{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9994879672299027,
  "eval_steps": 100,
  "global_step": 488,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 839.4509201049805,
      "epoch": 0.0005049548696585243,
      "grad_norm": 0.2548963129520416,
      "kl": 0.0,
      "learning_rate": 1.5151515151515152e-08,
      "loss": 0.1299,
      "reward": 0.3950892947614193,
      "reward_std": 0.3455488122999668,
      "rewards/accuracy_reward": 0.3950892947614193,
      "rewards/format_reward": 0.0,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 847.5413093566895,
      "epoch": 0.0025247743482926213,
      "grad_norm": 0.17471390962600708,
      "kl": 0.0001666247844696045,
      "learning_rate": 7.575757575757576e-08,
      "loss": 0.1225,
      "reward": 0.34858631575480103,
      "reward_std": 0.3048761789686978,
      "rewards/accuracy_reward": 0.34858631575480103,
      "rewards/format_reward": 0.0,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 862.8038803100586,
      "epoch": 0.005049548696585243,
      "grad_norm": 0.2465265542268753,
      "kl": 0.00022046566009521485,
      "learning_rate": 1.5151515151515152e-07,
      "loss": 0.138,
      "reward": 0.36235119625926016,
      "reward_std": 0.31897980310022833,
      "rewards/accuracy_reward": 0.36235119625926016,
      "rewards/format_reward": 0.0,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 858.6939163208008,
      "epoch": 0.0075743230448778644,
      "grad_norm": 0.16985075175762177,
      "kl": 0.00023434162139892578,
      "learning_rate": 2.2727272727272729e-07,
      "loss": 0.1388,
      "reward": 0.3501488171517849,
      "reward_std": 0.3107242323458195,
      "rewards/accuracy_reward": 0.3500000074505806,
      "rewards/format_reward": 0.0001488095265813172,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 871.340788269043,
      "epoch": 0.010099097393170485,
      "grad_norm": 0.19429659843444824,
      "kl": 0.0002512931823730469,
      "learning_rate": 3.0303030303030305e-07,
      "loss": 0.1322,
      "reward": 0.3501488149166107,
      "reward_std": 0.3262017160654068,
      "rewards/accuracy_reward": 0.3501488149166107,
      "rewards/format_reward": 0.0,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 865.7901931762696,
      "epoch": 0.012623871741463106,
      "grad_norm": 0.14376680552959442,
      "kl": 0.0004102230072021484,
      "learning_rate": 3.787878787878788e-07,
      "loss": 0.1202,
      "reward": 0.3611607223749161,
      "reward_std": 0.31936271116137505,
      "rewards/accuracy_reward": 0.3611607223749161,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 864.0622192382813,
      "epoch": 0.015148646089755729,
      "grad_norm": 0.18397963047027588,
      "kl": 0.0009374618530273438,
      "learning_rate": 4.5454545454545457e-07,
      "loss": 0.1329,
      "reward": 0.3620535783469677,
      "reward_std": 0.3106327898800373,
      "rewards/accuracy_reward": 0.3620535783469677,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 869.1050796508789,
      "epoch": 0.01767342043804835,
      "grad_norm": 0.14349086582660675,
      "kl": 0.005994081497192383,
      "learning_rate": 5.303030303030303e-07,
      "loss": 0.1382,
      "reward": 0.37440476976335046,
      "reward_std": 0.3009735390543938,
      "rewards/accuracy_reward": 0.3742559600621462,
      "rewards/format_reward": 0.0001488095265813172,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 856.7110305786133,
      "epoch": 0.02019819478634097,
      "grad_norm": 1.4588767290115356,
      "kl": 0.0035955429077148436,
      "learning_rate": 6.060606060606061e-07,
      "loss": 0.1527,
      "reward": 0.413244054839015,
      "reward_std": 0.301710956916213,
      "rewards/accuracy_reward": 0.413244054839015,
      "rewards/format_reward": 0.0,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 896.8750152587891,
      "epoch": 0.02272296913463359,
      "grad_norm": 0.15681959688663483,
      "kl": 0.009355926513671875,
      "learning_rate": 6.818181818181818e-07,
      "loss": 0.1496,
      "reward": 0.4224702470004559,
      "reward_std": 0.3048379421234131,
      "rewards/accuracy_reward": 0.4224702470004559,
      "rewards/format_reward": 0.0,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 843.5924240112305,
      "epoch": 0.025247743482926212,
      "grad_norm": 0.14472152292728424,
      "kl": 0.00784759521484375,
      "learning_rate": 7.575757575757576e-07,
      "loss": 0.1563,
      "reward": 0.44523810148239135,
      "reward_std": 0.28487018793821334,
      "rewards/accuracy_reward": 0.44523810148239135,
      "rewards/format_reward": 0.0,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 849.950163269043,
      "epoch": 0.027772517831218833,
      "grad_norm": 2.2360219955444336,
      "kl": 0.026047897338867188,
      "learning_rate": 8.333333333333334e-07,
      "loss": 0.1656,
      "reward": 0.48050596117973327,
      "reward_std": 0.2785939615219831,
      "rewards/accuracy_reward": 0.48050596117973327,
      "rewards/format_reward": 0.0,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 835.623080444336,
      "epoch": 0.030297292179511458,
      "grad_norm": 0.1463390588760376,
      "kl": 0.025413894653320314,
      "learning_rate": 9.090909090909091e-07,
      "loss": 0.1391,
      "reward": 0.49255953207612035,
      "reward_std": 0.2822702344506979,
      "rewards/accuracy_reward": 0.49255953207612035,
      "rewards/format_reward": 0.0,
      "step": 60
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 822.6875152587891,
      "epoch": 0.032822066527804075,
      "grad_norm": 0.23362481594085693,
      "kl": 0.012601470947265625,
      "learning_rate": 9.84848484848485e-07,
      "loss": 0.1345,
      "reward": 0.4882440596818924,
      "reward_std": 0.27538601867854595,
      "rewards/accuracy_reward": 0.4882440596818924,
      "rewards/format_reward": 0.0,
      "step": 65
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 799.619660949707,
      "epoch": 0.0353468408760967,
      "grad_norm": 0.4000104069709778,
      "kl": 0.008473968505859375,
      "learning_rate": 1.0606060606060606e-06,
      "loss": 0.1164,
      "reward": 0.4852678671479225,
      "reward_std": 0.2619207199662924,
      "rewards/accuracy_reward": 0.4852678671479225,
      "rewards/format_reward": 0.0,
      "step": 70
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 772.0263519287109,
      "epoch": 0.03787161522438932,
      "grad_norm": 2.9015414714813232,
      "kl": 0.016247940063476563,
      "learning_rate": 1.1363636363636364e-06,
      "loss": 0.1033,
      "reward": 0.504464291036129,
      "reward_std": 0.24284302350133657,
      "rewards/accuracy_reward": 0.504464291036129,
      "rewards/format_reward": 0.0,
      "step": 75
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 781.6562637329101,
      "epoch": 0.04039638957268194,
      "grad_norm": 0.23751582205295563,
      "kl": 0.008544158935546876,
      "learning_rate": 1.2121212121212122e-06,
      "loss": 0.0983,
      "reward": 0.5026785813271999,
      "reward_std": 0.24498660415410994,
      "rewards/accuracy_reward": 0.5026785813271999,
      "rewards/format_reward": 0.0,
      "step": 80
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 759.9604278564453,
      "epoch": 0.042921163920974566,
      "grad_norm": 4.397524356842041,
      "kl": 0.011675262451171875,
      "learning_rate": 1.2878787878787878e-06,
      "loss": 0.0924,
      "reward": 0.5069940604269505,
      "reward_std": 0.2519028801470995,
      "rewards/accuracy_reward": 0.5069940604269505,
      "rewards/format_reward": 0.0,
      "step": 85
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 776.5525421142578,
      "epoch": 0.04544593826926718,
      "grad_norm": 0.5079767107963562,
      "kl": 0.0114349365234375,
      "learning_rate": 1.3636363636363636e-06,
      "loss": 0.0933,
      "reward": 0.5169642955064774,
      "reward_std": 0.26492451392114164,
      "rewards/accuracy_reward": 0.5169642955064774,
      "rewards/format_reward": 0.0,
      "step": 90
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 734.737516784668,
      "epoch": 0.04797071261755981,
      "grad_norm": 11.932097434997559,
      "kl": 0.019676971435546874,
      "learning_rate": 1.4393939393939394e-06,
      "loss": 0.0666,
      "reward": 0.530654776096344,
      "reward_std": 0.23760274276137353,
      "rewards/accuracy_reward": 0.530654776096344,
      "rewards/format_reward": 0.0,
      "step": 95
    },
    {
      "epoch": 0.050495486965852425,
      "grad_norm": 1.00216805934906,
      "learning_rate": 1.5151515151515152e-06,
      "loss": 0.0734,
      "step": 100
    },
    {
      "epoch": 0.050495486965852425,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 795.8742891868449,
      "eval_kl": 0.01560433903095885,
      "eval_loss": 0.05890418216586113,
      "eval_reward": 0.4977447599727915,
      "eval_reward_std": 0.228113556685655,
      "eval_rewards/accuracy_reward": 0.4977447599727915,
      "eval_rewards/format_reward": 0.0,
      "eval_runtime": 5093.0388,
      "eval_samples_per_second": 0.754,
      "eval_steps_per_second": 0.005,
      "step": 100
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 733.3500892639161,
      "epoch": 0.05302026131414505,
      "grad_norm": 0.22215726971626282,
      "kl": 0.013599395751953125,
      "learning_rate": 1.5909090909090908e-06,
      "loss": 0.0612,
      "reward": 0.5424851287156344,
      "reward_std": 0.23987281071022154,
      "rewards/accuracy_reward": 0.5424851287156344,
      "rewards/format_reward": 0.0,
      "step": 105
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 753.378581237793,
      "epoch": 0.05554503566243767,
      "grad_norm": 0.3153633177280426,
      "kl": 0.01254425048828125,
      "learning_rate": 1.6666666666666669e-06,
      "loss": 0.0519,
      "reward": 0.5129464380443096,
      "reward_std": 0.23213699869811535,
      "rewards/accuracy_reward": 0.5129464380443096,
      "rewards/format_reward": 0.0,
      "step": 110
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 728.7988174438476,
      "epoch": 0.05806981001073029,
      "grad_norm": 0.5135952234268188,
      "kl": 0.0145751953125,
      "learning_rate": 1.7424242424242423e-06,
      "loss": 0.0601,
      "reward": 0.5419642932713031,
      "reward_std": 0.23888240195810795,
      "rewards/accuracy_reward": 0.5419642932713031,
      "rewards/format_reward": 0.0,
      "step": 115
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 718.3470321655274,
      "epoch": 0.060594584359022916,
      "grad_norm": 32253.96484375,
      "kl": 8.86512451171875,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 0.462,
      "reward": 0.5446428656578064,
      "reward_std": 0.21705915424972771,
      "rewards/accuracy_reward": 0.5446428656578064,
      "rewards/format_reward": 0.0,
      "step": 120
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 719.036474609375,
      "epoch": 0.06311935870731554,
      "grad_norm": 0.20379751920700073,
      "kl": 0.01573638916015625,
      "learning_rate": 1.893939393939394e-06,
      "loss": 0.066,
      "reward": 0.5467261992394924,
      "reward_std": 0.22874139733612536,
      "rewards/accuracy_reward": 0.5467261992394924,
      "rewards/format_reward": 0.0,
      "step": 125
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 766.3154907226562,
      "epoch": 0.06564413305560815,
      "grad_norm": 0.21040405333042145,
      "kl": 0.02525787353515625,
      "learning_rate": 1.96969696969697e-06,
      "loss": 0.0674,
      "reward": 0.514732152968645,
      "reward_std": 0.2296198945492506,
      "rewards/accuracy_reward": 0.514732152968645,
      "rewards/format_reward": 0.0,
      "step": 130
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 772.217121887207,
      "epoch": 0.06816890740390077,
      "grad_norm": 0.392875999212265,
      "kl": 0.033099365234375,
      "learning_rate": 2.0454545454545453e-06,
      "loss": 0.1037,
      "reward": 0.5169642955064774,
      "reward_std": 0.2294642563909292,
      "rewards/accuracy_reward": 0.5169642955064774,
      "rewards/format_reward": 0.0,
      "step": 135
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 845.8943603515625,
      "epoch": 0.0706936817521934,
      "grad_norm": 19866.705078125,
      "kl": 3.577685546875,
      "learning_rate": 2.121212121212121e-06,
      "loss": 0.6622,
      "reward": 0.48511905670166017,
      "reward_std": 0.2477981884032488,
      "rewards/accuracy_reward": 0.48511905670166017,
      "rewards/format_reward": 0.0,
      "step": 140
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 844.8823059082031,
      "epoch": 0.07321845610048602,
      "grad_norm": 0.47434961795806885,
      "kl": 0.17655029296875,
      "learning_rate": 2.1969696969696974e-06,
      "loss": 0.146,
      "reward": 0.4336309637874365,
      "reward_std": 0.2661529861390591,
      "rewards/accuracy_reward": 0.4336309637874365,
      "rewards/format_reward": 0.0,
      "step": 145
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 758.0837905883789,
      "epoch": 0.07574323044877863,
      "grad_norm": 9.017133712768555,
      "kl": 0.224951171875,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.1059,
      "reward": 0.44047619961202145,
      "reward_std": 0.2847708839923143,
      "rewards/accuracy_reward": 0.44047619961202145,
      "rewards/format_reward": 0.0,
      "step": 150
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 679.5257614135742,
      "epoch": 0.07826800479707126,
      "grad_norm": 15.091608047485352,
      "kl": 0.21767578125,
      "learning_rate": 2.3484848484848486e-06,
      "loss": 0.0639,
      "reward": 0.4681547723710537,
      "reward_std": 0.26657171882689,
      "rewards/accuracy_reward": 0.4681547723710537,
      "rewards/format_reward": 0.0,
      "step": 155
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 594.3864700317383,
      "epoch": 0.08079277914536388,
      "grad_norm": 18.641502380371094,
      "kl": 0.4121337890625,
      "learning_rate": 2.4242424242424244e-06,
      "loss": 0.0134,
      "reward": 0.4694940596818924,
      "reward_std": 0.26797937490046025,
      "rewards/accuracy_reward": 0.4694940596818924,
      "rewards/format_reward": 0.0,
      "step": 160
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 573.9558166503906,
      "epoch": 0.08331755349365651,
      "grad_norm": 3.902953624725342,
      "kl": 0.3707763671875,
      "learning_rate": 2.5e-06,
      "loss": -0.0055,
      "reward": 0.4639881007373333,
      "reward_std": 0.24555196836590767,
      "rewards/accuracy_reward": 0.4639881007373333,
      "rewards/format_reward": 0.0,
      "step": 165
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 567.443019104004,
      "epoch": 0.08584232784194913,
      "grad_norm": 4.564860820770264,
      "kl": 21.7295166015625,
      "learning_rate": 2.5757575757575756e-06,
      "loss": 1.4361,
      "reward": 0.5007440589368344,
      "reward_std": 0.25035002939403056,
      "rewards/accuracy_reward": 0.5007440589368344,
      "rewards/format_reward": 0.0,
      "step": 170
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 573.968465423584,
      "epoch": 0.08836710219024174,
      "grad_norm": 2.453469753265381,
      "kl": 0.58544921875,
      "learning_rate": 2.651515151515152e-06,
      "loss": -0.0241,
      "reward": 0.479761915653944,
      "reward_std": 0.24731975235044956,
      "rewards/accuracy_reward": 0.479761915653944,
      "rewards/format_reward": 0.0,
      "step": 175
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 592.1180152893066,
      "epoch": 0.09089187653853437,
      "grad_norm": 7.268795967102051,
      "kl": 1.03837890625,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.0296,
      "reward": 0.5111607246100902,
      "reward_std": 0.25646619591861963,
      "rewards/accuracy_reward": 0.5111607246100902,
      "rewards/format_reward": 0.0,
      "step": 180
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 609.6851280212402,
      "epoch": 0.09341665088682699,
      "grad_norm": 3.061584234237671,
      "kl": 0.851953125,
      "learning_rate": 2.803030303030303e-06,
      "loss": 0.017,
      "reward": 0.4995535835623741,
      "reward_std": 0.23969481624662875,
      "rewards/accuracy_reward": 0.4995535835623741,
      "rewards/format_reward": 0.0,
      "step": 185
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 605.6526893615722,
      "epoch": 0.09594142523511962,
      "grad_norm": 15.86893367767334,
      "kl": 2.0413330078125,
      "learning_rate": 2.878787878787879e-06,
      "loss": 0.0528,
      "reward": 0.5168154835700989,
      "reward_std": 0.24480598308146,
      "rewards/accuracy_reward": 0.5168154835700989,
      "rewards/format_reward": 0.0,
      "step": 190
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 589.4162292480469,
      "epoch": 0.09846619958341224,
      "grad_norm": 51.3820915222168,
      "kl": 1.005224609375,
      "learning_rate": 2.9545454545454547e-06,
      "loss": 0.0164,
      "reward": 0.5026785813271999,
      "reward_std": 0.24141334034502507,
      "rewards/accuracy_reward": 0.5026785813271999,
      "rewards/format_reward": 0.0,
      "step": 195
    },
    {
      "epoch": 0.10099097393170485,
      "grad_norm": 2.8177883625030518,
      "learning_rate": 2.999990675938041e-06,
      "loss": 0.0187,
      "step": 200
    },
    {
      "epoch": 0.10099097393170485,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 755.0453377480833,
      "eval_kl": 2.4657899844720497,
      "eval_loss": 0.17787984013557434,
      "eval_reward": 0.4527506747786303,
      "eval_reward_std": 0.23182548258615576,
      "eval_rewards/accuracy_reward": 0.4527506747786303,
      "eval_rewards/format_reward": 0.0,
      "eval_runtime": 5012.4785,
      "eval_samples_per_second": 0.766,
      "eval_steps_per_second": 0.005,
      "step": 200
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1894.196615600586,
      "epoch": 0.4198668714797747,
      "grad_norm": 2.7988369464874268,
      "kl": 1.30494384765625,
      "learning_rate": 2.1584141861097513e-06,
      "loss": 0.0839,
      "reward": 0.07008928679861129,
      "reward_std": 0.05558663532137871,
      "rewards/accuracy_reward": 0.07008928679861129,
      "rewards/format_reward": 0.0,
      "step": 205
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1115.846598815918,
      "epoch": 0.43010752688172044,
      "grad_norm": 4.760381698608398,
      "kl": 0.363568115234375,
      "learning_rate": 2.1097780693886946e-06,
      "loss": 0.03,
      "reward": 0.11488095466047525,
      "reward_std": 0.09430552553385496,
      "rewards/accuracy_reward": 0.11488095466047525,
      "rewards/format_reward": 0.0,
      "step": 210
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 741.814892578125,
      "epoch": 0.4403481822836662,
      "grad_norm": 1.493429183959961,
      "kl": 0.41728515625,
      "learning_rate": 2.060361338950506e-06,
      "loss": 0.0096,
      "reward": 0.10699404974002391,
      "reward_std": 0.08532515289261937,
      "rewards/accuracy_reward": 0.10699404974002391,
      "rewards/format_reward": 0.0,
      "step": 215
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 757.3077529907226,
      "epoch": 0.45058883768561186,
      "grad_norm": 0.6104231476783752,
      "kl": 0.28935546875,
      "learning_rate": 2.010227256134962e-06,
      "loss": 0.0056,
      "reward": 0.10029762159101666,
      "reward_std": 0.08028191016055644,
      "rewards/accuracy_reward": 0.10029762159101666,
      "rewards/format_reward": 0.0,
      "step": 220
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 796.1696578979493,
      "epoch": 0.4608294930875576,
      "grad_norm": 3.8080809116363525,
      "kl": 0.3931396484375,
      "learning_rate": 1.9594400006079216e-06,
      "loss": 0.0098,
      "reward": 0.11964285904541612,
      "reward_std": 0.09335053358227015,
      "rewards/accuracy_reward": 0.11964285904541612,
      "rewards/format_reward": 0.0,
      "step": 225
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 872.2796279907227,
      "epoch": 0.47107014848950335,
      "grad_norm": 2.8400909900665283,
      "kl": 0.3711669921875,
      "learning_rate": 1.9080645882010618e-06,
      "loss": 0.0085,
      "reward": 0.1058035738300532,
      "reward_std": 0.08510495922528208,
      "rewards/accuracy_reward": 0.1058035738300532,
      "rewards/format_reward": 0.0,
      "step": 230
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 950.6646041870117,
      "epoch": 0.48131080389144903,
      "grad_norm": 1.133971929550171,
      "kl": 0.3427490234375,
      "learning_rate": 1.8561667876811873e-06,
      "loss": 0.0169,
      "reward": 0.1235119073651731,
      "reward_std": 0.08968696529045701,
      "rewards/accuracy_reward": 0.1235119073651731,
      "rewards/format_reward": 0.0,
      "step": 235
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 973.8951126098633,
      "epoch": 0.4915514592933948,
      "grad_norm": 0.8387630581855774,
      "kl": 0.418408203125,
      "learning_rate": 1.8038130365556615e-06,
      "loss": 0.0219,
      "reward": 0.12857143091969192,
      "reward_std": 0.10588475931435823,
      "rewards/accuracy_reward": 0.12857143091969192,
      "rewards/format_reward": 0.0,
      "step": 240
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 962.3290390014648,
      "epoch": 0.5017921146953405,
      "grad_norm": 2.0839455127716064,
      "kl": 0.6023193359375,
      "learning_rate": 1.751070356021742e-06,
      "loss": 0.0248,
      "reward": 0.12485119318589569,
      "reward_std": 0.10606874478980899,
      "rewards/accuracy_reward": 0.12485119318589569,
      "rewards/format_reward": 0.0,
      "step": 245
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1087.3797821044923,
      "epoch": 0.5120327700972862,
      "grad_norm": 4.897603988647461,
      "kl": 0.297705078125,
      "learning_rate": 1.6980062651687004e-06,
      "loss": 0.0245,
      "reward": 0.12886905036866664,
      "reward_std": 0.10398078355938196,
      "rewards/accuracy_reward": 0.12886905036866664,
      "rewards/format_reward": 0.0,
      "step": 250
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1061.2479339599608,
      "epoch": 0.522273425499232,
      "grad_norm": 10.860794067382812,
      "kl": 0.60859375,
      "learning_rate": 1.6446886945425608e-06,
      "loss": 0.0344,
      "reward": 0.14122024178504944,
      "reward_std": 0.10650016171857715,
      "rewards/accuracy_reward": 0.14122024178504944,
      "rewards/format_reward": 0.0,
      "step": 255
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1152.5089492797852,
      "epoch": 0.5325140809011777,
      "grad_norm": 1.4874508380889893,
      "kl": 0.3950927734375,
      "learning_rate": 1.5911858991841063e-06,
      "loss": 0.0344,
      "reward": 0.1450892877765,
      "reward_std": 0.10981706054881216,
      "rewards/accuracy_reward": 0.1450892877765,
      "rewards/format_reward": 0.0,
      "step": 260
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1092.9384155273438,
      "epoch": 0.5427547363031234,
      "grad_norm": 1.4266042709350586,
      "kl": 0.861279296875,
      "learning_rate": 1.537566371251482e-06,
      "loss": 0.0509,
      "reward": 0.14776785927824676,
      "reward_std": 0.11378104742616416,
      "rewards/accuracy_reward": 0.14776785927824676,
      "rewards/format_reward": 0.0,
      "step": 265
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1142.4348419189453,
      "epoch": 0.5529953917050692,
      "grad_norm": 4.305612564086914,
      "kl": 0.542431640625,
      "learning_rate": 1.4838987523392473e-06,
      "loss": 0.0374,
      "reward": 0.149107145704329,
      "reward_std": 0.10749768582172692,
      "rewards/accuracy_reward": 0.149107145704329,
      "rewards/format_reward": 0.0,
      "step": 270
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1178.0220428466796,
      "epoch": 0.5632360471070148,
      "grad_norm": 1.8183989524841309,
      "kl": 0.477099609375,
      "learning_rate": 1.4302517456061282e-06,
      "loss": 0.0336,
      "reward": 0.14880952723324298,
      "reward_std": 0.115222292765975,
      "rewards/accuracy_reward": 0.14880952723324298,
      "rewards/format_reward": 0.0,
      "step": 275
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1173.8620773315429,
      "epoch": 0.5734767025089605,
      "grad_norm": 3.585653781890869,
      "kl": 0.5623046875,
      "learning_rate": 1.3766940278239555e-06,
      "loss": 0.0363,
      "reward": 0.13660714430734516,
      "reward_std": 0.10423116614110768,
      "rewards/accuracy_reward": 0.13660714430734516,
      "rewards/format_reward": 0.0,
      "step": 280
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1201.4131134033203,
      "epoch": 0.5837173579109063,
      "grad_norm": 1.5504530668258667,
      "kl": 0.548388671875,
      "learning_rate": 1.3232941614603844e-06,
      "loss": 0.0346,
      "reward": 0.14375000298023224,
      "reward_std": 0.12105894400738179,
      "rewards/accuracy_reward": 0.14375000298023224,
      "rewards/format_reward": 0.0,
      "step": 285
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1276.2178771972656,
      "epoch": 0.593958013312852,
      "grad_norm": 1.5955944061279297,
      "kl": 0.5833984375,
      "learning_rate": 1.270120506907939e-06,
      "loss": 0.0347,
      "reward": 0.13065476389601827,
      "reward_std": 0.0991502583026886,
      "rewards/accuracy_reward": 0.13065476389601827,
      "rewards/format_reward": 0.0,
      "step": 290
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1367.1632751464845,
      "epoch": 0.6041986687147978,
      "grad_norm": 6.252760410308838,
      "kl": 0.5171875,
      "learning_rate": 1.2172411349717491e-06,
      "loss": 0.0357,
      "reward": 0.133630954567343,
      "reward_std": 0.11369595509022475,
      "rewards/accuracy_reward": 0.133630954567343,
      "rewards/format_reward": 0.0,
      "step": 295
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1294.0861907958983,
      "epoch": 0.6144393241167435,
      "grad_norm": 1.7657579183578491,
      "kl": 0.9818359375,
      "learning_rate": 1.1647237397280027e-06,
      "loss": 0.0553,
      "reward": 0.11979166907258332,
      "reward_std": 0.10067843548022211,
      "rewards/accuracy_reward": 0.11979166907258332,
      "rewards/format_reward": 0.0,
      "step": 300
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1419.5473419189452,
      "epoch": 0.6246799795186891,
      "grad_norm": 2.0053577423095703,
      "kl": 0.32607421875,
      "learning_rate": 1.112635551864672e-06,
      "loss": 0.0249,
      "reward": 0.1291666693985462,
      "reward_std": 0.09413350112736225,
      "rewards/accuracy_reward": 0.1291666693985462,
      "rewards/format_reward": 0.0,
      "step": 305
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1386.4494354248047,
      "epoch": 0.6349206349206349,
      "grad_norm": 3.0205366611480713,
      "kl": 0.508544921875,
      "learning_rate": 1.061043252615451e-06,
      "loss": 0.0343,
      "reward": 0.13586309785023332,
      "reward_std": 0.10630648182705045,
      "rewards/accuracy_reward": 0.13586309785023332,
      "rewards/format_reward": 0.0,
      "step": 310
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1371.6342407226562,
      "epoch": 0.6451612903225806,
      "grad_norm": 1.5079602003097534,
      "kl": 0.572265625,
      "learning_rate": 1.0100128883970838e-06,
      "loss": 0.0347,
      "reward": 0.13020833493210376,
      "reward_std": 0.104894710611552,
      "rewards/accuracy_reward": 0.13020833493210376,
      "rewards/format_reward": 0.0,
      "step": 315
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1415.7412506103515,
      "epoch": 0.6554019457245264,
      "grad_norm": 1.1102583408355713,
      "kl": 0.466064453125,
      "learning_rate": 9.596097862593594e-07,
      "loss": 0.0346,
      "reward": 0.11889881128445268,
      "reward_std": 0.10287130940705538,
      "rewards/accuracy_reward": 0.11889881128445268,
      "rewards/format_reward": 0.0,
      "step": 320
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1400.208950805664,
      "epoch": 0.6656426011264721,
      "grad_norm": 1.7711780071258545,
      "kl": 0.5921875,
      "learning_rate": 9.098984702560116e-07,
      "loss": 0.038,
      "reward": 0.12872024006210267,
      "reward_std": 0.10296571510843933,
      "rewards/accuracy_reward": 0.12872024006210267,
      "rewards/format_reward": 0.0,
      "step": 325
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1419.4535919189452,
      "epoch": 0.6758832565284179,
      "grad_norm": 1.5115975141525269,
      "kl": 0.464794921875,
      "learning_rate": 8.609425788435818e-07,
      "loss": 0.0352,
      "reward": 0.11815476433839649,
      "reward_std": 0.09560534581542016,
      "rewards/accuracy_reward": 0.11815476433839649,
      "rewards/format_reward": 0.0,
      "step": 330
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1371.5564270019531,
      "epoch": 0.6861239119303636,
      "grad_norm": 1.2688425779342651,
      "kl": 0.593115234375,
      "learning_rate": 8.128047834139907e-07,
      "loss": 0.0373,
      "reward": 0.1520833361428231,
      "reward_std": 0.10912553807720542,
      "rewards/accuracy_reward": 0.1520833361428231,
      "rewards/format_reward": 0.0,
      "step": 335
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1391.9670043945312,
      "epoch": 0.6963645673323092,
      "grad_norm": 1.829864740371704,
      "kl": 0.538720703125,
      "learning_rate": 7.65546708065104e-07,
      "loss": 0.0385,
      "reward": 0.13824404990300537,
      "reward_std": 0.11598918633535504,
      "rewards/accuracy_reward": 0.13824404990300537,
      "rewards/format_reward": 0.0,
      "step": 340
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1407.3750274658203,
      "epoch": 0.706605222734255,
      "grad_norm": 2.415090322494507,
      "kl": 0.52529296875,
      "learning_rate": 7.192288507120073e-07,
      "loss": 0.0326,
      "reward": 0.12991071683354677,
      "reward_std": 0.09881549612618983,
      "rewards/accuracy_reward": 0.12991071683354677,
      "rewards/format_reward": 0.0,
      "step": 345
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1416.4004791259765,
      "epoch": 0.7168458781362007,
      "grad_norm": 2.3220245838165283,
      "kl": 0.397216796875,
      "learning_rate": 6.739105056399712e-07,
      "loss": 0.0236,
      "reward": 0.14464285979047417,
      "reward_std": 0.1005579138174653,
      "rewards/accuracy_reward": 0.14464285979047417,
      "rewards/format_reward": 0.0,
      "step": 350
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1480.5619354248047,
      "epoch": 0.7270865335381465,
      "grad_norm": 2.364630699157715,
      "kl": 0.56025390625,
      "learning_rate": 6.296496875982565e-07,
      "loss": 0.0327,
      "reward": 0.11845238371752202,
      "reward_std": 0.089572025090456,
      "rewards/accuracy_reward": 0.11845238371752202,
      "rewards/format_reward": 0.0,
      "step": 355
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1497.1189331054688,
      "epoch": 0.7373271889400922,
      "grad_norm": 1.7453265190124512,
      "kl": 0.359423828125,
      "learning_rate": 5.86503057531935e-07,
      "loss": 0.0196,
      "reward": 0.1288690499495715,
      "reward_std": 0.09691634746268392,
      "rewards/accuracy_reward": 0.1288690499495715,
      "rewards/format_reward": 0.0,
      "step": 360
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1490.572946166992,
      "epoch": 0.7475678443420379,
      "grad_norm": 2.5578675270080566,
      "kl": 0.44873046875,
      "learning_rate": 5.445258500467891e-07,
      "loss": 0.0306,
      "reward": 0.13690476482734085,
      "reward_std": 0.09427342982962728,
      "rewards/accuracy_reward": 0.13690476482734085,
      "rewards/format_reward": 0.0,
      "step": 365
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1448.5931884765625,
      "epoch": 0.7578084997439836,
      "grad_norm": 1.473059058189392,
      "kl": 0.623779296875,
      "learning_rate": 5.037718027001601e-07,
      "loss": 0.0447,
      "reward": 0.12886904999613763,
      "reward_std": 0.10351952081546187,
      "rewards/accuracy_reward": 0.12886904999613763,
      "rewards/format_reward": 0.0,
      "step": 370
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1525.3043518066406,
      "epoch": 0.7680491551459293,
      "grad_norm": 2.5979256629943848,
      "kl": 0.438134765625,
      "learning_rate": 4.642930872082603e-07,
      "loss": 0.0294,
      "reward": 0.1340773832052946,
      "reward_std": 0.09693685229867696,
      "rewards/accuracy_reward": 0.1340773832052946,
      "rewards/format_reward": 0.0,
      "step": 375
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1548.1570190429688,
      "epoch": 0.778289810547875,
      "grad_norm": 1.8802986145019531,
      "kl": 0.41083984375,
      "learning_rate": 4.261402426580037e-07,
      "loss": 0.0239,
      "reward": 0.1302083361428231,
      "reward_std": 0.09426894560456275,
      "rewards/accuracy_reward": 0.1302083361428231,
      "rewards/format_reward": 0.0,
      "step": 380
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1499.0522583007812,
      "epoch": 0.7885304659498208,
      "grad_norm": 1.5898677110671997,
      "kl": 0.464453125,
      "learning_rate": 3.893621108088737e-07,
      "loss": 0.0377,
      "reward": 0.13675595503300428,
      "reward_std": 0.1057541430927813,
      "rewards/accuracy_reward": 0.13675595503300428,
      "rewards/format_reward": 0.0,
      "step": 385
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1490.6623809814453,
      "epoch": 0.7987711213517665,
      "grad_norm": 1.5012719631195068,
      "kl": 0.517529296875,
      "learning_rate": 3.540057735676388e-07,
      "loss": 0.0362,
      "reward": 0.13720238325186074,
      "reward_std": 0.11000550435855985,
      "rewards/accuracy_reward": 0.13720238325186074,
      "rewards/format_reward": 0.0,
      "step": 390
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1487.6908081054687,
      "epoch": 0.8090117767537123,
      "grad_norm": 4.934905529022217,
      "kl": 0.600732421875,
      "learning_rate": 3.20116492715959e-07,
      "loss": 0.0425,
      "reward": 0.13824405027553438,
      "reward_std": 0.10485180700197816,
      "rewards/accuracy_reward": 0.13824405027553438,
      "rewards/format_reward": 0.0,
      "step": 395
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1447.5931762695313,
      "epoch": 0.819252432155658,
      "grad_norm": 1.9514836072921753,
      "kl": 0.618359375,
      "learning_rate": 2.8773765196804834e-07,
      "loss": 0.0428,
      "reward": 0.12217262103222311,
      "reward_std": 0.10348854968324303,
      "rewards/accuracy_reward": 0.12217262103222311,
      "rewards/format_reward": 0.0,
      "step": 400
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1489.2579193115234,
      "epoch": 0.8294930875576036,
      "grad_norm": 4.070176601409912,
      "kl": 0.48359375,
      "learning_rate": 2.5691070143256447e-07,
      "loss": 0.0375,
      "reward": 0.13288690745830536,
      "reward_std": 0.10663077728822827,
      "rewards/accuracy_reward": 0.13288690745830536,
      "rewards/format_reward": 0.0,
      "step": 405
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1444.0242858886718,
      "epoch": 0.8397337429595494,
      "grad_norm": 2.602212429046631,
      "kl": 0.635986328125,
      "learning_rate": 2.276751045498197e-07,
      "loss": 0.0414,
      "reward": 0.14747024136595427,
      "reward_std": 0.10692896069958806,
      "rewards/accuracy_reward": 0.14747024136595427,
      "rewards/format_reward": 0.0,
      "step": 410
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1452.9589599609376,
      "epoch": 0.8499743983614951,
      "grad_norm": 3.2816104888916016,
      "kl": 0.64755859375,
      "learning_rate": 2.0006828757224982e-07,
      "loss": 0.0374,
      "reward": 0.14821428790455685,
      "reward_std": 0.10495366291143,
      "rewards/accuracy_reward": 0.14821428790455685,
      "rewards/format_reward": 0.0,
      "step": 415
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1461.2986938476563,
      "epoch": 0.8602150537634409,
      "grad_norm": 1.5426130294799805,
      "kl": 0.415869140625,
      "learning_rate": 1.7412559165281155e-07,
      "loss": 0.0259,
      "reward": 0.1513392882887274,
      "reward_std": 0.099711880274117,
      "rewards/accuracy_reward": 0.1513392882887274,
      "rewards/format_reward": 0.0,
      "step": 420
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1480.4108947753907,
      "epoch": 0.8704557091653866,
      "grad_norm": 1.577873706817627,
      "kl": 0.3802734375,
      "learning_rate": 1.4988022760263492e-07,
      "loss": 0.027,
      "reward": 0.12901785969734192,
      "reward_std": 0.10641210693866014,
      "rewards/accuracy_reward": 0.12901785969734192,
      "rewards/format_reward": 0.0,
      "step": 425
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1430.3998809814452,
      "epoch": 0.8806963645673324,
      "grad_norm": 2.2523367404937744,
      "kl": 0.503662109375,
      "learning_rate": 1.2736323337586142e-07,
      "loss": 0.0365,
      "reward": 0.13869047933258116,
      "reward_std": 0.1244213636033237,
      "rewards/accuracy_reward": 0.13869047933258116,
      "rewards/format_reward": 0.0,
      "step": 430
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1426.7138580322267,
      "epoch": 0.890937019969278,
      "grad_norm": 4.5399346351623535,
      "kl": 0.675048828125,
      "learning_rate": 1.0660343433608627e-07,
      "loss": 0.0481,
      "reward": 0.1412202403997071,
      "reward_std": 0.10602434603497386,
      "rewards/accuracy_reward": 0.1412202403997071,
      "rewards/format_reward": 0.0,
      "step": 435
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1438.0576110839843,
      "epoch": 0.9011776753712237,
      "grad_norm": 5.20224142074585,
      "kl": 0.7138671875,
      "learning_rate": 8.762740635526756e-08,
      "loss": 0.048,
      "reward": 0.13943452620878816,
      "reward_std": 0.11082206377759576,
      "rewards/accuracy_reward": 0.13943452620878816,
      "rewards/format_reward": 0.0,
      "step": 440
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1476.9360321044921,
      "epoch": 0.9114183307731695,
      "grad_norm": 2.1484456062316895,
      "kl": 0.59794921875,
      "learning_rate": 7.04594417923507e-08,
      "loss": 0.0401,
      "reward": 0.13556547947227954,
      "reward_std": 0.10029366509988905,
      "rewards/accuracy_reward": 0.13556547947227954,
      "rewards/format_reward": 0.0,
      "step": 445
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1484.7338073730468,
      "epoch": 0.9216589861751152,
      "grad_norm": 3.5503857135772705,
      "kl": 0.50009765625,
      "learning_rate": 5.512151839515234e-08,
      "loss": 0.036,
      "reward": 0.14583333618938923,
      "reward_std": 0.11404564101248979,
      "rewards/accuracy_reward": 0.14583333618938923,
      "rewards/format_reward": 0.0,
      "step": 450
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1441.9092498779296,
      "epoch": 0.931899641577061,
      "grad_norm": 4.833263397216797,
      "kl": 0.46787109375,
      "learning_rate": 4.16332711653184e-08,
      "loss": 0.0332,
      "reward": 0.14017857438884676,
      "reward_std": 0.10374454082921147,
      "rewards/accuracy_reward": 0.14017857438884676,
      "rewards/format_reward": 0.0,
      "step": 455
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1402.6604400634765,
      "epoch": 0.9421402969790067,
      "grad_norm": 2.325770378112793,
      "kl": 0.434716796875,
      "learning_rate": 3.0011967222373595e-08,
      "loss": 0.0299,
      "reward": 0.13675595447421074,
      "reward_std": 0.1053824592847377,
      "rewards/accuracy_reward": 0.13675595447421074,
      "rewards/format_reward": 0.0,
      "step": 460
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1410.3284545898437,
      "epoch": 0.9523809523809523,
      "grad_norm": 1.8935805559158325,
      "kl": 0.473193359375,
      "learning_rate": 2.0272483699039334e-08,
      "loss": 0.0327,
      "reward": 0.14107143036089836,
      "reward_std": 0.11868918919935822,
      "rewards/accuracy_reward": 0.14107143036089836,
      "rewards/format_reward": 0.0,
      "step": 465
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1431.0174346923827,
      "epoch": 0.9626216077828981,
      "grad_norm": 1.980904459953308,
      "kl": 0.47861328125,
      "learning_rate": 1.2427288696117589e-08,
      "loss": 0.0294,
      "reward": 0.14642857415601612,
      "reward_std": 0.10726110143586993,
      "rewards/accuracy_reward": 0.14642857415601612,
      "rewards/format_reward": 0.0,
      "step": 470
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1449.0582153320313,
      "epoch": 0.9728622631848438,
      "grad_norm": 4.871621131896973,
      "kl": 0.472021484375,
      "learning_rate": 6.486425321322987e-09,
      "loss": 0.0256,
      "reward": 0.1415178598370403,
      "reward_std": 0.10020972546190024,
      "rewards/accuracy_reward": 0.1415178598370403,
      "rewards/format_reward": 0.0,
      "step": 475
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1416.2077575683593,
      "epoch": 0.9831029185867896,
      "grad_norm": 3.306790590286255,
      "kl": 0.481982421875,
      "learning_rate": 2.457498832494187e-09,
      "loss": 0.0289,
      "reward": 0.13943452583625912,
      "reward_std": 0.10717600891366601,
      "rewards/accuracy_reward": 0.13943452583625912,
      "rewards/format_reward": 0.0,
      "step": 480
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1434.3193756103515,
      "epoch": 0.9933435739887353,
      "grad_norm": 3.3276352882385254,
      "kl": 0.465087890625,
      "learning_rate": 3.456669016438374e-10,
      "loss": 0.032,
      "reward": 0.13824404957704245,
      "reward_std": 0.10434536337852478,
      "rewards/accuracy_reward": 0.13824404957704245,
      "rewards/format_reward": 0.0,
      "step": 485
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 1441.995569864909,
      "epoch": 0.9994879672299027,
      "kl": 0.4940592447916667,
      "reward": 0.14682539909457168,
      "reward_std": 0.11686915632647772,
      "rewards/accuracy_reward": 0.14682539909457168,
      "rewards/format_reward": 0.0,
      "step": 488,
      "total_flos": 0.0,
      "train_loss": 0.01953804415467455,
      "train_runtime": 102141.6698,
      "train_samples_per_second": 0.918,
      "train_steps_per_second": 0.005
    }
  ],
  "logging_steps": 5,
  "max_steps": 488,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 24,
  "trial_name": null,
  "trial_params": null
}