high_mean": 0.0010317661872250028,
      "clip_ratio/low_mean": 0.0011683286265906645,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002200094830186572,
      "epoch": 3.335958005249344,
      "grad_norm": 0.3731497824192047,
      "learning_rate": 1e-06,
      "loss": -0.1228,
      "step": 1428
    },
    {
      "clip_ratio/high_max": 0.002318192884558812,
      "clip_ratio/high_mean": 0.0009074763056560187,
      "clip_ratio/low_mean": 0.0005459625608636998,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014534388537867926,
      "completions/clipped_ratio": 0.1417410714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3460.0,
      "completions/mean_length": 1155.704345703125,
      "completions/mean_terminated_length": 670.11572265625,
      "completions/min_length": 162.0,
      "completions/min_terminated_length": 162.0,
      "epoch": 3.338291046952464,
      "grad_norm": 0.29542016983032227,
      "learning_rate": 1e-06,
      "loss": -0.0759,
      "num_tokens": 209816866.0,
      "reward": 0.6116071939468384,
      "reward_std": 0.1679650843143463,
      "rewards/verify_math_reward/mean": 0.6116071343421936,
      "rewards/verify_math_reward/std": 0.48765692114830017,
      "step": 1429
    },
    {
      "clip_ratio/high_max": 0.0026791319833137095,
      "clip_ratio/high_mean": 0.0010337810726923635,
      "clip_ratio/low_mean": 0.0007834240414013038,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018172051168221515,
      "epoch": 3.3406240886555847,
      "grad_norm": 0.3144708573818207,
      "learning_rate": 1e-06,
      "loss": -0.0762,
      "step": 1430
    },
    {
      "clip_ratio/high_max": 0.0030275606841314584,
      "clip_ratio/high_mean": 0.0010740918824012624,
      "clip_ratio/low_mean": 0.00085768296776223,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019317748447065242,
      "epoch": 3.3429571303587053,
      "grad_norm": 0.22165705263614655,
      "learning_rate": 1e-06,
      "loss": -0.0764,
      "step": 1431
    },
    {
      "clip_ratio/high_max": 0.0026318026721128263,
      "clip_ratio/high_mean": 0.0010720016362029128,
      "clip_ratio/low_mean": 0.0010904667506110854,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002162468394089956,
      "epoch": 3.3452901720618256,
      "grad_norm": 0.23472611606121063,
      "learning_rate": 1e-06,
      "loss": -0.0764,
      "step": 1432
    },
    {
      "clip_ratio/high_max": 0.0021777022557216696,
      "clip_ratio/high_mean": 0.0008209644402086269,
      "clip_ratio/low_mean": 0.00048828055128069536,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013092449844407383,
      "completions/clipped_ratio": 0.1841517857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3955.0,
      "completions/mean_length": 1342.7098388671875,
      "completions/mean_terminated_length": 721.2421875,
      "completions/min_length": 125.0,
      "completions/min_terminated_length": 125.0,
      "epoch": 3.347623213764946,
      "grad_norm": 0.3252272307872772,
      "learning_rate": 1e-06,
      "loss": -0.0597,
      "num_tokens": 210438814.0,
      "reward": 0.559151828289032,
      "reward_std": 0.1498156636953354,
      "rewards/verify_math_reward/mean": 0.5591517686843872,
      "rewards/verify_math_reward/std": 0.496766060590744,
      "step": 1433
    },
    {
      "clip_ratio/high_max": 0.0021948493376839906,
      "clip_ratio/high_mean": 0.0007772110275254818,
      "clip_ratio/low_mean": 0.0007795270157657797,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015567380032734945,
      "epoch": 3.3499562554680664,
      "grad_norm": 0.2555447816848755,
      "learning_rate": 1e-06,
      "loss": -0.0599,
      "step": 1434
    },
    {
      "clip_ratio/high_max": 0.0026419895366416313,
      "clip_ratio/high_mean": 0.0009608177933841944,
      "clip_ratio/low_mean": 0.0008887026015145238,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018495203839847818,
      "epoch": 3.352289297171187,
      "grad_norm": 0.2451346218585968,
      "learning_rate": 1e-06,
      "loss": -0.0601,
      "step": 1435
    },
    {
      "clip_ratio/high_max": 0.0022972883889451623,
      "clip_ratio/high_mean": 0.000838000169096631,
      "clip_ratio/low_mean": 0.0010169958750338992,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018549960732343607,
      "epoch": 3.3546223388743073,
      "grad_norm": 0.19539831578731537,
      "learning_rate": 1e-06,
      "loss": -0.0601,
      "step": 1436
    },
    {
      "clip_ratio/high_max": 0.0023833828745409846,
      "clip_ratio/high_mean": 0.0009572358248988166,
      "clip_ratio/low_mean": 0.000545118511581677,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015023543310235254,
      "completions/clipped_ratio": 0.1573660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3901.0,
      "completions/mean_length": 1180.9810791015625,
      "completions/mean_terminated_length": 636.5867919921875,
      "completions/min_length": 163.0,
      "completions/min_terminated_length": 163.0,
      "epoch": 3.356955380577428,
      "grad_norm": 0.2828822731971741,
      "learning_rate": 1e-06,
      "loss": -0.0351,
      "num_tokens": 211010349.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.15639057755470276,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 1437
    },
    {
      "clip_ratio/high_max": 0.002974518807604909,
      "clip_ratio/high_mean": 0.0011843534666695632,
      "clip_ratio/low_mean": 0.0007223030552268028,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019066565146204084,
      "epoch": 3.359288422280548,
      "grad_norm": 0.2754281759262085,
      "learning_rate": 1e-06,
      "loss": -0.0353,
      "step": 1438
    },
    {
      "clip_ratio/high_max": 0.0027547199715627357,
      "clip_ratio/high_mean": 0.0010415829019621015,
      "clip_ratio/low_mean": 0.0009017273655445024,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001943310235219542,
      "epoch": 3.361621463983669,
      "grad_norm": 0.24657297134399414,
      "learning_rate": 1e-06,
      "loss": -0.0355,
      "step": 1439
    },
    {
      "clip_ratio/high_max": 0.0024943129246821627,
      "clip_ratio/high_mean": 0.0010055546554212924,
      "clip_ratio/low_mean": 0.0010134600152014173,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020190146860841196,
      "epoch": 3.363954505686789,
      "grad_norm": 0.24397654831409454,
      "learning_rate": 1e-06,
      "loss": -0.0355,
      "step": 1440
    },
    {
      "clip_ratio/high_max": 0.002338218291697558,
      "clip_ratio/high_mean": 0.0008907749215723015,
      "clip_ratio/low_mean": 0.0007159109391068341,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016066858697740827,
      "completions/clipped_ratio": 0.1852678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2982.0,
      "completions/mean_length": 1248.5023193359375,
      "completions/mean_terminated_length": 600.9890747070312,
      "completions/min_length": 175.0,
      "completions/min_terminated_length": 175.0,
      "epoch": 3.3662875473899097,
      "grad_norm": 0.5020262598991394,
      "learning_rate": 1e-06,
      "loss": -0.0937,
      "num_tokens": 211549791.0,
      "reward": 0.5837053656578064,
      "reward_std": 0.16735847294330597,
      "rewards/verify_math_reward/mean": 0.5837053656578064,
      "rewards/verify_math_reward/std": 0.49321892857551575,
      "step": 1441
    },
    {
      "clip_ratio/high_max": 0.003099871944868937,
      "clip_ratio/high_mean": 0.0011718298719642917,
      "clip_ratio/low_mean": 0.0009420182432222646,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002113848146109376,
      "epoch": 3.36862058909303,
      "grad_norm": 0.2910059690475464,
      "learning_rate": 1e-06,
      "loss": -0.0941,
      "step": 1442
    },
    {
      "clip_ratio/high_max": 0.0036351317903609015,
      "clip_ratio/high_mean": 0.0012362859579297947,
      "clip_ratio/low_mean": 0.0012269529543118551,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002463238903146703,
      "epoch": 3.3709536307961505,
      "grad_norm": 0.25305771827697754,
      "learning_rate": 1e-06,
      "loss": -0.0944,
      "step": 1443
    },
    {
      "clip_ratio/high_max": 0.00290341027721297,
      "clip_ratio/high_mean": 0.0010353881625633221,
      "clip_ratio/low_mean": 0.001412170080584474,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002447558203130029,
      "epoch": 3.3732866724992707,
      "grad_norm": 0.2992576062679291,
      "learning_rate": 1e-06,
      "loss": -0.0943,
      "step": 1444
    },
    {
      "clip_ratio/high_max": 0.0026444131799507886,
      "clip_ratio/high_mean": 0.0012394175937515683,
      "clip_ratio/low_mean": 0.0006469405016105156,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018863580698962323,
      "completions/clipped_ratio": 0.1607142857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2743.0,
      "completions/mean_length": 1167.6273193359375,
      "completions/mean_terminated_length": 606.875,
      "completions/min_length": 113.0,
      "completions/min_terminated_length": 113.0,
      "epoch": 3.3756197142023914,
      "grad_norm": 0.3741614520549774,
      "learning_rate": 1e-06,
      "loss": -0.0885,
      "num_tokens": 212092313.0,
      "reward": 0.6037946939468384,
      "reward_std": 0.19392429292201996,
      "rewards/verify_math_reward/mean": 0.6037946343421936,
      "rewards/verify_math_reward/std": 0.48938122391700745,
      "step": 1445
    },
    {
      "clip_ratio/high_max": 0.0033751697483239695,
      "clip_ratio/high_mean": 0.0014848263417661656,
      "clip_ratio/low_mean": 0.0009012938253363245,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00238612021348672,
      "epoch": 3.377952755905512,
      "grad_norm": 0.4091361463069916,
      "learning_rate": 1e-06,
      "loss": -0.0888,
      "step": 1446
    },
    {
      "clip_ratio/high_max": 0.002985111394082196,
      "clip_ratio/high_mean": 0.0014381614200829063,
      "clip_ratio/low_mean": 0.0011305916850687936,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0025687531087896787,
      "epoch": 3.3802857976086322,
      "grad_norm": 0.27056244015693665,
      "learning_rate": 1e-06,
      "loss": -0.089,
      "step": 1447
    },
    {
      "clip_ratio/high_max": 0.0030088677958701737,
      "clip_ratio/high_mean": 0.001426322076440556,
      "clip_ratio/low_mean": 0.001348124345895485,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0027744463659473695,
      "epoch": 3.382618839311753,
      "grad_norm": 0.4235474169254303,
      "learning_rate": 1e-06,
      "loss": -0.0891,
      "step": 1448
    },
    {
      "clip_ratio/high_max": 0.002184545337513555,
      "clip_ratio/high_mean": 0.0008029766613617539,
      "clip_ratio/low_mean": 0.0007079693550622324,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015109460146049969,
      "completions/clipped_ratio": 0.1741071428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2959.0,
      "completions/mean_length": 1263.3070068359375,
      "completions/mean_terminated_length": 666.1445922851562,
      "completions/min_length": 174.0,
      "completions/min_terminated_length": 174.0,
      "epoch": 3.384951881014873,
      "grad_norm": 0.3509569764137268,
      "learning_rate": 1e-06,
      "loss": -0.0516,
      "num_tokens": 212675356.0,
      "reward": 0.5022321939468384,
      "reward_std": 0.14789676666259766,
      "rewards/verify_math_reward/mean": 0.5022321343421936,
      "rewards/verify_math_reward/std": 0.5002743005752563,
      "step": 1449
    },
    {
      "clip_ratio/high_max": 0.0027263353185844608,
      "clip_ratio/high_mean": 0.0009901039575197501,
      "clip_ratio/low_mean": 0.0009751847173902206,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019652886549010873,
      "epoch": 3.3872849227179938,
      "grad_norm": 0.2886752188205719,
      "learning_rate": 1e-06,
      "loss": -0.0518,
      "step": 1450
    },
    {
      "clip_ratio/high_max": 0.0028788219424313866,
      "clip_ratio/high_mean": 0.0010585665240796516,
      "clip_ratio/low_mean": 0.0011499497886688914,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022085162927396595,
      "epoch": 3.389617964421114,
      "grad_norm": 0.2655644714832306,
      "learning_rate": 1e-06,
      "loss": -0.0521,
      "step": 1451
    },
    {
      "clip_ratio/high_max": 0.0024294789545820095,
      "clip_ratio/high_mean": 0.0009403186340932734,
      "clip_ratio/low_mean": 0.0013969198989798315,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002337238511245232,
      "epoch": 3.3919510061242346,
      "grad_norm": 0.41138431429862976,
      "learning_rate": 1e-06,
      "loss": -0.0521,
      "step": 1452
    },
    {
      "clip_ratio/high_max": 0.003638951056927908,
      "clip_ratio/high_mean": 0.0010821499108715216,
      "clip_ratio/low_mean": 0.0006731365037921933,
      "clip_ratio/low_min": 3.533069684635848e-05,
      "clip_ratio/region_mean": 0.0017552864010212943,
      "completions/clipped_ratio": 0.2053571428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4087.0,
      "completions/mean_length": 1338.818115234375,
      "completions/mean_terminated_length": 626.2879028320312,
      "completions/min_length": 171.0,
      "completions/min_terminated_length": 171.0,
      "epoch": 3.394284047827355,
      "grad_norm": 0.5266408324241638,
      "learning_rate": 1e-06,
      "loss": -0.0852,
      "num_tokens": 213219121.0,
      "reward": 0.5301339626312256,
      "reward_std": 0.1286219209432602,
      "rewards/verify_math_reward/mean": 0.5301339030265808,
      "rewards/verify_math_reward/std": 0.49936985969543457,
      "step": 1453
    },
    {
      "clip_ratio/high_max": 0.0034111940694856457,
      "clip_ratio/high_mean": 0.0011253326847509015,
      "clip_ratio/low_mean": 0.0011274409289399046,
      "clip_ratio/low_min": 3.533069684635848e-05,
      "clip_ratio/region_mean": 0.002252773614600301,
      "epoch": 3.3966170895304755,
      "grad_norm": 0.3869584798812866,
      "learning_rate": 1e-06,
      "loss": -0.0854,
      "step": 1454
    },
    {
      "clip_ratio/high_max": 0.003642870804469567,
      "clip_ratio/high_mean": 0.001153471666839323,
      "clip_ratio/low_mean": 0.0013703641252504895,
      "clip_ratio/low_min": 3.533069684635848e-05,
      "clip_ratio/region_mean": 0.0025238358066417277,
      "epoch": 3.3989501312335957,
      "grad_norm": 0.25054648518562317,
      "learning_rate": 1e-06,
      "loss": -0.0856,
      "step": 1455
    },
    {
      "clip_ratio/high_max": 0.003084875839704182,
      "clip_ratio/high_mean": 0.0010314236951671774,
      "clip_ratio/low_mean": 0.0015251716185957775,
      "clip_ratio/low_min": 5.299604163155891e-05,
      "clip_ratio/region_mean": 0.002556595398345962,
      "epoch": 3.4012831729367163,
      "grad_norm": 0.29132816195487976,
      "learning_rate": 1e-06,
      "loss": -0.0857,
      "step": 1456
    },
    {
      "clip_ratio/high_max": 0.0021597528975689784,
      "clip_ratio/high_mean": 0.0007703071896685287,
      "clip_ratio/low_mean": 0.0005968487203062978,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013671559136128053,
      "completions/clipped_ratio": 0.1595982142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3225.0,
      "completions/mean_length": 1205.3270263671875,
      "completions/mean_terminated_length": 656.3678588867188,
      "completions/min_length": 112.0,
      "completions/min_terminated_length": 112.0,
      "epoch": 3.4036162146398365,
      "grad_norm": 0.3055138885974884,
      "learning_rate": 1e-06,
      "loss": -0.0559,
      "num_tokens": 213798326.0,
      "reward": 0.5703125,
      "reward_std": 0.14917626976966858,
      "rewards/verify_math_reward/mean": 0.5703125,
      "rewards/verify_math_reward/std": 0.49530795216560364,
      "step": 1457
    },
    {
      "clip_ratio/high_max": 0.0023967510569491424,
      "clip_ratio/high_mean": 0.000890681307282648,
      "clip_ratio/low_mean": 0.000815241770396824,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017059230849554297,
      "epoch": 3.405949256342957,
      "grad_norm": 0.31193846464157104,
      "learning_rate": 1e-06,
      "loss": -0.056,
      "step": 1458
    },
    {
      "clip_ratio/high_max": 0.0029225418911664747,
      "clip_ratio/high_mean": 0.0009370192456117366,
      "clip_ratio/low_mean": 0.0010417152334412094,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001978734500880819,
      "epoch": 3.4082822980460774,
      "grad_norm": 0.2549133002758026,
      "learning_rate": 1e-06,
      "loss": -0.0562,
      "step": 1459
    },
    {
      "clip_ratio/high_max": 0.002612090807815548,
      "clip_ratio/high_mean": 0.0008295100005852873,
      "clip_ratio/low_mean": 0.001182372754556127,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020118827233090997,
      "epoch": 3.410615339749198,
      "grad_norm": 0.22129330039024353,
      "learning_rate": 1e-06,
      "loss": -0.0562,
      "step": 1460
    },
    {
      "clip_ratio/high_max": 0.0024314190086442977,
      "clip_ratio/high_mean": 0.0008742830959818093,
      "clip_ratio/low_mean": 0.0006333228957373649,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015076059935381636,
      "completions/clipped_ratio": 0.1819196428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3286.0,
      "completions/mean_length": 1304.188720703125,
      "completions/mean_terminated_length": 683.3629150390625,
      "completions/min_length": 188.0,
      "completions/min_terminated_length": 188.0,
      "epoch": 3.4129483814523183,
      "grad_norm": 0.33513981103897095,
      "learning_rate": 1e-06,
      "loss": -0.0537,
      "num_tokens": 214401231.0,
      "reward": 0.5267857313156128,
      "reward_std": 0.15262722969055176,
      "rewards/verify_math_reward/mean": 0.5267857313156128,
      "rewards/verify_math_reward/std": 0.4995608627796173,
      "step": 1461
    },
    {
      "clip_ratio/high_max": 0.0030209238320821896,
      "clip_ratio/high_mean": 0.0010743962702690624,
      "clip_ratio/low_mean": 0.000757148836783017,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018315451088710688,
      "epoch": 3.415281423155439,
      "grad_norm": 0.32869017124176025,
      "learning_rate": 1e-06,
      "loss": -0.0539,
      "step": 1462
    },
    {
      "clip_ratio/high_max": 0.0027078243874711916,
      "clip_ratio/high_mean": 0.0010203564634139184,
      "clip_ratio/low_mean": 0.0009456528478040127,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001966009229363408,
      "epoch": 3.417614464858559,
      "grad_norm": 0.31274259090423584,
      "learning_rate": 1e-06,
      "loss": -0.054,
      "step": 1463
    },
    {
      "clip_ratio/high_max": 0.0029160712656448595,
      "clip_ratio/high_mean": 0.0009818917023949325,
      "clip_ratio/low_mean": 0.001110161920223618,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002092053597152699,
      "epoch": 3.41994750656168,
      "grad_norm": 0.3171895146369934,
      "learning_rate": 1e-06,
      "loss": -0.0542,
      "step": 1464
    },
    {
      "clip_ratio/high_max": 0.0025632300530560315,
      "clip_ratio/high_mean": 0.0010922487563220784,
      "clip_ratio/low_mean": 0.0004209585167700425,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015132072949199937,
      "completions/clipped_ratio": 0.2287946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3044.0,
      "completions/mean_length": 1456.368408203125,
      "completions/mean_terminated_length": 673.2648315429688,
      "completions/min_length": 153.0,
      "completions/min_terminated_length": 153.0,
      "epoch": 3.4222805482648004,
      "grad_norm": 0.2822405993938446,
      "learning_rate": 1e-06,
      "loss": -0.1051,
      "num_tokens": 214964665.0,
      "reward": 0.5189732313156128,
      "reward_std": 0.16979841887950897,
      "rewards/verify_math_reward/mean": 0.5189732313156128,
      "rewards/verify_math_reward/std": 0.49991893768310547,
      "step": 1465
    },
    {
      "clip_ratio/high_max": 0.0032525687711313367,
      "clip_ratio/high_mean": 0.00120888634774019,
      "clip_ratio/low_mean": 0.0005372365794755751,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001746122958138585,
      "epoch": 3.4246135899679206,
      "grad_norm": 0.26969727873802185,
      "learning_rate": 1e-06,
      "loss": -0.1052,
      "step": 1466
    },
    {
      "clip_ratio/high_max": 0.0027927601477131248,
      "clip_ratio/high_mean": 0.0012141498791606864,
      "clip_ratio/low_mean": 0.0008037213856368908,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020178713093628176,
      "epoch": 3.4269466316710413,
      "grad_norm": 0.22524547576904297,
      "learning_rate": 1e-06,
      "loss": -0.1054,
      "step": 1467
    },
    {
      "clip_ratio/high_max": 0.0033165725326398388,
      "clip_ratio/high_mean": 0.0013566705092671327,
      "clip_ratio/low_mean": 0.0008546340759494342,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022113045706646517,
      "epoch": 3.4292796733741615,
      "grad_norm": 0.2359517365694046,
      "learning_rate": 1e-06,
      "loss": -0.1055,
      "step": 1468
    },
    {
      "clip_ratio/high_max": 0.002432909059280064,
      "clip_ratio/high_mean": 0.0009736033844092162,
      "clip_ratio/low_mean": 0.0008165398685378022,
      "clip_ratio/low_min": 3.7224537663860247e-05,
      "clip_ratio/region_mean": 0.0017901432293001562,
      "completions/clipped_ratio": 0.1339285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2316.0,
      "completions/mean_length": 1084.165283203125,
      "completions/mean_terminated_length": 618.41748046875,
      "completions/min_length": 126.0,
      "completions/min_terminated_length": 126.0,
      "epoch": 3.431612715077282,
      "grad_norm": 0.3691757023334503,
      "learning_rate": 1e-06,
      "loss": -0.0603,
      "num_tokens": 215528637.0,
      "reward": 0.5970982313156128,
      "reward_std": 0.1787768304347992,
      "rewards/verify_math_reward/mean": 0.5970982313156128,
      "rewards/verify_math_reward/std": 0.49075523018836975,
      "step": 1469
    },
    {
      "clip_ratio/high_max": 0.003240038473450113,
      "clip_ratio/high_mean": 0.001244753459104686,
      "clip_ratio/low_mean": 0.0010759066644823179,
      "clip_ratio/low_min": 2.810251862683799e-05,
      "clip_ratio/region_mean": 0.002320660096302163,
      "epoch": 3.4339457567804024,
      "grad_norm": 0.2718658745288849,
      "learning_rate": 1e-06,
      "loss": -0.0606,
      "step": 1470
    },
    {
      "clip_ratio/high_max": 0.003535356947395485,
      "clip_ratio/high_mean": 0.0012196841926197521,
      "clip_ratio/low_mean": 0.0012764189232257195,
      "clip_ratio/low_min": 6.488450890174136e-05,
      "clip_ratio/region_mean": 0.0024961030794656835,
      "epoch": 3.436278798483523,
      "grad_norm": 0.2634088099002838,
      "learning_rate": 1e-06,
      "loss": -0.0608,
      "step": 1471
    },
    {
      "clip_ratio/high_max": 0.003507355591864325,
      "clip_ratio/high_mean": 0.001267346080567222,
      "clip_ratio/low_mean": 0.0015165622462518513,
      "clip_ratio/low_min": 9.048136416822672e-05,
      "clip_ratio/region_mean": 0.002783908297715243,
      "epoch": 3.4386118401866432,
      "grad_norm": 0.24711427092552185,
      "learning_rate": 1e-06,
      "loss": -0.061,
      "step": 1472
    },
    {
      "clip_ratio/high_max": 0.0026092431653523818,
      "clip_ratio/high_mean": 0.0009640817843319383,
      "clip_ratio/low_mean": 0.0006873779857414775,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016514597373316064,
      "completions/clipped_ratio": 0.1729910714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3145.0,
      "completions/mean_length": 1235.1942138671875,
      "completions/mean_terminated_length": 636.780029296875,
      "completions/min_length": 143.0,
      "completions/min_terminated_length": 143.0,
      "epoch": 3.440944881889764,
      "grad_norm": 0.33380481600761414,
      "learning_rate": 1e-06,
      "loss": -0.0653,
      "num_tokens": 216100843.0,
      "reward": 0.5345982313156128,
      "reward_std": 0.1698751002550125,
      "rewards/verify_math_reward/mean": 0.5345982313156128,
      "rewards/verify_math_reward/std": 0.4990801215171814,
      "step": 1473
    },
    {
      "clip_ratio/high_max": 0.0029111165495123714,
      "clip_ratio/high_mean": 0.001066847351467004,
      "clip_ratio/low_mean": 0.0009605682098481338,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020274155322113074,
      "epoch": 3.443277923592884,
      "grad_norm": 0.3211491107940674,
      "learning_rate": 1e-06,
      "loss": -0.0656,
      "step": 1474
    },
    {
      "clip_ratio/high_max": 0.0030401530166273005,
      "clip_ratio/high_mean": 0.0011105716566817136,
      "clip_ratio/low_mean": 0.001179943916213233,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002290515556524042,
      "epoch": 3.4456109652960047,
      "grad_norm": 0.2712908983230591,
      "learning_rate": 1e-06,
      "loss": -0.0657,
      "step": 1475
    },
    {
      "clip_ratio/high_max": 0.0028097581598558463,
      "clip_ratio/high_mean": 0.0010535451365285553,
      "clip_ratio/low_mean": 0.0012918122920382302,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002345357359445188,
      "epoch": 3.447944006999125,
      "grad_norm": 0.35899391770362854,
      "learning_rate": 1e-06,
      "loss": -0.0658,
      "step": 1476
    },
    {
      "clip_ratio/high_max": 0.0027615407088887878,
      "clip_ratio/high_mean": 0.0009777019477041904,
      "clip_ratio/low_mean": 0.0004897192902717507,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014674212216050364,
      "completions/clipped_ratio": 0.1953125,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3851.0,
      "completions/mean_length": 1360.4732666015625,
      "completions/mean_terminated_length": 696.5104370117188,
      "completions/min_length": 191.0,
      "completions/min_terminated_length": 191.0,
      "epoch": 3.4502770487022456,
      "grad_norm": 0.30021557211875916,
      "learning_rate": 1e-06,
      "loss": -0.0882,
      "num_tokens": 216688891.0,
      "reward": 0.5401785969734192,
      "reward_std": 0.13301397860050201,
      "rewards/verify_math_reward/mean": 0.5401785969734192,
      "rewards/verify_math_reward/std": 0.49866142868995667,
      "step": 1477
    },
    {
      "clip_ratio/high_max": 0.003418921187403612,
      "clip_ratio/high_mean": 0.0011423223222664092,
      "clip_ratio/low_mean": 0.0006081819453811477,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001750504299707245,
      "epoch": 3.452610090405366,
      "grad_norm": 0.2633775472640991,
      "learning_rate": 1e-06,
      "loss": -0.0884,
      "step": 1478
    },
    {
      "clip_ratio/high_max": 0.002891820127842948,
      "clip_ratio/high_mean": 0.0011583926898310892,
      "clip_ratio/low_mean": 0.0007072868293107604,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001865679514594376,
      "epoch": 3.4549431321084865,
      "grad_norm": 0.25078412890434265,
      "learning_rate": 1e-06,
      "loss": -0.0885,
      "step": 1479
    },
    {
      "clip_ratio/high_max": 0.003491372503049206,
      "clip_ratio/high_mean": 0.0011569949419936165,
      "clip_ratio/low_mean": 0.0009172863428830169,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020742812994285487,
      "epoch": 3.457276173811607,
      "grad_norm": 0.21249748766422272,
      "learning_rate": 1e-06,
      "loss": -0.0886,
      "step": 1480
    },
    {
      "clip_ratio/high_max": 0.002392310043433099,
      "clip_ratio/high_mean": 0.0007180312068157946,
      "clip_ratio/low_mean": 0.0007742367833998287,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014922680238669273,
      "completions/clipped_ratio": 0.1462053571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3920.0,
      "completions/mean_length": 1145.55810546875,
      "completions/mean_terminated_length": 640.3189697265625,
      "completions/min_length": 148.0,
      "completions/min_terminated_length": 148.0,
      "epoch": 3.4596092155147273,
      "grad_norm": 0.35187554359436035,
      "learning_rate": 1e-06,
      "loss": -0.0425,
      "num_tokens": 217263943.0,
      "reward": 0.6183035969734192,
      "reward_std": 0.1450003683567047,
      "rewards/verify_math_reward/mean": 0.6183035969734192,
      "rewards/verify_math_reward/std": 0.4860740303993225,
      "step": 1481
    },
    {
      "clip_ratio/high_max": 0.0031581436760461656,
      "clip_ratio/high_mean": 0.0009571097762091085,
      "clip_ratio/low_mean": 0.0010297510352756944,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019868608069373295,
      "epoch": 3.4619422572178475,
      "grad_norm": 0.27443844079971313,
      "learning_rate": 1e-06,
      "loss": -0.0428,
      "step": 1482
    },
    {
      "clip_ratio/high_max": 0.002752281063294504,
      "clip_ratio/high_mean": 0.0008675640815454244,
      "clip_ratio/low_mean": 0.0012707108498943853,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021382749073382,
      "epoch": 3.464275298920968,
      "grad_norm": 0.32418301701545715,
      "learning_rate": 1e-06,
      "loss": -0.0429,
      "step": 1483
    },
    {
      "clip_ratio/high_max": 0.0028385285258991644,
      "clip_ratio/high_mean": 0.00083394997909636,
      "clip_ratio/low_mean": 0.0014543270044669043,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002288276948092971,
      "epoch": 3.466608340624089,
      "grad_norm": 0.2492295652627945,
      "learning_rate": 1e-06,
      "loss": -0.043,
      "step": 1484
    },
    {
      "clip_ratio/high_max": 0.002522495015000459,
      "clip_ratio/high_mean": 0.0011079816522396868,
      "clip_ratio/low_mean": 0.0007437406347889919,
      "clip_ratio/low_min": 3.380270391062368e-05,
      "clip_ratio/region_mean": 0.0018517222561058588,
      "completions/clipped_ratio": 0.1540178571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2589.0,
      "completions/mean_length": 1192.966552734375,
      "completions/mean_terminated_length": 664.4459228515625,
      "completions/min_length": 147.0,
      "completions/min_terminated_length": 147.0,
      "epoch": 3.468941382327209,
      "grad_norm": 0.3722512423992157,
      "learning_rate": 1e-06,
      "loss": -0.0637,
      "num_tokens": 217863585.0,
      "reward": 0.5680803656578064,
      "reward_std": 0.19704709947109222,
      "rewards/verify_math_reward/mean": 0.5680803656578064,
      "rewards/verify_math_reward/std": 0.4956200420856476,
      "step": 1485
    },
    {
      "clip_ratio/high_max": 0.0029757852462353185,
      "clip_ratio/high_mean": 0.0013127740094205365,
      "clip_ratio/low_mean": 0.001094865772756748,
      "clip_ratio/low_min": 0.00011356317554600537,
      "clip_ratio/region_mean": 0.0024076397821772844,
      "epoch": 3.4712744240303297,
      "grad_norm": 0.31583505868911743,
      "learning_rate": 1e-06,
      "loss": -0.064,
      "step": 1486
    },
    {
      "clip_ratio/high_max": 0.0030220092594390735,
      "clip_ratio/high_mean": 0.0013473187464114744,
      "clip_ratio/low_mean": 0.0012209015203552553,
      "clip_ratio/low_min": 0.0001388404725730652,
      "clip_ratio/region_mean": 0.00256822032679338,
      "epoch": 3.47360746573345,
      "grad_norm": 0.3032079339027405,
      "learning_rate": 1e-06,
      "loss": -0.0642,
      "step": 1487
    },
    {
      "clip_ratio/high_max": 0.002851707373338286,
      "clip_ratio/high_mean": 0.0012149710855737794,
      "clip_ratio/low_mean": 0.0015116343056433834,
      "clip_ratio/low_min": 0.00014014556109032128,
      "clip_ratio/region_mean": 0.0027266053948551416,
      "epoch": 3.4759405074365706,
      "grad_norm": 0.2756774127483368,
      "learning_rate": 1e-06,
      "loss": -0.0643,
      "step": 1488
    },
    {
      "clip_ratio/high_max": 0.0022743008157704026,
      "clip_ratio/high_mean": 0.0008011747177079087,
      "clip_ratio/low_mean": 0.000605744857693935,
      "clip_ratio/low_min": 1.3760458386968821e-05,
      "clip_ratio/region_mean": 0.0014069195785850752,
      "completions/clipped_ratio": 0.2064732142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3992.0,
      "completions/mean_length": 1348.6273193359375,
      "completions/mean_terminated_length": 633.7693481445312,
      "completions/min_length": 197.0,
      "completions/min_terminated_length": 197.0,
      "epoch": 3.478273549139691,
      "grad_norm": 0.40627923607826233,
      "learning_rate": 1e-06,
      "loss": -0.0719,
      "num_tokens": 218403563.0,
      "reward": 0.5479910969734192,
      "reward_std": 0.14673490822315216,
      "rewards/verify_math_reward/mean": 0.5479910969734192,
      "rewards/verify_math_reward/std": 0.49796950817108154,
      "step": 1489
    },
    {
      "clip_ratio/high_max": 0.003021533993887715,
      "clip_ratio/high_mean": 0.0011017732085747411,
      "clip_ratio/low_mean": 0.0009259931696306012,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002027766386163421,
      "epoch": 3.4806065908428114,
      "grad_norm": 0.3433954119682312,
      "learning_rate": 1e-06,
      "loss": -0.0722,
      "step": 1490
    },
    {
      "clip_ratio/high_max": 0.002794168824038934,
      "clip_ratio/high_mean": 0.0009876603235170478,
      "clip_ratio/low_mean": 0.0010355198619436123,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002023180182732176,
      "epoch": 3.4829396325459316,
      "grad_norm": 0.3085142970085144,
      "learning_rate": 1e-06,
      "loss": -0.0724,
      "step": 1491
    },
    {
      "clip_ratio/high_max": 0.0029426438704831526,
      "clip_ratio/high_mean": 0.000971582294369,
      "clip_ratio/low_mean": 0.0011516471972754516,
      "clip_ratio/low_min": 2.1566596842603758e-05,
      "clip_ratio/region_mean": 0.0021232294602668844,
      "epoch": 3.4852726742490523,
      "grad_norm": 0.25115662813186646,
      "learning_rate": 1e-06,
      "loss": -0.0724,
      "step": 1492
    },
    {
      "clip_ratio/high_max": 0.0029007212397118565,
      "clip_ratio/high_mean": 0.0010477241830813,
      "clip_ratio/low_mean": 0.0006721846602886217,
      "clip_ratio/low_min": 1.7053205738193356e-05,
      "clip_ratio/region_mean": 0.0017199088288180064,
      "completions/clipped_ratio": 0.1808035714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3476.0,
      "completions/mean_length": 1296.5703125,
      "completions/mean_terminated_length": 678.7125244140625,
      "completions/min_length": 173.0,
      "completions/min_terminated_length": 173.0,
      "epoch": 3.4876057159521725,
      "grad_norm": 0.3308228552341461,
      "learning_rate": 1e-06,
      "loss": -0.083,
      "num_tokens": 218999322.0,
      "reward": 0.5234375,
      "reward_std": 0.17002595961093903,
      "rewards/verify_math_reward/mean": 0.5234375,
      "rewards/verify_math_reward/std": 0.49972933530807495,
      "step": 1493
    },
    {
      "clip_ratio/high_max": 0.0030103138415142894,
      "clip_ratio/high_mean": 0.0012242803904882749,
      "clip_ratio/low_mean": 0.0009718649816932157,
      "clip_ratio/low_min": 1.2447719200281426e-05,
      "clip_ratio/region_mean": 0.0021961453821859322,
      "epoch": 3.489938757655293,
      "grad_norm": 0.3048311471939087,
      "learning_rate": 1e-06,
      "loss": -0.0832,
      "step": 1494
    },
    {
      "clip_ratio/high_max": 0.0033319630529149435,
      "clip_ratio/high_mean": 0.0012709472885035211,
      "clip_ratio/low_mean": 0.001170280753285624,
      "clip_ratio/low_min": 2.4895438400562853e-05,
      "clip_ratio/region_mean": 0.0024412280108663253,
      "epoch": 3.4922717993584134,
      "grad_norm": 0.26479119062423706,
      "learning_rate": 1e-06,
      "loss": -0.0834,
      "step": 1495
    },
    {
      "clip_ratio/high_max": 0.0031683956112829037,
      "clip_ratio/high_mean": 0.0011402380714571336,
      "clip_ratio/low_mean": 0.001393181777530117,
      "clip_ratio/low_min": 3.734315760084428e-05,
      "clip_ratio/region_mean": 0.0025334198217024095,
      "epoch": 3.494604841061534,
      "grad_norm": 0.25681331753730774,
      "learning_rate": 1e-06,
      "loss": -0.0835,
      "step": 1496
    },
    {
      "clip_ratio/high_max": 0.0023501041723648086,
      "clip_ratio/high_mean": 0.000847990909278451,
      "clip_ratio/low_mean": 0.00043196730780437065,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00127995821821969,
      "completions/clipped_ratio": 0.1986607142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3779.0,
      "completions/mean_length": 1313.32373046875,
      "completions/mean_terminated_length": 623.4679565429688,
      "completions/min_length": 154.0,
      "completions/min_terminated_length": 154.0,
      "epoch": 3.4969378827646542,
      "grad_norm": 0.23563139140605927,
      "learning_rate": 1e-06,
      "loss": -0.0622,
      "num_tokens": 219524948.0,
      "reward": 0.5345982313156128,
      "reward_std": 0.12831631302833557,
      "rewards/verify_math_reward/mean": 0.5345982313156128,
      "rewards/verify_math_reward/std": 0.4990801215171814,
      "step": 1497
    },
    {
      "clip_ratio/high_max": 0.0026061613607453182,
      "clip_ratio/high_mean": 0.0009985405467887176,
      "clip_ratio/low_mean": 0.0006235589321477164,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016220994584728032,
      "epoch": 3.499270924467775,
      "grad_norm": 0.23240315914154053,
      "learning_rate": 1e-06,
      "loss": -0.0623,
      "step": 1498
    },
    {
      "clip_ratio/high_max": 0.002646317963808542,
      "clip_ratio/high_mean": 0.0009712394812595448,
      "clip_ratio/low_mean": 0.0006868788418614713,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016581182972004171,
      "epoch": 3.5016039661708955,
      "grad_norm": 0.2139987349510193,
      "learning_rate": 1e-06,
      "loss": -0.0624,
      "step": 1499
    },
    {
      "clip_ratio/high_max": 0.0025528997921355767,
      "clip_ratio/high_mean": 0.0009007496255435399,
      "clip_ratio/low_mean": 0.0009149990187324875,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018157487211283296,
      "epoch": 3.5039370078740157,
      "grad_norm": 0.22334018349647522,
      "learning_rate": 1e-06,
      "loss": -0.0625,
      "step": 1500
    },
    {
      "clip_ratio/high_max": 0.0022797048441134393,
      "clip_ratio/high_mean": 0.0009540160008327803,
      "clip_ratio/low_mean": 0.0005550076903091394,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015090236738615204,
      "completions/clipped_ratio": 0.1841517857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4064.0,
      "completions/mean_length": 1354.9676513671875,
      "completions/mean_terminated_length": 736.2667846679688,
      "completions/min_length": 181.0,
      "completions/min_terminated_length": 181.0,
      "epoch": 3.506270049577136,
      "grad_norm": 0.3129500448703766,
      "learning_rate": 1e-06,
      "loss": -0.0965,
      "num_tokens": 220151575.0,
      "reward": 0.5792410969734192,
      "reward_std": 0.17836888134479523,
      "rewards/verify_math_reward/mean": 0.5792410969734192,
      "rewards/verify_math_reward/std": 0.49395665526390076,
      "step": 1501
    },
    {
      "clip_ratio/high_max": 0.0026659188079065643,
      "clip_ratio/high_mean": 0.0010767554540507263,
      "clip_ratio/low_mean": 0.0008429652934864862,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019197206638636999,
      "epoch": 3.5086030912802566,
      "grad_norm": 0.2578331232070923,
      "learning_rate": 1e-06,
      "loss": -0.0969,
      "step": 1502
    },
    {
      "clip_ratio/high_max": 0.0027675622841343284,
      "clip_ratio/high_mean": 0.0011148411813337589,
      "clip_ratio/low_mean": 0.0009087679827644024,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020236091877450235,
      "epoch": 3.5109361329833773,
      "grad_norm": 0.24845904111862183,
      "learning_rate": 1e-06,
      "loss": -0.097,
      "step": 1503
    },
    {
      "clip_ratio/high_max": 0.002540670393500477,
      "clip_ratio/high_mean": 0.001044667356836726,
      "clip_ratio/low_mean": 0.0011574263771763071,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002202093724918086,
      "epoch": 3.5132691746864975,
      "grad_norm": 0.292165607213974,
      "learning_rate": 1e-06,
      "loss": -0.0971,
      "step": 1504
    },
    {
      "clip_ratio/high_max": 0.0026543206186033785,
      "clip_ratio/high_mean": 0.000993739169643959,
      "clip_ratio/low_mean": 0.0007413307284878101,
      "clip_ratio/low_min": 4.1377028537681326e-05,
      "clip_ratio/region_mean": 0.001735069883579854,
      "completions/clipped_ratio": 0.1964285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4063.0,
      "completions/mean_length": 1401.6429443359375,
      "completions/mean_terminated_length": 743.022216796875,
      "completions/min_length": 130.0,
      "completions/min_terminated_length": 130.0,
      "epoch": 3.515602216389618,
      "grad_norm": 0.35746651887893677,
      "learning_rate": 1e-06,
      "loss": -0.0986,
      "num_tokens": 220782847.0,
      "reward": 0.5323660969734192,
      "reward_std": 0.15777993202209473,
      "rewards/verify_math_reward/mean": 0.5323660969734192,
      "rewards/verify_math_reward/std": 0.4992299973964691,
      "step": 1505
    },
    {
      "clip_ratio/high_max": 0.0028942258359165862,
      "clip_ratio/high_mean": 0.0011398786791687598,
      "clip_ratio/low_mean": 0.0008631971413706196,
      "clip_ratio/low_min": 1.7467858924646862e-05,
      "clip_ratio/region_mean": 0.0020030757877975702,
      "epoch": 3.5179352580927383,
      "grad_norm": 0.3008044958114624,
      "learning_rate": 1e-06,
      "loss": -0.0988,
      "step": 1506
    },
    {
      "clip_ratio/high_max": 0.003281142056948738,
      "clip_ratio/high_mean": 0.0011600244151850347,
      "clip_ratio/low_mean": 0.0010390899451522273,
      "clip_ratio/low_min": 1.840942604758311e-05,
      "clip_ratio/region_mean": 0.002199114329414442,
      "epoch": 3.520268299795859,
      "grad_norm": 0.2596571147441864,
      "learning_rate": 1e-06,
      "loss": -0.099,
      "step": 1507
    },
    {
      "clip_ratio/high_max": 0.002689667609956814,
      "clip_ratio/high_mean": 0.0010504342744752648,
      "clip_ratio/low_mean": 0.0013327549550012918,
      "clip_ratio/low_min": 3.4935717849293724e-05,
      "clip_ratio/region_mean": 0.0023831892031012103,
      "epoch": 3.522601341498979,
      "grad_norm": 0.21419784426689148,
      "learning_rate": 1e-06,
      "loss": -0.0991,
      "step": 1508
    },
    {
      "clip_ratio/high_max": 0.002025470144872088,
      "clip_ratio/high_mean": 0.000934464811507496,
      "clip_ratio/low_mean": 0.0007760484650134458,
      "clip_ratio/low_min": 1.0796337846841197e-05,
      "clip_ratio/region_mean": 0.0017105132646975107,
      "completions/clipped_ratio": 0.1640625,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4023.0,
      "completions/mean_length": 1251.680908203125,
      "completions/mean_terminated_length": 693.449951171875,
      "completions/min_length": 179.0,
      "completions/min_terminated_length": 179.0,
      "epoch": 3.5249343832021,
      "grad_norm": 0.3148497939109802,
      "learning_rate": 1e-06,
      "loss": -0.0672,
      "num_tokens": 221397097.0,
      "reward": 0.566964328289032,
      "reward_std": 0.1840411275625229,
      "rewards/verify_math_reward/mean": 0.5669642686843872,
      "rewards/verify_math_reward/std": 0.49577224254608154,
      "step": 1509
    },
    {
      "clip_ratio/high_max": 0.002676955111382995,
      "clip_ratio/high_mean": 0.0011807695045717992,
      "clip_ratio/low_mean": 0.0010588440036372049,
      "clip_ratio/low_min": 1.955977495526895e-05,
      "clip_ratio/region_mean": 0.002239613422716502,
      "epoch": 3.52726742490522,
      "grad_norm": 0.3090699017047882,
      "learning_rate": 1e-06,
      "loss": -0.0675,
      "step": 1510
    },
    {
      "clip_ratio/high_max": 0.0025940429040929303,
      "clip_ratio/high_mean": 0.001103127473470522,
      "clip_ratio/low_mean": 0.0012223327503306791,
      "clip_ratio/low_min": 1.955977495526895e-05,
      "clip_ratio/region_mean": 0.0023254601765074767,
      "epoch": 3.5296004666083407,
      "grad_norm": 0.2467300146818161,
      "learning_rate": 1e-06,
      "loss": -0.0676,
      "step": 1511
    },
    {
      "clip_ratio/high_max": 0.002437049952277448,
      "clip_ratio/high_mean": 0.0010338622214476345,
      "clip_ratio/low_mean": 0.0014347818942042068,
      "clip_ratio/low_min": 1.3039849363849498e-05,
      "clip_ratio/region_mean": 0.0024686441174708307,
      "epoch": 3.531933508311461,
      "grad_norm": 0.2955998182296753,
      "learning_rate": 1e-06,
      "loss": -0.0677,
      "step": 1512
    },
    {
      "clip_ratio/high_max": 0.0023475160778616555,
      "clip_ratio/high_mean": 0.000987039549727342,
      "clip_ratio/low_mean": 0.000704028707332327,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016910682679736055,
      "completions/clipped_ratio": 0.1707589285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3899.0,
      "completions/mean_length": 1262.3326416015625,
      "completions/mean_terminated_length": 678.8182983398438,
      "completions/min_length": 170.0,
      "completions/min_terminated_length": 170.0,
      "epoch": 3.5342665500145816,
      "grad_norm": 0.35950562357902527,
      "learning_rate": 1e-06,
      "loss": -0.1186,
      "num_tokens": 221986099.0,
      "reward": 0.598214328289032,
      "reward_std": 0.16904954612255096,
      "rewards/verify_math_reward/mean": 0.5982142686843872,
      "rewards/verify_math_reward/std": 0.49053287506103516,
      "step": 1513
    },
    {
      "clip_ratio/high_max": 0.0035973012636532076,
      "clip_ratio/high_mean": 0.001298691571719246,
      "clip_ratio/low_mean": 0.0008401642062381143,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021388557652244344,
      "epoch": 3.536599591717702,
      "grad_norm": 0.27253785729408264,
      "learning_rate": 1e-06,
      "loss": -0.1188,
      "step": 1514
    },
    {
      "clip_ratio/high_max": 0.0032188785189646296,
      "clip_ratio/high_mean": 0.0012338145006651757,
      "clip_ratio/low_mean": 0.0010743178299890133,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002308132352482062,
      "epoch": 3.5389326334208224,
      "grad_norm": 0.22632579505443573,
      "learning_rate": 1e-06,
      "loss": -0.119,
      "step": 1515
    },
    {
      "clip_ratio/high_max": 0.00277203215227928,
      "clip_ratio/high_mean": 0.0011150104546686634,
      "clip_ratio/low_mean": 0.0012494994443841279,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023645098917768337,
      "epoch": 3.5412656751239426,
      "grad_norm": 0.233441561460495,
      "learning_rate": 1e-06,
      "loss": -0.119,
      "step": 1516
    },
    {
      "clip_ratio/high_max": 0.0033738151541911066,
      "clip_ratio/high_mean": 0.0010480080181878293,
      "clip_ratio/low_mean": 0.000549867203062604,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015978752744558733,
      "completions/clipped_ratio": 0.1964285714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3058.0,
      "completions/mean_length": 1312.8984375,
      "completions/mean_terminated_length": 632.584716796875,
      "completions/min_length": 123.0,
      "completions/min_terminated_length": 123.0,
      "epoch": 3.5435987168270633,
      "grad_norm": 0.3846149146556854,
      "learning_rate": 1e-06,
      "loss": -0.0588,
      "num_tokens": 222541640.0,
      "reward": 0.53125,
      "reward_std": 0.13775329291820526,
      "rewards/verify_math_reward/mean": 0.53125,
      "rewards/verify_math_reward/std": 0.4993011951446533,
      "step": 1517
    },
    {
      "clip_ratio/high_max": 0.003223258572688792,
      "clip_ratio/high_mean": 0.0011083738590969006,
      "clip_ratio/low_mean": 0.0009958546870620921,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021042284861323424,
      "epoch": 3.545931758530184,
      "grad_norm": 0.3291001319885254,
      "learning_rate": 1e-06,
      "loss": -0.059,
      "step": 1518
    },
    {
      "clip_ratio/high_max": 0.0034833777172025293,
      "clip_ratio/high_mean": 0.0011976627065450884,
      "clip_ratio/low_mean": 0.001046523768309271,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022441865148721263,
      "epoch": 3.548264800233304,
      "grad_norm": 0.4383046329021454,
      "learning_rate": 1e-06,
      "loss": -0.0591,
      "step": 1519
    },
    {
      "clip_ratio/high_max": 0.0036275723250582814,
      "clip_ratio/high_mean": 0.0011557021680346224,
      "clip_ratio/low_mean": 0.0012577502566273324,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002413452424661955,
      "epoch": 3.5505978419364244,
      "grad_norm": 0.2779461145401001,
      "learning_rate": 1e-06,
      "loss": -0.0592,
      "step": 1520
    },
    {
      "clip_ratio/high_max": 0.0026107412777491845,
      "clip_ratio/high_mean": 0.0011020833007933106,
      "clip_ratio/low_mean": 0.0007288959131983574,
      "clip_ratio/low_min": 1.5644554878235795e-05,
      "clip_ratio/region_mean": 0.00183097917761188,
      "completions/clipped_ratio": 0.1819196428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3292.0,
      "completions/mean_length": 1295.583740234375,
      "completions/mean_terminated_length": 672.844482421875,
      "completions/min_length": 170.0,
      "completions/min_terminated_length": 170.0,
      "epoch": 3.552930883639545,
      "grad_norm": 0.35715851187705994,
      "learning_rate": 1e-06,
      "loss": -0.1101,
      "num_tokens": 223118067.0,
      "reward": 0.5714285969734192,
      "reward_std": 0.20850077271461487,
      "rewards/verify_math_reward/mean": 0.5714285969734192,
      "rewards/verify_math_reward/std": 0.49514803290367126,
      "step": 1521
    },
    {
      "clip_ratio/high_max": 0.0031851034073042683,
      "clip_ratio/high_mean": 0.0013549956274800934,
      "clip_ratio/low_mean": 0.0009374552846566075,
      "clip_ratio/low_min": 3.128910975647159e-05,
      "clip_ratio/region_mean": 0.0022924509103177115,
      "epoch": 3.5552639253426657,
      "grad_norm": 0.30691832304000854,
      "learning_rate": 1e-06,
      "loss": -0.1104,
      "step": 1522
    },
    {
      "clip_ratio/high_max": 0.003452163393376395,
      "clip_ratio/high_mean": 0.0013758698623860255,
      "clip_ratio/low_mean": 0.0010093992796100792,
      "clip_ratio/low_min": 1.7370761270285584e-05,
      "clip_ratio/region_mean": 0.002385269144724589,
      "epoch": 3.557596967045786,
      "grad_norm": 0.2864389717578888,
      "learning_rate": 1e-06,
      "loss": -0.1106,
      "step": 1523
    },
    {
      "clip_ratio/high_max": 0.003090730773692485,
      "clip_ratio/high_mean": 0.0013269843148009386,
      "clip_ratio/low_mean": 0.0013503560730896425,
      "clip_ratio/low_min": 3.098661545664072e-05,
      "clip_ratio/region_mean": 0.002677340409718454,
      "epoch": 3.5599300087489065,
      "grad_norm": 0.25327908992767334,
      "learning_rate": 1e-06,
      "loss": -0.1107,
      "step": 1524
    },
    {
      "clip_ratio/high_max": 0.0025775858885026537,
      "clip_ratio/high_mean": 0.000957845215452835,
      "clip_ratio/low_mean": 0.0007984565493188711,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017563017754582688,
      "completions/clipped_ratio": 0.1685267857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3018.0,
      "completions/mean_length": 1227.01904296875,
      "completions/mean_terminated_length": 645.5208129882812,
      "completions/min_length": 158.0,
      "completions/min_terminated_length": 158.0,
      "epoch": 3.5622630504520267,
      "grad_norm": 0.4949190318584442,
      "learning_rate": 1e-06,
      "loss": -0.0815,
      "num_tokens": 223689812.0,
      "reward": 0.5959821939468384,
      "reward_std": 0.1830640286207199,
      "rewards/verify_math_reward/mean": 0.5959821343421936,
      "rewards/verify_math_reward/std": 0.490975022315979,
      "step": 1525
    },
    {
      "clip_ratio/high_max": 0.0029955473728477955,
      "clip_ratio/high_mean": 0.0011499087486299686,
      "clip_ratio/low_mean": 0.0010720030668380787,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022219118181965314,
      "epoch": 3.5645960921551474,
      "grad_norm": 0.28687843680381775,
      "learning_rate": 1e-06,
      "loss": -0.0817,
      "step": 1526
    },
    {
      "clip_ratio/high_max": 0.0034787955810315907,
      "clip_ratio/high_mean": 0.0011736647647921927,
      "clip_ratio/low_mean": 0.0011168585679115495,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002290523327246774,
      "epoch": 3.5669291338582676,
      "grad_norm": 0.25637078285217285,
      "learning_rate": 1e-06,
      "loss": -0.0819,
      "step": 1527
    },
    {
      "clip_ratio/high_max": 0.00323119483073242,
      "clip_ratio/high_mean": 0.0011336969037074596,
      "clip_ratio/low_mean": 0.0014666480810774374,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002600344938400667,
      "epoch": 3.5692621755613883,
      "grad_norm": 0.32233622670173645,
      "learning_rate": 1e-06,
      "loss": -0.0821,
      "step": 1528
    },
    {
      "clip_ratio/high_max": 0.002661401405930519,
      "clip_ratio/high_mean": 0.0010299417026544688,
      "clip_ratio/low_mean": 0.0004914766668662196,
      "clip_ratio/low_min": 1.4328289580589626e-05,
      "clip_ratio/region_mean": 0.0015214183513307944,
      "completions/clipped_ratio": 0.1819196428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3200.0,
      "completions/mean_length": 1296.469970703125,
      "completions/mean_terminated_length": 673.9276733398438,
      "completions/min_length": 142.0,
      "completions/min_terminated_length": 142.0,
      "epoch": 3.5715952172645085,
      "grad_norm": 0.3157018721103668,
      "learning_rate": 1e-06,
      "loss": -0.0676,
      "num_tokens": 224279809.0,
      "reward": 0.5491071939468384,
      "reward_std": 0.16270402073860168,
      "rewards/verify_math_reward/mean": 0.5491071343421936,
      "rewards/verify_math_reward/std": 0.49786055088043213,
      "step": 1529
    },
    {
      "clip_ratio/high_max": 0.0030244411318562925,
      "clip_ratio/high_mean": 0.0011270798877376365,
      "clip_ratio/low_mean": 0.0007608474184053193,
      "clip_ratio/low_min": 1.735147088766098e-05,
      "clip_ratio/region_mean": 0.0018879273447964806,
      "epoch": 3.573928258967629,
      "grad_norm": 0.28625938296318054,
      "learning_rate": 1e-06,
      "loss": -0.0679,
      "step": 1530
    },
    {
      "clip_ratio/high_max": 0.0030160953974700533,
      "clip_ratio/high_mean": 0.0011641476585282362,
      "clip_ratio/low_mean": 0.0009189475094899535,
      "clip_ratio/low_min": 3.470294177532196e-05,
      "clip_ratio/region_mean": 0.0020830951689276844,
      "epoch": 3.5762613006707493,
      "grad_norm": 0.23650334775447845,
      "learning_rate": 1e-06,
      "loss": -0.068,
      "step": 1531
    },
    {
      "clip_ratio/high_max": 0.0031486430452787317,
      "clip_ratio/high_mean": 0.0011302901166345691,
      "clip_ratio/low_mean": 0.0010223674125882098,
      "clip_ratio/low_min": 8.596973930252716e-05,
      "clip_ratio/region_mean": 0.002152657529222779,
      "epoch": 3.57859434237387,
      "grad_norm": 0.21814867854118347,
      "learning_rate": 1e-06,
      "loss": -0.0681,
      "step": 1532
    },
    {
      "clip_ratio/high_max": 0.0026995788648491725,
      "clip_ratio/high_mean": 0.0010108518345077755,
      "clip_ratio/low_mean": 0.0006342021315504098,
      "clip_ratio/low_min": 2.504006442904938e-05,
      "clip_ratio/region_mean": 0.0016450539478682913,
      "completions/clipped_ratio": 0.1830357142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2487.0,
      "completions/mean_length": 1274.7288818359375,
      "completions/mean_terminated_length": 642.6406860351562,
      "completions/min_length": 141.0,
      "completions/min_terminated_length": 141.0,
      "epoch": 3.5809273840769906,
      "grad_norm": 0.42358139157295227,
      "learning_rate": 1e-06,
      "loss": -0.0879,
      "num_tokens": 224843742.0,
      "reward": 0.5881696939468384,
      "reward_std": 0.16555650532245636,
      "rewards/verify_math_reward/mean": 0.5881696343421936,
      "rewards/verify_math_reward/std": 0.4924395978450775,
      "step": 1533
    },
    {
      "clip_ratio/high_max": 0.0031136035977397114,
      "clip_ratio/high_mean": 0.0012501615965447854,
      "clip_ratio/low_mean": 0.0009599393051757943,
      "clip_ratio/low_min": 1.252003221452469e-05,
      "clip_ratio/region_mean": 0.0022101009089965373,
      "epoch": 3.583260425780111,
      "grad_norm": 0.31278711557388306,
      "learning_rate": 1e-06,
      "loss": -0.0883,
      "step": 1534
    },
    {
      "clip_ratio/high_max": 0.0035869555504177697,
      "clip_ratio/high_mean": 0.001279306146898307,
      "clip_ratio/low_mean": 0.0011774325430451427,
      "clip_ratio/low_min": 5.008012885809876e-05,
      "clip_ratio/region_mean": 0.0024567386863054708,
      "epoch": 3.585593467483231,
      "grad_norm": 0.26393356919288635,
      "learning_rate": 1e-06,
      "loss": -0.0886,
      "step": 1535
    },
    {
      "clip_ratio/high_max": 0.00319365114410175,
      "clip_ratio/high_mean": 0.0011895089846802875,
      "clip_ratio/low_mean": 0.0013025746593484655,
      "clip_ratio/low_min": 3.052503234357573e-05,
      "clip_ratio/region_mean": 0.0024920836149249226,
      "epoch": 3.5879265091863517,
      "grad_norm": 0.27731215953826904,
      "learning_rate": 1e-06,
      "loss": -0.0885,
      "step": 1536
    },
    {
      "clip_ratio/high_max": 0.003128183823719155,
      "clip_ratio/high_mean": 0.0011135051026940346,
      "clip_ratio/low_mean": 0.0005207825106481323,
      "clip_ratio/low_min": 1.7472742911195382e-05,
      "clip_ratio/region_mean": 0.0016342875969712622,
      "completions/clipped_ratio": 0.1819196428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3815.0,
      "completions/mean_length": 1286.3895263671875,
      "completions/mean_terminated_length": 661.605712890625,
      "completions/min_length": 166.0,
      "completions/min_terminated_length": 166.0,
      "epoch": 3.5902595508894724,
      "grad_norm": 0.3832108676433563,
      "learning_rate": 1e-06,
      "loss": -0.0734,
      "num_tokens": 225421971.0,
      "reward": 0.5301339626312256,
      "reward_std": 0.17659901082515717,
      "rewards/verify_math_reward/mean": 0.5301339030265808,
      "rewards/verify_math_reward/std": 0.49936985969543457,
      "step": 1537
    },
    {
      "clip_ratio/high_max": 0.003468630740826484,
      "clip_ratio/high_mean": 0.0012430049791873898,
      "clip_ratio/low_mean": 0.0006966114424358238,
      "clip_ratio/low_min": 2.518130531825591e-05,
      "clip_ratio/region_mean": 0.0019396164425415918,
      "epoch": 3.5925925925925926,
      "grad_norm": 0.28782737255096436,
      "learning_rate": 1e-06,
      "loss": -0.0736,
      "step": 1538
    },
    {
      "clip_ratio/high_max": 0.00373454752843827,
      "clip_ratio/high_mean": 0.0013299306010594591,
      "clip_ratio/low_mean": 0.000834752927403315,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0021646835084538907,
      "epoch": 3.5949256342957128,
      "grad_norm": 0.2644355893135071,
      "learning_rate": 1e-06,
      "loss": -0.0738,
      "step": 1539
    },
    {
      "clip_ratio/high_max": 0.0035358406603336334,
      "clip_ratio/high_mean": 0.0012311937171034515,
      "clip_ratio/low_mean": 0.0010186804520344594,
      "clip_ratio/low_min": 4.148689185967669e-05,
      "clip_ratio/region_mean": 0.002249874160042964,
      "epoch": 3.5972586759988334,
      "grad_norm": 0.29489627480506897,
      "learning_rate": 1e-06,
      "loss": -0.0739,
      "step": 1540
    },
    {
      "clip_ratio/high_max": 0.003314832487376407,
      "clip_ratio/high_mean": 0.0010940357988147298,
      "clip_ratio/low_mean": 0.0006898802121213521,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017839160209405236,
      "completions/clipped_ratio": 0.2299107142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2849.0,
      "completions/mean_length": 1440.69091796875,
      "completions/mean_terminated_length": 647.9464111328125,
      "completions/min_length": 177.0,
      "completions/min_terminated_length": 177.0,
      "epoch": 3.599591717701954,
      "grad_norm": 0.39426520466804504,
      "learning_rate": 1e-06,
      "loss": -0.1001,
      "num_tokens": 225952342.0,
      "reward": 0.5412946939468384,
      "reward_std": 0.1713072657585144,
      "rewards/verify_math_reward/mean": 0.5412946343421936,
      "rewards/verify_math_reward/std": 0.49857014417648315,
      "step": 1541
    },
    {
      "clip_ratio/high_max": 0.0036451994528761134,
      "clip_ratio/high_mean": 0.0013332387898117304,
      "clip_ratio/low_mean": 0.0009492857570876367,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002282524590555113,
      "epoch": 3.6019247594050743,
      "grad_norm": 0.46961215138435364,
      "learning_rate": 1e-06,
      "loss": -0.1004,
      "step": 1542
    },
    {
      "clip_ratio/high_max": 0.003487508642137982,
      "clip_ratio/high_mean": 0.0013010401034989627,
      "clip_ratio/low_mean": 0.0011625834558799397,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002463623481162358,
      "epoch": 3.604257801108195,
      "grad_norm": 0.26113542914390564,
      "learning_rate": 1e-06,
      "loss": -0.1006,
      "step": 1543
    },
    {
      "clip_ratio/high_max": 0.003831864560197573,
      "clip_ratio/high_mean": 0.0012798725147149526,
      "clip_ratio/low_mean": 0.0014721531297254842,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0027520255971467122,
      "epoch": 3.606590842811315,
      "grad_norm": 0.2875174283981323,
      "learning_rate": 1e-06,
      "loss": -0.1007,
      "step": 1544
    },
    {
      "clip_ratio/high_max": 0.003023206598300021,
      "clip_ratio/high_mean": 0.001271988690859871,
      "clip_ratio/low_mean": 0.0006859922905277926,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019579810250434093,
      "completions/clipped_ratio": 0.2198660714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2363.0,
      "completions/mean_length": 1441.9923095703125,
      "completions/mean_terminated_length": 694.010009765625,
      "completions/min_length": 147.0,
      "completions/min_terminated_length": 147.0,
      "epoch": 3.608923884514436,
      "grad_norm": 0.4443901777267456,
      "learning_rate": 1e-06,
      "loss": -0.1147,
      "num_tokens": 226527839.0,
      "reward": 0.5691964626312256,
      "reward_std": 0.20523346960544586,
      "rewards/verify_math_reward/mean": 0.5691964030265808,
      "rewards/verify_math_reward/std": 0.4954652488231659,
      "step": 1545
    },
    {
      "clip_ratio/high_max": 0.0034600492435856722,
      "clip_ratio/high_mean": 0.0014443448417296167,
      "clip_ratio/low_mean": 0.0009484658476139884,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023928107184474356,
      "epoch": 3.611256926217556,
      "grad_norm": 0.27703621983528137,
      "learning_rate": 1e-06,
      "loss": -0.115,
      "step": 1546
    },
    {
      "clip_ratio/high_max": 0.003560012446541805,
      "clip_ratio/high_mean": 0.0013946929029771127,
      "clip_ratio/low_mean": 0.0011226687602174934,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0025173616813845,
      "epoch": 3.6135899679206767,
      "grad_norm": 0.32288092374801636,
      "learning_rate": 1e-06,
      "loss": -0.1152,
      "step": 1547
    },
    {
      "clip_ratio/high_max": 0.003413889651710633,
      "clip_ratio/high_mean": 0.0014856893867545296,
      "clip_ratio/low_mean": 0.0013747827833867632,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002860472144675441,
      "epoch": 3.615923009623797,
      "grad_norm": 0.30773279070854187,
      "learning_rate": 1e-06,
      "loss": -0.1154,
      "step": 1548
    },
    {
      "clip_ratio/high_max": 0.0021399538090918213,
      "clip_ratio/high_mean": 0.0008620997614343651,
      "clip_ratio/low_mean": 0.0005568014967138879,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0014189012617862318,
      "completions/clipped_ratio": 0.2477678571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2899.0,
      "completions/mean_length": 1520.0301513671875,
      "completions/mean_terminated_length": 671.5653076171875,
      "completions/min_length": 182.0,
      "completions/min_terminated_length": 182.0,
      "epoch": 3.6182560513269175,
      "grad_norm": 0.38228416442871094,
      "learning_rate": 1e-06,
      "loss": -0.0771,
      "num_tokens": 227069810.0,
      "reward": 0.5189732313156128,
      "reward_std": 0.17058978974819183,
      "rewards/verify_math_reward/mean": 0.5189732313156128,
      "rewards/verify_math_reward/std": 0.49991893768310547,
      "step": 1549
    },
    {
      "clip_ratio/high_max": 0.0030548020695277955,
      "clip_ratio/high_mean": 0.001068041374310269,
      "clip_ratio/low_mean": 0.0009025194267451297,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001970560821064282,
      "epoch": 3.6205890930300377,
      "grad_norm": 0.32157832384109497,
      "learning_rate": 1e-06,
      "loss": -0.0774,
      "step": 1550
    },
    {
      "clip_ratio/high_max": 0.002751321553660091,
      "clip_ratio/high_mean": 0.0010448119501234032,
      "clip_ratio/low_mean": 0.001041971154336352,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020867831117357127,
      "epoch": 3.6229221347331584,
      "grad_norm": 0.24815063178539276,
      "learning_rate": 1e-06,
      "loss": -0.0776,
      "step": 1551
    },
    {
      "clip_ratio/high_max": 0.0026036685085273348,
      "clip_ratio/high_mean": 0.0009824264798226068,
      "clip_ratio/low_mean": 0.0012956580212630797,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022780845465604216,
      "epoch": 3.625255176436279,
      "grad_norm": 0.3354860544204712,
      "learning_rate": 1e-06,
      "loss": -0.0776,
      "step": 1552
    },
    {
      "clip_ratio/high_max": 0.0030948232670198195,
      "clip_ratio/high_mean": 0.001258082458662102,
      "clip_ratio/low_mean": 0.0007376344910881016,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019957169497502036,
      "completions/clipped_ratio": 0.1662946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2397.0,
      "completions/mean_length": 1174.458740234375,
      "completions/mean_terminated_length": 591.71484375,
      "completions/min_length": 143.0,
      "completions/min_terminated_length": 143.0,
      "epoch": 3.6275882181393992,
      "grad_norm": 0.37742355465888977,
      "learning_rate": 1e-06,
      "loss": -0.0925,
      "num_tokens": 227601141.0,
      "reward": 0.606026828289032,
      "reward_std": 0.15725818276405334,
      "rewards/verify_math_reward/mean": 0.6060267686843872,
      "rewards/verify_math_reward/std": 0.48890194296836853,
      "step": 1553
    },
    {
      "clip_ratio/high_max": 0.003571727851522155,
      "clip_ratio/high_mean": 0.0013547524649766274,
      "clip_ratio/low_mean": 0.0009574865052854875,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002312238997546956,
      "epoch": 3.6299212598425195,
      "grad_norm": 0.364230751991272,
      "learning_rate": 1e-06,
      "loss": -0.0927,
      "step": 1554
    },
    {
      "clip_ratio/high_max": 0.003034077410120517,
      "clip_ratio/high_mean": 0.0012617486208910123,
      "clip_ratio/low_mean": 0.0011363514995537116,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0023981001213542186,
      "epoch": 3.63225430154564,
      "grad_norm": 0.3032190203666687,
      "learning_rate": 1e-06,
      "loss": -0.093,
      "step": 1555
    },
    {
      "clip_ratio/high_max": 0.0028370428626658395,
      "clip_ratio/high_mean": 0.001281575325265294,
      "clip_ratio/low_mean": 0.0013283705375215504,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002609945928270463,
      "epoch": 3.6345873432487608,
      "grad_norm": 0.27822694182395935,
      "learning_rate": 1e-06,
      "loss": -0.0931,
      "step": 1556
    },
    {
      "clip_ratio/high_max": 0.002680474444787251,
      "clip_ratio/high_mean": 0.0010216867940471275,
      "clip_ratio/low_mean": 0.0005964219117231551,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016181086975848302,
      "completions/clipped_ratio": 0.2142857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3919.0,
      "completions/mean_length": 1436.2857666015625,
      "completions/mean_terminated_length": 710.9091186523438,
      "completions/min_length": 135.0,
      "completions/min_terminated_length": 135.0,
      "epoch": 3.636920384951881,
      "grad_norm": 0.3404932916164398,
      "learning_rate": 1e-06,
      "loss": -0.0837,
      "num_tokens": 228181853.0,
      "reward": 0.4988839626312256,
      "reward_std": 0.18201345205307007,
      "rewards/verify_math_reward/mean": 0.4988839328289032,
      "rewards/verify_math_reward/std": 0.5002779960632324,
      "step": 1557
    },
    {
      "clip_ratio/high_max": 0.0031010887505544815,
      "clip_ratio/high_mean": 0.00122041084432567,
      "clip_ratio/low_mean": 0.0008648375960547128,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002085248452203814,
      "epoch": 3.6392534266550016,
      "grad_norm": 0.31997114419937134,
      "learning_rate": 1e-06,
      "loss": -0.0839,
      "step": 1558
    },
    {
      "clip_ratio/high_max": 0.003221661929273978,
      "clip_ratio/high_mean": 0.001245005041710101,
      "clip_ratio/low_mean": 0.001028126946039265,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0022731319768354297,
      "epoch": 3.641586468358122,
      "grad_norm": 0.23597189784049988,
      "learning_rate": 1e-06,
      "loss": -0.0841,
      "step": 1559
    },
    {
      "clip_ratio/high_max": 0.0029438552883220837,
      "clip_ratio/high_mean": 0.0011887395639860188,
      "clip_ratio/low_mean": 0.0012269534418010153,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002415692993963603,
      "epoch": 3.6439195100612425,
      "grad_norm": 0.2510083317756653,
      "learning_rate": 1e-06,
      "loss": -0.0842,
      "step": 1560
    },
    {
      "clip_ratio/high_max": 0.0028251345502212644,
      "clip_ratio/high_mean": 0.0012001818868156988,
      "clip_ratio/low_mean": 0.0006806998262618436,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0018808816967066377,
      "completions/clipped_ratio": 0.2142857142857143,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3896.0,
      "completions/mean_length": 1424.5101318359375,
      "completions/mean_terminated_length": 695.921875,
      "completions/min_length": 153.0,
      "completions/min_terminated_length": 153.0,
      "epoch": 3.6462525517643627,
      "grad_norm": 0.3832797408103943,
      "learning_rate": 1e-06,
      "loss": -0.1433,
      "num_tokens": 228758326.0,
      "reward": 0.574776828289032,
      "reward_std": 0.2134273499250412,
      "rewards/verify_math_reward/mean": 0.5747767686843872,
      "rewards/verify_math_reward/std": 0.49465295672416687,
      "step": 1561
    },
    {
      "clip_ratio/high_max": 0.0038279997825156897,
      "clip_ratio/high_mean": 0.001497450426541036,
      "clip_ratio/low_mean": 0.0009243143595085712,
      "clip_ratio/low_min": 1.1888910194102209e-05,
      "clip_ratio/region_mean": 0.0024217647878685966,
      "epoch": 3.6485855934674833,
      "grad_norm": 0.33032625913619995,
      "learning_rate": 1e-06,
      "loss": -0.1436,
      "step": 1562
    },
    {
      "clip_ratio/high_max": 0.0033771845774026588,
      "clip_ratio/high_mean": 0.001481091407185886,
      "clip_ratio/low_mean": 0.001131790140789235,
      "clip_ratio/low_min": 7.760118933219928e-06,
      "clip_ratio/region_mean": 0.0026128815370611846,
      "epoch": 3.6509186351706036,
      "grad_norm": 0.41828563809394836,
      "learning_rate": 1e-06,
      "loss": -0.1437,
      "step": 1563
    },
    {
      "clip_ratio/high_max": 0.003442059605731629,
      "clip_ratio/high_mean": 0.0013597105171356816,
      "clip_ratio/low_mean": 0.0012348913332971279,
      "clip_ratio/low_min": 2.3280357709154487e-05,
      "clip_ratio/region_mean": 0.002594601915916428,
      "epoch": 3.653251676873724,
      "grad_norm": 0.2683136463165283,
      "learning_rate": 1e-06,
      "loss": -0.144,
      "step": 1564
    },
    {
      "clip_ratio/high_max": 0.003076203509408515,
      "clip_ratio/high_mean": 0.001119939750424237,
      "clip_ratio/low_mean": 0.0006350224448397057,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001754962228005752,
      "completions/clipped_ratio": 0.2433035714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3473.0,
      "completions/mean_length": 1496.946533203125,
      "completions/mean_terminated_length": 661.2625122070312,
      "completions/min_length": 163.0,
      "completions/min_terminated_length": 163.0,
      "epoch": 3.6555847185768444,
      "grad_norm": 0.3296455144882202,
      "learning_rate": 1e-06,
      "loss": -0.069,
      "num_tokens": 229305646.0,
      "reward": 0.4843750298023224,
      "reward_std": 0.14563976228237152,
      "rewards/verify_math_reward/mean": 0.484375,
      "rewards/verify_math_reward/std": 0.5000349283218384,
      "step": 1565
    },
    {
      "clip_ratio/high_max": 0.00340685666014906,
      "clip_ratio/high_mean": 0.0012808764458895894,
      "clip_ratio/low_mean": 0.0007630479349245434,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020439243817236274,
      "epoch": 3.657917760279965,
      "grad_norm": 0.3247814178466797,
      "learning_rate": 1e-06,
      "loss": -0.0693,
      "step": 1566
    },
    {
      "clip_ratio/high_max": 0.004076259312569164,
      "clip_ratio/high_mean": 0.001424640493496554,
      "clip_ratio/low_mean": 0.0010457216612849152,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002470362167514395,
      "epoch": 3.6602508019830857,
      "grad_norm": 0.27823442220687866,
      "learning_rate": 1e-06,
      "loss": -0.0695,
      "step": 1567
    },
    {
      "clip_ratio/high_max": 0.003177139748004265,
      "clip_ratio/high_mean": 0.00116790375977871,
      "clip_ratio/low_mean": 0.0011514383477333467,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002319342085684184,
      "epoch": 3.662583843686206,
      "grad_norm": 0.2684524953365326,
      "learning_rate": 1e-06,
      "loss": -0.0695,
      "step": 1568
    },
    {
      "clip_ratio/high_max": 0.002009153824474197,
      "clip_ratio/high_mean": 0.0007501048912672559,
      "clip_ratio/low_mean": 0.00029923092245098815,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0010493357949599158,
      "completions/clipped_ratio": 0.1495535714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3410.0,
      "completions/mean_length": 1177.0045166015625,
      "completions/mean_terminated_length": 663.6903076171875,
      "completions/min_length": 148.0,
      "completions/min_terminated_length": 148.0,
      "epoch": 3.664916885389326,
      "grad_norm": 0.24503754079341888,
      "learning_rate": 1e-06,
      "loss": -0.0425,
      "num_tokens": 229906562.0,
      "reward": 0.6584821939468384,
      "reward_std": 0.12223109602928162,
      "rewards/verify_math_reward/mean": 0.6584821343421936,
      "rewards/verify_math_reward/std": 0.4744836091995239,
      "step": 1569
    },
    {
      "clip_ratio/high_max": 0.00233134716108907,
      "clip_ratio/high_mean": 0.00090423029178055,
      "clip_ratio/low_mean": 0.0003734052588697523,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0012776355688401964,
      "epoch": 3.667249927092447,
      "grad_norm": 0.3895801603794098,
      "learning_rate": 1e-06,
      "loss": -0.0425,
      "step": 1570
    },
    {
      "clip_ratio/high_max": 0.002474146938766353,
      "clip_ratio/high_mean": 0.0009299981356889475,
      "clip_ratio/low_mean": 0.0004818480730364172,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001411846220435109,
      "epoch": 3.6695829687955674,
      "grad_norm": 0.21874651312828064,
      "learning_rate": 1e-06,
      "loss": -0.0427,
      "step": 1571
    },
    {
      "clip_ratio/high_max": 0.002170715044485405,
      "clip_ratio/high_mean": 0.0007921155311123584,
      "clip_ratio/low_mean": 0.0005451086276480055,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013372241410252172,
      "epoch": 3.6719160104986877,
      "grad_norm": 0.21428361535072327,
      "learning_rate": 1e-06,
      "loss": -0.0427,
      "step": 1572
    },
    {
      "clip_ratio/high_max": 0.0026036849012598395,
      "clip_ratio/high_mean": 0.0010126106444658944,
      "clip_ratio/low_mean": 0.0006757626333637745,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016883732459973544,
      "completions/clipped_ratio": 0.1785714285714286,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3878.0,
      "completions/mean_length": 1322.5179443359375,
      "completions/mean_terminated_length": 719.5869750976562,
      "completions/min_length": 191.0,
      "completions/min_terminated_length": 191.0,
      "epoch": 3.674249052201808,
      "grad_norm": 0.3311329185962677,
      "learning_rate": 1e-06,
      "loss": -0.0896,
      "num_tokens": 230528706.0,
      "reward": 0.5580357313156128,
      "reward_std": 0.1582677811384201,
      "rewards/verify_math_reward/mean": 0.5580357313156128,
      "rewards/verify_math_reward/std": 0.49689781665802,
      "step": 1573
    },
    {
      "clip_ratio/high_max": 0.003360018708917778,
      "clip_ratio/high_mean": 0.0012589432990353089,
      "clip_ratio/low_mean": 0.0008136436954373494,
      "clip_ratio/low_min": 1.4357913642015774e-05,
      "clip_ratio/region_mean": 0.0020725869471789338,
      "epoch": 3.6765820939049285,
      "grad_norm": 0.2822306156158447,
      "learning_rate": 1e-06,
      "loss": -0.0898,
      "step": 1574
    },
    {
      "clip_ratio/high_max": 0.0033367287542205304,
      "clip_ratio/high_mean": 0.0012043147580698133,
      "clip_ratio/low_mean": 0.0009804068076846306,
      "clip_ratio/low_min": 2.8715827284031548e-05,
      "clip_ratio/region_mean": 0.002184721546655055,
      "epoch": 3.678915135608049,
      "grad_norm": 0.24652273952960968,
      "learning_rate": 1e-06,
      "loss": -0.09,
      "step": 1575
    },
    {
      "clip_ratio/high_max": 0.0031179989455267787,
      "clip_ratio/high_mean": 0.0011686939578794409,
      "clip_ratio/low_mean": 0.001148701980127953,
      "clip_ratio/low_min": 1.4357913642015774e-05,
      "clip_ratio/region_mean": 0.002317395934369415,
      "epoch": 3.6812481773111694,
      "grad_norm": 0.22814157605171204,
      "learning_rate": 1e-06,
      "loss": -0.0901,
      "step": 1576
    },
    {
      "clip_ratio/high_max": 0.0022585531005461235,
      "clip_ratio/high_mean": 0.0008721222602616763,
      "clip_ratio/low_mean": 0.0006358251857818686,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0015079474505910184,
      "completions/clipped_ratio": 0.1975446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4028.0,
      "completions/mean_length": 1368.5313720703125,
      "completions/mean_terminated_length": 697.095947265625,
      "completions/min_length": 161.0,
      "completions/min_terminated_length": 161.0,
      "epoch": 3.68358121901429,
      "grad_norm": 0.32418620586395264,
      "learning_rate": 1e-06,
      "loss": -0.0899,
      "num_tokens": 231114142.0,
      "reward": 0.520089328289032,
      "reward_std": 0.16070912778377533,
      "rewards/verify_math_reward/mean": 0.5200892686843872,
      "rewards/verify_math_reward/std": 0.4998753070831299,
      "step": 1577
    },
    {
      "clip_ratio/high_max": 0.0029885207768529654,
      "clip_ratio/high_mean": 0.0011393665590730961,
      "clip_ratio/low_mean": 0.0008626439630461391,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002002010471187532,
      "epoch": 3.6859142607174102,
      "grad_norm": 0.2751808166503906,
      "learning_rate": 1e-06,
      "loss": -0.09,
      "step": 1578
    },
    {
      "clip_ratio/high_max": 0.003105989428149769,
      "clip_ratio/high_mean": 0.0011380136656953255,
      "clip_ratio/low_mean": 0.0009521870260869036,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0020902006945107132,
      "epoch": 3.688247302420531,
      "grad_norm": 0.24647995829582214,
      "learning_rate": 1e-06,
      "loss": -0.0902,
      "step": 1579
    },
    {
      "clip_ratio/high_max": 0.0028495530496002175,
      "clip_ratio/high_mean": 0.0010594399755063932,
      "clip_ratio/low_mean": 0.0010839170718099922,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002143357072782237,
      "epoch": 3.690580344123651,
      "grad_norm": 0.23350109159946442,
      "learning_rate": 1e-06,
      "loss": -0.0903,
      "step": 1580
    },
    {
      "clip_ratio/high_max": 0.0023942124462337233,
      "clip_ratio/high_mean": 0.0007833131221559597,
      "clip_ratio/low_mean": 0.000566296829674684,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0013496099709300324,
      "completions/clipped_ratio": 0.1975446428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3932.0,
      "completions/mean_length": 1337.771240234375,
      "completions/mean_terminated_length": 658.7635498046875,
      "completions/min_length": 193.0,
      "completions/min_terminated_length": 193.0,
      "epoch": 3.6929133858267718,
      "grad_norm": 0.3115820586681366,
      "learning_rate": 1e-06,
      "loss": -0.0788,
      "num_tokens": 231676737.0,
      "reward": 0.5758928656578064,
      "reward_std": 0.1375589668750763,
      "rewards/verify_math_reward/mean": 0.5758928656578064,
      "rewards/verify_math_reward/std": 0.49448272585868835,
      "step": 1581
    },
    {
      "clip_ratio/high_max": 0.0033269945488427766,
      "clip_ratio/high_mean": 0.001019602606902481,
      "clip_ratio/low_mean": 0.0007508156904805219,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.00177041827555513,
      "epoch": 3.695246427529892,
      "grad_norm": 0.2711862623691559,
      "learning_rate": 1e-06,
      "loss": -0.079,
      "step": 1582
    },
    {
      "clip_ratio/high_max": 0.003074318909057183,
      "clip_ratio/high_mean": 0.0009527142501610797,
      "clip_ratio/low_mean": 0.0008890186472854111,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001841732933826279,
      "epoch": 3.6975794692330126,
      "grad_norm": 0.3059765100479126,
      "learning_rate": 1e-06,
      "loss": -0.0791,
      "step": 1583
    },
    {
      "clip_ratio/high_max": 0.0028426043718354777,
      "clip_ratio/high_mean": 0.0008659506893309299,
      "clip_ratio/low_mean": 0.0010804437115439214,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0019463943535811268,
      "epoch": 3.699912510936133,
      "grad_norm": 0.24489635229110718,
      "learning_rate": 1e-06,
      "loss": -0.0793,
      "step": 1584
    },
    {
      "clip_ratio/high_max": 0.002322709835425485,
      "clip_ratio/high_mean": 0.0008757356263231486,
      "clip_ratio/low_mean": 0.0006089560674809036,
      "clip_ratio/low_min": 1.4637002095696516e-05,
      "clip_ratio/region_mean": 0.001484691645600833,
      "completions/clipped_ratio": 0.1696428571428571,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 2399.0,
      "completions/mean_length": 1211.009033203125,
      "completions/mean_terminated_length": 621.6021728515625,
      "completions/min_length": 213.0,
      "completions/min_terminated_length": 213.0,
      "epoch": 3.7022455526392535,
      "grad_norm": 0.3546612858772278,
      "learning_rate": 1e-06,
      "loss": -0.0814,
      "num_tokens": 232225569.0,
      "reward": 0.5703125,
      "reward_std": 0.17615394294261932,
      "rewards/verify_math_reward/mean": 0.5703125,
      "rewards/verify_math_reward/std": 0.49530795216560364,
      "step": 1585
    },
    {
      "clip_ratio/high_max": 0.0026163434667978436,
      "clip_ratio/high_mean": 0.0010972397249133792,
      "clip_ratio/low_mean": 0.0008886452833394287,
      "clip_ratio/low_min": 4.30421805504011e-05,
      "clip_ratio/region_mean": 0.0019858849700540304,
      "epoch": 3.704578594342374,
      "grad_norm": 0.3230176866054535,
      "learning_rate": 1e-06,
      "loss": -0.0816,
      "step": 1586
    },
    {
      "clip_ratio/high_max": 0.0028723169307340868,
      "clip_ratio/high_mean": 0.0011486682378745172,
      "clip_ratio/low_mean": 0.00103460737682326,
      "clip_ratio/low_min": 7.179314161476213e-05,
      "clip_ratio/region_mean": 0.0021832755810464732,
      "epoch": 3.7069116360454943,
      "grad_norm": 0.2995162308216095,
      "learning_rate": 1e-06,
      "loss": -0.0817,
      "step": 1587
    },
    {
      "clip_ratio/high_max": 0.0025985515967477113,
      "clip_ratio/high_mean": 0.0009519480572635075,
      "clip_ratio/low_mean": 0.0012239921852597035,
      "clip_ratio/low_min": 2.805626172630582e-05,
      "clip_ratio/region_mean": 0.0021759402225143276,
      "epoch": 3.7092446777486145,
      "grad_norm": 0.2810906171798706,
      "learning_rate": 1e-06,
      "loss": -0.0818,
      "step": 1588
    },
    {
      "clip_ratio/high_max": 0.002692543566809036,
      "clip_ratio/high_mean": 0.0009204212274198653,
      "clip_ratio/low_mean": 0.0006138828317716616,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001534304075903492,
      "completions/clipped_ratio": 0.2075892857142857,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3768.0,
      "completions/mean_length": 1386.204345703125,
      "completions/mean_terminated_length": 676.3140869140625,
      "completions/min_length": 127.0,
      "completions/min_terminated_length": 127.0,
      "epoch": 3.711577719451735,
      "grad_norm": 0.32845979928970337,
      "learning_rate": 1e-06,
      "loss": -0.0701,
      "num_tokens": 232796368.0,
      "reward": 0.5301339626312256,
      "reward_std": 0.14135508239269257,
      "rewards/verify_math_reward/mean": 0.5301339030265808,
      "rewards/verify_math_reward/std": 0.49936985969543457,
      "step": 1589
    },
    {
      "clip_ratio/high_max": 0.0029532299013226293,
      "clip_ratio/high_mean": 0.0010062977435154608,
      "clip_ratio/low_mean": 0.0007842654104024405,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017905630884342827,
      "epoch": 3.713910761154856,
      "grad_norm": 0.29079484939575195,
      "learning_rate": 1e-06,
      "loss": -0.0703,
      "step": 1590
    },
    {
      "clip_ratio/high_max": 0.0029964288987684995,
      "clip_ratio/high_mean": 0.0010623050002322998,
      "clip_ratio/low_mean": 0.0009586893656887696,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002020994397753384,
      "epoch": 3.716243802857976,
      "grad_norm": 0.2608084976673126,
      "learning_rate": 1e-06,
      "loss": -0.0704,
      "step": 1591
    },
    {
      "clip_ratio/high_max": 0.0024345178389921784,
      "clip_ratio/high_mean": 0.0009444899878872093,
      "clip_ratio/low_mean": 0.0011719619114956004,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.002116451876645442,
      "epoch": 3.7185768445610963,
      "grad_norm": 0.281854510307312,
      "learning_rate": 1e-06,
      "loss": -0.0705,
      "step": 1592
    },
    {
      "clip_ratio/high_max": 0.002234755262179533,
      "clip_ratio/high_mean": 0.0007076127913023811,
      "clip_ratio/low_mean": 0.0004429770510796516,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.001150589836470317,
      "completions/clipped_ratio": 0.1495535714285714,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 3125.0,
      "completions/mean_length": 1113.5804443359375,
      "completions/mean_terminated_length": 589.1128540039062,
      "completions/min_length": 168.0,
      "completions/min_terminated_length": 168.0,
      "epoch": 3.720909886264217,
      "grad_norm": 0.2862772047519684,
      "learning_rate": 1e-06,
      "loss": -0.0584,
      "num_tokens": 233325464.0,
      "reward": 0.645089328289032,
      "reward_std": 0.12062598019838333,
      "rewards/verify_math_reward/mean": 0.6450892686843872,
      "rewards/verify_math_reward/std": 0.4787535071372986,
      "step": 1593
    },
    {
      "clip_ratio/high_max": 0.0031345177776529454,
      "clip_ratio/high_mean": 0.0010075391801365186,
      "clip_ratio/low_mean": 0.0006057004723061254,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0016132396885950584,
      "epoch": 3.7232429279673376,
      "grad_norm": 0.25686419010162354,
      "learning_rate": 1e-06,
      "loss": -0.0586,
      "step": 1594
    },
    {
      "clip_ratio/high_max": 0.0028180151057313196,
      "clip_ratio/high_mean": 0.0009742606889631134,
      "clip_ratio/low_mean": 0.0008117809757095529,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017860416701296344,
      "epoch": 3.725575969670458,
      "grad_norm": 0.1940649300813675,
      "learning_rate": 1e-06,
      "loss": -0.0587,
      "step": 1595
    },
    {
      "clip_ratio/high_max": 0.002606392386951484,
      "clip_ratio/high_mean": 0.0008690365175425541,
      "clip_ratio/low_mean": 0.0008804468684502353,
      "clip_ratio/low_min": 0.0,
      "clip_ratio/region_mean": 0.0017494833955424838,
      "epoch": 3.7279090113735784,
      "grad_norm": 0.26634952425956726,
      "learning_rate": 1e-06,
      "loss": -0.0588,
      "step": 1596
    },
    {
      "clip_ratio/high_max": 0.0021255537612887565,
      "clip_ratio/high_mean": 0.0009493224206380546,
      "clip_ratio/low_mean": 0.0006909248835427206,
      "clip_ratio/low_min": 1.2359106221992988e-05,
      "clip_ratio/region_mean": 0.0016402473047492094,
      "completions/clipped_ratio": 0.1662946428571429,
      "completions/max_length": 4096.0,
      "completions/max_terminated_length": 4079.0,
      "completions/mean_length": 1277.224365234375,
      "completions/mean_terminated_length": 714.9785766601562,
      "completions/min_length": 124.0,
      "completions/min_terminated_length": 124.0,
      "epoch": 3.7302420530766986,
      "grad_norm": 0.399694561958313,
      "learning_rate": 1e-06,
      "loss": -0.0923,
      "num_tokens": 233939537.0,
      "reward": 0.5602678656578064,
      "reward_std": 0.19261160492897034,
      "rewards/verify_math_reward/mean": 0.5602678656578064,
      "rewards/verify_math_reward/std": 0.4966317117214203,
      "step": 1597
    },
    {
      "clip_ratio/high_max": 0.0033188991001225077,
      "clip_ratio/high_mean": 0.001244036258867709,
      "clip_ratio/low_mean": 0.0009532424633107439,
      "clip_ratio/low_min": 4.3214809920755215e-05,
      "clip_ratio/region_mean": 0.002197278750827536,
      "epoch": 3.7325750947798193,
      "grad_norm": 0.3414709270000458,
      "learning_rate": 1e-06,
      "loss": -0.0926,
      "step": 1598
    },
    {
      "clip_ratio/high_max": 0.0031182927050394937,
      "clip_ratio/high_mean": 0.0011967702739639208,
      "clip_ratio/low_mean": 0.0010417346588837972,
      "clip_ratio/low_min": 5.5573917052242905e-05,
      "clip_ratio/region_mean": 0.0022385049087461084,
      "epoch": 3.7349081364829395,
      "grad_norm": 0.2714030146598816,
      "learning_rate": 1e-06,
      "loss": -0.0927,
      "step": 1599
    },
    {
      "clip_ratio/high_max": 0.002907983485783916,
      "clip_ratio/high_mean": 0.0012127297868573805,
      "clip_ratio/low_mean": 0.001272044322831789,
      "clip_ratio/low_min": 3.710917189891916e-05,
      "clip_ratio/region_mean": 0.002484774093318265,
      "epoch": 3.73724117818606,
      "grad_norm": 0.3983018696308136,
      "learning_rate": 1e-06,
      "loss": -0.0928,
      "step": 1600
    },
    {
      "epoch": 3.73724117818606,
      "step": 1600,
      "total_flos": 0.0,
      "train_loss": -0.05325791612935007,
      "train_runtime": 64519.0326,
      "train_samples_per_second": 22.22,
      "train_steps_per_second": 0.025
    }
  ],
  "logging_steps": 1,
  "max_steps": 1600,
  "num_input_tokens_seen": 233939537,
  "num_train_epochs": 4,
  "save_steps": 160,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}