{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9994815966822186,
  "eval_steps": 10,
  "global_step": 2892,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 426.8250160217285,
      "epoch": 0.0010368066355624676,
      "grad_norm": 14.147396087646484,
      "kl": 0.0,
      "learning_rate": 1.0344827586206896e-08,
      "loss": 0.1143,
      "reward": 0.30000000447034836,
      "reward_std": 0.353316530585289,
      "rewards/accuracy_reward": 0.22500000894069672,
      "rewards/format_reward": 0.07500000111758709,
      "step": 1
    },
    {
      "completion_length": 345.9749946594238,
      "epoch": 0.002073613271124935,
      "grad_norm": 0.5298608541488647,
      "kl": 0.0,
      "learning_rate": 2.0689655172413793e-08,
      "loss": 0.0471,
      "reward": 0.3000000026077032,
      "reward_std": 0.34968600049614906,
      "rewards/accuracy_reward": 0.27500000409781933,
      "rewards/format_reward": 0.02500000037252903,
      "step": 2
    },
    {
      "completion_length": 321.2000045776367,
      "epoch": 0.003110419906687403,
      "grad_norm": 3.317950487136841,
      "kl": 0.00021088123321533203,
      "learning_rate": 3.103448275862069e-08,
      "loss": 0.0899,
      "reward": 0.4750000014901161,
      "reward_std": 0.42900275811553,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.02500000037252903,
      "step": 3
    },
    {
      "completion_length": 283.5250053405762,
      "epoch": 0.00414722654224987,
      "grad_norm": 0.7831366062164307,
      "kl": 0.0001735687255859375,
      "learning_rate": 4.1379310344827585e-08,
      "loss": -0.0534,
      "reward": 0.5250000096857548,
      "reward_std": 0.4415663778781891,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.0,
      "step": 4
    },
    {
      "completion_length": 311.0750045776367,
      "epoch": 0.005184033177812338,
      "grad_norm": 0.6258169412612915,
      "kl": 0.00017344951629638672,
      "learning_rate": 5.172413793103448e-08,
      "loss": -0.0563,
      "reward": 0.24999999813735485,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.20000000670552254,
      "rewards/format_reward": 0.05000000074505806,
      "step": 5
    },
    {
      "completion_length": 363.50000381469727,
      "epoch": 0.006220839813374806,
      "grad_norm": 3.521343946456909,
      "kl": 0.0002512931823730469,
      "learning_rate": 6.206896551724137e-08,
      "loss": 0.1373,
      "reward": 0.550000011920929,
      "reward_std": 0.517391100525856,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.07500000111758709,
      "step": 6
    },
    {
      "completion_length": 387.9500045776367,
      "epoch": 0.007257646448937273,
      "grad_norm": 3.267216205596924,
      "kl": 0.0003764629364013672,
      "learning_rate": 7.241379310344827e-08,
      "loss": 0.0081,
      "reward": 0.2750000059604645,
      "reward_std": 0.36053743213415146,
      "rewards/accuracy_reward": 0.20000000670552254,
      "rewards/format_reward": 0.07500000111758709,
      "step": 7
    },
    {
      "completion_length": 340.6500053405762,
      "epoch": 0.00829445308449974,
      "grad_norm": 2.6857998371124268,
      "kl": 0.0002391338348388672,
      "learning_rate": 8.275862068965517e-08,
      "loss": 0.055,
      "reward": 0.375,
      "reward_std": 0.39822831004858017,
      "rewards/accuracy_reward": 0.34999999962747097,
      "rewards/format_reward": 0.02500000037252903,
      "step": 8
    },
    {
      "completion_length": 488.42501068115234,
      "epoch": 0.00933125972006221,
      "grad_norm": 0.6219092607498169,
      "kl": 0.00023365020751953125,
      "learning_rate": 9.310344827586207e-08,
      "loss": -0.011,
      "reward": 0.22500000521540642,
      "reward_std": 0.3297629803419113,
      "rewards/accuracy_reward": 0.22500000521540642,
      "rewards/format_reward": 0.0,
      "step": 9
    },
    {
      "completion_length": 360.625,
      "epoch": 0.010368066355624676,
      "grad_norm": 0.5460860133171082,
      "kl": 0.000164031982421875,
      "learning_rate": 1.0344827586206897e-07,
      "loss": 0.0558,
      "reward": 0.25000000558793545,
      "reward_std": 0.3856646828353405,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.02500000037252903,
      "step": 10
    },
    {
      "completion_length": 341.00000762939453,
      "epoch": 0.011404872991187144,
      "grad_norm": 1.1156655550003052,
      "kl": 0.00022840499877929688,
      "learning_rate": 1.1379310344827586e-07,
      "loss": 0.0461,
      "reward": 0.4000000059604645,
      "reward_std": 0.497468076646328,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.0,
      "step": 11
    },
    {
      "completion_length": 354.0999984741211,
      "epoch": 0.012441679626749611,
      "grad_norm": 0.694392740726471,
      "kl": 0.00033974647521972656,
      "learning_rate": 1.2413793103448275e-07,
      "loss": 0.0749,
      "reward": 0.3250000011175871,
      "reward_std": 0.3046357296407223,
      "rewards/accuracy_reward": 0.3250000011175871,
      "rewards/format_reward": 0.0,
      "step": 12
    },
    {
      "completion_length": 371.7750015258789,
      "epoch": 0.01347848626231208,
      "grad_norm": 1.2974847555160522,
      "kl": 0.0002536773681640625,
      "learning_rate": 1.3448275862068965e-07,
      "loss": 0.0034,
      "reward": 0.42500000447034836,
      "reward_std": 0.5263643711805344,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.05000000074505806,
      "step": 13
    },
    {
      "completion_length": 377.55001068115234,
      "epoch": 0.014515292897874546,
      "grad_norm": 0.8971682786941528,
      "kl": 0.0002269744873046875,
      "learning_rate": 1.4482758620689654e-07,
      "loss": 0.0266,
      "reward": 0.4000000078231096,
      "reward_std": 0.4415663816034794,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 0.0,
      "step": 14
    },
    {
      "completion_length": 382.6500053405762,
      "epoch": 0.015552099533437015,
      "grad_norm": 1.676780104637146,
      "kl": 0.00034689903259277344,
      "learning_rate": 1.5517241379310344e-07,
      "loss": 0.0303,
      "reward": 0.25000000558793545,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.02500000037252903,
      "step": 15
    },
    {
      "completion_length": 237.32500457763672,
      "epoch": 0.01658890616899948,
      "grad_norm": 0.6732295155525208,
      "kl": 0.0001780986785888672,
      "learning_rate": 1.6551724137931034e-07,
      "loss": 0.0088,
      "reward": 0.2000000048428774,
      "reward_std": 0.31719936057925224,
      "rewards/accuracy_reward": 0.1750000026077032,
      "rewards/format_reward": 0.02500000037252903,
      "step": 16
    },
    {
      "completion_length": 246.77500343322754,
      "epoch": 0.01762571280456195,
      "grad_norm": 0.8578572273254395,
      "kl": 0.00020766258239746094,
      "learning_rate": 1.7586206896551724e-07,
      "loss": -0.0679,
      "reward": 0.2500000037252903,
      "reward_std": 0.23617041110992432,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.0,
      "step": 17
    },
    {
      "completion_length": 287.7250061035156,
      "epoch": 0.01866251944012442,
      "grad_norm": 1.177450180053711,
      "kl": 0.0002694129943847656,
      "learning_rate": 1.8620689655172414e-07,
      "loss": 0.1109,
      "reward": 0.5750000178813934,
      "reward_std": 0.5335852541029453,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.07500000111758709,
      "step": 18
    },
    {
      "completion_length": 276.7000045776367,
      "epoch": 0.019699326075686883,
      "grad_norm": 0.9230583906173706,
      "kl": 0.0002455711364746094,
      "learning_rate": 1.9655172413793106e-07,
      "loss": -0.0118,
      "reward": 0.6000000163912773,
      "reward_std": 0.3731010667979717,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.0,
      "step": 19
    },
    {
      "completion_length": 389.7750015258789,
      "epoch": 0.020736132711249352,
      "grad_norm": 0.8271303772926331,
      "kl": 0.000316619873046875,
      "learning_rate": 2.0689655172413793e-07,
      "loss": -0.0191,
      "reward": 0.450000012293458,
      "reward_std": 0.25609347224235535,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.05000000074505806,
      "step": 20
    },
    {
      "completion_length": 313.29999923706055,
      "epoch": 0.02177293934681182,
      "grad_norm": 3.3089773654937744,
      "kl": 0.0006225109100341797,
      "learning_rate": 2.1724137931034486e-07,
      "loss": -0.0103,
      "reward": 0.4000000096857548,
      "reward_std": 0.26129765808582306,
      "rewards/accuracy_reward": 0.4000000096857548,
      "rewards/format_reward": 0.0,
      "step": 21
    },
    {
      "completion_length": 342.8750114440918,
      "epoch": 0.02280974598237429,
      "grad_norm": 0.981069803237915,
      "kl": 0.0006477832794189453,
      "learning_rate": 2.2758620689655173e-07,
      "loss": 0.0766,
      "reward": 0.5250000059604645,
      "reward_std": 0.44892579317092896,
      "rewards/accuracy_reward": 0.44999999925494194,
      "rewards/format_reward": 0.07500000111758709,
      "step": 22
    },
    {
      "completion_length": 429.30001068115234,
      "epoch": 0.023846552617936754,
      "grad_norm": 0.4381193518638611,
      "kl": 0.0002999305725097656,
      "learning_rate": 2.3793103448275865e-07,
      "loss": -0.018,
      "reward": 0.3750000037252903,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.0,
      "step": 23
    },
    {
      "completion_length": 341.90000915527344,
      "epoch": 0.024883359253499222,
      "grad_norm": 0.8023965954780579,
      "kl": 0.0002903938293457031,
      "learning_rate": 2.482758620689655e-07,
      "loss": 0.0741,
      "reward": 0.5750000048428774,
      "reward_std": 0.36053744703531265,
      "rewards/accuracy_reward": 0.5000000167638063,
      "rewards/format_reward": 0.07500000111758709,
      "step": 24
    },
    {
      "completion_length": 414.6750183105469,
      "epoch": 0.02592016588906169,
      "grad_norm": 1.5727014541625977,
      "kl": 0.00022554397583007812,
      "learning_rate": 2.5862068965517245e-07,
      "loss": -0.0566,
      "reward": 0.42500000819563866,
      "reward_std": 0.5225953236222267,
      "rewards/accuracy_reward": 0.42500000819563866,
      "rewards/format_reward": 0.0,
      "step": 25
    },
    {
      "completion_length": 285.1750030517578,
      "epoch": 0.02695697252462416,
      "grad_norm": 0.7682381272315979,
      "kl": 0.0005116462707519531,
      "learning_rate": 2.689655172413793e-07,
      "loss": -0.0143,
      "reward": 0.35000000335276127,
      "reward_std": 0.45413001254200935,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.02500000037252903,
      "step": 26
    },
    {
      "completion_length": 340.00000762939453,
      "epoch": 0.027993779160186624,
      "grad_norm": 0.6389257907867432,
      "kl": 0.00046062469482421875,
      "learning_rate": 2.7931034482758624e-07,
      "loss": -0.0795,
      "reward": 0.20000000670552254,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.20000000670552254,
      "rewards/format_reward": 0.0,
      "step": 27
    },
    {
      "completion_length": 273.35000228881836,
      "epoch": 0.029030585795749093,
      "grad_norm": 1.1553552150726318,
      "kl": 0.00037384033203125,
      "learning_rate": 2.896551724137931e-07,
      "loss": -0.035,
      "reward": 0.30000000447034836,
      "reward_std": 0.3046357333660126,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.0,
      "step": 28
    },
    {
      "completion_length": 294.79999923706055,
      "epoch": 0.03006739243131156,
      "grad_norm": 2.068281650543213,
      "kl": 0.0003008842468261719,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.0542,
      "reward": 0.8250000029802322,
      "reward_std": 0.3371223881840706,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.05000000074505806,
      "step": 29
    },
    {
      "completion_length": 307.32500076293945,
      "epoch": 0.03110419906687403,
      "grad_norm": 4348.3486328125,
      "kl": 6.1257734298706055,
      "learning_rate": 3.103448275862069e-07,
      "loss": 0.1656,
      "reward": 0.4750000163912773,
      "reward_std": 0.39822830632328987,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.0,
      "step": 30
    },
    {
      "completion_length": 387.575008392334,
      "epoch": 0.0321410057024365,
      "grad_norm": 1.4455534219741821,
      "kl": 0.00043702125549316406,
      "learning_rate": 3.2068965517241384e-07,
      "loss": 0.0563,
      "reward": 0.32500000298023224,
      "reward_std": 0.3966545984148979,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.02500000037252903,
      "step": 31
    },
    {
      "completion_length": 281.0250015258789,
      "epoch": 0.03317781233799896,
      "grad_norm": 1.362587332725525,
      "kl": 0.0006682872772216797,
      "learning_rate": 3.310344827586207e-07,
      "loss": -0.0157,
      "reward": 0.5750000104308128,
      "reward_std": 0.5587125271558762,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.02500000037252903,
      "step": 32
    },
    {
      "completion_length": 451.30001068115234,
      "epoch": 0.03421461897356143,
      "grad_norm": 0.9912571310997009,
      "kl": 0.0003376007080078125,
      "learning_rate": 3.4137931034482763e-07,
      "loss": 0.0678,
      "reward": 0.5000000074505806,
      "reward_std": 0.45413000881671906,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.02500000037252903,
      "step": 33
    },
    {
      "completion_length": 398.0,
      "epoch": 0.0352514256091239,
      "grad_norm": 2.6609644889831543,
      "kl": 0.0003285408020019531,
      "learning_rate": 3.517241379310345e-07,
      "loss": 0.0505,
      "reward": 0.3500000089406967,
      "reward_std": 0.37310105934739113,
      "rewards/accuracy_reward": 0.32500001043081284,
      "rewards/format_reward": 0.02500000037252903,
      "step": 34
    },
    {
      "completion_length": 316.0250015258789,
      "epoch": 0.036288232244686365,
      "grad_norm": 0.6625248193740845,
      "kl": 0.000457763671875,
      "learning_rate": 3.6206896551724143e-07,
      "loss": 0.0455,
      "reward": 0.5500000007450581,
      "reward_std": 0.3856646716594696,
      "rewards/accuracy_reward": 0.47499999590218067,
      "rewards/format_reward": 0.07500000111758709,
      "step": 35
    },
    {
      "completion_length": 335.325008392334,
      "epoch": 0.03732503888024884,
      "grad_norm": 9.054551124572754,
      "kl": 0.0006432533264160156,
      "learning_rate": 3.7241379310344827e-07,
      "loss": 0.0427,
      "reward": 0.40000000037252903,
      "reward_std": 0.32818928733468056,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.05000000074505806,
      "step": 36
    },
    {
      "completion_length": 380.825008392334,
      "epoch": 0.0383618455158113,
      "grad_norm": 0.7966611385345459,
      "kl": 0.0003523826599121094,
      "learning_rate": 3.827586206896552e-07,
      "loss": -0.0201,
      "reward": 0.3250000085681677,
      "reward_std": 0.3856646828353405,
      "rewards/accuracy_reward": 0.25000000931322575,
      "rewards/format_reward": 0.07500000111758709,
      "step": 37
    },
    {
      "completion_length": 338.87500381469727,
      "epoch": 0.03939865215137377,
      "grad_norm": 0.723584771156311,
      "kl": 0.00029540061950683594,
      "learning_rate": 3.931034482758621e-07,
      "loss": 0.0589,
      "reward": 0.4000000059604645,
      "reward_std": 0.553369790315628,
      "rewards/accuracy_reward": 0.32500000670552254,
      "rewards/format_reward": 0.07500000111758709,
      "step": 38
    },
    {
      "completion_length": 285.6000061035156,
      "epoch": 0.04043545878693624,
      "grad_norm": 1.385939121246338,
      "kl": 0.0009665489196777344,
      "learning_rate": 4.0344827586206897e-07,
      "loss": 0.0266,
      "reward": 0.3000000063329935,
      "reward_std": 0.3731010630726814,
      "rewards/accuracy_reward": 0.3000000063329935,
      "rewards/format_reward": 0.0,
      "step": 39
    },
    {
      "completion_length": 342.0750045776367,
      "epoch": 0.041472265422498704,
      "grad_norm": 1.392383098602295,
      "kl": 0.0010743141174316406,
      "learning_rate": 4.1379310344827586e-07,
      "loss": 0.037,
      "reward": 0.42500001564621925,
      "reward_std": 0.4469091147184372,
      "rewards/accuracy_reward": 0.4000000096857548,
      "rewards/format_reward": 0.02500000037252903,
      "step": 40
    },
    {
      "completion_length": 321.325008392334,
      "epoch": 0.04250907205806117,
      "grad_norm": 0.7412877678871155,
      "kl": 0.0005884170532226562,
      "learning_rate": 4.2413793103448276e-07,
      "loss": 0.0443,
      "reward": 0.6000000014901161,
      "reward_std": 0.3856646791100502,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.10000000335276127,
      "step": 41
    },
    {
      "completion_length": 302.37500762939453,
      "epoch": 0.04354587869362364,
      "grad_norm": 0.8022830486297607,
      "kl": 0.0013997554779052734,
      "learning_rate": 4.344827586206897e-07,
      "loss": 0.1065,
      "reward": 0.20000000670552254,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.1750000063329935,
      "rewards/format_reward": 0.02500000037252903,
      "step": 42
    },
    {
      "completion_length": 348.3499984741211,
      "epoch": 0.044582685329186106,
      "grad_norm": 0.5473265647888184,
      "kl": 0.0008740425109863281,
      "learning_rate": 4.4482758620689656e-07,
      "loss": -0.0136,
      "reward": 0.6000000089406967,
      "reward_std": 0.3297629877924919,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.05000000074505806,
      "step": 43
    },
    {
      "completion_length": 475.4000015258789,
      "epoch": 0.04561949196474858,
      "grad_norm": 0.7117322683334351,
      "kl": 0.0007257461547851562,
      "learning_rate": 4.5517241379310346e-07,
      "loss": 0.0024,
      "reward": 0.25000000558793545,
      "reward_std": 0.31719936057925224,
      "rewards/accuracy_reward": 0.25000000558793545,
      "rewards/format_reward": 0.0,
      "step": 44
    },
    {
      "completion_length": 364.4750099182129,
      "epoch": 0.04665629860031104,
      "grad_norm": 0.810202419757843,
      "kl": 0.0007882118225097656,
      "learning_rate": 4.6551724137931035e-07,
      "loss": 0.0617,
      "reward": 0.22500000335276127,
      "reward_std": 0.3479738272726536,
      "rewards/accuracy_reward": 0.2000000048428774,
      "rewards/format_reward": 0.02500000037252903,
      "step": 45
    },
    {
      "completion_length": 276.65000915527344,
      "epoch": 0.04769310523587351,
      "grad_norm": 0.6206042766571045,
      "kl": 0.0006685256958007812,
      "learning_rate": 4.758620689655173e-07,
      "loss": 0.0294,
      "reward": 0.2750000022351742,
      "reward_std": 0.41643913090229034,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.02500000037252903,
      "step": 46
    },
    {
      "completion_length": 358.450008392334,
      "epoch": 0.04872991187143598,
      "grad_norm": 3.410537004470825,
      "kl": 0.002124786376953125,
      "learning_rate": 4.862068965517241e-07,
      "loss": 0.0181,
      "reward": 0.3500000052154064,
      "reward_std": 0.4415663778781891,
      "rewards/accuracy_reward": 0.30000000819563866,
      "rewards/format_reward": 0.05000000074505806,
      "step": 47
    },
    {
      "completion_length": 307.2500114440918,
      "epoch": 0.049766718506998445,
      "grad_norm": 2.8998336791992188,
      "kl": 0.0014314651489257812,
      "learning_rate": 4.96551724137931e-07,
      "loss": 0.0167,
      "reward": 0.6250000074505806,
      "reward_std": 0.558712512254715,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.10000000149011612,
      "step": 48
    },
    {
      "completion_length": 303.54999923706055,
      "epoch": 0.05080352514256091,
      "grad_norm": 0.7844371199607849,
      "kl": 0.0009660720825195312,
      "learning_rate": 5.068965517241379e-07,
      "loss": 0.0628,
      "reward": 0.42500001192092896,
      "reward_std": 0.47768356651067734,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.02500000037252903,
      "step": 49
    },
    {
      "completion_length": 461.3499984741211,
      "epoch": 0.05184033177812338,
      "grad_norm": 0.5534600615501404,
      "kl": 0.002307891845703125,
      "learning_rate": 5.172413793103449e-07,
      "loss": -0.0613,
      "reward": 0.3750000037252903,
      "reward_std": 0.35704541951417923,
      "rewards/accuracy_reward": 0.32500000670552254,
      "rewards/format_reward": 0.05000000074505806,
      "step": 50
    },
    {
      "completion_length": 301.3500061035156,
      "epoch": 0.05287713841368585,
      "grad_norm": 0.9939333200454712,
      "kl": 0.0011472702026367188,
      "learning_rate": 5.275862068965517e-07,
      "loss": -0.0638,
      "reward": 0.5500000081956387,
      "reward_std": 0.38409100100398064,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.05000000074505806,
      "step": 51
    },
    {
      "completion_length": 293.6750030517578,
      "epoch": 0.05391394504924832,
      "grad_norm": 0.6740232110023499,
      "kl": 0.0012159347534179688,
      "learning_rate": 5.379310344827586e-07,
      "loss": 0.0925,
      "reward": 0.5250000152736902,
      "reward_std": 0.5210216604173183,
      "rewards/accuracy_reward": 0.4500000048428774,
      "rewards/format_reward": 0.07500000111758709,
      "step": 52
    },
    {
      "completion_length": 326.12500762939453,
      "epoch": 0.054950751684810784,
      "grad_norm": 0.8232463002204895,
      "kl": 0.0014095306396484375,
      "learning_rate": 5.482758620689655e-07,
      "loss": 0.0658,
      "reward": 0.5250000171363354,
      "reward_std": 0.4415663816034794,
      "rewards/accuracy_reward": 0.45000001415610313,
      "rewards/format_reward": 0.07500000298023224,
      "step": 53
    },
    {
      "completion_length": 285.4000053405762,
      "epoch": 0.05598755832037325,
      "grad_norm": 1.9464935064315796,
      "kl": 0.002483367919921875,
      "learning_rate": 5.586206896551725e-07,
      "loss": -0.1223,
      "reward": 0.32500000670552254,
      "reward_std": 0.4723408371210098,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.07500000111758709,
      "step": 54
    },
    {
      "completion_length": 331.77500915527344,
      "epoch": 0.05702436495593572,
      "grad_norm": 3.9341742992401123,
      "kl": 0.0035686492919921875,
      "learning_rate": 5.689655172413793e-07,
      "loss": 0.0667,
      "reward": 0.4750000163912773,
      "reward_std": 0.36053743213415146,
      "rewards/accuracy_reward": 0.42500000819563866,
      "rewards/format_reward": 0.05000000074505806,
      "step": 55
    },
    {
      "completion_length": 326.8500099182129,
      "epoch": 0.058061171591498185,
      "grad_norm": 1.525079369544983,
      "kl": 0.0031280517578125,
      "learning_rate": 5.793103448275862e-07,
      "loss": -0.0055,
      "reward": 0.37500000186264515,
      "reward_std": 0.37310105934739113,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.02500000037252903,
      "step": 56
    },
    {
      "completion_length": 267.37500762939453,
      "epoch": 0.05909797822706065,
      "grad_norm": 0.823444128036499,
      "kl": 0.002056121826171875,
      "learning_rate": 5.896551724137931e-07,
      "loss": 0.0713,
      "reward": 0.6500000059604645,
      "reward_std": 0.46669362485408783,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.10000000149011612,
      "step": 57
    },
    {
      "completion_length": 222.75,
      "epoch": 0.06013478486262312,
      "grad_norm": 1.267547845840454,
      "kl": 0.0028076171875,
      "learning_rate": 6.000000000000001e-07,
      "loss": 0.0757,
      "reward": 0.600000012665987,
      "reward_std": 0.4092182368040085,
      "rewards/accuracy_reward": 0.5000000167638063,
      "rewards/format_reward": 0.10000000149011612,
      "step": 58
    },
    {
      "completion_length": 252.62500762939453,
      "epoch": 0.06117159149818559,
      "grad_norm": 1.6207536458969116,
      "kl": 0.0028018951416015625,
      "learning_rate": 6.103448275862069e-07,
      "loss": 0.0935,
      "reward": 0.3750000074505806,
      "reward_std": 0.4614894315600395,
      "rewards/accuracy_reward": 0.3000000026077032,
      "rewards/format_reward": 0.07500000111758709,
      "step": 59
    },
    {
      "completion_length": 295.5250053405762,
      "epoch": 0.06220839813374806,
      "grad_norm": 1.536206603050232,
      "kl": 0.010120391845703125,
      "learning_rate": 6.206896551724138e-07,
      "loss": 0.0602,
      "reward": 0.6750000156462193,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.05000000074505806,
      "step": 60
    },
    {
      "completion_length": 265.5250015258789,
      "epoch": 0.06324520476931052,
      "grad_norm": 14.682554244995117,
      "kl": 0.020404815673828125,
      "learning_rate": 6.310344827586207e-07,
      "loss": -0.0303,
      "reward": 0.6000000089406967,
      "reward_std": 0.5896254852414131,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.12500000186264515,
      "step": 61
    },
    {
      "completion_length": 288.37500762939453,
      "epoch": 0.064282011404873,
      "grad_norm": 5.919084548950195,
      "kl": 0.006771087646484375,
      "learning_rate": 6.413793103448277e-07,
      "loss": 0.0497,
      "reward": 0.5500000044703484,
      "reward_std": 0.4704626724123955,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.10000000149011612,
      "step": 62
    },
    {
      "completion_length": 301.6250114440918,
      "epoch": 0.06531881804043546,
      "grad_norm": 16.421817779541016,
      "kl": 0.005756378173828125,
      "learning_rate": 6.517241379310345e-07,
      "loss": 0.1564,
      "reward": 0.6250000149011612,
      "reward_std": 0.5894869714975357,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.17500000447034836,
      "step": 63
    },
    {
      "completion_length": 284.77500915527344,
      "epoch": 0.06635562467599793,
      "grad_norm": 7.291428565979004,
      "kl": 0.0056972503662109375,
      "learning_rate": 6.620689655172414e-07,
      "loss": 0.0605,
      "reward": 0.5250000022351742,
      "reward_std": 0.46148940920829773,
      "rewards/accuracy_reward": 0.37499999813735485,
      "rewards/format_reward": 0.15000000223517418,
      "step": 64
    },
    {
      "completion_length": 343.8000030517578,
      "epoch": 0.06739243131156039,
      "grad_norm": 6.103272914886475,
      "kl": 0.004486083984375,
      "learning_rate": 6.724137931034483e-07,
      "loss": -0.0342,
      "reward": 0.4750000089406967,
      "reward_std": 0.5461488887667656,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.02500000037252903,
      "step": 65
    },
    {
      "completion_length": 276.20001220703125,
      "epoch": 0.06842923794712286,
      "grad_norm": 0.6578114032745361,
      "kl": 0.004322052001953125,
      "learning_rate": 6.827586206896553e-07,
      "loss": 0.1695,
      "reward": 0.3000000063329935,
      "reward_std": 0.47234082594513893,
      "rewards/accuracy_reward": 0.15000000409781933,
      "rewards/format_reward": 0.15000000223517418,
      "step": 66
    },
    {
      "completion_length": 246.37500381469727,
      "epoch": 0.06946604458268534,
      "grad_norm": 7.143874645233154,
      "kl": 0.004947662353515625,
      "learning_rate": 6.931034482758621e-07,
      "loss": 0.0603,
      "reward": 0.45000001043081284,
      "reward_std": 0.4902471899986267,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.10000000149011612,
      "step": 67
    },
    {
      "completion_length": 179.375,
      "epoch": 0.0705028512182478,
      "grad_norm": 0.9650272727012634,
      "kl": 0.00789642333984375,
      "learning_rate": 7.03448275862069e-07,
      "loss": 0.1452,
      "reward": 0.75,
      "reward_std": 0.5985585674643517,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.27500000409781933,
      "step": 68
    },
    {
      "completion_length": 284.3250045776367,
      "epoch": 0.07153965785381027,
      "grad_norm": 1.4530069828033447,
      "kl": 0.0058746337890625,
      "learning_rate": 7.137931034482759e-07,
      "loss": -0.0084,
      "reward": 0.775000024586916,
      "reward_std": 0.5697024464607239,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 0.2000000048428774,
      "step": 69
    },
    {
      "completion_length": 290.325008392334,
      "epoch": 0.07257646448937273,
      "grad_norm": 0.7363834381103516,
      "kl": 0.01473236083984375,
      "learning_rate": 7.241379310344829e-07,
      "loss": -0.0061,
      "reward": 0.4000000078231096,
      "reward_std": 0.41643914952874184,
      "rewards/accuracy_reward": 0.3250000011175871,
      "rewards/format_reward": 0.07500000111758709,
      "step": 70
    },
    {
      "completion_length": 194.20000076293945,
      "epoch": 0.0736132711249352,
      "grad_norm": 1.855769395828247,
      "kl": 0.01556396484375,
      "learning_rate": 7.344827586206896e-07,
      "loss": 0.0088,
      "reward": 0.8250000216066837,
      "reward_std": 0.5607292056083679,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.35000001080334187,
      "step": 71
    },
    {
      "completion_length": 244.00000381469727,
      "epoch": 0.07465007776049767,
      "grad_norm": 1.9801136255264282,
      "kl": 0.0140533447265625,
      "learning_rate": 7.448275862068965e-07,
      "loss": 0.0587,
      "reward": 0.9500000327825546,
      "reward_std": 0.5624414160847664,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.30000000819563866,
      "step": 72
    },
    {
      "completion_length": 223.0250015258789,
      "epoch": 0.07568688439606014,
      "grad_norm": 0.857162594795227,
      "kl": 0.01824951171875,
      "learning_rate": 7.551724137931035e-07,
      "loss": 0.0944,
      "reward": 0.7750000059604645,
      "reward_std": 0.5263643562793732,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.350000012665987,
      "step": 73
    },
    {
      "completion_length": 267.02500915527344,
      "epoch": 0.0767236910316226,
      "grad_norm": 1.467808723449707,
      "kl": 0.0139923095703125,
      "learning_rate": 7.655172413793104e-07,
      "loss": 0.0959,
      "reward": 0.8000000268220901,
      "reward_std": 0.4850429780781269,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.4000000096857548,
      "step": 74
    },
    {
      "completion_length": 222.25000381469727,
      "epoch": 0.07776049766718507,
      "grad_norm": 1.6067920923233032,
      "kl": 0.0166168212890625,
      "learning_rate": 7.758620689655173e-07,
      "loss": 0.1242,
      "reward": 0.7500000037252903,
      "reward_std": 0.6815058067440987,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.3500000014901161,
      "step": 75
    },
    {
      "completion_length": 199.7750015258789,
      "epoch": 0.07879730430274753,
      "grad_norm": 1.2611151933670044,
      "kl": 0.03570556640625,
      "learning_rate": 7.862068965517242e-07,
      "loss": 0.0485,
      "reward": 0.8250000178813934,
      "reward_std": 0.582266092300415,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.4000000134110451,
      "step": 76
    },
    {
      "completion_length": 173.07500457763672,
      "epoch": 0.07983411093831,
      "grad_norm": 3.898858070373535,
      "kl": 0.02545166015625,
      "learning_rate": 7.96551724137931e-07,
      "loss": 0.2017,
      "reward": 1.0250000208616257,
      "reward_std": 0.5948297046124935,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.5750000029802322,
      "step": 77
    },
    {
      "completion_length": 187.725004196167,
      "epoch": 0.08087091757387248,
      "grad_norm": 0.8916759490966797,
      "kl": 0.040130615234375,
      "learning_rate": 8.068965517241379e-07,
      "loss": 0.0355,
      "reward": 1.1250000149011612,
      "reward_std": 0.6099044531583786,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.45000000670552254,
      "step": 78
    },
    {
      "completion_length": 177.35000610351562,
      "epoch": 0.08190772420943494,
      "grad_norm": 1.327777624130249,
      "kl": 0.04638671875,
      "learning_rate": 8.172413793103448e-07,
      "loss": 0.0438,
      "reward": 0.9500000178813934,
      "reward_std": 0.7595372572541237,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.6000000089406967,
      "step": 79
    },
    {
      "completion_length": 183.1750030517578,
      "epoch": 0.08294453084499741,
      "grad_norm": 1.4906061887741089,
      "kl": 0.054718017578125,
      "learning_rate": 8.275862068965517e-07,
      "loss": 0.1092,
      "reward": 0.8500000089406967,
      "reward_std": 0.6130405142903328,
      "rewards/accuracy_reward": 0.15000000223517418,
      "rewards/format_reward": 0.7000000029802322,
      "step": 80
    },
    {
      "completion_length": 199.35000228881836,
      "epoch": 0.08398133748055987,
      "grad_norm": 3.892275333404541,
      "kl": 0.0404052734375,
      "learning_rate": 8.379310344827587e-07,
      "loss": 0.1296,
      "reward": 1.175000011920929,
      "reward_std": 0.6471008583903313,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.7250000089406967,
      "step": 81
    },
    {
      "completion_length": 141.80000114440918,
      "epoch": 0.08501814411612234,
      "grad_norm": 0.9681069254875183,
      "kl": 0.091522216796875,
      "learning_rate": 8.482758620689655e-07,
      "loss": 0.0576,
      "reward": 1.0500000417232513,
      "reward_std": 0.5101702138781548,
      "rewards/accuracy_reward": 0.32500000670552254,
      "rewards/format_reward": 0.7249999940395355,
      "step": 82
    },
    {
      "completion_length": 141.00000190734863,
      "epoch": 0.08605495075168482,
      "grad_norm": 1.6170580387115479,
      "kl": 0.0736083984375,
      "learning_rate": 8.586206896551725e-07,
      "loss": 0.091,
      "reward": 1.1250000149011612,
      "reward_std": 0.5112976878881454,
      "rewards/accuracy_reward": 0.3500000052154064,
      "rewards/format_reward": 0.7750000059604645,
      "step": 83
    },
    {
      "completion_length": 226.37499809265137,
      "epoch": 0.08709175738724728,
      "grad_norm": 2.232787847518921,
      "kl": 0.05548095703125,
      "learning_rate": 8.689655172413794e-07,
      "loss": 0.1663,
      "reward": 1.2000000327825546,
      "reward_std": 0.5984200611710548,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 0.7500000149011612,
      "step": 84
    },
    {
      "completion_length": 137.9500026702881,
      "epoch": 0.08812856402280975,
      "grad_norm": 1.994882583618164,
      "kl": 0.058135986328125,
      "learning_rate": 8.793103448275862e-07,
      "loss": 0.0491,
      "reward": 1.4500000029802322,
      "reward_std": 0.44892580062150955,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.800000011920929,
      "step": 85
    },
    {
      "completion_length": 121.20000076293945,
      "epoch": 0.08916537065837221,
      "grad_norm": 1.0910072326660156,
      "kl": 0.056640625,
      "learning_rate": 8.896551724137931e-07,
      "loss": 0.054,
      "reward": 1.175000011920929,
      "reward_std": 0.40558772534132004,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.800000011920929,
      "step": 86
    },
    {
      "completion_length": 148.9000015258789,
      "epoch": 0.09020217729393468,
      "grad_norm": 2.2213571071624756,
      "kl": 0.220458984375,
      "learning_rate": 9e-07,
      "loss": 0.0789,
      "reward": 1.3250000327825546,
      "reward_std": 0.5624815374612808,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.7750000059604645,
      "step": 87
    },
    {
      "completion_length": 157.7750015258789,
      "epoch": 0.09123898392949716,
      "grad_norm": 1.2787593603134155,
      "kl": 0.067138671875,
      "learning_rate": 9.103448275862069e-07,
      "loss": 0.0193,
      "reward": 1.2250000536441803,
      "reward_std": 0.510664701461792,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.7500000149011612,
      "step": 88
    },
    {
      "completion_length": 141.32500076293945,
      "epoch": 0.09227579056505962,
      "grad_norm": 1.5759501457214355,
      "kl": 0.109130859375,
      "learning_rate": 9.206896551724139e-07,
      "loss": 0.1089,
      "reward": 1.4500000476837158,
      "reward_std": 0.3804604858160019,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.8500000238418579,
      "step": 89
    },
    {
      "completion_length": 142.72500228881836,
      "epoch": 0.09331259720062209,
      "grad_norm": 0.9323010444641113,
      "kl": 0.079803466796875,
      "learning_rate": 9.310344827586207e-07,
      "loss": 0.0734,
      "reward": 1.3000000417232513,
      "reward_std": 0.4657529257237911,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.8500000238418579,
      "step": 90
    },
    {
      "completion_length": 122.37500190734863,
      "epoch": 0.09434940383618455,
      "grad_norm": 1.3933780193328857,
      "kl": 0.1185302734375,
      "learning_rate": 9.413793103448277e-07,
      "loss": 0.0271,
      "reward": 1.2000000476837158,
      "reward_std": 0.6617213189601898,
      "rewards/accuracy_reward": 0.3500000052154064,
      "rewards/format_reward": 0.8500000238418579,
      "step": 91
    },
    {
      "completion_length": 126.70000457763672,
      "epoch": 0.09538621047174702,
      "grad_norm": 7.927647113800049,
      "kl": 0.4715576171875,
      "learning_rate": 9.517241379310346e-07,
      "loss": 0.0897,
      "reward": 1.2000000178813934,
      "reward_std": 0.6092714741826057,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 0.75,
      "step": 92
    },
    {
      "completion_length": 172.30000114440918,
      "epoch": 0.09642301710730948,
      "grad_norm": 1.3404741287231445,
      "kl": 0.078857421875,
      "learning_rate": 9.620689655172413e-07,
      "loss": 0.0283,
      "reward": 1.3750000298023224,
      "reward_std": 0.46511995419859886,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 93
    },
    {
      "completion_length": 113.35000038146973,
      "epoch": 0.09745982374287196,
      "grad_norm": 7.312382221221924,
      "kl": 0.0780029296875,
      "learning_rate": 9.724137931034483e-07,
      "loss": 0.0145,
      "reward": 1.4000000059604645,
      "reward_std": 0.4092182442545891,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 94
    },
    {
      "completion_length": 109.42500114440918,
      "epoch": 0.09849663037843442,
      "grad_norm": 1.404718041419983,
      "kl": 0.08563232421875,
      "learning_rate": 9.827586206896552e-07,
      "loss": 0.072,
      "reward": 1.4250000417232513,
      "reward_std": 0.4740530401468277,
      "rewards/accuracy_reward": 0.45000001788139343,
      "rewards/format_reward": 0.9750000089406967,
      "step": 95
    },
    {
      "completion_length": 117.55000305175781,
      "epoch": 0.09953343701399689,
      "grad_norm": 1.4174925088882446,
      "kl": 0.1512451171875,
      "learning_rate": 9.93103448275862e-07,
      "loss": -0.0339,
      "reward": 1.4000000357627869,
      "reward_std": 0.578635536134243,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.8250000178813934,
      "step": 96
    },
    {
      "completion_length": 105.30000114440918,
      "epoch": 0.10057024364955935,
      "grad_norm": 6.377532005310059,
      "kl": 0.1229248046875,
      "learning_rate": 1.003448275862069e-06,
      "loss": -0.0262,
      "reward": 1.3500000536441803,
      "reward_std": 0.4958944097161293,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.8500000238418579,
      "step": 97
    },
    {
      "completion_length": 122.80000114440918,
      "epoch": 0.10160705028512182,
      "grad_norm": 3.010532855987549,
      "kl": 0.09405517578125,
      "learning_rate": 1.0137931034482759e-06,
      "loss": 0.0275,
      "reward": 1.3250000178813934,
      "reward_std": 0.2795085050165653,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9250000268220901,
      "step": 98
    },
    {
      "completion_length": 117.97500038146973,
      "epoch": 0.1026438569206843,
      "grad_norm": 1.6175012588500977,
      "kl": 0.09100341796875,
      "learning_rate": 1.0241379310344828e-06,
      "loss": 0.0431,
      "reward": 1.2500000447034836,
      "reward_std": 0.5535082966089249,
      "rewards/accuracy_reward": 0.3750000111758709,
      "rewards/format_reward": 0.875,
      "step": 99
    },
    {
      "completion_length": 123.85000419616699,
      "epoch": 0.10368066355624676,
      "grad_norm": 1.486450433731079,
      "kl": 0.08172607421875,
      "learning_rate": 1.0344827586206898e-06,
      "loss": 0.0611,
      "reward": 1.475000023841858,
      "reward_std": 0.4525563009083271,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.8500000089406967,
      "step": 100
    },
    {
      "completion_length": 118.80000114440918,
      "epoch": 0.10471747019180923,
      "grad_norm": 1.8546826839447021,
      "kl": 0.07427978515625,
      "learning_rate": 1.0448275862068965e-06,
      "loss": -0.0648,
      "reward": 1.2500000298023224,
      "reward_std": 0.45412999391555786,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.8500000089406967,
      "step": 101
    },
    {
      "completion_length": 107.62500190734863,
      "epoch": 0.1057542768273717,
      "grad_norm": 1.0969343185424805,
      "kl": 0.1190185546875,
      "learning_rate": 1.0551724137931035e-06,
      "loss": 0.0655,
      "reward": 1.4000000059604645,
      "reward_std": 0.3804604709148407,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9500000178813934,
      "step": 102
    },
    {
      "completion_length": 125.30000114440918,
      "epoch": 0.10679108346293416,
      "grad_norm": 2.0286576747894287,
      "kl": 0.1263427734375,
      "learning_rate": 1.0655172413793104e-06,
      "loss": 0.1098,
      "reward": 1.350000023841858,
      "reward_std": 0.44156638905406,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.9000000357627869,
      "step": 103
    },
    {
      "completion_length": 115.07500076293945,
      "epoch": 0.10782789009849664,
      "grad_norm": 2.2312369346618652,
      "kl": 0.1448974609375,
      "learning_rate": 1.0758620689655172e-06,
      "loss": 0.0629,
      "reward": 1.3250000178813934,
      "reward_std": 0.47234081476926804,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.9500000178813934,
      "step": 104
    },
    {
      "completion_length": 119.22500228881836,
      "epoch": 0.1088646967340591,
      "grad_norm": 0.6519957184791565,
      "kl": 0.0863037109375,
      "learning_rate": 1.0862068965517241e-06,
      "loss": -0.0487,
      "reward": 1.3500000089406967,
      "reward_std": 0.33353203162550926,
      "rewards/accuracy_reward": 0.4500000011175871,
      "rewards/format_reward": 0.9000000059604645,
      "step": 105
    },
    {
      "completion_length": 123.47500228881836,
      "epoch": 0.10990150336962157,
      "grad_norm": 1.1476188898086548,
      "kl": 0.10223388671875,
      "learning_rate": 1.096551724137931e-06,
      "loss": 0.0576,
      "reward": 1.2500000149011612,
      "reward_std": 0.3533165417611599,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.9000000208616257,
      "step": 106
    },
    {
      "completion_length": 113.65000343322754,
      "epoch": 0.11093831000518403,
      "grad_norm": 1.9051084518432617,
      "kl": 0.08203125,
      "learning_rate": 1.106896551724138e-06,
      "loss": 0.0674,
      "reward": 1.4000000357627869,
      "reward_std": 0.5806522481143475,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.800000011920929,
      "step": 107
    },
    {
      "completion_length": 123.80000305175781,
      "epoch": 0.1119751166407465,
      "grad_norm": 1.0134373903274536,
      "kl": 0.1263427734375,
      "learning_rate": 1.117241379310345e-06,
      "loss": 0.1178,
      "reward": 1.6500000357627869,
      "reward_std": 0.47234082967042923,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.8750000298023224,
      "step": 108
    },
    {
      "completion_length": 189.0250072479248,
      "epoch": 0.11301192327630896,
      "grad_norm": 2.021437883377075,
      "kl": 0.06475830078125,
      "learning_rate": 1.1275862068965517e-06,
      "loss": 0.046,
      "reward": 1.2750000357627869,
      "reward_std": 0.6544602662324905,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.8000000268220901,
      "step": 109
    },
    {
      "completion_length": 125.30000495910645,
      "epoch": 0.11404872991187144,
      "grad_norm": 1.5949187278747559,
      "kl": 0.07537841796875,
      "learning_rate": 1.1379310344827587e-06,
      "loss": 0.1211,
      "reward": 1.2750000059604645,
      "reward_std": 0.5409446507692337,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.800000011920929,
      "step": 110
    },
    {
      "completion_length": 112.60000228881836,
      "epoch": 0.1150855365474339,
      "grad_norm": 0.9887787103652954,
      "kl": 0.09454345703125,
      "learning_rate": 1.1482758620689656e-06,
      "loss": 0.0565,
      "reward": 1.225000023841858,
      "reward_std": 0.5337237864732742,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.8750000298023224,
      "step": 111
    },
    {
      "completion_length": 113.90000534057617,
      "epoch": 0.11612234318299637,
      "grad_norm": 1.2379708290100098,
      "kl": 0.1116943359375,
      "learning_rate": 1.1586206896551724e-06,
      "loss": 0.0541,
      "reward": 1.225000023841858,
      "reward_std": 0.5572371743619442,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.7750000208616257,
      "step": 112
    },
    {
      "completion_length": 140.9500026702881,
      "epoch": 0.11715914981855884,
      "grad_norm": 2.4870309829711914,
      "kl": 0.076171875,
      "learning_rate": 1.1689655172413793e-06,
      "loss": 0.0255,
      "reward": 1.4750000536441803,
      "reward_std": 0.4415663853287697,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.8250000178813934,
      "step": 113
    },
    {
      "completion_length": 143.625,
      "epoch": 0.1181959564541213,
      "grad_norm": 2.575742721557617,
      "kl": 0.05718994140625,
      "learning_rate": 1.1793103448275863e-06,
      "loss": 0.071,
      "reward": 1.225000023841858,
      "reward_std": 0.562441386282444,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.800000011920929,
      "step": 114
    },
    {
      "completion_length": 150.72500228881836,
      "epoch": 0.11923276308968378,
      "grad_norm": 4.119329929351807,
      "kl": 0.07586669921875,
      "learning_rate": 1.1896551724137932e-06,
      "loss": 0.0335,
      "reward": 1.2000000178813934,
      "reward_std": 0.5210216380655766,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.800000011920929,
      "step": 115
    },
    {
      "completion_length": 129.87500190734863,
      "epoch": 0.12026956972524624,
      "grad_norm": 0.9842166304588318,
      "kl": 0.0792236328125,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.0267,
      "reward": 1.3750000298023224,
      "reward_std": 0.6417581215500832,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.8250000178813934,
      "step": 116
    },
    {
      "completion_length": 125.87500381469727,
      "epoch": 0.12130637636080871,
      "grad_norm": 1.5760966539382935,
      "kl": 0.0994873046875,
      "learning_rate": 1.210344827586207e-06,
      "loss": 0.0867,
      "reward": 1.2000000029802322,
      "reward_std": 0.6293330267071724,
      "rewards/accuracy_reward": 0.4500000011175871,
      "rewards/format_reward": 0.75,
      "step": 117
    },
    {
      "completion_length": 144.975004196167,
      "epoch": 0.12234318299637117,
      "grad_norm": 0.8227043747901917,
      "kl": 0.0821533203125,
      "learning_rate": 1.2206896551724138e-06,
      "loss": 0.0002,
      "reward": 1.3500000089406967,
      "reward_std": 0.472973819822073,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.800000011920929,
      "step": 118
    },
    {
      "completion_length": 151.1500015258789,
      "epoch": 0.12337998963193364,
      "grad_norm": 3.4477927684783936,
      "kl": 0.11090087890625,
      "learning_rate": 1.2310344827586206e-06,
      "loss": 0.057,
      "reward": 1.5000000596046448,
      "reward_std": 0.5665664002299309,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.7750000059604645,
      "step": 119
    },
    {
      "completion_length": 133.72500228881836,
      "epoch": 0.12441679626749612,
      "grad_norm": 1.916924238204956,
      "kl": 0.14306640625,
      "learning_rate": 1.2413793103448275e-06,
      "loss": 0.1373,
      "reward": 1.4749999940395355,
      "reward_std": 0.41643914580345154,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.800000011920929,
      "step": 120
    },
    {
      "completion_length": 144.70000457763672,
      "epoch": 0.12545360290305857,
      "grad_norm": 1.4124659299850464,
      "kl": 0.11669921875,
      "learning_rate": 1.2517241379310345e-06,
      "loss": 0.0701,
      "reward": 1.600000023841858,
      "reward_std": 0.5210216268897057,
      "rewards/accuracy_reward": 0.7500000298023224,
      "rewards/format_reward": 0.8500000238418579,
      "step": 121
    },
    {
      "completion_length": 143.50000381469727,
      "epoch": 0.12649040953862103,
      "grad_norm": 2.2720749378204346,
      "kl": 0.09759521484375,
      "learning_rate": 1.2620689655172414e-06,
      "loss": 0.0201,
      "reward": 1.449999988079071,
      "reward_std": 0.5462873950600624,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.7750000059604645,
      "step": 122
    },
    {
      "completion_length": 125.87500190734863,
      "epoch": 0.12752721617418353,
      "grad_norm": 1.0222535133361816,
      "kl": 0.07147216796875,
      "learning_rate": 1.2724137931034484e-06,
      "loss": 0.0006,
      "reward": 1.4500000476837158,
      "reward_std": 0.3856646865606308,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.9500000029802322,
      "step": 123
    },
    {
      "completion_length": 145.07500648498535,
      "epoch": 0.128564022809746,
      "grad_norm": 4.159966945648193,
      "kl": 0.08453369140625,
      "learning_rate": 1.2827586206896553e-06,
      "loss": 0.0968,
      "reward": 1.300000011920929,
      "reward_std": 0.5659333989024162,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.7750000208616257,
      "step": 124
    },
    {
      "completion_length": 118.10000038146973,
      "epoch": 0.12960082944530846,
      "grad_norm": 1.1424363851547241,
      "kl": 0.0953369140625,
      "learning_rate": 1.293103448275862e-06,
      "loss": -0.0023,
      "reward": 1.3250000178813934,
      "reward_std": 0.3245587684214115,
      "rewards/accuracy_reward": 0.3999999985098839,
      "rewards/format_reward": 0.925000011920929,
      "step": 125
    },
    {
      "completion_length": 108.20000076293945,
      "epoch": 0.13063763608087092,
      "grad_norm": 3.2891056537628174,
      "kl": 0.100830078125,
      "learning_rate": 1.303448275862069e-06,
      "loss": -0.036,
      "reward": 1.2750000208616257,
      "reward_std": 0.37323955446481705,
      "rewards/accuracy_reward": 0.40000000409781933,
      "rewards/format_reward": 0.8750000149011612,
      "step": 126
    },
    {
      "completion_length": 128.57500457763672,
      "epoch": 0.1316744427164334,
      "grad_norm": 1.2664824724197388,
      "kl": 0.07080078125,
      "learning_rate": 1.3137931034482758e-06,
      "loss": 0.0349,
      "reward": 1.6750000417232513,
      "reward_std": 0.36053745076060295,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.925000011920929,
      "step": 127
    },
    {
      "completion_length": 145.00000190734863,
      "epoch": 0.13271124935199585,
      "grad_norm": 1.661323070526123,
      "kl": 0.08135986328125,
      "learning_rate": 1.3241379310344827e-06,
      "loss": 0.0568,
      "reward": 1.5000000298023224,
      "reward_std": 0.4850430078804493,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.8750000298023224,
      "step": 128
    },
    {
      "completion_length": 129.42500114440918,
      "epoch": 0.13374805598755832,
      "grad_norm": 1.649943232536316,
      "kl": 0.0888671875,
      "learning_rate": 1.3344827586206897e-06,
      "loss": -0.0105,
      "reward": 1.3750000298023224,
      "reward_std": 0.36053745076060295,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 129
    },
    {
      "completion_length": 119.15000343322754,
      "epoch": 0.13478486262312078,
      "grad_norm": 6.876035690307617,
      "kl": 0.19091796875,
      "learning_rate": 1.3448275862068966e-06,
      "loss": -0.0036,
      "reward": 1.475000023841858,
      "reward_std": 0.4849044345319271,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8250000178813934,
      "step": 130
    },
    {
      "completion_length": 108.72500038146973,
      "epoch": 0.13582166925868325,
      "grad_norm": 1.3705264329910278,
      "kl": 0.1119384765625,
      "learning_rate": 1.3551724137931036e-06,
      "loss": 0.0473,
      "reward": 1.1250000149011612,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.17500000074505806,
      "rewards/format_reward": 0.9500000029802322,
      "step": 131
    },
    {
      "completion_length": 132.72500038146973,
      "epoch": 0.1368584758942457,
      "grad_norm": 9.587868690490723,
      "kl": 0.10333251953125,
      "learning_rate": 1.3655172413793105e-06,
      "loss": 0.1013,
      "reward": 1.3750000596046448,
      "reward_std": 0.5335852727293968,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.8750000298023224,
      "step": 132
    },
    {
      "completion_length": 122.40000343322754,
      "epoch": 0.13789528252980818,
      "grad_norm": 5.267590045928955,
      "kl": 0.120849609375,
      "learning_rate": 1.3758620689655173e-06,
      "loss": -0.0325,
      "reward": 1.475000023841858,
      "reward_std": 0.49746808409690857,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 133
    },
    {
      "completion_length": 121.22500228881836,
      "epoch": 0.13893208916537067,
      "grad_norm": 2.1384124755859375,
      "kl": 0.184814453125,
      "learning_rate": 1.3862068965517242e-06,
      "loss": 0.0501,
      "reward": 1.2750000357627869,
      "reward_std": 0.4291412718594074,
      "rewards/accuracy_reward": 0.4250000137835741,
      "rewards/format_reward": 0.8500000089406967,
      "step": 134
    },
    {
      "completion_length": 108.37500381469727,
      "epoch": 0.13996889580093314,
      "grad_norm": 4.733362674713135,
      "kl": 0.0623779296875,
      "learning_rate": 1.396551724137931e-06,
      "loss": 0.0074,
      "reward": 1.550000011920929,
      "reward_std": 0.23617041110992432,
      "rewards/accuracy_reward": 0.5749999973922968,
      "rewards/format_reward": 0.9750000089406967,
      "step": 135
    },
    {
      "completion_length": 133.85000038146973,
      "epoch": 0.1410057024364956,
      "grad_norm": 1.9882097244262695,
      "kl": 0.07232666015625,
      "learning_rate": 1.406896551724138e-06,
      "loss": 0.048,
      "reward": 1.574999988079071,
      "reward_std": 0.4290027506649494,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 136
    },
    {
      "completion_length": 142.67500114440918,
      "epoch": 0.14204250907205807,
      "grad_norm": 2.703753709793091,
      "kl": 0.1512451171875,
      "learning_rate": 1.4172413793103449e-06,
      "loss": 0.0586,
      "reward": 1.375,
      "reward_std": 0.4578990526497364,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.800000011920929,
      "step": 137
    },
    {
      "completion_length": 119.35000228881836,
      "epoch": 0.14307931570762053,
      "grad_norm": 2.8184866905212402,
      "kl": 0.10125732421875,
      "learning_rate": 1.4275862068965518e-06,
      "loss": -0.0393,
      "reward": 1.5000000298023224,
      "reward_std": 0.4290027655661106,
      "rewards/accuracy_reward": 0.5250000208616257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 138
    },
    {
      "completion_length": 111.85000228881836,
      "epoch": 0.144116122343183,
      "grad_norm": 4.2106242179870605,
      "kl": 0.1044921875,
      "learning_rate": 1.4379310344827588e-06,
      "loss": 0.0386,
      "reward": 1.0750000178813934,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.25000000186264515,
      "rewards/format_reward": 0.8250000178813934,
      "step": 139
    },
    {
      "completion_length": 117.80000305175781,
      "epoch": 0.14515292897874546,
      "grad_norm": 4.182285785675049,
      "kl": 0.0821533203125,
      "learning_rate": 1.4482758620689657e-06,
      "loss": -0.0056,
      "reward": 1.3500000536441803,
      "reward_std": 0.5786355212330818,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.925000011920929,
      "step": 140
    },
    {
      "completion_length": 135.22500228881836,
      "epoch": 0.14618973561430793,
      "grad_norm": 1.8042055368423462,
      "kl": 0.1070556640625,
      "learning_rate": 1.4586206896551724e-06,
      "loss": 0.0296,
      "reward": 1.5,
      "reward_std": 0.3856646940112114,
      "rewards/accuracy_reward": 0.5500000063329935,
      "rewards/format_reward": 0.9500000178813934,
      "step": 141
    },
    {
      "completion_length": 126.97500228881836,
      "epoch": 0.1472265422498704,
      "grad_norm": 4.371006011962891,
      "kl": 0.09185791015625,
      "learning_rate": 1.4689655172413792e-06,
      "loss": 0.0361,
      "reward": 1.75,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 142
    },
    {
      "completion_length": 131.6750030517578,
      "epoch": 0.14826334888543286,
      "grad_norm": 17.638200759887695,
      "kl": 0.11883544921875,
      "learning_rate": 1.4793103448275861e-06,
      "loss": 0.0239,
      "reward": 1.375,
      "reward_std": 0.24873404577374458,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9500000178813934,
      "step": 143
    },
    {
      "completion_length": 135.87500381469727,
      "epoch": 0.14930015552099535,
      "grad_norm": 10.907958984375,
      "kl": 0.119384765625,
      "learning_rate": 1.489655172413793e-06,
      "loss": 0.0428,
      "reward": 1.2250000089406967,
      "reward_std": 0.3930240869522095,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.8500000089406967,
      "step": 144
    },
    {
      "completion_length": 147.6250057220459,
      "epoch": 0.1503369621565578,
      "grad_norm": 921.281005859375,
      "kl": 3.2681884765625,
      "learning_rate": 1.5e-06,
      "loss": 0.2005,
      "reward": 1.225000038743019,
      "reward_std": 0.30463574081659317,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 145
    },
    {
      "completion_length": 148.72500228881836,
      "epoch": 0.15137376879212028,
      "grad_norm": 14.504183769226074,
      "kl": 0.0855712890625,
      "learning_rate": 1.510344827586207e-06,
      "loss": 0.0652,
      "reward": 1.675000011920929,
      "reward_std": 0.34448179230093956,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 146
    },
    {
      "completion_length": 173.975004196167,
      "epoch": 0.15241057542768274,
      "grad_norm": 251.97561645507812,
      "kl": 0.8280029296875,
      "learning_rate": 1.520689655172414e-06,
      "loss": 0.0497,
      "reward": 1.5250000059604645,
      "reward_std": 0.4343454875051975,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8750000149011612,
      "step": 147
    },
    {
      "completion_length": 180.57500457763672,
      "epoch": 0.1534473820632452,
      "grad_norm": 206.3589630126953,
      "kl": 0.494384765625,
      "learning_rate": 1.5310344827586209e-06,
      "loss": 0.1159,
      "reward": 1.2500000149011612,
      "reward_std": 0.37310104444622993,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.800000011920929,
      "step": 148
    },
    {
      "completion_length": 170.9250030517578,
      "epoch": 0.15448418869880767,
      "grad_norm": 61797.4140625,
      "kl": 1144.1375732421875,
      "learning_rate": 1.5413793103448276e-06,
      "loss": 27.6538,
      "reward": 0.9000000208616257,
      "reward_std": 0.5481655634939671,
      "rewards/accuracy_reward": 0.1750000026077032,
      "rewards/format_reward": 0.7250000089406967,
      "step": 149
    },
    {
      "completion_length": 165.55000495910645,
      "epoch": 0.15552099533437014,
      "grad_norm": 886645.125,
      "kl": 669.49169921875,
      "learning_rate": 1.5517241379310346e-06,
      "loss": 31.4775,
      "reward": 1.2750000208616257,
      "reward_std": 0.49226387962698936,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.800000011920929,
      "step": 150
    },
    {
      "completion_length": 118.1500015258789,
      "epoch": 0.1565578019699326,
      "grad_norm": 120.34538269042969,
      "kl": 1.54296875,
      "learning_rate": 1.5620689655172415e-06,
      "loss": 0.0637,
      "reward": 1.425000011920929,
      "reward_std": 0.26865706592798233,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.9500000178813934,
      "step": 151
    },
    {
      "completion_length": 126.35000419616699,
      "epoch": 0.15759460860549507,
      "grad_norm": 93.3667984008789,
      "kl": 1.257568359375,
      "learning_rate": 1.5724137931034485e-06,
      "loss": -0.0012,
      "reward": 1.5,
      "reward_std": 0.4291412904858589,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 152
    },
    {
      "completion_length": 185.10000610351562,
      "epoch": 0.15863141524105753,
      "grad_norm": 607.76611328125,
      "kl": 1.4993896484375,
      "learning_rate": 1.5827586206896554e-06,
      "loss": 0.1878,
      "reward": 1.5500000417232513,
      "reward_std": 0.4902471825480461,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 153
    },
    {
      "completion_length": 150.1500015258789,
      "epoch": 0.15966822187662,
      "grad_norm": 621.0995483398438,
      "kl": 0.4017333984375,
      "learning_rate": 1.593103448275862e-06,
      "loss": 0.0861,
      "reward": 1.5750000178813934,
      "reward_std": 0.27228758856654167,
      "rewards/accuracy_reward": 0.6750000007450581,
      "rewards/format_reward": 0.9000000208616257,
      "step": 154
    },
    {
      "completion_length": 148.475004196167,
      "epoch": 0.1607050285121825,
      "grad_norm": 1060.59033203125,
      "kl": 5.067626953125,
      "learning_rate": 1.603448275862069e-06,
      "loss": 0.3634,
      "reward": 1.2999999970197678,
      "reward_std": 0.48553744331002235,
      "rewards/accuracy_reward": 0.4500000011175871,
      "rewards/format_reward": 0.8500000238418579,
      "step": 155
    },
    {
      "completion_length": 134.1500015258789,
      "epoch": 0.16174183514774496,
      "grad_norm": 100.48238372802734,
      "kl": 0.8262939453125,
      "learning_rate": 1.6137931034482759e-06,
      "loss": 0.0642,
      "reward": 1.2750000059604645,
      "reward_std": 0.248734038323164,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.8500000089406967,
      "step": 156
    },
    {
      "completion_length": 147.5750026702881,
      "epoch": 0.16277864178330742,
      "grad_norm": 62.48579788208008,
      "kl": 0.651611328125,
      "learning_rate": 1.6241379310344828e-06,
      "loss": 0.151,
      "reward": 1.25,
      "reward_std": 0.41643912717700005,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 0.8500000238418579,
      "step": 157
    },
    {
      "completion_length": 220.97500610351562,
      "epoch": 0.16381544841886989,
      "grad_norm": 575.1019287109375,
      "kl": 2.5919189453125,
      "learning_rate": 1.6344827586206896e-06,
      "loss": 0.2918,
      "reward": 1.4000000357627869,
      "reward_std": 0.36053742468357086,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.9000000059604645,
      "step": 158
    },
    {
      "completion_length": 144.225004196167,
      "epoch": 0.16485225505443235,
      "grad_norm": 7.095518112182617,
      "kl": 0.1251220703125,
      "learning_rate": 1.6448275862068965e-06,
      "loss": 0.0249,
      "reward": 1.5,
      "reward_std": 0.39728759601712227,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9000000059604645,
      "step": 159
    },
    {
      "completion_length": 142.92500114440918,
      "epoch": 0.16588906168999482,
      "grad_norm": 33.05251693725586,
      "kl": 0.8707275390625,
      "learning_rate": 1.6551724137931035e-06,
      "loss": 0.0677,
      "reward": 1.2750000357627869,
      "reward_std": 0.3371223881840706,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 160
    },
    {
      "completion_length": 154.0750026702881,
      "epoch": 0.16692586832555728,
      "grad_norm": 65645.4609375,
      "kl": 342.103515625,
      "learning_rate": 1.6655172413793104e-06,
      "loss": 13.3002,
      "reward": 1.5000000149011612,
      "reward_std": 0.41643913090229034,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 161
    },
    {
      "completion_length": 143.15000534057617,
      "epoch": 0.16796267496111975,
      "grad_norm": 230979.875,
      "kl": 1004.058837890625,
      "learning_rate": 1.6758620689655174e-06,
      "loss": 41.8463,
      "reward": 1.4500000476837158,
      "reward_std": 0.3099784553050995,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9000000059604645,
      "step": 162
    },
    {
      "completion_length": 162.97500228881836,
      "epoch": 0.1689994815966822,
      "grad_norm": 104.41748809814453,
      "kl": 0.559814453125,
      "learning_rate": 1.6862068965517243e-06,
      "loss": 0.0662,
      "reward": 1.2750000208616257,
      "reward_std": 0.45977721735835075,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.800000011920929,
      "step": 163
    },
    {
      "completion_length": 125.02500534057617,
      "epoch": 0.17003628823224468,
      "grad_norm": 422910.1875,
      "kl": 7556.533203125,
      "learning_rate": 1.696551724137931e-06,
      "loss": 274.9914,
      "reward": 1.4500000178813934,
      "reward_std": 0.39302410185337067,
      "rewards/accuracy_reward": 0.5500000081956387,
      "rewards/format_reward": 0.9000000208616257,
      "step": 164
    },
    {
      "completion_length": 136.90000534057617,
      "epoch": 0.17107309486780714,
      "grad_norm": 39.6871452331543,
      "kl": 0.236572265625,
      "learning_rate": 1.706896551724138e-06,
      "loss": 0.0168,
      "reward": 1.324999988079071,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 165
    },
    {
      "completion_length": 190.02500343322754,
      "epoch": 0.17210990150336963,
      "grad_norm": 2343.3095703125,
      "kl": 3.414794921875,
      "learning_rate": 1.717241379310345e-06,
      "loss": 0.3358,
      "reward": 1.4000000357627869,
      "reward_std": 0.5968463532626629,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.8250000178813934,
      "step": 166
    },
    {
      "completion_length": 123.57500267028809,
      "epoch": 0.1731467081389321,
      "grad_norm": 1024.3455810546875,
      "kl": 4.07666015625,
      "learning_rate": 1.727586206896552e-06,
      "loss": 0.2434,
      "reward": 1.550000011920929,
      "reward_std": 0.39665462449193,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.8500000238418579,
      "step": 167
    },
    {
      "completion_length": 127.15000343322754,
      "epoch": 0.17418351477449456,
      "grad_norm": 11.419363021850586,
      "kl": 0.1678466796875,
      "learning_rate": 1.7379310344827588e-06,
      "loss": 0.0772,
      "reward": 1.475000023841858,
      "reward_std": 0.5659334138035774,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.8500000238418579,
      "step": 168
    },
    {
      "completion_length": 158.8250026702881,
      "epoch": 0.17522032141005703,
      "grad_norm": 812.2886962890625,
      "kl": 10.5908203125,
      "learning_rate": 1.7482758620689658e-06,
      "loss": 0.3363,
      "reward": 1.7500000298023224,
      "reward_std": 0.1928323283791542,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 169
    },
    {
      "completion_length": 142.80000114440918,
      "epoch": 0.1762571280456195,
      "grad_norm": 114.91448974609375,
      "kl": 0.28302001953125,
      "learning_rate": 1.7586206896551723e-06,
      "loss": 0.0374,
      "reward": 1.5000000596046448,
      "reward_std": 0.373101070523262,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 170
    },
    {
      "completion_length": 175.60000038146973,
      "epoch": 0.17729393468118196,
      "grad_norm": 19.830774307250977,
      "kl": 0.1810302734375,
      "learning_rate": 1.7689655172413793e-06,
      "loss": 0.0822,
      "reward": 1.5250000059604645,
      "reward_std": 0.324558787047863,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9750000089406967,
      "step": 171
    },
    {
      "completion_length": 159.2500057220459,
      "epoch": 0.17833074131674442,
      "grad_norm": 154360.703125,
      "kl": 380.17138671875,
      "learning_rate": 1.7793103448275862e-06,
      "loss": 27.3536,
      "reward": 1.4000000655651093,
      "reward_std": 0.6183832511305809,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.8250000178813934,
      "step": 172
    },
    {
      "completion_length": 135.850004196167,
      "epoch": 0.1793675479523069,
      "grad_norm": 10.023521423339844,
      "kl": 0.11395263671875,
      "learning_rate": 1.789655172413793e-06,
      "loss": 0.0159,
      "reward": 1.4500000178813934,
      "reward_std": 0.22894952446222305,
      "rewards/accuracy_reward": 0.4749999977648258,
      "rewards/format_reward": 0.9750000089406967,
      "step": 173
    },
    {
      "completion_length": 179.9250030517578,
      "epoch": 0.18040435458786935,
      "grad_norm": 25.44389533996582,
      "kl": 0.44427490234375,
      "learning_rate": 1.8e-06,
      "loss": 0.0691,
      "reward": 1.350000023841858,
      "reward_std": 0.5138007514178753,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8000000268220901,
      "step": 174
    },
    {
      "completion_length": 157.3000030517578,
      "epoch": 0.18144116122343182,
      "grad_norm": 648.756103515625,
      "kl": 1.504638671875,
      "learning_rate": 1.8103448275862069e-06,
      "loss": 0.1569,
      "reward": 1.4500000476837158,
      "reward_std": 0.5697024464607239,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.8750000298023224,
      "step": 175
    },
    {
      "completion_length": 140.4250030517578,
      "epoch": 0.1824779678589943,
      "grad_norm": 16.518726348876953,
      "kl": 0.1947021484375,
      "learning_rate": 1.8206896551724138e-06,
      "loss": 0.0508,
      "reward": 1.3499999940395355,
      "reward_std": 0.30463574081659317,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 176
    },
    {
      "completion_length": 168.4000015258789,
      "epoch": 0.18351477449455678,
      "grad_norm": 12.404158592224121,
      "kl": 0.2431640625,
      "learning_rate": 1.8310344827586208e-06,
      "loss": -0.0125,
      "reward": 1.4250000417232513,
      "reward_std": 0.26129765436053276,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 1.0,
      "step": 177
    },
    {
      "completion_length": 164.92500114440918,
      "epoch": 0.18455158113011924,
      "grad_norm": 29.254444122314453,
      "kl": 0.3502197265625,
      "learning_rate": 1.8413793103448277e-06,
      "loss": 0.1379,
      "reward": 1.4000000357627869,
      "reward_std": 0.5660719126462936,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.8750000149011612,
      "step": 178
    },
    {
      "completion_length": 149.7750015258789,
      "epoch": 0.1855883877656817,
      "grad_norm": 6.365128517150879,
      "kl": 0.097900390625,
      "learning_rate": 1.8517241379310345e-06,
      "loss": 0.0811,
      "reward": 1.2750000208616257,
      "reward_std": 0.39302409440279007,
      "rewards/accuracy_reward": 0.32500001043081284,
      "rewards/format_reward": 0.9500000178813934,
      "step": 179
    },
    {
      "completion_length": 123.5,
      "epoch": 0.18662519440124417,
      "grad_norm": 19.429393768310547,
      "kl": 0.1754150390625,
      "learning_rate": 1.8620689655172414e-06,
      "loss": 0.0367,
      "reward": 1.4500000327825546,
      "reward_std": 0.3804604858160019,
      "rewards/accuracy_reward": 0.5500000063329935,
      "rewards/format_reward": 0.9000000208616257,
      "step": 180
    },
    {
      "completion_length": 149.12500381469727,
      "epoch": 0.18766200103680664,
      "grad_norm": 14.26073932647705,
      "kl": 0.134765625,
      "learning_rate": 1.8724137931034484e-06,
      "loss": 0.0374,
      "reward": 1.5750000178813934,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.6000000182539225,
      "rewards/format_reward": 0.9750000089406967,
      "step": 181
    },
    {
      "completion_length": 144.2750072479248,
      "epoch": 0.1886988076723691,
      "grad_norm": 4.313653945922852,
      "kl": 0.19671630859375,
      "learning_rate": 1.8827586206896553e-06,
      "loss": 0.0236,
      "reward": 1.4000000208616257,
      "reward_std": 0.28485122695565224,
      "rewards/accuracy_reward": 0.4749999940395355,
      "rewards/format_reward": 0.925000011920929,
      "step": 182
    },
    {
      "completion_length": 132.47500228881836,
      "epoch": 0.18973561430793157,
      "grad_norm": 4.134154796600342,
      "kl": 0.1151123046875,
      "learning_rate": 1.8931034482758623e-06,
      "loss": 0.0525,
      "reward": 1.4250000417232513,
      "reward_std": 0.37310105189681053,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.8750000298023224,
      "step": 183
    },
    {
      "completion_length": 171.25000381469727,
      "epoch": 0.19077242094349403,
      "grad_norm": 7.449805736541748,
      "kl": 0.124755859375,
      "learning_rate": 1.9034482758620692e-06,
      "loss": 0.0377,
      "reward": 1.1250000447034836,
      "reward_std": 0.4688488021492958,
      "rewards/accuracy_reward": 0.2750000022351742,
      "rewards/format_reward": 0.8500000238418579,
      "step": 184
    },
    {
      "completion_length": 134.2750015258789,
      "epoch": 0.1918092275790565,
      "grad_norm": 4.306768417358398,
      "kl": 0.11016845703125,
      "learning_rate": 1.913793103448276e-06,
      "loss": 0.0353,
      "reward": 1.4000000059604645,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.4499999973922968,
      "rewards/format_reward": 0.9500000178813934,
      "step": 185
    },
    {
      "completion_length": 162.60000610351562,
      "epoch": 0.19284603421461896,
      "grad_norm": 4.144135475158691,
      "kl": 0.0718994140625,
      "learning_rate": 1.9241379310344827e-06,
      "loss": 0.0482,
      "reward": 1.4250000417232513,
      "reward_std": 0.31719936802983284,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.925000011920929,
      "step": 186
    },
    {
      "completion_length": 168.5,
      "epoch": 0.19388284085018145,
      "grad_norm": 2.6224422454833984,
      "kl": 0.1402587890625,
      "learning_rate": 1.9344827586206894e-06,
      "loss": -0.0576,
      "reward": 1.4500000476837158,
      "reward_std": 0.5263643860816956,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.8750000149011612,
      "step": 187
    },
    {
      "completion_length": 155.87500190734863,
      "epoch": 0.19491964748574392,
      "grad_norm": 2.5991694927215576,
      "kl": 0.1173095703125,
      "learning_rate": 1.9448275862068966e-06,
      "loss": 0.0525,
      "reward": 1.425000011920929,
      "reward_std": 0.4129471443593502,
      "rewards/accuracy_reward": 0.5250000040978193,
      "rewards/format_reward": 0.9000000357627869,
      "step": 188
    },
    {
      "completion_length": 157.60000228881836,
      "epoch": 0.19595645412130638,
      "grad_norm": 11.603816986083984,
      "kl": 0.08740234375,
      "learning_rate": 1.9551724137931033e-06,
      "loss": 0.0086,
      "reward": 1.5999999940395355,
      "reward_std": 0.4525563307106495,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 189
    },
    {
      "completion_length": 163.6750030517578,
      "epoch": 0.19699326075686885,
      "grad_norm": 2.4964864253997803,
      "kl": 0.08258056640625,
      "learning_rate": 1.9655172413793105e-06,
      "loss": 0.007,
      "reward": 1.5250000357627869,
      "reward_std": 0.329762976616621,
      "rewards/accuracy_reward": 0.5500000063329935,
      "rewards/format_reward": 0.9750000089406967,
      "step": 190
    },
    {
      "completion_length": 163.37500381469727,
      "epoch": 0.1980300673924313,
      "grad_norm": 2.807250499725342,
      "kl": 0.0880126953125,
      "learning_rate": 1.9758620689655172e-06,
      "loss": -0.0182,
      "reward": 1.3000000417232513,
      "reward_std": 0.4849044606089592,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.8499999940395355,
      "step": 191
    },
    {
      "completion_length": 113.47500228881836,
      "epoch": 0.19906687402799378,
      "grad_norm": 0.9696298241615295,
      "kl": 0.1170654296875,
      "learning_rate": 1.986206896551724e-06,
      "loss": -0.0394,
      "reward": 1.725000023841858,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 1.0,
      "step": 192
    },
    {
      "completion_length": 139.8250026702881,
      "epoch": 0.20010368066355624,
      "grad_norm": 1.4813657999038696,
      "kl": 0.0897216796875,
      "learning_rate": 1.996551724137931e-06,
      "loss": 0.0159,
      "reward": 1.3750000298023224,
      "reward_std": 0.41643912717700005,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9250000268220901,
      "step": 193
    },
    {
      "completion_length": 186.35000228881836,
      "epoch": 0.2011404872991187,
      "grad_norm": 2.646305561065674,
      "kl": 0.1492919921875,
      "learning_rate": 2.006896551724138e-06,
      "loss": 0.0906,
      "reward": 1.5250000357627869,
      "reward_std": 0.40387552604079247,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9000000357627869,
      "step": 194
    },
    {
      "completion_length": 178.600004196167,
      "epoch": 0.20217729393468117,
      "grad_norm": 3.9357028007507324,
      "kl": 0.1038818359375,
      "learning_rate": 2.017241379310345e-06,
      "loss": 0.013,
      "reward": 1.2750000059604645,
      "reward_std": 0.3966546133160591,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.8750000149011612,
      "step": 195
    },
    {
      "completion_length": 175.0500030517578,
      "epoch": 0.20321410057024364,
      "grad_norm": 6.768120288848877,
      "kl": 0.12518310546875,
      "learning_rate": 2.0275862068965518e-06,
      "loss": 0.0436,
      "reward": 1.4749999940395355,
      "reward_std": 0.472340852022171,
      "rewards/accuracy_reward": 0.6250000223517418,
      "rewards/format_reward": 0.8499999940395355,
      "step": 196
    },
    {
      "completion_length": 145.22500038146973,
      "epoch": 0.2042509072058061,
      "grad_norm": 2.82489275932312,
      "kl": 0.0924072265625,
      "learning_rate": 2.037931034482759e-06,
      "loss": 0.0287,
      "reward": 1.475000023841858,
      "reward_std": 0.4290027730166912,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.8750000298023224,
      "step": 197
    },
    {
      "completion_length": 186.82500076293945,
      "epoch": 0.2052877138413686,
      "grad_norm": 6.81854772567749,
      "kl": 0.0621337890625,
      "learning_rate": 2.0482758620689657e-06,
      "loss": -0.1108,
      "reward": 1.225000038743019,
      "reward_std": 0.373101070523262,
      "rewards/accuracy_reward": 0.32500001043081284,
      "rewards/format_reward": 0.9000000059604645,
      "step": 198
    },
    {
      "completion_length": 160.30000114440918,
      "epoch": 0.20632452047693106,
      "grad_norm": 3.4426076412200928,
      "kl": 0.06744384765625,
      "learning_rate": 2.0586206896551724e-06,
      "loss": 0.0542,
      "reward": 1.1750000268220901,
      "reward_std": 0.4614894203841686,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.8250000178813934,
      "step": 199
    },
    {
      "completion_length": 190.45000457763672,
      "epoch": 0.20736132711249353,
      "grad_norm": 2.7901272773742676,
      "kl": 0.095458984375,
      "learning_rate": 2.0689655172413796e-06,
      "loss": 0.0244,
      "reward": 1.5000000298023224,
      "reward_std": 0.5173911228775978,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 200
    },
    {
      "completion_length": 175.1500015258789,
      "epoch": 0.208398133748056,
      "grad_norm": 5.208185195922852,
      "kl": 0.0733642578125,
      "learning_rate": 2.0793103448275863e-06,
      "loss": 0.0926,
      "reward": 1.449999988079071,
      "reward_std": 0.4525563232600689,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 201
    },
    {
      "completion_length": 162.62500190734863,
      "epoch": 0.20943494038361846,
      "grad_norm": 1.4480936527252197,
      "kl": 0.1016845703125,
      "learning_rate": 2.089655172413793e-06,
      "loss": 0.0509,
      "reward": 1.4000000059604645,
      "reward_std": 0.45977722853422165,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 202
    },
    {
      "completion_length": 164.6750030517578,
      "epoch": 0.21047174701918092,
      "grad_norm": 4.689480304718018,
      "kl": 0.0806884765625,
      "learning_rate": 2.1e-06,
      "loss": 0.0337,
      "reward": 1.425000011920929,
      "reward_std": 0.37310105934739113,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.8500000238418579,
      "step": 203
    },
    {
      "completion_length": 165.60000228881836,
      "epoch": 0.2115085536547434,
      "grad_norm": 5.075602054595947,
      "kl": 0.078369140625,
      "learning_rate": 2.110344827586207e-06,
      "loss": 0.0524,
      "reward": 1.3250000327825546,
      "reward_std": 0.3407529257237911,
      "rewards/accuracy_reward": 0.4500000011175871,
      "rewards/format_reward": 0.875,
      "step": 204
    },
    {
      "completion_length": 155.70000457763672,
      "epoch": 0.21254536029030585,
      "grad_norm": 1.853857398033142,
      "kl": 0.09381103515625,
      "learning_rate": 2.1206896551724137e-06,
      "loss": -0.0339,
      "reward": 1.2250000089406967,
      "reward_std": 0.5588510297238827,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 0.8250000029802322,
      "step": 205
    },
    {
      "completion_length": 183.0250015258789,
      "epoch": 0.21358216692586832,
      "grad_norm": 0.6376509666442871,
      "kl": 0.068603515625,
      "learning_rate": 2.131034482758621e-06,
      "loss": 0.082,
      "reward": 1.3750000149011612,
      "reward_std": 0.36067599430680275,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.8000000044703484,
      "step": 206
    },
    {
      "completion_length": 186.22500228881836,
      "epoch": 0.21461897356143078,
      "grad_norm": 0.9813095927238464,
      "kl": 0.0640869140625,
      "learning_rate": 2.1413793103448276e-06,
      "loss": 0.0382,
      "reward": 1.3000000417232513,
      "reward_std": 0.45413000881671906,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.8750000149011612,
      "step": 207
    },
    {
      "completion_length": 151.05000114440918,
      "epoch": 0.21565578019699327,
      "grad_norm": 0.8648532032966614,
      "kl": 0.0977783203125,
      "learning_rate": 2.1517241379310343e-06,
      "loss": 0.0119,
      "reward": 1.4749999940395355,
      "reward_std": 0.3804604560136795,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.925000011920929,
      "step": 208
    },
    {
      "completion_length": 152.5250015258789,
      "epoch": 0.21669258683255574,
      "grad_norm": 0.7379907965660095,
      "kl": 0.09222412109375,
      "learning_rate": 2.1620689655172415e-06,
      "loss": 0.0774,
      "reward": 1.300000011920929,
      "reward_std": 0.44892578572034836,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9250000268220901,
      "step": 209
    },
    {
      "completion_length": 143.6500015258789,
      "epoch": 0.2177293934681182,
      "grad_norm": 0.6392645835876465,
      "kl": 0.07421875,
      "learning_rate": 2.1724137931034482e-06,
      "loss": 0.0565,
      "reward": 1.4249999821186066,
      "reward_std": 0.23617041110992432,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 1.0,
      "step": 210
    },
    {
      "completion_length": 156.05000495910645,
      "epoch": 0.21876620010368067,
      "grad_norm": 0.8323525786399841,
      "kl": 0.0712890625,
      "learning_rate": 2.1827586206896554e-06,
      "loss": -0.0088,
      "reward": 1.5250000357627869,
      "reward_std": 0.5156789347529411,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 211
    },
    {
      "completion_length": 198.5500030517578,
      "epoch": 0.21980300673924313,
      "grad_norm": 0.6546644568443298,
      "kl": 0.07940673828125,
      "learning_rate": 2.193103448275862e-06,
      "loss": 0.0794,
      "reward": 1.5750000178813934,
      "reward_std": 0.42178186774253845,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.925000011920929,
      "step": 212
    },
    {
      "completion_length": 146.5000057220459,
      "epoch": 0.2208398133748056,
      "grad_norm": 0.8499487042427063,
      "kl": 0.11083984375,
      "learning_rate": 2.203448275862069e-06,
      "loss": 0.0293,
      "reward": 1.2999999970197678,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9250000268220901,
      "step": 213
    },
    {
      "completion_length": 145.57500076293945,
      "epoch": 0.22187662001036806,
      "grad_norm": 0.687872588634491,
      "kl": 0.07757568359375,
      "learning_rate": 2.213793103448276e-06,
      "loss": 0.0198,
      "reward": 1.3000000417232513,
      "reward_std": 0.3856646753847599,
      "rewards/accuracy_reward": 0.4500000048428774,
      "rewards/format_reward": 0.8500000089406967,
      "step": 214
    },
    {
      "completion_length": 133.89999961853027,
      "epoch": 0.22291342664593053,
      "grad_norm": 0.5972610116004944,
      "kl": 0.09869384765625,
      "learning_rate": 2.2241379310344828e-06,
      "loss": -0.0411,
      "reward": 1.6250000298023224,
      "reward_std": 0.1928323432803154,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 215
    },
    {
      "completion_length": 121.60000419616699,
      "epoch": 0.223950233281493,
      "grad_norm": 0.7216851711273193,
      "kl": 0.14794921875,
      "learning_rate": 2.23448275862069e-06,
      "loss": -0.0187,
      "reward": 1.5750000178813934,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 216
    },
    {
      "completion_length": 178.8499984741211,
      "epoch": 0.22498703991705546,
      "grad_norm": 0.6720969676971436,
      "kl": 0.113037109375,
      "learning_rate": 2.2448275862068967e-06,
      "loss": -0.0698,
      "reward": 1.2500000447034836,
      "reward_std": 0.40558772161602974,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.8750000149011612,
      "step": 217
    },
    {
      "completion_length": 189.72500228881836,
      "epoch": 0.22602384655261792,
      "grad_norm": 0.719994068145752,
      "kl": 0.1602783203125,
      "learning_rate": 2.2551724137931034e-06,
      "loss": 0.0713,
      "reward": 1.4250000417232513,
      "reward_std": 0.45413000136613846,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.925000011920929,
      "step": 218
    },
    {
      "completion_length": 133.8250026702881,
      "epoch": 0.22706065318818042,
      "grad_norm": 0.4400331377983093,
      "kl": 0.1085205078125,
      "learning_rate": 2.26551724137931e-06,
      "loss": 0.0333,
      "reward": 1.5250000208616257,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 219
    },
    {
      "completion_length": 154.80000686645508,
      "epoch": 0.22809745982374288,
      "grad_norm": 0.6450486779212952,
      "kl": 0.131103515625,
      "learning_rate": 2.2758620689655173e-06,
      "loss": 0.0126,
      "reward": 1.3250000327825546,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 220
    },
    {
      "completion_length": 145.72500610351562,
      "epoch": 0.22913426645930535,
      "grad_norm": 1.3167037963867188,
      "kl": 0.135986328125,
      "learning_rate": 2.286206896551724e-06,
      "loss": 0.0465,
      "reward": 1.8250000178813934,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.9750000089406967,
      "rewards/format_reward": 0.8500000238418579,
      "step": 221
    },
    {
      "completion_length": 172.0500030517578,
      "epoch": 0.2301710730948678,
      "grad_norm": 0.7341930866241455,
      "kl": 0.148193359375,
      "learning_rate": 2.2965517241379312e-06,
      "loss": 0.0681,
      "reward": 1.0500000417232513,
      "reward_std": 0.393024068325758,
      "rewards/accuracy_reward": 0.22500000149011612,
      "rewards/format_reward": 0.8250000178813934,
      "step": 222
    },
    {
      "completion_length": 146.42500495910645,
      "epoch": 0.23120787973043028,
      "grad_norm": 0.8762367367744446,
      "kl": 0.123046875,
      "learning_rate": 2.306896551724138e-06,
      "loss": 0.1031,
      "reward": 1.550000011920929,
      "reward_std": 0.5210216194391251,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 223
    },
    {
      "completion_length": 157.7750072479248,
      "epoch": 0.23224468636599274,
      "grad_norm": 0.8583253026008606,
      "kl": 0.13037109375,
      "learning_rate": 2.3172413793103447e-06,
      "loss": 0.1291,
      "reward": 1.6250000298023224,
      "reward_std": 0.41643912717700005,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.875,
      "step": 224
    },
    {
      "completion_length": 146.5500030517578,
      "epoch": 0.2332814930015552,
      "grad_norm": 0.8889981508255005,
      "kl": 0.113525390625,
      "learning_rate": 2.327586206896552e-06,
      "loss": -0.024,
      "reward": 1.5999999940395355,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.6250000223517418,
      "rewards/format_reward": 0.9750000089406967,
      "step": 225
    },
    {
      "completion_length": 137.5500030517578,
      "epoch": 0.23431829963711767,
      "grad_norm": 0.7300592064857483,
      "kl": 0.1185302734375,
      "learning_rate": 2.3379310344827586e-06,
      "loss": 0.0438,
      "reward": 1.3499999940395355,
      "reward_std": 0.46511995419859886,
      "rewards/accuracy_reward": 0.4000000096857548,
      "rewards/format_reward": 0.9500000029802322,
      "step": 226
    },
    {
      "completion_length": 119.70000457763672,
      "epoch": 0.23535510627268014,
      "grad_norm": 0.727929413318634,
      "kl": 0.137451171875,
      "learning_rate": 2.3482758620689658e-06,
      "loss": 0.0458,
      "reward": 1.350000023841858,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.3999999985098839,
      "rewards/format_reward": 0.9500000178813934,
      "step": 227
    },
    {
      "completion_length": 147.70000076293945,
      "epoch": 0.2363919129082426,
      "grad_norm": 0.9192894101142883,
      "kl": 0.139404296875,
      "learning_rate": 2.3586206896551725e-06,
      "loss": 0.0975,
      "reward": 1.350000023841858,
      "reward_std": 0.39302409067749977,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 0.9250000268220901,
      "step": 228
    },
    {
      "completion_length": 156.4000015258789,
      "epoch": 0.23742871954380507,
      "grad_norm": 0.7950097918510437,
      "kl": 0.1527099609375,
      "learning_rate": 2.3689655172413793e-06,
      "loss": 0.0755,
      "reward": 1.600000023841858,
      "reward_std": 0.3245587833225727,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 229
    },
    {
      "completion_length": 109.77500343322754,
      "epoch": 0.23846552617936756,
      "grad_norm": 0.933161735534668,
      "kl": 0.142333984375,
      "learning_rate": 2.3793103448275864e-06,
      "loss": 0.0125,
      "reward": 1.449999988079071,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 230
    },
    {
      "completion_length": 125.25,
      "epoch": 0.23950233281493002,
      "grad_norm": 0.715398371219635,
      "kl": 0.1328125,
      "learning_rate": 2.389655172413793e-06,
      "loss": -0.0194,
      "reward": 1.5500000417232513,
      "reward_std": 0.3407529182732105,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 231
    },
    {
      "completion_length": 106.65000343322754,
      "epoch": 0.2405391394504925,
      "grad_norm": 1.061570644378662,
      "kl": 0.205810546875,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.0093,
      "reward": 1.5250000059604645,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 1.0,
      "step": 232
    },
    {
      "completion_length": 107.42500305175781,
      "epoch": 0.24157594608605495,
      "grad_norm": 0.8809791207313538,
      "kl": 0.1182861328125,
      "learning_rate": 2.410344827586207e-06,
      "loss": 0.0425,
      "reward": 1.4250000417232513,
      "reward_std": 0.40558769553899765,
      "rewards/accuracy_reward": 0.44999999552965164,
      "rewards/format_reward": 0.9750000089406967,
      "step": 233
    },
    {
      "completion_length": 118.6500015258789,
      "epoch": 0.24261275272161742,
      "grad_norm": 0.6223216652870178,
      "kl": 0.1490478515625,
      "learning_rate": 2.420689655172414e-06,
      "loss": 0.1029,
      "reward": 1.475000023841858,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 234
    },
    {
      "completion_length": 116.7750015258789,
      "epoch": 0.24364955935717988,
      "grad_norm": 1.0359702110290527,
      "kl": 0.21630859375,
      "learning_rate": 2.4310344827586205e-06,
      "loss": 0.1095,
      "reward": 1.4500000178813934,
      "reward_std": 0.5409446731209755,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 235
    },
    {
      "completion_length": 101.80000114440918,
      "epoch": 0.24468636599274235,
      "grad_norm": 0.6835160255432129,
      "kl": 0.20654296875,
      "learning_rate": 2.4413793103448277e-06,
      "loss": 0.0014,
      "reward": 1.4500000476837158,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 1.0,
      "step": 236
    },
    {
      "completion_length": 110.22500038146973,
      "epoch": 0.24572317262830481,
      "grad_norm": 0.50115966796875,
      "kl": 0.1844482421875,
      "learning_rate": 2.4517241379310344e-06,
      "loss": 0.0505,
      "reward": 1.7249999940395355,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 237
    },
    {
      "completion_length": 113.75000190734863,
      "epoch": 0.24675997926386728,
      "grad_norm": 7.211286544799805,
      "kl": 0.42578125,
      "learning_rate": 2.462068965517241e-06,
      "loss": 0.0999,
      "reward": 1.550000011920929,
      "reward_std": 0.3966546207666397,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000298023224,
      "step": 238
    },
    {
      "completion_length": 112.39999961853027,
      "epoch": 0.24779678589942974,
      "grad_norm": 0.547843873500824,
      "kl": 0.156494140625,
      "learning_rate": 2.4724137931034483e-06,
      "loss": -0.0177,
      "reward": 1.2750000059604645,
      "reward_std": 0.1928323470056057,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 1.0,
      "step": 239
    },
    {
      "completion_length": 111.02500343322754,
      "epoch": 0.24883359253499224,
      "grad_norm": 0.553968608379364,
      "kl": 0.1614990234375,
      "learning_rate": 2.482758620689655e-06,
      "loss": -0.0008,
      "reward": 1.5,
      "reward_std": 0.20019177719950676,
      "rewards/accuracy_reward": 0.5500000100582838,
      "rewards/format_reward": 0.9500000178813934,
      "step": 240
    },
    {
      "completion_length": 101.9000015258789,
      "epoch": 0.2498703991705547,
      "grad_norm": 0.6699236631393433,
      "kl": 0.17431640625,
      "learning_rate": 2.4931034482758622e-06,
      "loss": 0.0162,
      "reward": 1.3500000089406967,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.37500000186264515,
      "rewards/format_reward": 0.9750000089406967,
      "step": 241
    },
    {
      "completion_length": 114.30000114440918,
      "epoch": 0.25090720580611714,
      "grad_norm": 0.8109186291694641,
      "kl": 0.145751953125,
      "learning_rate": 2.503448275862069e-06,
      "loss": -0.009,
      "reward": 1.475000023841858,
      "reward_std": 0.3371223919093609,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 242
    },
    {
      "completion_length": 150.85000228881836,
      "epoch": 0.25194401244167963,
      "grad_norm": 0.7055056691169739,
      "kl": 0.14892578125,
      "learning_rate": 2.513793103448276e-06,
      "loss": 0.0376,
      "reward": 1.2500000298023224,
      "reward_std": 0.3930240757763386,
      "rewards/accuracy_reward": 0.32500000670552254,
      "rewards/format_reward": 0.925000011920929,
      "step": 243
    },
    {
      "completion_length": 116.4000015258789,
      "epoch": 0.25298081907724207,
      "grad_norm": 0.8164229393005371,
      "kl": 0.1978759765625,
      "learning_rate": 2.524137931034483e-06,
      "loss": -0.0283,
      "reward": 1.5500000417232513,
      "reward_std": 0.36053745076060295,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 244
    },
    {
      "completion_length": 149.54999923706055,
      "epoch": 0.25401762571280456,
      "grad_norm": 0.702789843082428,
      "kl": 0.1356201171875,
      "learning_rate": 2.5344827586206896e-06,
      "loss": 0.0248,
      "reward": 1.4500000476837158,
      "reward_std": 0.3731010667979717,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 245
    },
    {
      "completion_length": 139.87500381469727,
      "epoch": 0.25505443234836706,
      "grad_norm": 0.8374885320663452,
      "kl": 0.1865234375,
      "learning_rate": 2.5448275862068968e-06,
      "loss": 0.0117,
      "reward": 1.3750000298023224,
      "reward_std": 0.31719936057925224,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 246
    },
    {
      "completion_length": 154.17500686645508,
      "epoch": 0.2560912389839295,
      "grad_norm": 0.5410239696502686,
      "kl": 0.1728515625,
      "learning_rate": 2.5551724137931035e-06,
      "loss": 0.0571,
      "reward": 1.5750000178813934,
      "reward_std": 0.26143617182970047,
      "rewards/accuracy_reward": 0.6249999944120646,
      "rewards/format_reward": 0.9500000178813934,
      "step": 247
    },
    {
      "completion_length": 132.79999923706055,
      "epoch": 0.257128045619492,
      "grad_norm": 0.733527660369873,
      "kl": 0.1473388671875,
      "learning_rate": 2.5655172413793107e-06,
      "loss": 0.0043,
      "reward": 1.5999999940395355,
      "reward_std": 0.3479738123714924,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 248
    },
    {
      "completion_length": 250.8000030517578,
      "epoch": 0.2581648522550544,
      "grad_norm": 0.6939677596092224,
      "kl": 0.1351318359375,
      "learning_rate": 2.5758620689655174e-06,
      "loss": 0.0938,
      "reward": 1.5750000476837158,
      "reward_std": 0.3856646791100502,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.8750000149011612,
      "step": 249
    },
    {
      "completion_length": 151.17500495910645,
      "epoch": 0.2592016588906169,
      "grad_norm": 0.3936190605163574,
      "kl": 0.166015625,
      "learning_rate": 2.586206896551724e-06,
      "loss": 0.0012,
      "reward": 1.5999999940395355,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 250
    },
    {
      "completion_length": 235.57499885559082,
      "epoch": 0.26023846552617935,
      "grad_norm": 1.5798455476760864,
      "kl": 0.26806640625,
      "learning_rate": 2.596551724137931e-06,
      "loss": 0.2016,
      "reward": 1.4000000357627869,
      "reward_std": 0.6146142035722733,
      "rewards/accuracy_reward": 0.5749999955296516,
      "rewards/format_reward": 0.8250000029802322,
      "step": 251
    },
    {
      "completion_length": 187.70000457763672,
      "epoch": 0.26127527216174184,
      "grad_norm": 0.5298973917961121,
      "kl": 0.16845703125,
      "learning_rate": 2.606896551724138e-06,
      "loss": 0.0335,
      "reward": 1.3250000029802322,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 252
    },
    {
      "completion_length": 160.1750030517578,
      "epoch": 0.2623120787973043,
      "grad_norm": 0.6976402401924133,
      "kl": 0.157958984375,
      "learning_rate": 2.617241379310345e-06,
      "loss": 0.0418,
      "reward": 1.5750000178813934,
      "reward_std": 0.5587124899029732,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 253
    },
    {
      "completion_length": 284.1750144958496,
      "epoch": 0.2633488854328668,
      "grad_norm": 0.6427121758460999,
      "kl": 0.1856689453125,
      "learning_rate": 2.6275862068965515e-06,
      "loss": 0.1826,
      "reward": 1.350000023841858,
      "reward_std": 0.4850429743528366,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.8750000298023224,
      "step": 254
    },
    {
      "completion_length": 304.4750061035156,
      "epoch": 0.2643856920684292,
      "grad_norm": 3.0576741695404053,
      "kl": 0.264404296875,
      "learning_rate": 2.6379310344827587e-06,
      "loss": 0.1303,
      "reward": 1.225000038743019,
      "reward_std": 0.6094100177288055,
      "rewards/accuracy_reward": 0.4249999988824129,
      "rewards/format_reward": 0.800000011920929,
      "step": 255
    },
    {
      "completion_length": 373.6000099182129,
      "epoch": 0.2654224987039917,
      "grad_norm": 0.5253204107284546,
      "kl": 0.1571044921875,
      "learning_rate": 2.6482758620689654e-06,
      "loss": 0.1922,
      "reward": 0.9750000089406967,
      "reward_std": 0.47768357396125793,
      "rewards/accuracy_reward": 0.3000000063329935,
      "rewards/format_reward": 0.6750000044703484,
      "step": 256
    },
    {
      "completion_length": 355.32501220703125,
      "epoch": 0.2664593053395542,
      "grad_norm": 0.9178584218025208,
      "kl": 0.2265625,
      "learning_rate": 2.6586206896551726e-06,
      "loss": 0.3166,
      "reward": 1.0000000298023224,
      "reward_std": 0.7397527396678925,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.6250000149011612,
      "step": 257
    },
    {
      "completion_length": 237.27500534057617,
      "epoch": 0.26749611197511663,
      "grad_norm": 0.6554264426231384,
      "kl": 0.17626953125,
      "learning_rate": 2.6689655172413793e-06,
      "loss": 0.1707,
      "reward": 1.0750000327825546,
      "reward_std": 0.6130405068397522,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.7000000178813934,
      "step": 258
    },
    {
      "completion_length": 530.1000061035156,
      "epoch": 0.2685329186106791,
      "grad_norm": 0.6588838696479797,
      "kl": 0.1446533203125,
      "learning_rate": 2.6793103448275865e-06,
      "loss": 0.1994,
      "reward": 1.0000000149011612,
      "reward_std": 0.6033165566623211,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.5750000104308128,
      "step": 259
    },
    {
      "completion_length": 372.25000762939453,
      "epoch": 0.26956972524624156,
      "grad_norm": 0.8617604970932007,
      "kl": 0.279052734375,
      "learning_rate": 2.6896551724137932e-06,
      "loss": 0.2631,
      "reward": 1.1500000059604645,
      "reward_std": 0.7571920454502106,
      "rewards/accuracy_reward": 0.5250000022351742,
      "rewards/format_reward": 0.625,
      "step": 260
    },
    {
      "completion_length": 370.0750045776367,
      "epoch": 0.27060653188180406,
      "grad_norm": 0.8590409755706787,
      "kl": 0.257080078125,
      "learning_rate": 2.7e-06,
      "loss": 0.1892,
      "reward": 1.2000000476837158,
      "reward_std": 0.7344100549817085,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.800000011920929,
      "step": 261
    },
    {
      "completion_length": 197.8250026702881,
      "epoch": 0.2716433385173665,
      "grad_norm": 0.921504020690918,
      "kl": 0.262939453125,
      "learning_rate": 2.710344827586207e-06,
      "loss": 0.2638,
      "reward": 1.4500000178813934,
      "reward_std": 0.4723408371210098,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.8500000089406967,
      "step": 262
    },
    {
      "completion_length": 230.94999885559082,
      "epoch": 0.272680145152929,
      "grad_norm": 0.7630376219749451,
      "kl": 0.238525390625,
      "learning_rate": 2.720689655172414e-06,
      "loss": 0.1966,
      "reward": 1.300000011920929,
      "reward_std": 0.6491576656699181,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.7750000059604645,
      "step": 263
    },
    {
      "completion_length": 208.65000343322754,
      "epoch": 0.2737169517884914,
      "grad_norm": 0.9253503680229187,
      "kl": 0.29345703125,
      "learning_rate": 2.731034482758621e-06,
      "loss": 0.2176,
      "reward": 1.1250000447034836,
      "reward_std": 0.5414391383528709,
      "rewards/accuracy_reward": 0.3250000011175871,
      "rewards/format_reward": 0.7999999970197678,
      "step": 264
    },
    {
      "completion_length": 141.70000457763672,
      "epoch": 0.2747537584240539,
      "grad_norm": 0.9654400944709778,
      "kl": 0.34521484375,
      "learning_rate": 2.7413793103448278e-06,
      "loss": 0.1135,
      "reward": 1.1250000298023224,
      "reward_std": 0.4170721210539341,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.8500000238418579,
      "step": 265
    },
    {
      "completion_length": 121.32500267028809,
      "epoch": 0.27579056505961635,
      "grad_norm": 1.2874776124954224,
      "kl": 0.447509765625,
      "learning_rate": 2.7517241379310345e-06,
      "loss": 0.1512,
      "reward": 1.3750000298023224,
      "reward_std": 0.41643913462758064,
      "rewards/accuracy_reward": 0.4500000048428774,
      "rewards/format_reward": 0.925000011920929,
      "step": 266
    },
    {
      "completion_length": 124.97500324249268,
      "epoch": 0.27682737169517885,
      "grad_norm": 0.7988603711128235,
      "kl": 0.448486328125,
      "learning_rate": 2.7620689655172413e-06,
      "loss": 0.1464,
      "reward": 1.5249999910593033,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9500000029802322,
      "step": 267
    },
    {
      "completion_length": 43.82500076293945,
      "epoch": 0.27786417833074134,
      "grad_norm": 1.037258505821228,
      "kl": 0.646484375,
      "learning_rate": 2.7724137931034484e-06,
      "loss": 0.0372,
      "reward": 1.4500000178813934,
      "reward_std": 0.2487340308725834,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 1.0,
      "step": 268
    },
    {
      "completion_length": 47.30000019073486,
      "epoch": 0.2789009849663038,
      "grad_norm": 0.7092971205711365,
      "kl": 0.41455078125,
      "learning_rate": 2.782758620689655e-06,
      "loss": 0.0284,
      "reward": 1.6500000059604645,
      "reward_std": 0.24151314795017242,
      "rewards/accuracy_reward": 0.6750000007450581,
      "rewards/format_reward": 0.9750000089406967,
      "step": 269
    },
    {
      "completion_length": 94.62500286102295,
      "epoch": 0.27993779160186627,
      "grad_norm": 0.4653095304965973,
      "kl": 0.7783203125,
      "learning_rate": 2.793103448275862e-06,
      "loss": 0.0589,
      "reward": 1.425000011920929,
      "reward_std": 0.15685367584228516,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9750000089406967,
      "step": 270
    },
    {
      "completion_length": 151.97500228881836,
      "epoch": 0.2809745982374287,
      "grad_norm": 0.9383624792098999,
      "kl": 0.50537109375,
      "learning_rate": 2.803448275862069e-06,
      "loss": 0.0766,
      "reward": 1.5500000417232513,
      "reward_std": 0.32818931341171265,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 271
    },
    {
      "completion_length": 193.1000051498413,
      "epoch": 0.2820114048729912,
      "grad_norm": 0.7510073781013489,
      "kl": 0.46533203125,
      "learning_rate": 2.813793103448276e-06,
      "loss": 0.1247,
      "reward": 1.4500000178813934,
      "reward_std": 0.41657765209674835,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 272
    },
    {
      "completion_length": 219.00000762939453,
      "epoch": 0.28304821150855364,
      "grad_norm": 0.8484097719192505,
      "kl": 0.505859375,
      "learning_rate": 2.824137931034483e-06,
      "loss": 0.2461,
      "reward": 1.475000023841858,
      "reward_std": 0.4783165566623211,
      "rewards/accuracy_reward": 0.5750000141561031,
      "rewards/format_reward": 0.9000000059604645,
      "step": 273
    },
    {
      "completion_length": 54.650001525878906,
      "epoch": 0.28408501814411613,
      "grad_norm": 0.9514493346214294,
      "kl": 0.64794921875,
      "learning_rate": 2.8344827586206897e-06,
      "loss": 0.041,
      "reward": 1.300000011920929,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 274
    },
    {
      "completion_length": 203.30000114440918,
      "epoch": 0.28512182477967857,
      "grad_norm": 0.680706262588501,
      "kl": 0.5048828125,
      "learning_rate": 2.8448275862068964e-06,
      "loss": 0.1976,
      "reward": 1.350000023841858,
      "reward_std": 0.3611704185605049,
      "rewards/accuracy_reward": 0.44999999925494194,
      "rewards/format_reward": 0.9000000059604645,
      "step": 275
    },
    {
      "completion_length": 150.50000381469727,
      "epoch": 0.28615863141524106,
      "grad_norm": 0.5486094355583191,
      "kl": 0.56982421875,
      "learning_rate": 2.8551724137931036e-06,
      "loss": 0.0903,
      "reward": 1.4000000059604645,
      "reward_std": 0.24936703965067863,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9500000029802322,
      "step": 276
    },
    {
      "completion_length": 193.65000534057617,
      "epoch": 0.2871954380508035,
      "grad_norm": 0.9081488847732544,
      "kl": 0.53564453125,
      "learning_rate": 2.8655172413793104e-06,
      "loss": 0.0827,
      "reward": 1.3000000417232513,
      "reward_std": 0.3407529219985008,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.925000011920929,
      "step": 277
    },
    {
      "completion_length": 118.72500324249268,
      "epoch": 0.288232244686366,
      "grad_norm": 0.8595836758613586,
      "kl": 0.46435546875,
      "learning_rate": 2.8758620689655175e-06,
      "loss": 0.096,
      "reward": 1.300000011920929,
      "reward_std": 0.3858031928539276,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9250000268220901,
      "step": 278
    },
    {
      "completion_length": 197.52500247955322,
      "epoch": 0.2892690513219285,
      "grad_norm": 0.6958838105201721,
      "kl": 0.708984375,
      "learning_rate": 2.8862068965517243e-06,
      "loss": 0.1442,
      "reward": 1.2000000029802322,
      "reward_std": 0.26129765436053276,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.9000000059604645,
      "step": 279
    },
    {
      "completion_length": 150.2000026702881,
      "epoch": 0.2903058579574909,
      "grad_norm": 0.8232119679450989,
      "kl": 0.640625,
      "learning_rate": 2.8965517241379314e-06,
      "loss": 0.136,
      "reward": 1.350000023841858,
      "reward_std": 0.335410188883543,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 280
    },
    {
      "completion_length": 104.10000324249268,
      "epoch": 0.2913426645930534,
      "grad_norm": 0.7683313488960266,
      "kl": 0.57568359375,
      "learning_rate": 2.906896551724138e-06,
      "loss": 0.084,
      "reward": 1.2500000447034836,
      "reward_std": 0.3046357296407223,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 281
    },
    {
      "completion_length": 179.05000019073486,
      "epoch": 0.29237947122861585,
      "grad_norm": 0.7294011116027832,
      "kl": 0.525390625,
      "learning_rate": 2.917241379310345e-06,
      "loss": 0.2052,
      "reward": 1.4000000357627869,
      "reward_std": 0.4237985461950302,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.925000011920929,
      "step": 282
    },
    {
      "completion_length": 99.7750015258789,
      "epoch": 0.29341627786417834,
      "grad_norm": 0.783625602722168,
      "kl": 0.57958984375,
      "learning_rate": 2.9275862068965516e-06,
      "loss": 0.0685,
      "reward": 1.425000011920929,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.4500000048428774,
      "rewards/format_reward": 0.9750000089406967,
      "step": 283
    },
    {
      "completion_length": 150.60000228881836,
      "epoch": 0.2944530844997408,
      "grad_norm": 0.9992136359214783,
      "kl": 0.68603515625,
      "learning_rate": 2.9379310344827584e-06,
      "loss": 0.1626,
      "reward": 1.4500000029802322,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000029802322,
      "step": 284
    },
    {
      "completion_length": 100.15000057220459,
      "epoch": 0.2954898911353033,
      "grad_norm": 0.5232768654823303,
      "kl": 0.70703125,
      "learning_rate": 2.9482758620689655e-06,
      "loss": 0.0727,
      "reward": 1.2500000149011612,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 285
    },
    {
      "completion_length": 146.7750015258789,
      "epoch": 0.2965266977708657,
      "grad_norm": 1.0567251443862915,
      "kl": 0.83203125,
      "learning_rate": 2.9586206896551723e-06,
      "loss": 0.1205,
      "reward": 1.225000023841858,
      "reward_std": 0.37323957309126854,
      "rewards/accuracy_reward": 0.3000000063329935,
      "rewards/format_reward": 0.925000011920929,
      "step": 286
    },
    {
      "completion_length": 96.50000286102295,
      "epoch": 0.2975635044064282,
      "grad_norm": 0.7808656096458435,
      "kl": 0.728515625,
      "learning_rate": 2.9689655172413794e-06,
      "loss": 0.0931,
      "reward": 1.4000000357627869,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.4250000100582838,
      "rewards/format_reward": 0.9750000089406967,
      "step": 287
    },
    {
      "completion_length": 196.0250015258789,
      "epoch": 0.2986003110419907,
      "grad_norm": 0.7538319230079651,
      "kl": 0.6689453125,
      "learning_rate": 2.979310344827586e-06,
      "loss": 0.1413,
      "reward": 1.4500000327825546,
      "reward_std": 0.26129764690995216,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.925000011920929,
      "step": 288
    },
    {
      "completion_length": 144.70000076293945,
      "epoch": 0.29963711767755313,
      "grad_norm": 1.2029356956481934,
      "kl": 1.0400390625,
      "learning_rate": 2.9896551724137933e-06,
      "loss": 0.1362,
      "reward": 1.175000011920929,
      "reward_std": 0.31199513748288155,
      "rewards/accuracy_reward": 0.22500000707805157,
      "rewards/format_reward": 0.9500000178813934,
      "step": 289
    },
    {
      "completion_length": 354.9250087738037,
      "epoch": 0.3006739243131156,
      "grad_norm": 0.8506849408149719,
      "kl": 0.8056640625,
      "learning_rate": 3e-06,
      "loss": 0.3898,
      "reward": 1.25,
      "reward_std": 0.5210216455161572,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.800000011920929,
      "step": 290
    },
    {
      "completion_length": 223.7750062942505,
      "epoch": 0.30171073094867806,
      "grad_norm": 0.479114294052124,
      "kl": 0.705078125,
      "learning_rate": 2.999998906682319e-06,
      "loss": 0.135,
      "reward": 1.2750000357627869,
      "reward_std": 0.22531899064779282,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.8750000149011612,
      "step": 291
    },
    {
      "completion_length": 140.07500171661377,
      "epoch": 0.30274753758424056,
      "grad_norm": 0.6931892037391663,
      "kl": 0.734375,
      "learning_rate": 2.9999956267308692e-06,
      "loss": 0.1173,
      "reward": 1.5000000149011612,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.9500000178813934,
      "step": 292
    },
    {
      "completion_length": 172.12500286102295,
      "epoch": 0.303784344219803,
      "grad_norm": 0.528679609298706,
      "kl": 0.580078125,
      "learning_rate": 2.9999901601504326e-06,
      "loss": 0.1734,
      "reward": 1.2749999910593033,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 293
    },
    {
      "completion_length": 313.8000030517578,
      "epoch": 0.3048211508553655,
      "grad_norm": 0.8700326681137085,
      "kl": 0.62646484375,
      "learning_rate": 2.9999825069489775e-06,
      "loss": 0.2371,
      "reward": 1.0500000268220901,
      "reward_std": 0.47768355160951614,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.8250000178813934,
      "step": 294
    },
    {
      "completion_length": 239.125,
      "epoch": 0.3058579574909279,
      "grad_norm": 0.5965785384178162,
      "kl": 0.740234375,
      "learning_rate": 2.999972667137661e-06,
      "loss": 0.2412,
      "reward": 0.9750000387430191,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.10000000149011612,
      "rewards/format_reward": 0.8750000298023224,
      "step": 295
    },
    {
      "completion_length": 161.44999980926514,
      "epoch": 0.3068947641264904,
      "grad_norm": 0.45877698063850403,
      "kl": 0.638671875,
      "learning_rate": 2.9999606407308272e-06,
      "loss": 0.1932,
      "reward": 1.3750000596046448,
      "reward_std": 0.3913118876516819,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.9250000268220901,
      "step": 296
    },
    {
      "completion_length": 230.45000648498535,
      "epoch": 0.30793157076205285,
      "grad_norm": 0.774716854095459,
      "kl": 0.755859375,
      "learning_rate": 2.999946427746007e-06,
      "loss": 0.2691,
      "reward": 1.2750000357627869,
      "reward_std": 0.472340852022171,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 0.8750000149011612,
      "step": 297
    },
    {
      "completion_length": 90.85000038146973,
      "epoch": 0.30896837739761535,
      "grad_norm": 0.41126278042793274,
      "kl": 0.732421875,
      "learning_rate": 2.99993002820392e-06,
      "loss": 0.0062,
      "reward": 1.1500000059604645,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.17500000447034836,
      "rewards/format_reward": 0.9750000089406967,
      "step": 298
    },
    {
      "completion_length": 37.85000038146973,
      "epoch": 0.31000518403317784,
      "grad_norm": 1.6783168315887451,
      "kl": 0.7119140625,
      "learning_rate": 2.999911442128473e-06,
      "loss": 0.0434,
      "reward": 1.2500000298023224,
      "reward_std": 0.1928323321044445,
      "rewards/accuracy_reward": 0.24999999813735485,
      "rewards/format_reward": 1.0,
      "step": 299
    },
    {
      "completion_length": 91.00000095367432,
      "epoch": 0.3110419906687403,
      "grad_norm": 0.8122076988220215,
      "kl": 0.76318359375,
      "learning_rate": 2.9998906695467587e-06,
      "loss": 0.1001,
      "reward": 1.475000023841858,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 300
    },
    {
      "completion_length": 55.82500076293945,
      "epoch": 0.31207879730430277,
      "grad_norm": 0.7224457859992981,
      "kl": 0.87890625,
      "learning_rate": 2.9998677104890597e-06,
      "loss": 0.1167,
      "reward": 1.5750000476837158,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.6250000186264515,
      "rewards/format_reward": 0.9500000178813934,
      "step": 301
    },
    {
      "completion_length": 138.60000133514404,
      "epoch": 0.3131156039398652,
      "grad_norm": 0.742752730846405,
      "kl": 0.798828125,
      "learning_rate": 2.9998425649888448e-06,
      "loss": 0.1063,
      "reward": 1.0750000178813934,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.15000000037252903,
      "rewards/format_reward": 0.925000011920929,
      "step": 302
    },
    {
      "completion_length": 148.55000114440918,
      "epoch": 0.3141524105754277,
      "grad_norm": 1.1690900325775146,
      "kl": 0.955078125,
      "learning_rate": 2.9998152330827693e-06,
      "loss": 0.1856,
      "reward": 1.4000000357627869,
      "reward_std": 0.4723408482968807,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.8750000149011612,
      "step": 303
    },
    {
      "completion_length": 225.50000381469727,
      "epoch": 0.31518921721099014,
      "grad_norm": 0.7336009740829468,
      "kl": 0.8759765625,
      "learning_rate": 2.9997857148106767e-06,
      "loss": 0.2877,
      "reward": 1.0750000476837158,
      "reward_std": 0.3479737937450409,
      "rewards/accuracy_reward": 0.22500000149011612,
      "rewards/format_reward": 0.8500000238418579,
      "step": 304
    },
    {
      "completion_length": 199.55000400543213,
      "epoch": 0.31622602384655263,
      "grad_norm": 1.3255078792572021,
      "kl": 1.025390625,
      "learning_rate": 2.9997540102155977e-06,
      "loss": 0.2364,
      "reward": 1.3000000417232513,
      "reward_std": 0.4723408408463001,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.8750000149011612,
      "step": 305
    },
    {
      "completion_length": 237.2249984741211,
      "epoch": 0.31726283048211507,
      "grad_norm": 0.7050570845603943,
      "kl": 0.7607421875,
      "learning_rate": 2.9997201193437498e-06,
      "loss": 0.2609,
      "reward": 1.3750000596046448,
      "reward_std": 0.5481655970215797,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8250000178813934,
      "step": 306
    },
    {
      "completion_length": 232.12500858306885,
      "epoch": 0.31829963711767756,
      "grad_norm": 0.6234104037284851,
      "kl": 0.64306640625,
      "learning_rate": 2.9996840422445373e-06,
      "loss": 0.2367,
      "reward": 1.0250000208616257,
      "reward_std": 0.3804604709148407,
      "rewards/accuracy_reward": 0.15000000223517418,
      "rewards/format_reward": 0.8750000149011612,
      "step": 307
    },
    {
      "completion_length": 286.7750015258789,
      "epoch": 0.31933644375324,
      "grad_norm": 0.5603132247924805,
      "kl": 0.68310546875,
      "learning_rate": 2.9996457789705527e-06,
      "loss": 0.3039,
      "reward": 1.1500000208616257,
      "reward_std": 0.38409100100398064,
      "rewards/accuracy_reward": 0.2750000022351742,
      "rewards/format_reward": 0.8750000149011612,
      "step": 308
    },
    {
      "completion_length": 97.02500247955322,
      "epoch": 0.3203732503888025,
      "grad_norm": 0.44136732816696167,
      "kl": 0.67578125,
      "learning_rate": 2.9996053295775733e-06,
      "loss": 0.0828,
      "reward": 1.25,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 309
    },
    {
      "completion_length": 253.27499961853027,
      "epoch": 0.321410057024365,
      "grad_norm": 0.6165421009063721,
      "kl": 0.64892578125,
      "learning_rate": 2.999562694124565e-06,
      "loss": 0.2532,
      "reward": 1.350000023841858,
      "reward_std": 0.347973819822073,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.8750000298023224,
      "step": 310
    },
    {
      "completion_length": 67.55000019073486,
      "epoch": 0.3224468636599274,
      "grad_norm": 0.5345252752304077,
      "kl": 0.552734375,
      "learning_rate": 2.999517872673681e-06,
      "loss": 0.0778,
      "reward": 1.5250000208616257,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 311
    },
    {
      "completion_length": 140.20000171661377,
      "epoch": 0.3234836702954899,
      "grad_norm": 1.1492671966552734,
      "kl": 0.56787109375,
      "learning_rate": 2.9994708652902586e-06,
      "loss": 0.1224,
      "reward": 1.2000000029802322,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.9500000178813934,
      "step": 312
    },
    {
      "completion_length": 95.85000228881836,
      "epoch": 0.32452047693105235,
      "grad_norm": 0.9709307551383972,
      "kl": 0.71728515625,
      "learning_rate": 2.999421672042824e-06,
      "loss": 0.0839,
      "reward": 1.449999988079071,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 313
    },
    {
      "completion_length": 92.50000190734863,
      "epoch": 0.32555728356661484,
      "grad_norm": 0.989802896976471,
      "kl": 0.6923828125,
      "learning_rate": 2.999370293003089e-06,
      "loss": 0.0873,
      "reward": 1.4250000268220901,
      "reward_std": 0.18026872351765633,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 314
    },
    {
      "completion_length": 140.47500038146973,
      "epoch": 0.3265940902021773,
      "grad_norm": 0.6137426495552063,
      "kl": 0.791015625,
      "learning_rate": 2.9993167282459508e-06,
      "loss": 0.1567,
      "reward": 1.5750000178813934,
      "reward_std": 0.3479738123714924,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 315
    },
    {
      "completion_length": 105.0000057220459,
      "epoch": 0.32763089683773977,
      "grad_norm": 0.6631371974945068,
      "kl": 0.58154296875,
      "learning_rate": 2.9992609778494945e-06,
      "loss": 0.0668,
      "reward": 1.3250000178813934,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.9500000178813934,
      "step": 316
    },
    {
      "completion_length": 57.30000114440918,
      "epoch": 0.3286677034733022,
      "grad_norm": 0.542193591594696,
      "kl": 0.53125,
      "learning_rate": 2.999203041894991e-06,
      "loss": 0.0198,
      "reward": 1.475000023841858,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 1.0,
      "step": 317
    },
    {
      "completion_length": 110.09999942779541,
      "epoch": 0.3297045101088647,
      "grad_norm": 0.8033139705657959,
      "kl": 0.7060546875,
      "learning_rate": 2.999142920466896e-06,
      "loss": 0.0931,
      "reward": 1.3750000298023224,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.4250000026077032,
      "rewards/format_reward": 0.9500000029802322,
      "step": 318
    },
    {
      "completion_length": 155.02500247955322,
      "epoch": 0.33074131674442714,
      "grad_norm": 0.7720692753791809,
      "kl": 0.59521484375,
      "learning_rate": 2.999080613652853e-06,
      "loss": 0.102,
      "reward": 1.600000023841858,
      "reward_std": 0.25609345361590385,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 319
    },
    {
      "completion_length": 125.42500495910645,
      "epoch": 0.33177812337998963,
      "grad_norm": 0.43371641635894775,
      "kl": 0.7421875,
      "learning_rate": 2.9990161215436888e-06,
      "loss": 0.1467,
      "reward": 1.0500000268220901,
      "reward_std": 0.22360678762197495,
      "rewards/accuracy_reward": 0.10000000335276127,
      "rewards/format_reward": 0.9500000178813934,
      "step": 320
    },
    {
      "completion_length": 63.07500076293945,
      "epoch": 0.3328149300155521,
      "grad_norm": 1.1089128255844116,
      "kl": 0.56591796875,
      "learning_rate": 2.9989494442334186e-06,
      "loss": 0.0028,
      "reward": 1.375,
      "reward_std": 0.27228759974241257,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.9750000089406967,
      "step": 321
    },
    {
      "completion_length": 69.57500171661377,
      "epoch": 0.33385173665111456,
      "grad_norm": 0.5004400610923767,
      "kl": 0.697265625,
      "learning_rate": 2.998880581819241e-06,
      "loss": 0.0053,
      "reward": 1.5250000357627869,
      "reward_std": 0.13693063706159592,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 1.0,
      "step": 322
    },
    {
      "completion_length": 85.10000228881836,
      "epoch": 0.33488854328667705,
      "grad_norm": 0.7260141968727112,
      "kl": 0.4267578125,
      "learning_rate": 2.9988095344015403e-06,
      "loss": 0.0783,
      "reward": 1.5499999821186066,
      "reward_std": 0.279508501291275,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 323
    },
    {
      "completion_length": 179.37500381469727,
      "epoch": 0.3359253499222395,
      "grad_norm": 0.7410564422607422,
      "kl": 0.70703125,
      "learning_rate": 2.9987363020838867e-06,
      "loss": 0.1431,
      "reward": 1.25,
      "reward_std": 0.3856646940112114,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.925000011920929,
      "step": 324
    },
    {
      "completion_length": 128.05000114440918,
      "epoch": 0.336962156557802,
      "grad_norm": 0.2647928297519684,
      "kl": 0.49267578125,
      "learning_rate": 2.9986608849730356e-06,
      "loss": 0.0786,
      "reward": 1.2500000298023224,
      "reward_std": 0.18026870489120483,
      "rewards/accuracy_reward": 0.2750000022351742,
      "rewards/format_reward": 0.9750000089406967,
      "step": 325
    },
    {
      "completion_length": 139.60000228881836,
      "epoch": 0.3379989631933644,
      "grad_norm": 0.6439674496650696,
      "kl": 0.62255859375,
      "learning_rate": 2.998583283178926e-06,
      "loss": 0.113,
      "reward": 1.3750000298023224,
      "reward_std": 0.31733787432312965,
      "rewards/accuracy_reward": 0.4250000100582838,
      "rewards/format_reward": 0.9500000178813934,
      "step": 326
    },
    {
      "completion_length": 122.85000038146973,
      "epoch": 0.3390357698289269,
      "grad_norm": 0.4776754677295685,
      "kl": 0.53173828125,
      "learning_rate": 2.9985034968146832e-06,
      "loss": 0.0487,
      "reward": 1.7000000178813934,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 327
    },
    {
      "completion_length": 324.1000061035156,
      "epoch": 0.34007257646448935,
      "grad_norm": 0.7620232701301575,
      "kl": 0.4736328125,
      "learning_rate": 2.998421525996616e-06,
      "loss": 0.2654,
      "reward": 1.2750000357627869,
      "reward_std": 0.5517961122095585,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 0.8750000298023224,
      "step": 328
    },
    {
      "completion_length": 468.9500198364258,
      "epoch": 0.34110938310005184,
      "grad_norm": 0.35654211044311523,
      "kl": 0.5458984375,
      "learning_rate": 2.9983373708442182e-06,
      "loss": 0.3329,
      "reward": 1.3500000089406967,
      "reward_std": 0.4958944171667099,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.7750000059604645,
      "step": 329
    },
    {
      "completion_length": 370.1750030517578,
      "epoch": 0.3421461897356143,
      "grad_norm": 0.5573723316192627,
      "kl": 0.564453125,
      "learning_rate": 2.9982510314801665e-06,
      "loss": 0.2937,
      "reward": 1.2500000149011612,
      "reward_std": 0.6527480855584145,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 0.8500000238418579,
      "step": 330
    },
    {
      "completion_length": 277.975004196167,
      "epoch": 0.3431829963711768,
      "grad_norm": 0.5659462213516235,
      "kl": 0.5029296875,
      "learning_rate": 2.998162508030324e-06,
      "loss": 0.2449,
      "reward": 1.5000000298023224,
      "reward_std": 0.5031153149902821,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9000000208616257,
      "step": 331
    },
    {
      "completion_length": 175.72500610351562,
      "epoch": 0.34421980300673927,
      "grad_norm": 0.402981698513031,
      "kl": 0.49267578125,
      "learning_rate": 2.998071800623736e-06,
      "loss": 0.132,
      "reward": 1.550000011920929,
      "reward_std": 0.15685366839170456,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 332
    },
    {
      "completion_length": 442.42501068115234,
      "epoch": 0.3452566096423017,
      "grad_norm": 0.40613406896591187,
      "kl": 0.59130859375,
      "learning_rate": 2.997978909392631e-06,
      "loss": 0.3856,
      "reward": 1.3750000298023224,
      "reward_std": 0.7138539552688599,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.7750000059604645,
      "step": 333
    },
    {
      "completion_length": 437.3250045776367,
      "epoch": 0.3462934162778642,
      "grad_norm": 0.4869440793991089,
      "kl": 0.6865234375,
      "learning_rate": 2.997883834472423e-06,
      "loss": 0.2613,
      "reward": 1.1000000089406967,
      "reward_std": 0.596707858145237,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.7750000208616257,
      "step": 334
    },
    {
      "completion_length": 175.32500648498535,
      "epoch": 0.34733022291342663,
      "grad_norm": 0.45390427112579346,
      "kl": 0.51025390625,
      "learning_rate": 2.997786576001708e-06,
      "loss": 0.1233,
      "reward": 1.600000023841858,
      "reward_std": 0.3245587684214115,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 335
    },
    {
      "completion_length": 263.725004196167,
      "epoch": 0.3483670295489891,
      "grad_norm": 0.42588937282562256,
      "kl": 0.72509765625,
      "learning_rate": 2.9976871341222644e-06,
      "loss": 0.2371,
      "reward": 1.3250000178813934,
      "reward_std": 0.3913118802011013,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9000000208616257,
      "step": 336
    },
    {
      "completion_length": 228.55000686645508,
      "epoch": 0.34940383618455156,
      "grad_norm": 0.6435961127281189,
      "kl": 0.6455078125,
      "learning_rate": 2.997585508979055e-06,
      "loss": 0.1453,
      "reward": 1.425000011920929,
      "reward_std": 0.3479738272726536,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 337
    },
    {
      "completion_length": 381.4500026702881,
      "epoch": 0.35044064282011406,
      "grad_norm": 0.48273399472236633,
      "kl": 0.6640625,
      "learning_rate": 2.9974817007202243e-06,
      "loss": 0.2484,
      "reward": 1.350000038743019,
      "reward_std": 0.48553743213415146,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.8250000029802322,
      "step": 338
    },
    {
      "completion_length": 181.50000381469727,
      "epoch": 0.3514774494556765,
      "grad_norm": 0.4054751694202423,
      "kl": 0.6533203125,
      "learning_rate": 2.9973757094970992e-06,
      "loss": 0.1321,
      "reward": 1.050000011920929,
      "reward_std": 0.2812207043170929,
      "rewards/accuracy_reward": 0.15000000223517418,
      "rewards/format_reward": 0.9000000059604645,
      "step": 339
    },
    {
      "completion_length": 564.7500076293945,
      "epoch": 0.352514256091239,
      "grad_norm": 0.4397445321083069,
      "kl": 0.755859375,
      "learning_rate": 2.9972675354641895e-06,
      "loss": 0.2461,
      "reward": 1.0500000268220901,
      "reward_std": 0.49746809527277946,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.7250000089406967,
      "step": 340
    },
    {
      "completion_length": 295.7500057220459,
      "epoch": 0.3535510627268014,
      "grad_norm": 0.6863256096839905,
      "kl": 0.6005859375,
      "learning_rate": 2.9971571787791866e-06,
      "loss": 0.0904,
      "reward": 1.550000011920929,
      "reward_std": 0.5533697679638863,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 341
    },
    {
      "completion_length": 529.0750045776367,
      "epoch": 0.3545878693623639,
      "grad_norm": 0.5725196003913879,
      "kl": 0.5673828125,
      "learning_rate": 2.9970446396029643e-06,
      "loss": 0.3386,
      "reward": 0.9500000178813934,
      "reward_std": 0.5535083040595055,
      "rewards/accuracy_reward": 0.1750000026077032,
      "rewards/format_reward": 0.7750000208616257,
      "step": 342
    },
    {
      "completion_length": 258.2249984741211,
      "epoch": 0.3556246759979264,
      "grad_norm": 0.4096214771270752,
      "kl": 0.51904296875,
      "learning_rate": 2.9969299180995763e-06,
      "loss": 0.2192,
      "reward": 1.2750000059604645,
      "reward_std": 0.3913118839263916,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.875,
      "step": 343
    },
    {
      "completion_length": 363.0500030517578,
      "epoch": 0.35666148263348885,
      "grad_norm": 0.6106461882591248,
      "kl": 0.5458984375,
      "learning_rate": 2.996813014436259e-06,
      "loss": 0.2965,
      "reward": 1.4000000357627869,
      "reward_std": 0.6172638833522797,
      "rewards/accuracy_reward": 0.5750000141561031,
      "rewards/format_reward": 0.8250000178813934,
      "step": 344
    },
    {
      "completion_length": 123.2500057220459,
      "epoch": 0.35769828926905134,
      "grad_norm": 0.5532033443450928,
      "kl": 0.46533203125,
      "learning_rate": 2.9966939287834297e-06,
      "loss": -0.0141,
      "reward": 1.5,
      "reward_std": 0.3731010667979717,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 345
    },
    {
      "completion_length": 330.5750045776367,
      "epoch": 0.3587350959046138,
      "grad_norm": 0.5643976330757141,
      "kl": 0.4521484375,
      "learning_rate": 2.9965726613146864e-06,
      "loss": 0.2691,
      "reward": 0.9000000208616257,
      "reward_std": 0.5625799112021923,
      "rewards/accuracy_reward": 0.1250000037252903,
      "rewards/format_reward": 0.7750000059604645,
      "step": 346
    },
    {
      "completion_length": 296.6499996185303,
      "epoch": 0.35977190254017627,
      "grad_norm": 0.5962963104248047,
      "kl": 0.5673828125,
      "learning_rate": 2.9964492122068074e-06,
      "loss": 0.1955,
      "reward": 1.2750000357627869,
      "reward_std": 0.5660719126462936,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.875,
      "step": 347
    },
    {
      "completion_length": 210.375,
      "epoch": 0.3608087091757387,
      "grad_norm": 0.4225082993507385,
      "kl": 0.61572265625,
      "learning_rate": 2.9963235816397516e-06,
      "loss": 0.1628,
      "reward": 1.350000023841858,
      "reward_std": 0.29270509257912636,
      "rewards/accuracy_reward": 0.4250000100582838,
      "rewards/format_reward": 0.925000011920929,
      "step": 348
    },
    {
      "completion_length": 552.0000076293945,
      "epoch": 0.3618455158113012,
      "grad_norm": 0.7490342855453491,
      "kl": 0.5478515625,
      "learning_rate": 2.996195769796658e-06,
      "loss": 0.2469,
      "reward": 1.1250000447034836,
      "reward_std": 0.5973408445715904,
      "rewards/accuracy_reward": 0.3500000052154064,
      "rewards/format_reward": 0.7750000208616257,
      "step": 349
    },
    {
      "completion_length": 236.82500839233398,
      "epoch": 0.36288232244686364,
      "grad_norm": 0.4927626848220825,
      "kl": 0.4287109375,
      "learning_rate": 2.996065776863845e-06,
      "loss": 0.0864,
      "reward": 1.475000023841858,
      "reward_std": 0.3966546319425106,
      "rewards/accuracy_reward": 0.5500000156462193,
      "rewards/format_reward": 0.925000011920929,
      "step": 350
    },
    {
      "completion_length": 193.45000457763672,
      "epoch": 0.36391912908242613,
      "grad_norm": 0.5020635724067688,
      "kl": 0.38720703125,
      "learning_rate": 2.9959336030308104e-06,
      "loss": 0.0558,
      "reward": 1.2500000298023224,
      "reward_std": 0.3299014940857887,
      "rewards/accuracy_reward": 0.30000000819563866,
      "rewards/format_reward": 0.9500000178813934,
      "step": 351
    },
    {
      "completion_length": 222.27500534057617,
      "epoch": 0.3649559357179886,
      "grad_norm": 0.4954478442668915,
      "kl": 0.48486328125,
      "learning_rate": 2.995799248490232e-06,
      "loss": 0.1715,
      "reward": 1.2000000178813934,
      "reward_std": 0.36053742095828056,
      "rewards/accuracy_reward": 0.27500000409781933,
      "rewards/format_reward": 0.9250000268220901,
      "step": 352
    },
    {
      "completion_length": 213.42500686645508,
      "epoch": 0.36599274235355106,
      "grad_norm": 0.42243900895118713,
      "kl": 0.34814453125,
      "learning_rate": 2.995662713437966e-06,
      "loss": 0.1177,
      "reward": 1.5500000417232513,
      "reward_std": 0.4290027320384979,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 353
    },
    {
      "completion_length": 118.6500015258789,
      "epoch": 0.36702954898911355,
      "grad_norm": 0.5565142035484314,
      "kl": 0.290283203125,
      "learning_rate": 2.995523998073047e-06,
      "loss": 0.0399,
      "reward": 1.675000011920929,
      "reward_std": 0.31719936057925224,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 1.0,
      "step": 354
    },
    {
      "completion_length": 167.67500495910645,
      "epoch": 0.368066355624676,
      "grad_norm": 0.425595223903656,
      "kl": 0.333984375,
      "learning_rate": 2.9953831025976885e-06,
      "loss": 0.0722,
      "reward": 1.5750000178813934,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.5999999940395355,
      "rewards/format_reward": 0.9750000089406967,
      "step": 355
    },
    {
      "completion_length": 263.0250053405762,
      "epoch": 0.3691031622602385,
      "grad_norm": 0.34887924790382385,
      "kl": 0.301025390625,
      "learning_rate": 2.9952400272172815e-06,
      "loss": 0.1951,
      "reward": 1.4250000268220901,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 356
    },
    {
      "completion_length": 161.350004196167,
      "epoch": 0.3701399688958009,
      "grad_norm": 0.5152186155319214,
      "kl": 0.34130859375,
      "learning_rate": 2.995094772140396e-06,
      "loss": 0.0896,
      "reward": 1.4000000357627869,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.4249999988824129,
      "rewards/format_reward": 0.9750000089406967,
      "step": 357
    },
    {
      "completion_length": 126.57500076293945,
      "epoch": 0.3711767755313634,
      "grad_norm": 0.5196745991706848,
      "kl": 0.33740234375,
      "learning_rate": 2.9949473375787776e-06,
      "loss": -0.0047,
      "reward": 1.2750000059604645,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.3000000063329935,
      "rewards/format_reward": 0.9750000089406967,
      "step": 358
    },
    {
      "completion_length": 160.05000686645508,
      "epoch": 0.37221358216692585,
      "grad_norm": 0.6606730222702026,
      "kl": 0.4091796875,
      "learning_rate": 2.994797723747351e-06,
      "loss": 0.1742,
      "reward": 1.5750000178813934,
      "reward_std": 0.4363621734082699,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 359
    },
    {
      "completion_length": 119.82500267028809,
      "epoch": 0.37325038880248834,
      "grad_norm": 0.3371819257736206,
      "kl": 0.364990234375,
      "learning_rate": 2.9946459308642155e-06,
      "loss": -0.0122,
      "reward": 1.6250000298023224,
      "reward_std": 0.13693063706159592,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 1.0,
      "step": 360
    },
    {
      "completion_length": 221.07500076293945,
      "epoch": 0.3742871954380508,
      "grad_norm": 0.536543607711792,
      "kl": 0.4248046875,
      "learning_rate": 2.9944919591506495e-06,
      "loss": 0.1191,
      "reward": 1.1500000208616257,
      "reward_std": 0.25609344989061356,
      "rewards/accuracy_reward": 0.20000000670552254,
      "rewards/format_reward": 0.9500000029802322,
      "step": 361
    },
    {
      "completion_length": 203.3250026702881,
      "epoch": 0.37532400207361327,
      "grad_norm": 0.3392835855484009,
      "kl": 0.365234375,
      "learning_rate": 2.994335808831106e-06,
      "loss": 0.0678,
      "reward": 1.5750000178813934,
      "reward_std": 0.19346532225608826,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000029802322,
      "step": 362
    },
    {
      "completion_length": 148.77499961853027,
      "epoch": 0.37636080870917576,
      "grad_norm": 0.15861853957176208,
      "kl": 0.36328125,
      "learning_rate": 2.994177480133214e-06,
      "loss": 0.0677,
      "reward": 1.3499999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9750000089406967,
      "step": 363
    },
    {
      "completion_length": 403.47501373291016,
      "epoch": 0.3773976153447382,
      "grad_norm": 0.5314847230911255,
      "kl": 0.5068359375,
      "learning_rate": 2.994016973287778e-06,
      "loss": 0.2355,
      "reward": 1.4000000059604645,
      "reward_std": 0.5173911042511463,
      "rewards/accuracy_reward": 0.5500000156462193,
      "rewards/format_reward": 0.8500000238418579,
      "step": 364
    },
    {
      "completion_length": 160.77500534057617,
      "epoch": 0.3784344219803007,
      "grad_norm": 0.3340126574039459,
      "kl": 0.288330078125,
      "learning_rate": 2.9938542885287786e-06,
      "loss": 0.0289,
      "reward": 1.350000023841858,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 365
    },
    {
      "completion_length": 215.65000534057617,
      "epoch": 0.37947122861586313,
      "grad_norm": 0.35049688816070557,
      "kl": 0.376953125,
      "learning_rate": 2.99368942609337e-06,
      "loss": 0.1184,
      "reward": 1.3000000268220901,
      "reward_std": 0.22894952073693275,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.9500000178813934,
      "step": 366
    },
    {
      "completion_length": 192.12500381469727,
      "epoch": 0.3805080352514256,
      "grad_norm": 0.4274624288082123,
      "kl": 0.3583984375,
      "learning_rate": 2.993522386221882e-06,
      "loss": -0.0248,
      "reward": 1.5999999940395355,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9500000029802322,
      "step": 367
    },
    {
      "completion_length": 148.375,
      "epoch": 0.38154484188698806,
      "grad_norm": 0.6505801677703857,
      "kl": 0.35107421875,
      "learning_rate": 2.9933531691578178e-06,
      "loss": 0.026,
      "reward": 1.550000011920929,
      "reward_std": 0.26865705102682114,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.9750000089406967,
      "step": 368
    },
    {
      "completion_length": 293.8000030517578,
      "epoch": 0.38258164852255055,
      "grad_norm": 0.30225127935409546,
      "kl": 0.44970703125,
      "learning_rate": 2.9931817751478547e-06,
      "loss": 0.2183,
      "reward": 1.4250000268220901,
      "reward_std": 0.25609344989061356,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9000000059604645,
      "step": 369
    },
    {
      "completion_length": 151.32499885559082,
      "epoch": 0.383618455158113,
      "grad_norm": 0.5463283658027649,
      "kl": 0.296875,
      "learning_rate": 2.9930082044418442e-06,
      "loss": 0.0759,
      "reward": 1.350000023841858,
      "reward_std": 0.24873404577374458,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 370
    },
    {
      "completion_length": 120.22500228881836,
      "epoch": 0.3846552617936755,
      "grad_norm": 0.6592673063278198,
      "kl": 0.37451171875,
      "learning_rate": 2.99283245729281e-06,
      "loss": 0.083,
      "reward": 1.4000000357627869,
      "reward_std": 0.37310105934739113,
      "rewards/accuracy_reward": 0.4000000152736902,
      "rewards/format_reward": 1.0,
      "step": 371
    },
    {
      "completion_length": 150.79999923706055,
      "epoch": 0.3856920684292379,
      "grad_norm": 0.5108554363250732,
      "kl": 0.326171875,
      "learning_rate": 2.9926545339569476e-06,
      "loss": 0.0444,
      "reward": 1.4500000178813934,
      "reward_std": 0.30463575199246407,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 372
    },
    {
      "completion_length": 87.57500267028809,
      "epoch": 0.3867288750648004,
      "grad_norm": 0.5240278840065002,
      "kl": 0.3642578125,
      "learning_rate": 2.992474434693627e-06,
      "loss": -0.0313,
      "reward": 1.600000023841858,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.6000000033527613,
      "rewards/format_reward": 1.0,
      "step": 373
    },
    {
      "completion_length": 108.4000015258789,
      "epoch": 0.3877656817003629,
      "grad_norm": 0.46822282671928406,
      "kl": 0.33447265625,
      "learning_rate": 2.9922921597653893e-06,
      "loss": -0.0006,
      "reward": 1.5750000178813934,
      "reward_std": 0.1928323470056057,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 1.0,
      "step": 374
    },
    {
      "completion_length": 137.70000457763672,
      "epoch": 0.38880248833592534,
      "grad_norm": 0.554831326007843,
      "kl": 0.4365234375,
      "learning_rate": 2.9921077094379466e-06,
      "loss": 0.0776,
      "reward": 1.375,
      "reward_std": 0.2812207005918026,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 0.9750000089406967,
      "step": 375
    },
    {
      "completion_length": 239.45000839233398,
      "epoch": 0.38983929497148784,
      "grad_norm": 0.5239160060882568,
      "kl": 0.412841796875,
      "learning_rate": 2.9919210839801827e-06,
      "loss": 0.1685,
      "reward": 1.475000023841858,
      "reward_std": 0.452556349337101,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.925000011920929,
      "step": 376
    },
    {
      "completion_length": 87.90000343322754,
      "epoch": 0.3908761016070503,
      "grad_norm": 0.6395469307899475,
      "kl": 0.35498046875,
      "learning_rate": 2.991732283664152e-06,
      "loss": 0.0022,
      "reward": 1.5250000357627869,
      "reward_std": 0.3046357184648514,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 1.0,
      "step": 377
    },
    {
      "completion_length": 200.35000038146973,
      "epoch": 0.39191290824261277,
      "grad_norm": 0.6615502834320068,
      "kl": 0.3955078125,
      "learning_rate": 2.99154130876508e-06,
      "loss": 0.1602,
      "reward": 1.425000011920929,
      "reward_std": 0.504966028034687,
      "rewards/accuracy_reward": 0.5250000152736902,
      "rewards/format_reward": 0.9000000357627869,
      "step": 378
    },
    {
      "completion_length": 243.87500190734863,
      "epoch": 0.3929497148781752,
      "grad_norm": 0.588063657283783,
      "kl": 0.41064453125,
      "learning_rate": 2.9913481595613613e-06,
      "loss": 0.1196,
      "reward": 1.2500000149011612,
      "reward_std": 0.4181513413786888,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.9000000059604645,
      "step": 379
    },
    {
      "completion_length": 92.82500076293945,
      "epoch": 0.3939865215137377,
      "grad_norm": 0.3841590881347656,
      "kl": 0.41552734375,
      "learning_rate": 2.99115283633456e-06,
      "loss": 0.022,
      "reward": 1.7000000178813934,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.7000000067055225,
      "rewards/format_reward": 1.0,
      "step": 380
    },
    {
      "completion_length": 167.8250026702881,
      "epoch": 0.39502332814930013,
      "grad_norm": 0.6007952094078064,
      "kl": 0.5205078125,
      "learning_rate": 2.9909553393694107e-06,
      "loss": 0.1088,
      "reward": 1.2999999970197678,
      "reward_std": 0.3245587833225727,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9000000059604645,
      "step": 381
    },
    {
      "completion_length": 150.350004196167,
      "epoch": 0.3960601347848626,
      "grad_norm": 0.4360502064228058,
      "kl": 0.29931640625,
      "learning_rate": 2.9907556689538158e-06,
      "loss": 0.0512,
      "reward": 1.4500000178813934,
      "reward_std": 0.22894952446222305,
      "rewards/accuracy_reward": 0.4749999940395355,
      "rewards/format_reward": 0.9750000089406967,
      "step": 382
    },
    {
      "completion_length": 145.3750057220459,
      "epoch": 0.39709694142042506,
      "grad_norm": 0.7903898358345032,
      "kl": 0.355712890625,
      "learning_rate": 2.990553825378846e-06,
      "loss": 0.127,
      "reward": 1.6250000298023224,
      "reward_std": 0.41643914207816124,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9750000089406967,
      "step": 383
    },
    {
      "completion_length": 194.225004196167,
      "epoch": 0.39813374805598756,
      "grad_norm": 0.7693355083465576,
      "kl": 0.4150390625,
      "learning_rate": 2.99034980893874e-06,
      "loss": 0.0364,
      "reward": 1.7000000476837158,
      "reward_std": 0.3479738086462021,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 384
    },
    {
      "completion_length": 145.20000076293945,
      "epoch": 0.39917055469155005,
      "grad_norm": 0.5979889631271362,
      "kl": 0.472900390625,
      "learning_rate": 2.9901436199309044e-06,
      "loss": 0.0716,
      "reward": 1.7000000178813934,
      "reward_std": 0.30463574081659317,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 385
    },
    {
      "completion_length": 244.1250114440918,
      "epoch": 0.4002073613271125,
      "grad_norm": 0.4943735897541046,
      "kl": 0.5654296875,
      "learning_rate": 2.9899352586559133e-06,
      "loss": 0.1302,
      "reward": 1.2000000476837158,
      "reward_std": 0.3856646865606308,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 386
    },
    {
      "completion_length": 99.10000228881836,
      "epoch": 0.401244167962675,
      "grad_norm": 0.41970863938331604,
      "kl": 0.36328125,
      "learning_rate": 2.9897247254175052e-06,
      "loss": -0.0189,
      "reward": 1.5750000178813934,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.5749999973922968,
      "rewards/format_reward": 1.0,
      "step": 387
    },
    {
      "completion_length": 187.52499961853027,
      "epoch": 0.4022809745982374,
      "grad_norm": 0.3286798894405365,
      "kl": 0.505859375,
      "learning_rate": 2.9895120205225883e-06,
      "loss": 0.1248,
      "reward": 1.175000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.22500000894069672,
      "rewards/format_reward": 0.9500000178813934,
      "step": 388
    },
    {
      "completion_length": 209.52500534057617,
      "epoch": 0.4033177812337999,
      "grad_norm": 0.5088163018226624,
      "kl": 0.3583984375,
      "learning_rate": 2.9892971442812333e-06,
      "loss": 0.1461,
      "reward": 1.2500000298023224,
      "reward_std": 0.31719937175512314,
      "rewards/accuracy_reward": 0.29999999701976776,
      "rewards/format_reward": 0.9500000178813934,
      "step": 389
    },
    {
      "completion_length": 158.02499961853027,
      "epoch": 0.40435458786936235,
      "grad_norm": 0.3768276274204254,
      "kl": 0.33251953125,
      "learning_rate": 2.989080097006678e-06,
      "loss": 0.0693,
      "reward": 1.2999999970197678,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 390
    },
    {
      "completion_length": 263.57500076293945,
      "epoch": 0.40539139450492484,
      "grad_norm": 12.056117057800293,
      "kl": 1.38525390625,
      "learning_rate": 2.988860879015324e-06,
      "loss": 0.2372,
      "reward": 1.475000023841858,
      "reward_std": 0.4363621547818184,
      "rewards/accuracy_reward": 0.5500000156462193,
      "rewards/format_reward": 0.925000011920929,
      "step": 391
    },
    {
      "completion_length": 165.475004196167,
      "epoch": 0.4064282011404873,
      "grad_norm": 0.36441394686698914,
      "kl": 0.35595703125,
      "learning_rate": 2.9886394906267387e-06,
      "loss": 0.0168,
      "reward": 1.25,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.2750000022351742,
      "rewards/format_reward": 0.9750000089406967,
      "step": 392
    },
    {
      "completion_length": 254.17500495910645,
      "epoch": 0.40746500777604977,
      "grad_norm": 0.32492274045944214,
      "kl": 0.48095703125,
      "learning_rate": 2.988415932163652e-06,
      "loss": 0.128,
      "reward": 1.125,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.20000000298023224,
      "rewards/format_reward": 0.925000011920929,
      "step": 393
    },
    {
      "completion_length": 247.7250099182129,
      "epoch": 0.4085018144116122,
      "grad_norm": 425.0500183105469,
      "kl": 15.10888671875,
      "learning_rate": 2.988190203951958e-06,
      "loss": 1.26,
      "reward": 1.4750000536441803,
      "reward_std": 0.5084580183029175,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 394
    },
    {
      "completion_length": 204.92500114440918,
      "epoch": 0.4095386210471747,
      "grad_norm": 0.4983512759208679,
      "kl": 0.404052734375,
      "learning_rate": 2.9879623063207136e-06,
      "loss": 0.0576,
      "reward": 1.4250000417232513,
      "reward_std": 0.3354102037847042,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.9500000178813934,
      "step": 395
    },
    {
      "completion_length": 252.95000648498535,
      "epoch": 0.4105754276827372,
      "grad_norm": 0.6655735373497009,
      "kl": 0.510986328125,
      "learning_rate": 2.987732239602138e-06,
      "loss": 0.059,
      "reward": 1.2499999850988388,
      "reward_std": 0.26129766926169395,
      "rewards/accuracy_reward": 0.32500001043081284,
      "rewards/format_reward": 0.9249999970197678,
      "step": 396
    },
    {
      "completion_length": 258.30000495910645,
      "epoch": 0.41161223431829963,
      "grad_norm": 17.00373077392578,
      "kl": 0.52880859375,
      "learning_rate": 2.9875000041316126e-06,
      "loss": 0.2363,
      "reward": 1.5500000417232513,
      "reward_std": 0.46511994302272797,
      "rewards/accuracy_reward": 0.6250000037252903,
      "rewards/format_reward": 0.9250000268220901,
      "step": 397
    },
    {
      "completion_length": 305.8000087738037,
      "epoch": 0.4126490409538621,
      "grad_norm": 5.655200004577637,
      "kl": 1.62890625,
      "learning_rate": 2.98726560024768e-06,
      "loss": 0.3134,
      "reward": 1.475000023841858,
      "reward_std": 0.5571388304233551,
      "rewards/accuracy_reward": 0.600000012665987,
      "rewards/format_reward": 0.8750000149011612,
      "step": 398
    },
    {
      "completion_length": 186.90000343322754,
      "epoch": 0.41368584758942456,
      "grad_norm": 67.16045379638672,
      "kl": 10.04833984375,
      "learning_rate": 2.987029028292045e-06,
      "loss": 0.84,
      "reward": 1.3000000268220901,
      "reward_std": 0.3245587907731533,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.9500000029802322,
      "step": 399
    },
    {
      "completion_length": 168.5250015258789,
      "epoch": 0.41472265422498705,
      "grad_norm": 17.609403610229492,
      "kl": 3.233154296875,
      "learning_rate": 2.986790288609571e-06,
      "loss": 0.265,
      "reward": 1.6250000298023224,
      "reward_std": 0.3533165529370308,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 400
    },
    {
      "completion_length": 184.3000030517578,
      "epoch": 0.4157594608605495,
      "grad_norm": 1.3612682819366455,
      "kl": 0.52490234375,
      "learning_rate": 2.9865493815482827e-06,
      "loss": 0.122,
      "reward": 1.6250000298023224,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9250000268220901,
      "step": 401
    },
    {
      "completion_length": 267.40000343322754,
      "epoch": 0.416796267496112,
      "grad_norm": 31.001413345336914,
      "kl": 4.72412109375,
      "learning_rate": 2.986306307459364e-06,
      "loss": 0.5375,
      "reward": 1.4500000029802322,
      "reward_std": 0.43636219948530197,
      "rewards/accuracy_reward": 0.6000000182539225,
      "rewards/format_reward": 0.8500000089406967,
      "step": 402
    },
    {
      "completion_length": 211.35000038146973,
      "epoch": 0.4178330741316744,
      "grad_norm": 1.4517061710357666,
      "kl": 1.05419921875,
      "learning_rate": 2.986061066697158e-06,
      "loss": 0.0987,
      "reward": 1.2500000149011612,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.32500000670552254,
      "rewards/format_reward": 0.925000011920929,
      "step": 403
    },
    {
      "completion_length": 142.9250030517578,
      "epoch": 0.4188698807672369,
      "grad_norm": 1.8906280994415283,
      "kl": 1.05908203125,
      "learning_rate": 2.9858136596191662e-06,
      "loss": 0.1521,
      "reward": 1.3250000327825546,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.3999999985098839,
      "rewards/format_reward": 0.9250000268220901,
      "step": 404
    },
    {
      "completion_length": 141.87500190734863,
      "epoch": 0.4199066874027994,
      "grad_norm": 23.721681594848633,
      "kl": 3.887939453125,
      "learning_rate": 2.985564086586048e-06,
      "loss": 0.3364,
      "reward": 1.475000023841858,
      "reward_std": 0.4740530550479889,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9000000059604645,
      "step": 405
    },
    {
      "completion_length": 162.39999961853027,
      "epoch": 0.42094349403836184,
      "grad_norm": 65.63410949707031,
      "kl": 7.84228515625,
      "learning_rate": 2.9853123479616196e-06,
      "loss": 0.5849,
      "reward": 1.5500000268220901,
      "reward_std": 0.3479738086462021,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 406
    },
    {
      "completion_length": 247.4749984741211,
      "epoch": 0.42198030067392434,
      "grad_norm": 10.55413818359375,
      "kl": 4.9853515625,
      "learning_rate": 2.9850584441128557e-06,
      "loss": 0.4714,
      "reward": 1.475000023841858,
      "reward_std": 0.4472135789692402,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.8500000238418579,
      "step": 407
    },
    {
      "completion_length": 220.52500534057617,
      "epoch": 0.4230171073094868,
      "grad_norm": 0.9092906713485718,
      "kl": 0.403076171875,
      "learning_rate": 2.984802375409885e-06,
      "loss": 0.0505,
      "reward": 1.2750000059604645,
      "reward_std": 0.4495588019490242,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9000000208616257,
      "step": 408
    },
    {
      "completion_length": 199.35000038146973,
      "epoch": 0.42405391394504927,
      "grad_norm": 1.2738308906555176,
      "kl": 0.4384765625,
      "learning_rate": 2.984544142225995e-06,
      "loss": 0.1137,
      "reward": 1.4000000059604645,
      "reward_std": 0.3046357296407223,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.9500000178813934,
      "step": 409
    },
    {
      "completion_length": 305.950008392334,
      "epoch": 0.4250907205806117,
      "grad_norm": 2.3068294525146484,
      "kl": 0.871826171875,
      "learning_rate": 2.9842837449376264e-06,
      "loss": 0.2054,
      "reward": 1.550000011920929,
      "reward_std": 0.4290027618408203,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.8750000298023224,
      "step": 410
    },
    {
      "completion_length": 237.6500015258789,
      "epoch": 0.4261275272161742,
      "grad_norm": 2.3890562057495117,
      "kl": 1.326171875,
      "learning_rate": 2.9840211839243747e-06,
      "loss": 0.2206,
      "reward": 1.5,
      "reward_std": 0.47247934713959694,
      "rewards/accuracy_reward": 0.5999999940395355,
      "rewards/format_reward": 0.9000000357627869,
      "step": 411
    },
    {
      "completion_length": 491.37500953674316,
      "epoch": 0.42716433385173663,
      "grad_norm": 0.7049680948257446,
      "kl": 0.55810546875,
      "learning_rate": 2.9837564595689896e-06,
      "loss": 0.0784,
      "reward": 1.3500000536441803,
      "reward_std": 0.3913119025528431,
      "rewards/accuracy_reward": 0.6000000182539225,
      "rewards/format_reward": 0.7500000149011612,
      "step": 412
    },
    {
      "completion_length": 219.30000114440918,
      "epoch": 0.4282011404872991,
      "grad_norm": 5.027242183685303,
      "kl": 1.90966796875,
      "learning_rate": 2.983489572257376e-06,
      "loss": 0.1836,
      "reward": 1.2000000178813934,
      "reward_std": 0.27228760719299316,
      "rewards/accuracy_reward": 0.4000000096857548,
      "rewards/format_reward": 0.800000011920929,
      "step": 413
    },
    {
      "completion_length": 601.2500057220459,
      "epoch": 0.42923794712286156,
      "grad_norm": 33.8704948425293,
      "kl": 11.8896484375,
      "learning_rate": 2.9832205223785905e-06,
      "loss": 1.1407,
      "reward": 1.1500000059604645,
      "reward_std": 0.5894869565963745,
      "rewards/accuracy_reward": 0.4999999962747097,
      "rewards/format_reward": 0.6499999985098839,
      "step": 414
    },
    {
      "completion_length": 697.1750144958496,
      "epoch": 0.43027475375842406,
      "grad_norm": 0.6441704034805298,
      "kl": 0.45263671875,
      "learning_rate": 2.9829493103248416e-06,
      "loss": 0.169,
      "reward": 1.0000000149011612,
      "reward_std": 0.4237985797226429,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.6500000208616257,
      "step": 415
    },
    {
      "completion_length": 626.6500091552734,
      "epoch": 0.43131156039398655,
      "grad_norm": 1.1363885402679443,
      "kl": 1.099609375,
      "learning_rate": 2.9826759364914915e-06,
      "loss": 0.283,
      "reward": 1.0250000059604645,
      "reward_std": 0.504827506840229,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.6500000134110451,
      "step": 416
    },
    {
      "completion_length": 217.35000610351562,
      "epoch": 0.432348367029549,
      "grad_norm": 0.9580169916152954,
      "kl": 0.4140625,
      "learning_rate": 2.9824004012770523e-06,
      "loss": 0.1531,
      "reward": 1.2750000059604645,
      "reward_std": 0.3533165343105793,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 417
    },
    {
      "completion_length": 687.4500045776367,
      "epoch": 0.4333851736651115,
      "grad_norm": 0.6423575282096863,
      "kl": 0.49609375,
      "learning_rate": 2.9821227050831874e-06,
      "loss": 0.1532,
      "reward": 0.9749999940395355,
      "reward_std": 0.41643914207816124,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.6250000074505806,
      "step": 418
    },
    {
      "completion_length": 339.0999984741211,
      "epoch": 0.4344219803006739,
      "grad_norm": 1.1070829629898071,
      "kl": 1.319580078125,
      "learning_rate": 2.9818428483147104e-06,
      "loss": 0.259,
      "reward": 1.1750000417232513,
      "reward_std": 0.5048274733126163,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.8250000029802322,
      "step": 419
    },
    {
      "completion_length": 391.54999923706055,
      "epoch": 0.4354587869362364,
      "grad_norm": 0.9474948048591614,
      "kl": 0.658203125,
      "learning_rate": 2.9815608313795845e-06,
      "loss": 0.1309,
      "reward": 1.2500000447034836,
      "reward_std": 0.44219936802983284,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.7750000059604645,
      "step": 420
    },
    {
      "completion_length": 418.125,
      "epoch": 0.43649559357179885,
      "grad_norm": 35.806827545166016,
      "kl": 12.6572265625,
      "learning_rate": 2.981276654688922e-06,
      "loss": 0.7257,
      "reward": 1.0750000029802322,
      "reward_std": 0.4723408445715904,
      "rewards/accuracy_reward": 0.30000000819563866,
      "rewards/format_reward": 0.7750000208616257,
      "step": 421
    },
    {
      "completion_length": 406.25000762939453,
      "epoch": 0.43753240020736134,
      "grad_norm": 4.112382411956787,
      "kl": 2.720703125,
      "learning_rate": 2.9809903186569826e-06,
      "loss": 0.4188,
      "reward": 1.1750000268220901,
      "reward_std": 0.5968463569879532,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.8250000178813934,
      "step": 422
    },
    {
      "completion_length": 389.7249984741211,
      "epoch": 0.4385692068429238,
      "grad_norm": 3.500101327896118,
      "kl": 4.51220703125,
      "learning_rate": 2.9807018237011756e-06,
      "loss": 0.4401,
      "reward": 1.2500000149011612,
      "reward_std": 0.3858031928539276,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.8500000238418579,
      "step": 423
    },
    {
      "completion_length": 309.550012588501,
      "epoch": 0.43960601347848627,
      "grad_norm": 3.486119031906128,
      "kl": 4.0400390625,
      "learning_rate": 2.980411170242056e-06,
      "loss": 0.4344,
      "reward": 1.3250000178813934,
      "reward_std": 0.5156789012253284,
      "rewards/accuracy_reward": 0.4500000160187483,
      "rewards/format_reward": 0.8750000149011612,
      "step": 424
    },
    {
      "completion_length": 142.0750026702881,
      "epoch": 0.4406428201140487,
      "grad_norm": 0.783508837223053,
      "kl": 1.0380859375,
      "learning_rate": 2.9801183587033266e-06,
      "loss": 0.1069,
      "reward": 1.5250000357627869,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.9750000089406967,
      "step": 425
    },
    {
      "completion_length": 118.95000076293945,
      "epoch": 0.4416796267496112,
      "grad_norm": 0.49163949489593506,
      "kl": 0.404296875,
      "learning_rate": 2.9798233895118343e-06,
      "loss": 0.0385,
      "reward": 1.350000023841858,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 426
    },
    {
      "completion_length": 482.90000915527344,
      "epoch": 0.4427164333851737,
      "grad_norm": 1.6972235441207886,
      "kl": 1.5712890625,
      "learning_rate": 2.9795262630975732e-06,
      "loss": 0.3296,
      "reward": 0.9750000387430191,
      "reward_std": 0.5535082966089249,
      "rewards/accuracy_reward": 0.20000000298023224,
      "rewards/format_reward": 0.7750000059604645,
      "step": 427
    },
    {
      "completion_length": 404.5250015258789,
      "epoch": 0.44375324002073613,
      "grad_norm": 1.4435909986495972,
      "kl": 2.01611328125,
      "learning_rate": 2.979226979893681e-06,
      "loss": 0.3348,
      "reward": 1.2500000298023224,
      "reward_std": 0.4349784702062607,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.8500000238418579,
      "step": 428
    },
    {
      "completion_length": 241.77500343322754,
      "epoch": 0.4447900466562986,
      "grad_norm": 1.073678970336914,
      "kl": 2.5791015625,
      "learning_rate": 2.978925540336441e-06,
      "loss": 0.2655,
      "reward": 1.3250000327825546,
      "reward_std": 0.40387552604079247,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 429
    },
    {
      "completion_length": 259.87500762939453,
      "epoch": 0.44582685329186106,
      "grad_norm": 3.7204360961914062,
      "kl": 5.47900390625,
      "learning_rate": 2.978621944865277e-06,
      "loss": 0.3859,
      "reward": 1.3250000327825546,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9000000208616257,
      "step": 430
    },
    {
      "completion_length": 210.2000026702881,
      "epoch": 0.44686365992742355,
      "grad_norm": 0.5499123930931091,
      "kl": 0.99951171875,
      "learning_rate": 2.9783161939227587e-06,
      "loss": 0.1771,
      "reward": 1.0999999940395355,
      "reward_std": 0.22360678017139435,
      "rewards/accuracy_reward": 0.15000000596046448,
      "rewards/format_reward": 0.9500000029802322,
      "step": 431
    },
    {
      "completion_length": 254.32501220703125,
      "epoch": 0.447900466562986,
      "grad_norm": 1.4711023569107056,
      "kl": 1.126220703125,
      "learning_rate": 2.9780082879545958e-06,
      "loss": 0.2227,
      "reward": 1.4500000029802322,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 432
    },
    {
      "completion_length": 149.57500076293945,
      "epoch": 0.4489372731985485,
      "grad_norm": 0.3505478799343109,
      "kl": 0.44189453125,
      "learning_rate": 2.9776982274096416e-06,
      "loss": 0.0277,
      "reward": 1.600000023841858,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 433
    },
    {
      "completion_length": 118.62500381469727,
      "epoch": 0.4499740798341109,
      "grad_norm": 1.0018848180770874,
      "kl": 0.70654296875,
      "learning_rate": 2.9773860127398872e-06,
      "loss": 0.0473,
      "reward": 1.550000011920929,
      "reward_std": 0.25609345361590385,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9500000178813934,
      "step": 434
    },
    {
      "completion_length": 149.22500610351562,
      "epoch": 0.4510108864696734,
      "grad_norm": 0.4854688048362732,
      "kl": 1.08154296875,
      "learning_rate": 2.9770716444004673e-06,
      "loss": 0.0993,
      "reward": 1.300000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 435
    },
    {
      "completion_length": 225.75,
      "epoch": 0.45204769310523585,
      "grad_norm": 1.2092115879058838,
      "kl": 1.08251953125,
      "learning_rate": 2.9767551228496537e-06,
      "loss": 0.0811,
      "reward": 1.300000011920929,
      "reward_std": 0.22894952446222305,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 436
    },
    {
      "completion_length": 191.77499961853027,
      "epoch": 0.45308449974079834,
      "grad_norm": 2.3160505294799805,
      "kl": 1.91552734375,
      "learning_rate": 2.9764364485488584e-06,
      "loss": 0.1791,
      "reward": 1.2000000029802322,
      "reward_std": 0.1604842059314251,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 0.925000011920929,
      "step": 437
    },
    {
      "completion_length": 445.32500648498535,
      "epoch": 0.45412130637636083,
      "grad_norm": 1.1174695491790771,
      "kl": 2.077392578125,
      "learning_rate": 2.9761156219626305e-06,
      "loss": 0.3718,
      "reward": 1.4749999940395355,
      "reward_std": 0.521021630614996,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.8250000029802322,
      "step": 438
    },
    {
      "completion_length": 422.75,
      "epoch": 0.45515811301192327,
      "grad_norm": 1.358939528465271,
      "kl": 2.649658203125,
      "learning_rate": 2.9757926435586572e-06,
      "loss": 0.1542,
      "reward": 1.1000000089406967,
      "reward_std": 0.37310105934739113,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.7750000059604645,
      "step": 439
    },
    {
      "completion_length": 638.4000091552734,
      "epoch": 0.45619491964748576,
      "grad_norm": 1.3631545305252075,
      "kl": 2.658203125,
      "learning_rate": 2.975467513807763e-06,
      "loss": 0.4623,
      "reward": 1.0000000447034836,
      "reward_std": 0.7521778345108032,
      "rewards/accuracy_reward": 0.37500000186264515,
      "rewards/format_reward": 0.6250000149011612,
      "step": 440
    },
    {
      "completion_length": 549.6250038146973,
      "epoch": 0.4572317262830482,
      "grad_norm": 1.7535611391067505,
      "kl": 1.56591796875,
      "learning_rate": 2.975140233183908e-06,
      "loss": 0.3238,
      "reward": 1.1500000059604645,
      "reward_std": 0.5408061631023884,
      "rewards/accuracy_reward": 0.44999999552965164,
      "rewards/format_reward": 0.7000000029802322,
      "step": 441
    },
    {
      "completion_length": 441.25000762939453,
      "epoch": 0.4582685329186107,
      "grad_norm": 2.6060733795166016,
      "kl": 1.5341796875,
      "learning_rate": 2.9748108021641875e-06,
      "loss": 0.3068,
      "reward": 1.3250000178813934,
      "reward_std": 0.6076978109776974,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.8250000029802322,
      "step": 442
    },
    {
      "completion_length": 449.1750068664551,
      "epoch": 0.45930533955417313,
      "grad_norm": 2.323462963104248,
      "kl": 3.5185546875,
      "learning_rate": 2.974479221228831e-06,
      "loss": 0.5005,
      "reward": 1.1499999910593033,
      "reward_std": 0.4981010742485523,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.7749999910593033,
      "step": 443
    },
    {
      "completion_length": 369.0999984741211,
      "epoch": 0.4603421461897356,
      "grad_norm": 9.374835968017578,
      "kl": 4.626953125,
      "learning_rate": 2.9741454908612047e-06,
      "loss": 0.4378,
      "reward": 1.1250000149011612,
      "reward_std": 0.4525563307106495,
      "rewards/accuracy_reward": 0.3000000026077032,
      "rewards/format_reward": 0.8250000178813934,
      "step": 444
    },
    {
      "completion_length": 254.1500072479248,
      "epoch": 0.46137895282529806,
      "grad_norm": 1.3832933902740479,
      "kl": 1.93603515625,
      "learning_rate": 2.9738096115478055e-06,
      "loss": 0.2301,
      "reward": 1.3000000268220901,
      "reward_std": 0.3479738235473633,
      "rewards/accuracy_reward": 0.40000000409781933,
      "rewards/format_reward": 0.9000000208616257,
      "step": 445
    },
    {
      "completion_length": 344.5,
      "epoch": 0.46241575946086055,
      "grad_norm": 1.9132006168365479,
      "kl": 2.30419921875,
      "learning_rate": 2.9734715837782635e-06,
      "loss": 0.2175,
      "reward": 1.225000038743019,
      "reward_std": 0.40387552976608276,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 446
    },
    {
      "completion_length": 238.82500648498535,
      "epoch": 0.463452566096423,
      "grad_norm": 1.1751809120178223,
      "kl": 1.783203125,
      "learning_rate": 2.973131408045341e-06,
      "loss": 0.3002,
      "reward": 1.5750000476837158,
      "reward_std": 0.3913118988275528,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 447
    },
    {
      "completion_length": 186.3499984741211,
      "epoch": 0.4644893727319855,
      "grad_norm": 0.9299616813659668,
      "kl": 0.9111328125,
      "learning_rate": 2.972789084844932e-06,
      "loss": 0.1332,
      "reward": 1.5750000178813934,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000178813934,
      "step": 448
    },
    {
      "completion_length": 307.1250171661377,
      "epoch": 0.465526179367548,
      "grad_norm": 1.2203192710876465,
      "kl": 2.68505859375,
      "learning_rate": 2.97244461467606e-06,
      "loss": 0.3509,
      "reward": 1.550000011920929,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 449
    },
    {
      "completion_length": 197.77500534057617,
      "epoch": 0.4665629860031104,
      "grad_norm": 0.5328684449195862,
      "kl": 1.17578125,
      "learning_rate": 2.9720979980408795e-06,
      "loss": 0.173,
      "reward": 1.4749999940395355,
      "reward_std": 0.14429005607962608,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9500000178813934,
      "step": 450
    },
    {
      "completion_length": 183.67500495910645,
      "epoch": 0.4675997926386729,
      "grad_norm": 2.1329684257507324,
      "kl": 1.76220703125,
      "learning_rate": 2.9717492354446727e-06,
      "loss": 0.1922,
      "reward": 1.2750000059604645,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 451
    },
    {
      "completion_length": 97.95000076293945,
      "epoch": 0.46863659927423534,
      "grad_norm": 0.40684643387794495,
      "kl": 0.35595703125,
      "learning_rate": 2.9713983273958507e-06,
      "loss": 0.027,
      "reward": 1.574999988079071,
      "reward_std": 0.18026871606707573,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 452
    },
    {
      "completion_length": 141.92500019073486,
      "epoch": 0.46967340590979784,
      "grad_norm": 3.87530517578125,
      "kl": 2.384521484375,
      "learning_rate": 2.971045274405952e-06,
      "loss": 0.2588,
      "reward": 1.600000023841858,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.6250000186264515,
      "rewards/format_reward": 0.9750000089406967,
      "step": 453
    },
    {
      "completion_length": 138.90000343322754,
      "epoch": 0.4707102125453603,
      "grad_norm": 4.4843430519104,
      "kl": 2.71875,
      "learning_rate": 2.9706900769896434e-06,
      "loss": 0.1234,
      "reward": 1.5500000417232513,
      "reward_std": 0.36789682134985924,
      "rewards/accuracy_reward": 0.650000000372529,
      "rewards/format_reward": 0.9000000059604645,
      "step": 454
    },
    {
      "completion_length": 337.850004196167,
      "epoch": 0.47174701918092277,
      "grad_norm": 25.936668395996094,
      "kl": 4.15185546875,
      "learning_rate": 2.970332735664714e-06,
      "loss": 0.3841,
      "reward": 1.350000023841858,
      "reward_std": 0.3856646902859211,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.8750000149011612,
      "step": 455
    },
    {
      "completion_length": 245.5,
      "epoch": 0.4727838258164852,
      "grad_norm": 1.4629117250442505,
      "kl": 0.405517578125,
      "learning_rate": 2.969973250952083e-06,
      "loss": 0.1739,
      "reward": 1.100000023841858,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.8500000089406967,
      "step": 456
    },
    {
      "completion_length": 551.0750122070312,
      "epoch": 0.4738206324520477,
      "grad_norm": 2.0260560512542725,
      "kl": 0.8857421875,
      "learning_rate": 2.96961162337579e-06,
      "loss": 0.2795,
      "reward": 1.050000011920929,
      "reward_std": 0.4651199281215668,
      "rewards/accuracy_reward": 0.3000000063329935,
      "rewards/format_reward": 0.7500000149011612,
      "step": 457
    },
    {
      "completion_length": 735.4250030517578,
      "epoch": 0.47485743908761013,
      "grad_norm": 5.954936981201172,
      "kl": 0.9501953125,
      "learning_rate": 2.969247853463e-06,
      "loss": 0.424,
      "reward": 0.8750000149011612,
      "reward_std": 0.582404587417841,
      "rewards/accuracy_reward": 0.24999999813735485,
      "rewards/format_reward": 0.6250000186264515,
      "step": 458
    },
    {
      "completion_length": 592.0750122070312,
      "epoch": 0.4758942457231726,
      "grad_norm": 4.934417247772217,
      "kl": 0.63671875,
      "learning_rate": 2.968881941744003e-06,
      "loss": 0.2356,
      "reward": 1.1500000208616257,
      "reward_std": 0.5342182517051697,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.7000000029802322,
      "step": 459
    },
    {
      "completion_length": 586.4000091552734,
      "epoch": 0.4769310523587351,
      "grad_norm": 2.2492358684539795,
      "kl": 0.421875,
      "learning_rate": 2.968513888752207e-06,
      "loss": 0.3741,
      "reward": 1.2500000298023224,
      "reward_std": 0.8159334361553192,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.6749999970197678,
      "step": 460
    },
    {
      "completion_length": 455.5500030517578,
      "epoch": 0.47796785899429756,
      "grad_norm": 2.0860514640808105,
      "kl": 0.35400390625,
      "learning_rate": 2.9681436950241457e-06,
      "loss": 0.2959,
      "reward": 1.2500000149011612,
      "reward_std": 0.6224680915474892,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.7750000208616257,
      "step": 461
    },
    {
      "completion_length": 485.87500762939453,
      "epoch": 0.47900466562986005,
      "grad_norm": 68.13333892822266,
      "kl": 13.9140625,
      "learning_rate": 2.9677713610994704e-06,
      "loss": 1.3759,
      "reward": 1.1500000208616257,
      "reward_std": 0.5533697977662086,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.7500000149011612,
      "step": 462
    },
    {
      "completion_length": 181.725004196167,
      "epoch": 0.4800414722654225,
      "grad_norm": 9.417370796203613,
      "kl": 1.84912109375,
      "learning_rate": 2.9673968875209545e-06,
      "loss": 0.1592,
      "reward": 1.4500000178813934,
      "reward_std": 0.5253834910690784,
      "rewards/accuracy_reward": 0.600000012665987,
      "rewards/format_reward": 0.8500000238418579,
      "step": 463
    },
    {
      "completion_length": 575.8250045776367,
      "epoch": 0.481078278900985,
      "grad_norm": 2.923924446105957,
      "kl": 0.75537109375,
      "learning_rate": 2.967020274834488e-06,
      "loss": 0.2154,
      "reward": 1.2000000178813934,
      "reward_std": 0.658585287630558,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.7250000238418579,
      "step": 464
    },
    {
      "completion_length": 235.97500228881836,
      "epoch": 0.4821150855365474,
      "grad_norm": 27.21836280822754,
      "kl": 8.6328125,
      "learning_rate": 2.9666415235890823e-06,
      "loss": 0.814,
      "reward": 1.3500000089406967,
      "reward_std": 0.3913119100034237,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9000000208616257,
      "step": 465
    },
    {
      "completion_length": 180.10000038146973,
      "epoch": 0.4831518921721099,
      "grad_norm": 1.1835033893585205,
      "kl": 0.5048828125,
      "learning_rate": 2.966260634336863e-06,
      "loss": 0.0695,
      "reward": 1.350000023841858,
      "reward_std": 0.26129766553640366,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 466
    },
    {
      "completion_length": 141.75,
      "epoch": 0.48418869880767235,
      "grad_norm": 0.5516549348831177,
      "kl": 0.375244140625,
      "learning_rate": 2.965877607633075e-06,
      "loss": 0.0181,
      "reward": 1.5999999940395355,
      "reward_std": 0.33353202417492867,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.925000011920929,
      "step": 467
    },
    {
      "completion_length": 151.57499885559082,
      "epoch": 0.48522550544323484,
      "grad_norm": 2.3915843963623047,
      "kl": 0.5625,
      "learning_rate": 2.965492444036078e-06,
      "loss": 0.0569,
      "reward": 1.3499999940395355,
      "reward_std": 0.281220693141222,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 468
    },
    {
      "completion_length": 89.47500228881836,
      "epoch": 0.48626231207879733,
      "grad_norm": 0.3632679879665375,
      "kl": 0.4560546875,
      "learning_rate": 2.965105144107347e-06,
      "loss": 0.0197,
      "reward": 1.2750000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 1.0,
      "step": 469
    },
    {
      "completion_length": 104.25000381469727,
      "epoch": 0.48729911871435977,
      "grad_norm": 2.3217339515686035,
      "kl": 0.4345703125,
      "learning_rate": 2.9647157084114704e-06,
      "loss": -0.0065,
      "reward": 1.5,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.9750000089406967,
      "step": 470
    },
    {
      "completion_length": 157.50000381469727,
      "epoch": 0.48833592534992226,
      "grad_norm": 1.2729434967041016,
      "kl": 0.7998046875,
      "learning_rate": 2.9643241375161517e-06,
      "loss": 0.1192,
      "reward": 1.375,
      "reward_std": 0.5100317001342773,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.875,
      "step": 471
    },
    {
      "completion_length": 656.1500015258789,
      "epoch": 0.4893727319854847,
      "grad_norm": 1.7226147651672363,
      "kl": 0.68115234375,
      "learning_rate": 2.963930431992206e-06,
      "loss": 0.4477,
      "reward": 0.8500000238418579,
      "reward_std": 0.5660719275474548,
      "rewards/accuracy_reward": 0.1750000026077032,
      "rewards/format_reward": 0.6749999970197678,
      "step": 472
    },
    {
      "completion_length": 406.3750057220459,
      "epoch": 0.4904095386210472,
      "grad_norm": 0.951695442199707,
      "kl": 0.376953125,
      "learning_rate": 2.96353459241356e-06,
      "loss": 0.2062,
      "reward": 1.225000023841858,
      "reward_std": 0.49088017642498016,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.7750000208616257,
      "step": 473
    },
    {
      "completion_length": 170.22500228881836,
      "epoch": 0.49144634525660963,
      "grad_norm": 0.9786409735679626,
      "kl": 0.43505859375,
      "learning_rate": 2.963136619357251e-06,
      "loss": 0.0848,
      "reward": 1.425000011920929,
      "reward_std": 0.4797002226114273,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.8750000149011612,
      "step": 474
    },
    {
      "completion_length": 180.82500076293945,
      "epoch": 0.4924831518921721,
      "grad_norm": 0.526511013507843,
      "kl": 0.50146484375,
      "learning_rate": 2.9627365134034276e-06,
      "loss": 0.0805,
      "reward": 1.2999999970197678,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9249999970197678,
      "step": 475
    },
    {
      "completion_length": 179.5000057220459,
      "epoch": 0.49351995852773456,
      "grad_norm": 1.3897082805633545,
      "kl": 0.3935546875,
      "learning_rate": 2.9623342751353473e-06,
      "loss": 0.1006,
      "reward": 1.2500000298023224,
      "reward_std": 0.23617039993405342,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9000000208616257,
      "step": 476
    },
    {
      "completion_length": 104.20000267028809,
      "epoch": 0.49455676516329705,
      "grad_norm": 1.2519752979278564,
      "kl": 0.62939453125,
      "learning_rate": 2.961929905139375e-06,
      "loss": 0.0275,
      "reward": 1.5750000178813934,
      "reward_std": 0.3856646902859211,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 1.0,
      "step": 477
    },
    {
      "completion_length": 136.77500534057617,
      "epoch": 0.4955935717988595,
      "grad_norm": 32.357139587402344,
      "kl": 8.801513671875,
      "learning_rate": 2.9615234040049845e-06,
      "loss": 0.5752,
      "reward": 1.1250000149011612,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.17500000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 478
    },
    {
      "completion_length": 151.99999809265137,
      "epoch": 0.496630378434422,
      "grad_norm": 4.595073699951172,
      "kl": 1.439453125,
      "learning_rate": 2.9611147723247555e-06,
      "loss": 0.154,
      "reward": 1.5250000059604645,
      "reward_std": 0.28485121950507164,
      "rewards/accuracy_reward": 0.5749999992549419,
      "rewards/format_reward": 0.9500000029802322,
      "step": 479
    },
    {
      "completion_length": 102.72500038146973,
      "epoch": 0.4976671850699845,
      "grad_norm": 0.591180682182312,
      "kl": 0.371337890625,
      "learning_rate": 2.9607040106943733e-06,
      "loss": 0.0109,
      "reward": 1.4250000268220901,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 480
    },
    {
      "completion_length": 152.975004196167,
      "epoch": 0.4987039917055469,
      "grad_norm": 1.7282302379608154,
      "kl": 1.240234375,
      "learning_rate": 2.9602911197126286e-06,
      "loss": 0.1355,
      "reward": 1.2750000208616257,
      "reward_std": 0.26129766181111336,
      "rewards/accuracy_reward": 0.30000000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 481
    },
    {
      "completion_length": 119.20000076293945,
      "epoch": 0.4997407983411094,
      "grad_norm": 1.2713792324066162,
      "kl": 1.33203125,
      "learning_rate": 2.9598760999814165e-06,
      "loss": 0.0979,
      "reward": 1.5000000298023224,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 482
    },
    {
      "completion_length": 133.9250030517578,
      "epoch": 0.5007776049766719,
      "grad_norm": 0.8651389479637146,
      "kl": 0.416259765625,
      "learning_rate": 2.9594589521057342e-06,
      "loss": 0.0718,
      "reward": 1.4000000208616257,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.9500000178813934,
      "step": 483
    },
    {
      "completion_length": 134.850004196167,
      "epoch": 0.5018144116122343,
      "grad_norm": 1.2900584936141968,
      "kl": 1.51318359375,
      "learning_rate": 2.9590396766936826e-06,
      "loss": 0.1183,
      "reward": 1.350000038743019,
      "reward_std": 0.3479738235473633,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 484
    },
    {
      "completion_length": 86.97500133514404,
      "epoch": 0.5028512182477968,
      "grad_norm": 0.4113655388355255,
      "kl": 0.3232421875,
      "learning_rate": 2.9586182743564627e-06,
      "loss": 0.0003,
      "reward": 1.675000011920929,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 1.0,
      "step": 485
    },
    {
      "completion_length": 138.35000038146973,
      "epoch": 0.5038880248833593,
      "grad_norm": 3.6755096912384033,
      "kl": 2.095703125,
      "learning_rate": 2.9581947457083773e-06,
      "loss": 0.1498,
      "reward": 1.550000011920929,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 486
    },
    {
      "completion_length": 87.00000381469727,
      "epoch": 0.5049248315189218,
      "grad_norm": 0.35109928250312805,
      "kl": 0.46728515625,
      "learning_rate": 2.957769091366828e-06,
      "loss": 0.0177,
      "reward": 1.550000011920929,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 1.0,
      "step": 487
    },
    {
      "completion_length": 77.22500228881836,
      "epoch": 0.5059616381544841,
      "grad_norm": 1.4628422260284424,
      "kl": 0.9501953125,
      "learning_rate": 2.9573413119523146e-06,
      "loss": 0.0246,
      "reward": 1.5750000178813934,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 1.0,
      "step": 488
    },
    {
      "completion_length": 168.70000457763672,
      "epoch": 0.5069984447900466,
      "grad_norm": 6.089077472686768,
      "kl": 2.49169921875,
      "learning_rate": 2.9569114080884365e-06,
      "loss": 0.2892,
      "reward": 1.7000000476837158,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9250000268220901,
      "step": 489
    },
    {
      "completion_length": 93.82500457763672,
      "epoch": 0.5080352514256091,
      "grad_norm": 1.7852115631103516,
      "kl": 0.6474609375,
      "learning_rate": 2.9564793804018885e-06,
      "loss": 0.0774,
      "reward": 1.5250000059604645,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.9750000089406967,
      "step": 490
    },
    {
      "completion_length": 70.42500114440918,
      "epoch": 0.5090720580611716,
      "grad_norm": 0.8128455281257629,
      "kl": 0.35009765625,
      "learning_rate": 2.9560452295224624e-06,
      "loss": 0.0278,
      "reward": 1.5000000298023224,
      "reward_std": 0.31719935312867165,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 1.0,
      "step": 491
    },
    {
      "completion_length": 125.45000076293945,
      "epoch": 0.5101088646967341,
      "grad_norm": 0.7844669222831726,
      "kl": 0.4482421875,
      "learning_rate": 2.955608956083044e-06,
      "loss": 0.05,
      "reward": 1.550000011920929,
      "reward_std": 0.15685366839170456,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 492
    },
    {
      "completion_length": 127.45000457763672,
      "epoch": 0.5111456713322965,
      "grad_norm": 1.1734215021133423,
      "kl": 1.258544921875,
      "learning_rate": 2.9551705607196147e-06,
      "loss": 0.0648,
      "reward": 1.3750000298023224,
      "reward_std": 0.3407529294490814,
      "rewards/accuracy_reward": 0.4500000048428774,
      "rewards/format_reward": 0.925000011920929,
      "step": 493
    },
    {
      "completion_length": 180.07500457763672,
      "epoch": 0.512182477967859,
      "grad_norm": 0.663034975528717,
      "kl": 0.40966796875,
      "learning_rate": 2.9547300440712477e-06,
      "loss": 0.0733,
      "reward": 1.5250000059604645,
      "reward_std": 0.17304783314466476,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 494
    },
    {
      "completion_length": 179.37500667572021,
      "epoch": 0.5132192846034215,
      "grad_norm": 0.9060681462287903,
      "kl": 0.76513671875,
      "learning_rate": 2.9542874067801094e-06,
      "loss": 0.1252,
      "reward": 1.5249999910593033,
      "reward_std": 0.2686570920050144,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9000000059604645,
      "step": 495
    },
    {
      "completion_length": 197.75000667572021,
      "epoch": 0.514256091238984,
      "grad_norm": 1.4791067838668823,
      "kl": 1.224609375,
      "learning_rate": 2.9538426494914574e-06,
      "loss": 0.1744,
      "reward": 1.475000023841858,
      "reward_std": 0.3913119062781334,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.925000011920929,
      "step": 496
    },
    {
      "completion_length": 82.92500305175781,
      "epoch": 0.5152928978745464,
      "grad_norm": 22.87507438659668,
      "kl": 4.554931640625,
      "learning_rate": 2.9533957728536397e-06,
      "loss": 0.3108,
      "reward": 1.6250000298023224,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 497
    },
    {
      "completion_length": 97.22500038146973,
      "epoch": 0.5163297045101088,
      "grad_norm": 2.5906031131744385,
      "kl": 0.286376953125,
      "learning_rate": 2.9529467775180935e-06,
      "loss": 0.0659,
      "reward": 1.5000000298023224,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.9750000089406967,
      "step": 498
    },
    {
      "completion_length": 160.65000343322754,
      "epoch": 0.5173665111456713,
      "grad_norm": 1.3695307970046997,
      "kl": 1.130859375,
      "learning_rate": 2.952495664139346e-06,
      "loss": 0.1592,
      "reward": 1.475000023841858,
      "reward_std": 0.4723408408463001,
      "rewards/accuracy_reward": 0.5750000141561031,
      "rewards/format_reward": 0.9000000208616257,
      "step": 499
    },
    {
      "completion_length": 145.62500667572021,
      "epoch": 0.5184033177812338,
      "grad_norm": 3.733125925064087,
      "kl": 4.13330078125,
      "learning_rate": 2.952042433375009e-06,
      "loss": 0.1987,
      "reward": 1.425000011920929,
      "reward_std": 0.28485122323036194,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 500
    },
    {
      "completion_length": 125.32500457763672,
      "epoch": 0.5194401244167963,
      "grad_norm": 0.9280956983566284,
      "kl": 0.7236328125,
      "learning_rate": 2.9515870858857846e-06,
      "loss": 0.0983,
      "reward": 1.3250000178813934,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9500000178813934,
      "step": 501
    },
    {
      "completion_length": 75.125,
      "epoch": 0.5204769310523587,
      "grad_norm": 0.5787338614463806,
      "kl": 0.39306640625,
      "learning_rate": 2.9511296223354573e-06,
      "loss": 0.0172,
      "reward": 1.300000011920929,
      "reward_std": 0.06846531480550766,
      "rewards/accuracy_reward": 0.30000000074505806,
      "rewards/format_reward": 1.0,
      "step": 502
    },
    {
      "completion_length": 75.7750015258789,
      "epoch": 0.5215137376879212,
      "grad_norm": 0.7337321639060974,
      "kl": 0.386962890625,
      "learning_rate": 2.9506700433908988e-06,
      "loss": 0.0384,
      "reward": 1.375,
      "reward_std": 0.26129766181111336,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 1.0,
      "step": 503
    },
    {
      "completion_length": 67.75000095367432,
      "epoch": 0.5225505443234837,
      "grad_norm": 0.9252765774726868,
      "kl": 0.3251953125,
      "learning_rate": 2.9502083497220632e-06,
      "loss": -0.0029,
      "reward": 1.675000011920929,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 1.0,
      "step": 504
    },
    {
      "completion_length": 69.30000114440918,
      "epoch": 0.5235873509590462,
      "grad_norm": 0.8446915745735168,
      "kl": 0.34375,
      "learning_rate": 2.949744542001987e-06,
      "loss": 0.0156,
      "reward": 1.3750000298023224,
      "reward_std": 0.32818931341171265,
      "rewards/accuracy_reward": 0.40000000409781933,
      "rewards/format_reward": 0.9750000089406967,
      "step": 505
    },
    {
      "completion_length": 127.27500343322754,
      "epoch": 0.5246241575946086,
      "grad_norm": 2.3860890865325928,
      "kl": 1.30126953125,
      "learning_rate": 2.94927862090679e-06,
      "loss": 0.1343,
      "reward": 1.4749999940395355,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9750000089406967,
      "step": 506
    },
    {
      "completion_length": 71.9250020980835,
      "epoch": 0.5256609642301711,
      "grad_norm": 0.4810761511325836,
      "kl": 0.458984375,
      "learning_rate": 2.9488105871156717e-06,
      "loss": 0.063,
      "reward": 1.3750000149011612,
      "reward_std": 0.20019174739718437,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 0.9500000029802322,
      "step": 507
    },
    {
      "completion_length": 117.42500686645508,
      "epoch": 0.5266977708657335,
      "grad_norm": 1.6127690076828003,
      "kl": 0.67724609375,
      "learning_rate": 2.948340441310911e-06,
      "loss": 0.0579,
      "reward": 1.3750000447034836,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 508
    },
    {
      "completion_length": 67.17500305175781,
      "epoch": 0.527734577501296,
      "grad_norm": 1.0923641920089722,
      "kl": 0.5478515625,
      "learning_rate": 2.9478681841778667e-06,
      "loss": 0.0588,
      "reward": 1.5000000298023224,
      "reward_std": 0.292072132229805,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.9750000089406967,
      "step": 509
    },
    {
      "completion_length": 91.72500228881836,
      "epoch": 0.5287713841368584,
      "grad_norm": 1.9580719470977783,
      "kl": 0.9287109375,
      "learning_rate": 2.9473938164049744e-06,
      "loss": 0.0741,
      "reward": 1.4250000417232513,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.9500000178813934,
      "step": 510
    },
    {
      "completion_length": 74.10000133514404,
      "epoch": 0.5298081907724209,
      "grad_norm": 0.6495768427848816,
      "kl": 0.37744140625,
      "learning_rate": 2.946917338683748e-06,
      "loss": 0.0301,
      "reward": 1.125,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.15000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 511
    },
    {
      "completion_length": 121.17500114440918,
      "epoch": 0.5308449974079834,
      "grad_norm": 0.9497795104980469,
      "kl": 0.634765625,
      "learning_rate": 2.946438751708775e-06,
      "loss": 0.0806,
      "reward": 1.4000000059604645,
      "reward_std": 0.37310105189681053,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9000000208616257,
      "step": 512
    },
    {
      "completion_length": 71.57499980926514,
      "epoch": 0.5318818040435459,
      "grad_norm": 0.5596094131469727,
      "kl": 0.39501953125,
      "learning_rate": 2.94595805617772e-06,
      "loss": 0.0342,
      "reward": 1.2000000029802322,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.22500000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 513
    },
    {
      "completion_length": 174.45000171661377,
      "epoch": 0.5329186106791084,
      "grad_norm": 5.549030303955078,
      "kl": 0.6337890625,
      "learning_rate": 2.9454752527913194e-06,
      "loss": 0.1269,
      "reward": 1.1750000268220901,
      "reward_std": 0.2920721061527729,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.9500000178813934,
      "step": 514
    },
    {
      "completion_length": 148.6500072479248,
      "epoch": 0.5339554173146708,
      "grad_norm": 36.280643463134766,
      "kl": 4.50537109375,
      "learning_rate": 2.9449903422533836e-06,
      "loss": 0.3704,
      "reward": 1.600000023841858,
      "reward_std": 0.347973819822073,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 515
    },
    {
      "completion_length": 205.25000762939453,
      "epoch": 0.5349922239502333,
      "grad_norm": 97.75532531738281,
      "kl": 7.87744140625,
      "learning_rate": 2.9445033252707943e-06,
      "loss": 0.7113,
      "reward": 1.2249999940395355,
      "reward_std": 0.26143617182970047,
      "rewards/accuracy_reward": 0.30000001192092896,
      "rewards/format_reward": 0.925000011920929,
      "step": 516
    },
    {
      "completion_length": 83.45000267028809,
      "epoch": 0.5360290305857958,
      "grad_norm": 0.8470296263694763,
      "kl": 0.443359375,
      "learning_rate": 2.944014202553503e-06,
      "loss": -0.0235,
      "reward": 1.600000023841858,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6000000033527613,
      "rewards/format_reward": 1.0,
      "step": 517
    },
    {
      "completion_length": 80.52500057220459,
      "epoch": 0.5370658372213583,
      "grad_norm": 4.682936191558838,
      "kl": 0.51123046875,
      "learning_rate": 2.943522974814533e-06,
      "loss": 0.0104,
      "reward": 1.7000000178813934,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 1.0,
      "step": 518
    },
    {
      "completion_length": 80.22500228881836,
      "epoch": 0.5381026438569206,
      "grad_norm": 0.6545246243476868,
      "kl": 0.56591796875,
      "learning_rate": 2.9430296427699743e-06,
      "loss": 0.0316,
      "reward": 1.550000011920929,
      "reward_std": 0.21275537461042404,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 519
    },
    {
      "completion_length": 139.82500457763672,
      "epoch": 0.5391394504924831,
      "grad_norm": 0.5691772699356079,
      "kl": 0.54736328125,
      "learning_rate": 2.9425342071389853e-06,
      "loss": 0.0921,
      "reward": 1.2750000357627869,
      "reward_std": 0.1928323432803154,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.9500000029802322,
      "step": 520
    },
    {
      "completion_length": 185.52500534057617,
      "epoch": 0.5401762571280456,
      "grad_norm": 1.905286192893982,
      "kl": 0.724609375,
      "learning_rate": 2.94203666864379e-06,
      "loss": 0.1483,
      "reward": 1.475000023841858,
      "reward_std": 0.4290027618408203,
      "rewards/accuracy_reward": 0.5249999947845936,
      "rewards/format_reward": 0.9500000178813934,
      "step": 521
    },
    {
      "completion_length": 80.97500038146973,
      "epoch": 0.5412130637636081,
      "grad_norm": 0.35716575384140015,
      "kl": 0.343994140625,
      "learning_rate": 2.9415370280096797e-06,
      "loss": 0.0305,
      "reward": 1.3499999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9750000089406967,
      "step": 522
    },
    {
      "completion_length": 81.00000190734863,
      "epoch": 0.5422498703991706,
      "grad_norm": 0.46804380416870117,
      "kl": 0.41943359375,
      "learning_rate": 2.941035285965008e-06,
      "loss": 0.0485,
      "reward": 1.449999988079071,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000178813934,
      "step": 523
    },
    {
      "completion_length": 241.47500610351562,
      "epoch": 0.543286677034733,
      "grad_norm": 150.73275756835938,
      "kl": 35.86083984375,
      "learning_rate": 2.9405314432411933e-06,
      "loss": 2.8672,
      "reward": 1.350000023841858,
      "reward_std": 0.30477424710989,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.925000011920929,
      "step": 524
    },
    {
      "completion_length": 128.45000457763672,
      "epoch": 0.5443234836702955,
      "grad_norm": 1.8508985042572021,
      "kl": 1.2958984375,
      "learning_rate": 2.9400255005727162e-06,
      "loss": 0.1372,
      "reward": 1.550000011920929,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 525
    },
    {
      "completion_length": 245.07500839233398,
      "epoch": 0.545360290305858,
      "grad_norm": 1146.21435546875,
      "kl": 151.87255859375,
      "learning_rate": 2.9395174586971174e-06,
      "loss": 11.6084,
      "reward": 1.4000000059604645,
      "reward_std": 0.3913118802011013,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9000000208616257,
      "step": 526
    },
    {
      "completion_length": 86.67500114440918,
      "epoch": 0.5463970969414205,
      "grad_norm": 0.6502148509025574,
      "kl": 0.30712890625,
      "learning_rate": 2.939007318354999e-06,
      "loss": -0.0365,
      "reward": 1.6500000357627869,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 1.0,
      "step": 527
    },
    {
      "completion_length": 163.04999923706055,
      "epoch": 0.5474339035769828,
      "grad_norm": 1.801808476448059,
      "kl": 0.300537109375,
      "learning_rate": 2.938495080290021e-06,
      "loss": 0.0618,
      "reward": 1.0250000357627869,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.07500000111758709,
      "rewards/format_reward": 0.9500000178813934,
      "step": 528
    },
    {
      "completion_length": 288.6750011444092,
      "epoch": 0.5484707102125453,
      "grad_norm": 6.281752109527588,
      "kl": 0.3994140625,
      "learning_rate": 2.937980745248902e-06,
      "loss": 0.1594,
      "reward": 1.7000000476837158,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.8000000193715096,
      "rewards/format_reward": 0.9000000208616257,
      "step": 529
    },
    {
      "completion_length": 155.25000381469727,
      "epoch": 0.5495075168481078,
      "grad_norm": 0.6555303931236267,
      "kl": 0.35302734375,
      "learning_rate": 2.9374643139814184e-06,
      "loss": 0.0639,
      "reward": 1.300000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 530
    },
    {
      "completion_length": 214.6250114440918,
      "epoch": 0.5505443234836703,
      "grad_norm": 2.5419232845306396,
      "kl": 0.55859375,
      "learning_rate": 2.936945787240401e-06,
      "loss": 0.1867,
      "reward": 1.3250000476837158,
      "reward_std": 0.44892582297325134,
      "rewards/accuracy_reward": 0.3999999985098839,
      "rewards/format_reward": 0.925000011920929,
      "step": 531
    },
    {
      "completion_length": 100.6500015258789,
      "epoch": 0.5515811301192327,
      "grad_norm": 0.7034846544265747,
      "kl": 0.30908203125,
      "learning_rate": 2.936425165781735e-06,
      "loss": 0.0153,
      "reward": 1.5500000417232513,
      "reward_std": 0.31719936802983284,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 1.0,
      "step": 532
    },
    {
      "completion_length": 256.25,
      "epoch": 0.5526179367547952,
      "grad_norm": 4.066845417022705,
      "kl": 0.75,
      "learning_rate": 2.935902450364361e-06,
      "loss": 0.1729,
      "reward": 1.6000000536441803,
      "reward_std": 0.2487340271472931,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000059604645,
      "step": 533
    },
    {
      "completion_length": 94.05000305175781,
      "epoch": 0.5536547433903577,
      "grad_norm": 0.6132490634918213,
      "kl": 0.25537109375,
      "learning_rate": 2.9353776417502704e-06,
      "loss": 0.0136,
      "reward": 1.550000011920929,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 1.0,
      "step": 534
    },
    {
      "completion_length": 213.6500015258789,
      "epoch": 0.5546915500259202,
      "grad_norm": 11.384795188903809,
      "kl": 2.087890625,
      "learning_rate": 2.934850740704506e-06,
      "loss": 0.3464,
      "reward": 1.6000000536441803,
      "reward_std": 0.5031152740120888,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000357627869,
      "step": 535
    },
    {
      "completion_length": 102.1500015258789,
      "epoch": 0.5557283566614827,
      "grad_norm": 1.437371850013733,
      "kl": 0.63671875,
      "learning_rate": 2.934321747995163e-06,
      "loss": 0.0334,
      "reward": 1.7250000536441803,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 536
    },
    {
      "completion_length": 164.0500030517578,
      "epoch": 0.5567651632970451,
      "grad_norm": 0.8429715037345886,
      "kl": 0.400634765625,
      "learning_rate": 2.933790664393383e-06,
      "loss": 0.093,
      "reward": 1.425000011920929,
      "reward_std": 0.3354102075099945,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.9500000178813934,
      "step": 537
    },
    {
      "completion_length": 169.07500076293945,
      "epoch": 0.5578019699326076,
      "grad_norm": 0.981852650642395,
      "kl": 0.619384765625,
      "learning_rate": 2.9332574906733574e-06,
      "loss": 0.1237,
      "reward": 1.7000000476837158,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 538
    },
    {
      "completion_length": 343.6750011444092,
      "epoch": 0.55883877656817,
      "grad_norm": 1.502892255783081,
      "kl": 0.77099609375,
      "learning_rate": 2.9327222276123232e-06,
      "loss": 0.2786,
      "reward": 1.5000000149011612,
      "reward_std": 0.3533165603876114,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.8750000149011612,
      "step": 539
    },
    {
      "completion_length": 167.00000190734863,
      "epoch": 0.5598755832037325,
      "grad_norm": 0.9482361674308777,
      "kl": 0.635009765625,
      "learning_rate": 2.9321848759905643e-06,
      "loss": 0.08,
      "reward": 1.725000023841858,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 540
    },
    {
      "completion_length": 113.20000267028809,
      "epoch": 0.5609123898392949,
      "grad_norm": 0.9624785780906677,
      "kl": 0.711669921875,
      "learning_rate": 2.931645436591409e-06,
      "loss": 0.1214,
      "reward": 1.4500000178813934,
      "reward_std": 0.2614361569285393,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 541
    },
    {
      "completion_length": 353.399995803833,
      "epoch": 0.5619491964748574,
      "grad_norm": 4.333626747131348,
      "kl": 2.579345703125,
      "learning_rate": 2.9311039102012277e-06,
      "loss": 0.2335,
      "reward": 1.2000000178813934,
      "reward_std": 0.28485121577978134,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.8500000089406967,
      "step": 542
    },
    {
      "completion_length": 177.00000381469727,
      "epoch": 0.5629860031104199,
      "grad_norm": 1.3422383069992065,
      "kl": 0.70458984375,
      "learning_rate": 2.9305602976094354e-06,
      "loss": 0.1408,
      "reward": 1.3750000149011612,
      "reward_std": 0.279508501291275,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9250000268220901,
      "step": 543
    },
    {
      "completion_length": 181.97500610351562,
      "epoch": 0.5640228097459824,
      "grad_norm": 1.8206522464752197,
      "kl": 1.82177734375,
      "learning_rate": 2.9300145996084865e-06,
      "loss": 0.2003,
      "reward": 1.4000000059604645,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 544
    },
    {
      "completion_length": 340.52500915527344,
      "epoch": 0.5650596163815449,
      "grad_norm": 77.96588134765625,
      "kl": 18.46630859375,
      "learning_rate": 2.929466816993877e-06,
      "loss": 1.4838,
      "reward": 1.4250000268220901,
      "reward_std": 0.38109346479177475,
      "rewards/accuracy_reward": 0.5500000063329935,
      "rewards/format_reward": 0.8750000149011612,
      "step": 545
    },
    {
      "completion_length": 107.55000114440918,
      "epoch": 0.5660964230171073,
      "grad_norm": 0.47127431631088257,
      "kl": 0.373779296875,
      "learning_rate": 2.928916950564139e-06,
      "loss": 0.0424,
      "reward": 1.2000000327825546,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.9750000089406967,
      "step": 546
    },
    {
      "completion_length": 187.20000076293945,
      "epoch": 0.5671332296526698,
      "grad_norm": 4.727897644042969,
      "kl": 3.77197265625,
      "learning_rate": 2.928365001120846e-06,
      "loss": 0.2137,
      "reward": 1.5000000298023224,
      "reward_std": 0.4723408445715904,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9000000208616257,
      "step": 547
    },
    {
      "completion_length": 212.32500457763672,
      "epoch": 0.5681700362882323,
      "grad_norm": 1.5274919271469116,
      "kl": 2.5712890625,
      "learning_rate": 2.927810969468605e-06,
      "loss": 0.3168,
      "reward": 1.3000000268220901,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.9250000268220901,
      "step": 548
    },
    {
      "completion_length": 250.85000038146973,
      "epoch": 0.5692068429237948,
      "grad_norm": 1.4735158681869507,
      "kl": 1.9208984375,
      "learning_rate": 2.9272548564150593e-06,
      "loss": 0.2983,
      "reward": 1.5,
      "reward_std": 0.3913119100034237,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 549
    },
    {
      "completion_length": 102.10000228881836,
      "epoch": 0.5702436495593571,
      "grad_norm": 0.5658621191978455,
      "kl": 0.34375,
      "learning_rate": 2.926696662770887e-06,
      "loss": -0.0251,
      "reward": 1.350000023841858,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 1.0,
      "step": 550
    },
    {
      "completion_length": 118.80000495910645,
      "epoch": 0.5712804561949196,
      "grad_norm": 0.3587953448295593,
      "kl": 0.31005859375,
      "learning_rate": 2.9261363893497988e-06,
      "loss": 0.0136,
      "reward": 1.425000011920929,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 1.0,
      "step": 551
    },
    {
      "completion_length": 147.57500076293945,
      "epoch": 0.5723172628304821,
      "grad_norm": 0.8031432032585144,
      "kl": 0.4482421875,
      "learning_rate": 2.9255740369685367e-06,
      "loss": 0.016,
      "reward": 1.3500000089406967,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 552
    },
    {
      "completion_length": 164.05000495910645,
      "epoch": 0.5733540694660446,
      "grad_norm": 1.5201157331466675,
      "kl": 1.3935546875,
      "learning_rate": 2.9250096064468745e-06,
      "loss": 0.1377,
      "reward": 1.3500000536441803,
      "reward_std": 0.37310105562210083,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 553
    },
    {
      "completion_length": 101.75000190734863,
      "epoch": 0.574390876101607,
      "grad_norm": 0.4400118887424469,
      "kl": 0.295166015625,
      "learning_rate": 2.9244430986076146e-06,
      "loss": -0.0034,
      "reward": 1.550000011920929,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 1.0,
      "step": 554
    },
    {
      "completion_length": 202.8499984741211,
      "epoch": 0.5754276827371695,
      "grad_norm": 2.103760004043579,
      "kl": 2.682373046875,
      "learning_rate": 2.923874514276587e-06,
      "loss": 0.2481,
      "reward": 1.300000011920929,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9250000268220901,
      "step": 555
    },
    {
      "completion_length": 103.17500114440918,
      "epoch": 0.576464489372732,
      "grad_norm": 0.5067854523658752,
      "kl": 0.31201171875,
      "learning_rate": 2.9233038542826496e-06,
      "loss": 0.0438,
      "reward": 1.5250000059604645,
      "reward_std": 0.1928323432803154,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 1.0,
      "step": 556
    },
    {
      "completion_length": 172.92499923706055,
      "epoch": 0.5775012960082945,
      "grad_norm": 1.1114473342895508,
      "kl": 0.310791015625,
      "learning_rate": 2.9227311194576865e-06,
      "loss": 0.0272,
      "reward": 1.4500000178813934,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.4749999977648258,
      "rewards/format_reward": 0.9750000089406967,
      "step": 557
    },
    {
      "completion_length": 156.7500057220459,
      "epoch": 0.578538102643857,
      "grad_norm": 1.3353294134140015,
      "kl": 1.7294921875,
      "learning_rate": 2.9221563106366057e-06,
      "loss": 0.1717,
      "reward": 1.225000023841858,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.25000000186264515,
      "rewards/format_reward": 0.9750000089406967,
      "step": 558
    },
    {
      "completion_length": 175.80000686645508,
      "epoch": 0.5795749092794193,
      "grad_norm": 3.629734516143799,
      "kl": 1.093505859375,
      "learning_rate": 2.9215794286573383e-06,
      "loss": 0.1307,
      "reward": 1.275000050663948,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.9500000178813934,
      "step": 559
    },
    {
      "completion_length": 111.82500267028809,
      "epoch": 0.5806117159149818,
      "grad_norm": 0.19621193408966064,
      "kl": 0.2568359375,
      "learning_rate": 2.921000474360838e-06,
      "loss": 0.0142,
      "reward": 1.4749999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 1.0,
      "step": 560
    },
    {
      "completion_length": 121.42500305175781,
      "epoch": 0.5816485225505443,
      "grad_norm": 0.4136292040348053,
      "kl": 0.21484375,
      "learning_rate": 2.9204194485910798e-06,
      "loss": -0.0062,
      "reward": 1.600000023841858,
      "reward_std": 0.1928323358297348,
      "rewards/accuracy_reward": 0.5999999940395355,
      "rewards/format_reward": 1.0,
      "step": 561
    },
    {
      "completion_length": 131.62500190734863,
      "epoch": 0.5826853291861068,
      "grad_norm": 0.49414509534835815,
      "kl": 0.26416015625,
      "learning_rate": 2.9198363521950583e-06,
      "loss": 0.0075,
      "reward": 1.4000000357627869,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 1.0,
      "step": 562
    },
    {
      "completion_length": 173.89999961853027,
      "epoch": 0.5837221358216692,
      "grad_norm": 0.8812211751937866,
      "kl": 0.484375,
      "learning_rate": 2.9192511860227855e-06,
      "loss": 0.0233,
      "reward": 1.3500000089406967,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 563
    },
    {
      "completion_length": 175.225004196167,
      "epoch": 0.5847589424572317,
      "grad_norm": 1.1985092163085938,
      "kl": 1.24365234375,
      "learning_rate": 2.9186639509272923e-06,
      "loss": 0.0881,
      "reward": 1.5000000298023224,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9500000178813934,
      "step": 564
    },
    {
      "completion_length": 134.4250030517578,
      "epoch": 0.5857957490927942,
      "grad_norm": 0.4528398811817169,
      "kl": 0.299072265625,
      "learning_rate": 2.918074647764624e-06,
      "loss": 0.0062,
      "reward": 1.699999988079071,
      "reward_std": 0.18026871606707573,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 1.0,
      "step": 565
    },
    {
      "completion_length": 120.1500015258789,
      "epoch": 0.5868325557283567,
      "grad_norm": 0.4748910069465637,
      "kl": 0.2451171875,
      "learning_rate": 2.9174832773938413e-06,
      "loss": 0.026,
      "reward": 1.725000023841858,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.725000012665987,
      "rewards/format_reward": 1.0,
      "step": 566
    },
    {
      "completion_length": 180.22500610351562,
      "epoch": 0.5878693623639192,
      "grad_norm": 0.43165019154548645,
      "kl": 0.501708984375,
      "learning_rate": 2.916889840677019e-06,
      "loss": 0.0587,
      "reward": 1.350000023841858,
      "reward_std": 0.248734038323164,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 567
    },
    {
      "completion_length": 176.225004196167,
      "epoch": 0.5889061689994816,
      "grad_norm": 0.5352876782417297,
      "kl": 0.711669921875,
      "learning_rate": 2.9162943384792435e-06,
      "loss": 0.0419,
      "reward": 1.6500000059604645,
      "reward_std": 0.21275537088513374,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9750000089406967,
      "step": 568
    },
    {
      "completion_length": 193.17500495910645,
      "epoch": 0.589942975635044,
      "grad_norm": 0.6368687152862549,
      "kl": 0.36376953125,
      "learning_rate": 2.915696771668612e-06,
      "loss": 0.0535,
      "reward": 1.5750000476837158,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 569
    },
    {
      "completion_length": 132.32500457763672,
      "epoch": 0.5909797822706065,
      "grad_norm": 0.36829519271850586,
      "kl": 0.236328125,
      "learning_rate": 2.9150971411162315e-06,
      "loss": -0.0057,
      "reward": 1.4500000178813934,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.44999999925494194,
      "rewards/format_reward": 1.0,
      "step": 570
    },
    {
      "completion_length": 152.65000343322754,
      "epoch": 0.592016588906169,
      "grad_norm": 0.4644145965576172,
      "kl": 0.241943359375,
      "learning_rate": 2.914495447696218e-06,
      "loss": -0.0219,
      "reward": 1.625,
      "reward_std": 0.24873404577374458,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 1.0,
      "step": 571
    },
    {
      "completion_length": 190.15000915527344,
      "epoch": 0.5930533955417314,
      "grad_norm": 0.4808287024497986,
      "kl": 0.278564453125,
      "learning_rate": 2.9138916922856946e-06,
      "loss": 0.0605,
      "reward": 1.5250000059604645,
      "reward_std": 0.21275536343455315,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9750000089406967,
      "step": 572
    },
    {
      "completion_length": 203.225004196167,
      "epoch": 0.5940902021772939,
      "grad_norm": 9.010258674621582,
      "kl": 5.313232421875,
      "learning_rate": 2.9132858757647884e-06,
      "loss": 0.4188,
      "reward": 1.6500000357627869,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 573
    },
    {
      "completion_length": 142.100004196167,
      "epoch": 0.5951270088128564,
      "grad_norm": 0.4201050400733948,
      "kl": 0.2431640625,
      "learning_rate": 2.912677999016634e-06,
      "loss": 0.0162,
      "reward": 1.350000023841858,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 1.0,
      "step": 574
    },
    {
      "completion_length": 174.1750030517578,
      "epoch": 0.5961638154484189,
      "grad_norm": 0.3380955755710602,
      "kl": 0.2294921875,
      "learning_rate": 2.9120680629273677e-06,
      "loss": -0.0121,
      "reward": 1.4000000357627869,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9750000089406967,
      "step": 575
    },
    {
      "completion_length": 153.12500381469727,
      "epoch": 0.5972006220839814,
      "grad_norm": 0.44647836685180664,
      "kl": 0.2392578125,
      "learning_rate": 2.9114560683861275e-06,
      "loss": 0.0226,
      "reward": 1.5250000357627869,
      "reward_std": 0.24873404577374458,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 1.0,
      "step": 576
    },
    {
      "completion_length": 262.6500015258789,
      "epoch": 0.5982374287195438,
      "grad_norm": 7.015067100524902,
      "kl": 0.88427734375,
      "learning_rate": 2.910842016285053e-06,
      "loss": 0.1761,
      "reward": 1.475000023841858,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.925000011920929,
      "step": 577
    },
    {
      "completion_length": 132.9000015258789,
      "epoch": 0.5992742353551063,
      "grad_norm": 0.2848128080368042,
      "kl": 0.248046875,
      "learning_rate": 2.910225907519283e-06,
      "loss": -0.002,
      "reward": 1.2750000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 1.0,
      "step": 578
    },
    {
      "completion_length": 303.50000190734863,
      "epoch": 0.6003110419906688,
      "grad_norm": 96.8685302734375,
      "kl": 25.11279296875,
      "learning_rate": 2.9096077429869543e-06,
      "loss": 1.9498,
      "reward": 1.4500000178813934,
      "reward_std": 0.2920721024274826,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.9000000059604645,
      "step": 579
    },
    {
      "completion_length": 245.5250072479248,
      "epoch": 0.6013478486262313,
      "grad_norm": 1.2218270301818848,
      "kl": 0.70263671875,
      "learning_rate": 2.9089875235892005e-06,
      "loss": 0.1386,
      "reward": 1.5250000208616257,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 580
    },
    {
      "completion_length": 188.2000026702881,
      "epoch": 0.6023846552617936,
      "grad_norm": 4.220678806304932,
      "kl": 2.01708984375,
      "learning_rate": 2.9083652502301506e-06,
      "loss": 0.1683,
      "reward": 1.600000023841858,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 581
    },
    {
      "completion_length": 269.4500045776367,
      "epoch": 0.6034214618973561,
      "grad_norm": 5.021411895751953,
      "kl": 4.0146484375,
      "learning_rate": 2.907740923816928e-06,
      "loss": 0.3696,
      "reward": 1.2500000298023224,
      "reward_std": 0.4038754962384701,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.9250000268220901,
      "step": 582
    },
    {
      "completion_length": 183.97500038146973,
      "epoch": 0.6044582685329186,
      "grad_norm": 0.7388213872909546,
      "kl": 0.39697265625,
      "learning_rate": 2.907114545259649e-06,
      "loss": 0.1162,
      "reward": 1.350000038743019,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.9000000208616257,
      "step": 583
    },
    {
      "completion_length": 198.9749984741211,
      "epoch": 0.6054950751684811,
      "grad_norm": 6.0927934646606445,
      "kl": 1.3525390625,
      "learning_rate": 2.9064861154714206e-06,
      "loss": 0.108,
      "reward": 1.3750000149011612,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 584
    },
    {
      "completion_length": 350.6000099182129,
      "epoch": 0.6065318818040435,
      "grad_norm": 3.432065010070801,
      "kl": 1.06787109375,
      "learning_rate": 2.9058556353683413e-06,
      "loss": 0.1571,
      "reward": 1.6500000059604645,
      "reward_std": 0.5715806148946285,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.8750000149011612,
      "step": 585
    },
    {
      "completion_length": 186.05000495910645,
      "epoch": 0.607568688439606,
      "grad_norm": 1.2789217233657837,
      "kl": 1.263916015625,
      "learning_rate": 2.9052231058694975e-06,
      "loss": 0.1736,
      "reward": 1.550000011920929,
      "reward_std": 0.38409099355340004,
      "rewards/accuracy_reward": 0.6500000152736902,
      "rewards/format_reward": 0.9000000059604645,
      "step": 586
    },
    {
      "completion_length": 140.97500228881836,
      "epoch": 0.6086054950751685,
      "grad_norm": 0.31413471698760986,
      "kl": 0.222900390625,
      "learning_rate": 2.9045885278969635e-06,
      "loss": -0.0107,
      "reward": 1.675000011920929,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 587
    },
    {
      "completion_length": 130.9000015258789,
      "epoch": 0.609642301710731,
      "grad_norm": 0.4696264863014221,
      "kl": 0.25244140625,
      "learning_rate": 2.9039519023758004e-06,
      "loss": 0.0824,
      "reward": 1.4000000059604645,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 1.0,
      "step": 588
    },
    {
      "completion_length": 196.32500839233398,
      "epoch": 0.6106791083462935,
      "grad_norm": 4.557334899902344,
      "kl": 0.811767578125,
      "learning_rate": 2.9033132302340523e-06,
      "loss": 0.1282,
      "reward": 1.5250000655651093,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 589
    },
    {
      "completion_length": 292.9749984741211,
      "epoch": 0.6117159149818558,
      "grad_norm": 9.806172370910645,
      "kl": 5.443359375,
      "learning_rate": 2.902672512402749e-06,
      "loss": 0.2563,
      "reward": 1.4000000059604645,
      "reward_std": 0.2974148504436016,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.9000000208616257,
      "step": 590
    },
    {
      "completion_length": 274.25,
      "epoch": 0.6127527216174183,
      "grad_norm": 2.256058692932129,
      "kl": 1.713134765625,
      "learning_rate": 2.9020297498159004e-06,
      "loss": 0.2101,
      "reward": 1.5750000476837158,
      "reward_std": 0.36588016152381897,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 591
    },
    {
      "completion_length": 241.0500030517578,
      "epoch": 0.6137895282529808,
      "grad_norm": 1.3324028253555298,
      "kl": 0.608154296875,
      "learning_rate": 2.901384943410499e-06,
      "loss": 0.1396,
      "reward": 1.5750000476837158,
      "reward_std": 0.403875507414341,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 592
    },
    {
      "completion_length": 247.00000953674316,
      "epoch": 0.6148263348885433,
      "grad_norm": 6.3438873291015625,
      "kl": 0.499267578125,
      "learning_rate": 2.900738094126515e-06,
      "loss": 0.1196,
      "reward": 1.600000023841858,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 593
    },
    {
      "completion_length": 296.42499923706055,
      "epoch": 0.6158631415241057,
      "grad_norm": 1.7976075410842896,
      "kl": 0.67138671875,
      "learning_rate": 2.9000892029068974e-06,
      "loss": 0.0767,
      "reward": 1.3000000268220901,
      "reward_std": 0.3297629915177822,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.9249999970197678,
      "step": 594
    },
    {
      "completion_length": 152.9000015258789,
      "epoch": 0.6168999481596682,
      "grad_norm": 0.41897207498550415,
      "kl": 0.201171875,
      "learning_rate": 2.899438270697573e-06,
      "loss": 0.0164,
      "reward": 1.4000000059604645,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.4000000115483999,
      "rewards/format_reward": 1.0,
      "step": 595
    },
    {
      "completion_length": 224.95000457763672,
      "epoch": 0.6179367547952307,
      "grad_norm": 1.1412527561187744,
      "kl": 1.325439453125,
      "learning_rate": 2.8987852984474413e-06,
      "loss": 0.1502,
      "reward": 1.4000000357627869,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9750000089406967,
      "step": 596
    },
    {
      "completion_length": 170.4000015258789,
      "epoch": 0.6189735614307932,
      "grad_norm": 0.8881285190582275,
      "kl": 0.442626953125,
      "learning_rate": 2.898130287108378e-06,
      "loss": 0.0808,
      "reward": 1.2500000447034836,
      "reward_std": 0.1928323321044445,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 597
    },
    {
      "completion_length": 282.5500030517578,
      "epoch": 0.6200103680663557,
      "grad_norm": 11.336788177490234,
      "kl": 6.169921875,
      "learning_rate": 2.8974732376352302e-06,
      "loss": 0.6068,
      "reward": 1.5250000357627869,
      "reward_std": 0.347973819822073,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9250000268220901,
      "step": 598
    },
    {
      "completion_length": 203.20000648498535,
      "epoch": 0.6210471747019181,
      "grad_norm": 1.8857403993606567,
      "kl": 1.327392578125,
      "learning_rate": 2.8968141509858157e-06,
      "loss": 0.1394,
      "reward": 1.3250000178813934,
      "reward_std": 0.21275536715984344,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.9750000089406967,
      "step": 599
    },
    {
      "completion_length": 165.30000114440918,
      "epoch": 0.6220839813374806,
      "grad_norm": 0.3913152515888214,
      "kl": 0.234619140625,
      "learning_rate": 2.896153028120924e-06,
      "loss": -0.0227,
      "reward": 1.625,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.6250000186264515,
      "rewards/format_reward": 1.0,
      "step": 600
    },
    {
      "completion_length": 190.225004196167,
      "epoch": 0.623120787973043,
      "grad_norm": 2.7854464054107666,
      "kl": 0.409423828125,
      "learning_rate": 2.89548987000431e-06,
      "loss": 0.0567,
      "reward": 1.2000000178813934,
      "reward_std": 0.28485122323036194,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.9750000089406967,
      "step": 601
    },
    {
      "completion_length": 253.82500076293945,
      "epoch": 0.6241575946086055,
      "grad_norm": 5.791003704071045,
      "kl": 0.375244140625,
      "learning_rate": 2.8948246776026985e-06,
      "loss": 0.1521,
      "reward": 1.4500000327825546,
      "reward_std": 0.31733787432312965,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 602
    },
    {
      "completion_length": 136.20000076293945,
      "epoch": 0.6251944012441679,
      "grad_norm": 0.45052769780158997,
      "kl": 0.23486328125,
      "learning_rate": 2.8941574518857763e-06,
      "loss": -0.0226,
      "reward": 1.449999988079071,
      "reward_std": 0.18026872351765633,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 1.0,
      "step": 603
    },
    {
      "completion_length": 189.35000610351562,
      "epoch": 0.6262312078797304,
      "grad_norm": 1.993778109550476,
      "kl": 0.494140625,
      "learning_rate": 2.8934881938261987e-06,
      "loss": 0.0845,
      "reward": 1.425000011920929,
      "reward_std": 0.21275537461042404,
      "rewards/accuracy_reward": 0.4499999973922968,
      "rewards/format_reward": 0.9750000089406967,
      "step": 604
    },
    {
      "completion_length": 299.3500099182129,
      "epoch": 0.6272680145152929,
      "grad_norm": 3.556016683578491,
      "kl": 1.59912109375,
      "learning_rate": 2.8928169043995793e-06,
      "loss": 0.2569,
      "reward": 1.2749999910593033,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9000000357627869,
      "step": 605
    },
    {
      "completion_length": 132.07500076293945,
      "epoch": 0.6283048211508554,
      "grad_norm": 0.5948020219802856,
      "kl": 0.26513671875,
      "learning_rate": 2.8921435845844963e-06,
      "loss": 0.0789,
      "reward": 1.7000000178813934,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 1.0,
      "step": 606
    },
    {
      "completion_length": 239.55000495910645,
      "epoch": 0.6293416277864178,
      "grad_norm": 11.134238243103027,
      "kl": 3.626220703125,
      "learning_rate": 2.8914682353624855e-06,
      "loss": 0.3149,
      "reward": 1.3750000596046448,
      "reward_std": 0.36053745448589325,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9500000178813934,
      "step": 607
    },
    {
      "completion_length": 217.1500015258789,
      "epoch": 0.6303784344219803,
      "grad_norm": 0.6335292458534241,
      "kl": 0.396484375,
      "learning_rate": 2.890790857718042e-06,
      "loss": 0.0686,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 608
    },
    {
      "completion_length": 252.00000381469727,
      "epoch": 0.6314152410575428,
      "grad_norm": 45.83848190307617,
      "kl": 8.284423828125,
      "learning_rate": 2.8901114526386176e-06,
      "loss": 0.6874,
      "reward": 1.5250000357627869,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.925000011920929,
      "step": 609
    },
    {
      "completion_length": 224.375,
      "epoch": 0.6324520476931053,
      "grad_norm": 6.116477012634277,
      "kl": 2.39501953125,
      "learning_rate": 2.8894300211146207e-06,
      "loss": 0.2235,
      "reward": 1.375,
      "reward_std": 0.22360679879784584,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 610
    },
    {
      "completion_length": 194.2000026702881,
      "epoch": 0.6334888543286677,
      "grad_norm": 0.7082158923149109,
      "kl": 0.6552734375,
      "learning_rate": 2.8887465641394114e-06,
      "loss": 0.1211,
      "reward": 1.425000011920929,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 611
    },
    {
      "completion_length": 181.80000686645508,
      "epoch": 0.6345256609642301,
      "grad_norm": 0.7598231434822083,
      "kl": 1.06689453125,
      "learning_rate": 2.8880610827093045e-06,
      "loss": 0.0976,
      "reward": 1.4000000059604645,
      "reward_std": 0.21638592705130577,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9750000089406967,
      "step": 612
    },
    {
      "completion_length": 165.0500030517578,
      "epoch": 0.6355624675997926,
      "grad_norm": 0.45146992802619934,
      "kl": 0.185302734375,
      "learning_rate": 2.8873735778235657e-06,
      "loss": 0.0104,
      "reward": 1.5,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.9750000089406967,
      "step": 613
    },
    {
      "completion_length": 252.3750114440918,
      "epoch": 0.6365992742353551,
      "grad_norm": 0.7813171148300171,
      "kl": 0.745849609375,
      "learning_rate": 2.8866840504844094e-06,
      "loss": 0.1231,
      "reward": 1.2500000298023224,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.925000011920929,
      "step": 614
    },
    {
      "completion_length": 289.6500015258789,
      "epoch": 0.6376360808709176,
      "grad_norm": 3.395939588546753,
      "kl": 3.01611328125,
      "learning_rate": 2.8859925016969995e-06,
      "loss": 0.2634,
      "reward": 1.4249999821186066,
      "reward_std": 0.3737340532243252,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.8750000149011612,
      "step": 615
    },
    {
      "completion_length": 462.9249954223633,
      "epoch": 0.63867288750648,
      "grad_norm": 2.6913440227508545,
      "kl": 3.1015625,
      "learning_rate": 2.885298932469445e-06,
      "loss": 0.4213,
      "reward": 1.3250000476837158,
      "reward_std": 0.7158706709742546,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.8250000029802322,
      "step": 616
    },
    {
      "completion_length": 311.7750129699707,
      "epoch": 0.6397096941420425,
      "grad_norm": 1.5979684591293335,
      "kl": 2.26220703125,
      "learning_rate": 2.884603343812802e-06,
      "loss": 0.1652,
      "reward": 1.2500000149011612,
      "reward_std": 0.39728759974241257,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9000000208616257,
      "step": 617
    },
    {
      "completion_length": 327.32501220703125,
      "epoch": 0.640746500777605,
      "grad_norm": 12.20130443572998,
      "kl": 6.74072265625,
      "learning_rate": 2.8839057367410696e-06,
      "loss": 0.6414,
      "reward": 1.600000023841858,
      "reward_std": 0.4237985163927078,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 618
    },
    {
      "completion_length": 207.82500076293945,
      "epoch": 0.6417833074131675,
      "grad_norm": 1.5905780792236328,
      "kl": 0.2890625,
      "learning_rate": 2.883206112271189e-06,
      "loss": 0.0787,
      "reward": 1.375,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 0.9750000089406967,
      "step": 619
    },
    {
      "completion_length": 218.95000076293945,
      "epoch": 0.64282011404873,
      "grad_norm": 3.3236258029937744,
      "kl": 1.61669921875,
      "learning_rate": 2.8825044714230427e-06,
      "loss": 0.1564,
      "reward": 1.5250000208616257,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 620
    },
    {
      "completion_length": 233.95000076293945,
      "epoch": 0.6438569206842923,
      "grad_norm": 2.439260721206665,
      "kl": 0.500732421875,
      "learning_rate": 2.881800815219453e-06,
      "loss": 0.1286,
      "reward": 1.3250000178813934,
      "reward_std": 0.22360678762197495,
      "rewards/accuracy_reward": 0.37500000186264515,
      "rewards/format_reward": 0.9500000178813934,
      "step": 621
    },
    {
      "completion_length": 200.4000015258789,
      "epoch": 0.6448937273198548,
      "grad_norm": 0.662738025188446,
      "kl": 0.79296875,
      "learning_rate": 2.881095144686179e-06,
      "loss": 0.0613,
      "reward": 1.4500000178813934,
      "reward_std": 0.3533165529370308,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.9500000029802322,
      "step": 622
    },
    {
      "completion_length": 300.60000228881836,
      "epoch": 0.6459305339554173,
      "grad_norm": 1.7969317436218262,
      "kl": 0.51513671875,
      "learning_rate": 2.880387460851917e-06,
      "loss": 0.108,
      "reward": 1.225000023841858,
      "reward_std": 0.6021890863776207,
      "rewards/accuracy_reward": 0.34999999962747097,
      "rewards/format_reward": 0.8750000298023224,
      "step": 623
    },
    {
      "completion_length": 242.57500076293945,
      "epoch": 0.6469673405909798,
      "grad_norm": 0.5022680759429932,
      "kl": 0.31103515625,
      "learning_rate": 2.879677764748298e-06,
      "loss": 0.0101,
      "reward": 1.475000023841858,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.9500000029802322,
      "step": 624
    },
    {
      "completion_length": 245.57500076293945,
      "epoch": 0.6480041472265422,
      "grad_norm": 0.5517757534980774,
      "kl": 0.272705078125,
      "learning_rate": 2.8789660574098863e-06,
      "loss": 0.0851,
      "reward": 1.25,
      "reward_std": 0.3696090541779995,
      "rewards/accuracy_reward": 0.30000000819563866,
      "rewards/format_reward": 0.9500000029802322,
      "step": 625
    },
    {
      "completion_length": 160.2750015258789,
      "epoch": 0.6490409538621047,
      "grad_norm": 0.43617773056030273,
      "kl": 0.24365234375,
      "learning_rate": 2.8782523398741783e-06,
      "loss": -0.0277,
      "reward": 1.3250000327825546,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 626
    },
    {
      "completion_length": 194.95000457763672,
      "epoch": 0.6500777604976672,
      "grad_norm": 0.6225988268852234,
      "kl": 0.365234375,
      "learning_rate": 2.8775366131816006e-06,
      "loss": 0.0594,
      "reward": 1.4250000268220901,
      "reward_std": 0.37323956564068794,
      "rewards/accuracy_reward": 0.5250000040978193,
      "rewards/format_reward": 0.9000000208616257,
      "step": 627
    },
    {
      "completion_length": 320.20000076293945,
      "epoch": 0.6511145671332297,
      "grad_norm": 5.313282489776611,
      "kl": 1.0943603515625,
      "learning_rate": 2.876818878375509e-06,
      "loss": 0.2469,
      "reward": 1.2750000208616257,
      "reward_std": 0.4415663853287697,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9000000059604645,
      "step": 628
    },
    {
      "completion_length": 197.725004196167,
      "epoch": 0.6521513737687922,
      "grad_norm": 2.7158896923065186,
      "kl": 1.244873046875,
      "learning_rate": 2.8760991365021856e-06,
      "loss": 0.2859,
      "reward": 1.4250000268220901,
      "reward_std": 0.5048274546861649,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.925000011920929,
      "step": 629
    },
    {
      "completion_length": 232.45000457763672,
      "epoch": 0.6531881804043546,
      "grad_norm": 16.022092819213867,
      "kl": 4.732421875,
      "learning_rate": 2.8753773886108405e-06,
      "loss": 0.3514,
      "reward": 1.6500000059604645,
      "reward_std": 0.44533541798591614,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 630
    },
    {
      "completion_length": 307.2250061035156,
      "epoch": 0.654224987039917,
      "grad_norm": 62.19873809814453,
      "kl": 17.984130859375,
      "learning_rate": 2.874653635753605e-06,
      "loss": 1.3583,
      "reward": 1.4250000417232513,
      "reward_std": 0.4525563195347786,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 631
    },
    {
      "completion_length": 149.2000026702881,
      "epoch": 0.6552617936754795,
      "grad_norm": 0.5755558013916016,
      "kl": 0.265869140625,
      "learning_rate": 2.8739278789855362e-06,
      "loss": -0.005,
      "reward": 1.3250000327825546,
      "reward_std": 0.40387552231550217,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 632
    },
    {
      "completion_length": 168.62499809265137,
      "epoch": 0.656298600311042,
      "grad_norm": 0.7181000113487244,
      "kl": 0.38720703125,
      "learning_rate": 2.8732001193646104e-06,
      "loss": 0.078,
      "reward": 1.300000011920929,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.32500001043081284,
      "rewards/format_reward": 0.9750000089406967,
      "step": 633
    },
    {
      "completion_length": 253.22500610351562,
      "epoch": 0.6573354069466044,
      "grad_norm": 77.6446304321289,
      "kl": 20.80859375,
      "learning_rate": 2.872470357951724e-06,
      "loss": 1.654,
      "reward": 1.225000038743019,
      "reward_std": 0.3245587572455406,
      "rewards/accuracy_reward": 0.2999999988824129,
      "rewards/format_reward": 0.925000011920929,
      "step": 634
    },
    {
      "completion_length": 251.20000076293945,
      "epoch": 0.6583722135821669,
      "grad_norm": 1.6625399589538574,
      "kl": 1.81103515625,
      "learning_rate": 2.871738595810692e-06,
      "loss": 0.212,
      "reward": 1.3999999910593033,
      "reward_std": 0.2686570808291435,
      "rewards/accuracy_reward": 0.47500001825392246,
      "rewards/format_reward": 0.925000011920929,
      "step": 635
    },
    {
      "completion_length": 499.15001678466797,
      "epoch": 0.6594090202177294,
      "grad_norm": 8.252129554748535,
      "kl": 2.3154296875,
      "learning_rate": 2.871004834008246e-06,
      "loss": 0.3376,
      "reward": 1.0500000417232513,
      "reward_std": 0.597340852022171,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.7750000208616257,
      "step": 636
    },
    {
      "completion_length": 207.90000534057617,
      "epoch": 0.6604458268532919,
      "grad_norm": 2.2362051010131836,
      "kl": 0.253662109375,
      "learning_rate": 2.8702690736140318e-06,
      "loss": 0.0666,
      "reward": 1.550000011920929,
      "reward_std": 0.43636220321059227,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 637
    },
    {
      "completion_length": 230.37500381469727,
      "epoch": 0.6614826334888543,
      "grad_norm": 1.240039587020874,
      "kl": 0.55322265625,
      "learning_rate": 2.8695313157006097e-06,
      "loss": 0.1479,
      "reward": 1.3250000029802322,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.8750000149011612,
      "step": 638
    },
    {
      "completion_length": 283.0250053405762,
      "epoch": 0.6625194401244168,
      "grad_norm": 3.555227041244507,
      "kl": 0.6796875,
      "learning_rate": 2.868791561343451e-06,
      "loss": 0.16,
      "reward": 1.475000023841858,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9000000059604645,
      "step": 639
    },
    {
      "completion_length": 114.42500305175781,
      "epoch": 0.6635562467599793,
      "grad_norm": 0.4519474506378174,
      "kl": 0.255126953125,
      "learning_rate": 2.868049811620938e-06,
      "loss": 0.0198,
      "reward": 1.6500000059604645,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6500000022351742,
      "rewards/format_reward": 1.0,
      "step": 640
    },
    {
      "completion_length": 236.2750015258789,
      "epoch": 0.6645930533955418,
      "grad_norm": 7.132613182067871,
      "kl": 4.499267578125,
      "learning_rate": 2.8673060676143616e-06,
      "loss": 0.4182,
      "reward": 1.4000000059604645,
      "reward_std": 0.620261438190937,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8500000238418579,
      "step": 641
    },
    {
      "completion_length": 224.92499923706055,
      "epoch": 0.6656298600311042,
      "grad_norm": 40.20530319213867,
      "kl": 15.702392578125,
      "learning_rate": 2.8665603304079197e-06,
      "loss": 1.1686,
      "reward": 1.0250000059604645,
      "reward_std": 0.3731010742485523,
      "rewards/accuracy_reward": 0.20000000298023224,
      "rewards/format_reward": 0.8250000029802322,
      "step": 642
    },
    {
      "completion_length": 186.72500038146973,
      "epoch": 0.6666666666666666,
      "grad_norm": 0.7823323011398315,
      "kl": 0.713623046875,
      "learning_rate": 2.865812601088715e-06,
      "loss": 0.154,
      "reward": 1.5500000417232513,
      "reward_std": 0.41643913835287094,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 643
    },
    {
      "completion_length": 192.20000457763672,
      "epoch": 0.6677034733022291,
      "grad_norm": 1.6138575077056885,
      "kl": 1.10498046875,
      "learning_rate": 2.865062880746756e-06,
      "loss": 0.1178,
      "reward": 1.5750000476837158,
      "reward_std": 0.2937843129038811,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 644
    },
    {
      "completion_length": 202.75000381469727,
      "epoch": 0.6687402799377916,
      "grad_norm": 1.0543968677520752,
      "kl": 1.532470703125,
      "learning_rate": 2.864311170474953e-06,
      "loss": 0.1468,
      "reward": 1.350000023841858,
      "reward_std": 0.29270511120557785,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.925000011920929,
      "step": 645
    },
    {
      "completion_length": 143.85000038146973,
      "epoch": 0.6697770865733541,
      "grad_norm": 0.39155247807502747,
      "kl": 0.27880859375,
      "learning_rate": 2.8635574713691158e-06,
      "loss": 0.017,
      "reward": 1.350000023841858,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.37500000186264515,
      "rewards/format_reward": 0.9750000089406967,
      "step": 646
    },
    {
      "completion_length": 149.47500228881836,
      "epoch": 0.6708138932089165,
      "grad_norm": 5.471457481384277,
      "kl": 2.15234375,
      "learning_rate": 2.8628017845279553e-06,
      "loss": 0.1969,
      "reward": 1.4250000268220901,
      "reward_std": 0.4415663853287697,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.9000000208616257,
      "step": 647
    },
    {
      "completion_length": 180.02500534057617,
      "epoch": 0.671850699844479,
      "grad_norm": 1.4774538278579712,
      "kl": 1.03564453125,
      "learning_rate": 2.862044111053079e-06,
      "loss": 0.1221,
      "reward": 1.350000038743019,
      "reward_std": 0.4291412830352783,
      "rewards/accuracy_reward": 0.4500000048428774,
      "rewards/format_reward": 0.9000000208616257,
      "step": 648
    },
    {
      "completion_length": 247.97500228881836,
      "epoch": 0.6728875064800415,
      "grad_norm": 0.93613600730896,
      "kl": 0.880126953125,
      "learning_rate": 2.8612844520489902e-06,
      "loss": 0.1469,
      "reward": 1.2000000178813934,
      "reward_std": 0.4129471369087696,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.9000000208616257,
      "step": 649
    },
    {
      "completion_length": 251.92500686645508,
      "epoch": 0.673924313115604,
      "grad_norm": 2.6798095703125,
      "kl": 1.1048583984375,
      "learning_rate": 2.8605228086230877e-06,
      "loss": 0.211,
      "reward": 1.475000023841858,
      "reward_std": 0.391311913728714,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9250000268220901,
      "step": 650
    },
    {
      "completion_length": 218.125,
      "epoch": 0.6749611197511665,
      "grad_norm": 2.4933316707611084,
      "kl": 1.405029296875,
      "learning_rate": 2.8597591818856625e-06,
      "loss": 0.18,
      "reward": 1.4500000178813934,
      "reward_std": 0.45977719500660896,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.875,
      "step": 651
    },
    {
      "completion_length": 183.3750057220459,
      "epoch": 0.6759979263867288,
      "grad_norm": 0.8898698091506958,
      "kl": 1.068603515625,
      "learning_rate": 2.858993572949896e-06,
      "loss": 0.1626,
      "reward": 1.6250000298023224,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 652
    },
    {
      "completion_length": 262.07501220703125,
      "epoch": 0.6770347330222913,
      "grad_norm": 2.555324077606201,
      "kl": 3.16259765625,
      "learning_rate": 2.8582259829318604e-06,
      "loss": 0.3425,
      "reward": 1.5000000149011612,
      "reward_std": 0.36053744331002235,
      "rewards/accuracy_reward": 0.6250000186264515,
      "rewards/format_reward": 0.8750000149011612,
      "step": 653
    },
    {
      "completion_length": 177.27500534057617,
      "epoch": 0.6780715396578538,
      "grad_norm": 1.2724261283874512,
      "kl": 2.198974609375,
      "learning_rate": 2.8574564129505154e-06,
      "loss": 0.2648,
      "reward": 1.3250000178813934,
      "reward_std": 0.452556312084198,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 0.9250000268220901,
      "step": 654
    },
    {
      "completion_length": 144.27500343322754,
      "epoch": 0.6791083462934163,
      "grad_norm": 0.3718305826187134,
      "kl": 0.251220703125,
      "learning_rate": 2.856684864127707e-06,
      "loss": 0.0359,
      "reward": 1.5000000298023224,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 1.0,
      "step": 655
    },
    {
      "completion_length": 239.875,
      "epoch": 0.6801451529289787,
      "grad_norm": 1.6971384286880493,
      "kl": 2.00830078125,
      "learning_rate": 2.8559113375881657e-06,
      "loss": 0.298,
      "reward": 1.0250000357627869,
      "reward_std": 0.31199515238404274,
      "rewards/accuracy_reward": 0.15000000223517418,
      "rewards/format_reward": 0.8750000149011612,
      "step": 656
    },
    {
      "completion_length": 223.1500015258789,
      "epoch": 0.6811819595645412,
      "grad_norm": 1.9796103239059448,
      "kl": 3.091796875,
      "learning_rate": 2.855135834459505e-06,
      "loss": 0.23,
      "reward": 1.5000000298023224,
      "reward_std": 0.546148881316185,
      "rewards/accuracy_reward": 0.5999999940395355,
      "rewards/format_reward": 0.9000000208616257,
      "step": 657
    },
    {
      "completion_length": 201.97500228881836,
      "epoch": 0.6822187662001037,
      "grad_norm": 1.6794363260269165,
      "kl": 1.33544921875,
      "learning_rate": 2.85435835587222e-06,
      "loss": 0.1351,
      "reward": 1.675000011920929,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 658
    },
    {
      "completion_length": 144.45000076293945,
      "epoch": 0.6832555728356662,
      "grad_norm": 0.49781835079193115,
      "kl": 0.26416015625,
      "learning_rate": 2.8535789029596854e-06,
      "loss": -0.0649,
      "reward": 1.324999988079071,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 1.0,
      "step": 659
    },
    {
      "completion_length": 242.22500038146973,
      "epoch": 0.6842923794712286,
      "grad_norm": 0.9583486318588257,
      "kl": 1.5,
      "learning_rate": 2.8527974768581547e-06,
      "loss": 0.1847,
      "reward": 1.4250000417232513,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9250000268220901,
      "step": 660
    },
    {
      "completion_length": 209.17499923706055,
      "epoch": 0.685329186106791,
      "grad_norm": 1.6011143922805786,
      "kl": 0.8896484375,
      "learning_rate": 2.8520140787067567e-06,
      "loss": 0.1899,
      "reward": 1.475000038743019,
      "reward_std": 0.4399927146732807,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9000000357627869,
      "step": 661
    },
    {
      "completion_length": 249.95000076293945,
      "epoch": 0.6863659927423535,
      "grad_norm": 1.6479218006134033,
      "kl": 0.660888671875,
      "learning_rate": 2.8512287096474955e-06,
      "loss": 0.0907,
      "reward": 1.375,
      "reward_std": 0.3731010630726814,
      "rewards/accuracy_reward": 0.42500000819563866,
      "rewards/format_reward": 0.9500000178813934,
      "step": 662
    },
    {
      "completion_length": 138.65000343322754,
      "epoch": 0.687402799377916,
      "grad_norm": 0.9126381278038025,
      "kl": 0.72998046875,
      "learning_rate": 2.8504413708252483e-06,
      "loss": 0.0053,
      "reward": 1.3750000149011612,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.4000000152736902,
      "rewards/format_reward": 0.9750000089406967,
      "step": 663
    },
    {
      "completion_length": 162.7750015258789,
      "epoch": 0.6884396060134785,
      "grad_norm": 0.6816839575767517,
      "kl": 1.11083984375,
      "learning_rate": 2.8496520633877635e-06,
      "loss": 0.1068,
      "reward": 1.600000023841858,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 664
    },
    {
      "completion_length": 163.40000534057617,
      "epoch": 0.6894764126490409,
      "grad_norm": 0.30915480852127075,
      "kl": 0.21044921875,
      "learning_rate": 2.84886078848566e-06,
      "loss": 0.0071,
      "reward": 1.574999988079071,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 665
    },
    {
      "completion_length": 138.15000534057617,
      "epoch": 0.6905132192846034,
      "grad_norm": 0.2813793420791626,
      "kl": 0.23046875,
      "learning_rate": 2.8480675472724236e-06,
      "loss": 0.0338,
      "reward": 1.6250000298023224,
      "reward_std": 0.1369306445121765,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 1.0,
      "step": 666
    },
    {
      "completion_length": 244.97500228881836,
      "epoch": 0.6915500259201659,
      "grad_norm": 8.316036224365234,
      "kl": 6.7216796875,
      "learning_rate": 2.847272340904408e-06,
      "loss": 0.6203,
      "reward": 1.4500000476837158,
      "reward_std": 0.5163118988275528,
      "rewards/accuracy_reward": 0.5750000141561031,
      "rewards/format_reward": 0.8750000298023224,
      "step": 667
    },
    {
      "completion_length": 216.9250030517578,
      "epoch": 0.6925868325557284,
      "grad_norm": 0.797091007232666,
      "kl": 1.13037109375,
      "learning_rate": 2.8464751705408294e-06,
      "loss": 0.1587,
      "reward": 1.2500000149011612,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.9500000178813934,
      "step": 668
    },
    {
      "completion_length": 279.1749973297119,
      "epoch": 0.6936236391912908,
      "grad_norm": 4.413926601409912,
      "kl": 4.05224609375,
      "learning_rate": 2.84567603734377e-06,
      "loss": 0.4332,
      "reward": 1.4500000178813934,
      "reward_std": 0.4525563158094883,
      "rewards/accuracy_reward": 0.5250000171363354,
      "rewards/format_reward": 0.9250000268220901,
      "step": 669
    },
    {
      "completion_length": 322.7249984741211,
      "epoch": 0.6946604458268533,
      "grad_norm": 2.120075225830078,
      "kl": 5.30810546875,
      "learning_rate": 2.8448749424781704e-06,
      "loss": 0.3886,
      "reward": 1.2750000208616257,
      "reward_std": 0.4958944171667099,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.8500000089406967,
      "step": 670
    },
    {
      "completion_length": 299.50000762939453,
      "epoch": 0.6956972524624158,
      "grad_norm": 0.918604850769043,
      "kl": 0.921630859375,
      "learning_rate": 2.844071887111833e-06,
      "loss": 0.196,
      "reward": 1.725000023841858,
      "reward_std": 0.3354102149605751,
      "rewards/accuracy_reward": 0.8000000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 671
    },
    {
      "completion_length": 267.4250068664551,
      "epoch": 0.6967340590979783,
      "grad_norm": 1.9089648723602295,
      "kl": 1.540283203125,
      "learning_rate": 2.843266872415417e-06,
      "loss": 0.193,
      "reward": 1.4500000178813934,
      "reward_std": 0.3354101926088333,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.925000011920929,
      "step": 672
    },
    {
      "completion_length": 206.4250030517578,
      "epoch": 0.6977708657335407,
      "grad_norm": 0.4940345883369446,
      "kl": 0.331787109375,
      "learning_rate": 2.8424598995624386e-06,
      "loss": 0.0692,
      "reward": 1.4250000417232513,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 673
    },
    {
      "completion_length": 187.12500381469727,
      "epoch": 0.6988076723691031,
      "grad_norm": 0.8684625029563904,
      "kl": 0.30908203125,
      "learning_rate": 2.8416509697292676e-06,
      "loss": 0.1091,
      "reward": 1.5000000298023224,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 674
    },
    {
      "completion_length": 191.79999923706055,
      "epoch": 0.6998444790046656,
      "grad_norm": 1.0280557870864868,
      "kl": 1.33935546875,
      "learning_rate": 2.8408400840951266e-06,
      "loss": 0.1251,
      "reward": 1.6250000298023224,
      "reward_std": 0.3479738049209118,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9500000178813934,
      "step": 675
    },
    {
      "completion_length": 267.15000915527344,
      "epoch": 0.7008812856402281,
      "grad_norm": 0.3398536741733551,
      "kl": 0.2470703125,
      "learning_rate": 2.8400272438420907e-06,
      "loss": 0.033,
      "reward": 1.2750000208616257,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 676
    },
    {
      "completion_length": 391.2000045776367,
      "epoch": 0.7019180922757906,
      "grad_norm": 1.9886372089385986,
      "kl": 1.83837890625,
      "learning_rate": 2.8392124501550828e-06,
      "loss": 0.3499,
      "reward": 1.1500000208616257,
      "reward_std": 0.5283810868859291,
      "rewards/accuracy_reward": 0.2749999985098839,
      "rewards/format_reward": 0.8750000298023224,
      "step": 677
    },
    {
      "completion_length": 174.3000030517578,
      "epoch": 0.702954898911353,
      "grad_norm": 0.7306681275367737,
      "kl": 0.58935546875,
      "learning_rate": 2.838395704221874e-06,
      "loss": 0.0423,
      "reward": 1.7249999940395355,
      "reward_std": 0.3533165641129017,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 678
    },
    {
      "completion_length": 258.9750099182129,
      "epoch": 0.7039917055469155,
      "grad_norm": 10.755302429199219,
      "kl": 6.359130859375,
      "learning_rate": 2.8375770072330817e-06,
      "loss": 0.3311,
      "reward": 1.3250000029802322,
      "reward_std": 0.2722875848412514,
      "rewards/accuracy_reward": 0.4500000011175871,
      "rewards/format_reward": 0.875,
      "step": 679
    },
    {
      "completion_length": 271.7000102996826,
      "epoch": 0.705028512182478,
      "grad_norm": 1.0601376295089722,
      "kl": 1.66357421875,
      "learning_rate": 2.8367563603821675e-06,
      "loss": 0.2495,
      "reward": 1.4750000536441803,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 680
    },
    {
      "completion_length": 313.95000076293945,
      "epoch": 0.7060653188180405,
      "grad_norm": 1.3850581645965576,
      "kl": 1.41650390625,
      "learning_rate": 2.8359337648654334e-06,
      "loss": 0.219,
      "reward": 1.3750000298023224,
      "reward_std": 0.3913118839263916,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.9000000208616257,
      "step": 681
    },
    {
      "completion_length": 259.2250061035156,
      "epoch": 0.7071021254536028,
      "grad_norm": 1.035261631011963,
      "kl": 1.794677734375,
      "learning_rate": 2.8351092218820252e-06,
      "loss": 0.2137,
      "reward": 1.625,
      "reward_std": 0.4290027618408203,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.9500000029802322,
      "step": 682
    },
    {
      "completion_length": 289.79999923706055,
      "epoch": 0.7081389320891653,
      "grad_norm": 2.4445176124572754,
      "kl": 0.818115234375,
      "learning_rate": 2.8342827326339255e-06,
      "loss": 0.0812,
      "reward": 1.175000011920929,
      "reward_std": 0.31733785569667816,
      "rewards/accuracy_reward": 0.2500000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 683
    },
    {
      "completion_length": 278.17499923706055,
      "epoch": 0.7091757387247278,
      "grad_norm": 0.5835633873939514,
      "kl": 0.353759765625,
      "learning_rate": 2.833454298325955e-06,
      "loss": 0.0398,
      "reward": 1.550000011920929,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9500000029802322,
      "step": 684
    },
    {
      "completion_length": 190.80000686645508,
      "epoch": 0.7102125453602903,
      "grad_norm": 1.9967930316925049,
      "kl": 0.5185546875,
      "learning_rate": 2.832623920165769e-06,
      "loss": 0.1104,
      "reward": 1.7750000357627869,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 685
    },
    {
      "completion_length": 240.6250114440918,
      "epoch": 0.7112493519958528,
      "grad_norm": 2.839909315109253,
      "kl": 0.76611328125,
      "learning_rate": 2.831791599363858e-06,
      "loss": 0.2053,
      "reward": 1.2750000208616257,
      "reward_std": 0.3245587721467018,
      "rewards/accuracy_reward": 0.32499999925494194,
      "rewards/format_reward": 0.9500000029802322,
      "step": 686
    },
    {
      "completion_length": 175.7750015258789,
      "epoch": 0.7122861586314152,
      "grad_norm": 0.2983464002609253,
      "kl": 0.295166015625,
      "learning_rate": 2.8309573371335426e-06,
      "loss": 0.0193,
      "reward": 1.675000011920929,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.6750000007450581,
      "rewards/format_reward": 1.0,
      "step": 687
    },
    {
      "completion_length": 150.92499923706055,
      "epoch": 0.7133229652669777,
      "grad_norm": 0.7637765407562256,
      "kl": 0.322998046875,
      "learning_rate": 2.830121134690974e-06,
      "loss": 0.0776,
      "reward": 1.3250000178813934,
      "reward_std": 0.1928323358297348,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.9750000089406967,
      "step": 688
    },
    {
      "completion_length": 268.8000011444092,
      "epoch": 0.7143597719025402,
      "grad_norm": 0.7986348271369934,
      "kl": 1.36376953125,
      "learning_rate": 2.8292829932551333e-06,
      "loss": 0.2448,
      "reward": 1.375,
      "reward_std": 0.2812207192182541,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.925000011920929,
      "step": 689
    },
    {
      "completion_length": 245.72500228881836,
      "epoch": 0.7153965785381027,
      "grad_norm": 0.7799951434135437,
      "kl": 0.99072265625,
      "learning_rate": 2.8284429140478255e-06,
      "loss": 0.1347,
      "reward": 1.475000023841858,
      "reward_std": 0.47234082594513893,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 690
    },
    {
      "completion_length": 282.9250068664551,
      "epoch": 0.7164333851736651,
      "grad_norm": 0.9315149784088135,
      "kl": 1.8046875,
      "learning_rate": 2.827600898293683e-06,
      "loss": 0.2766,
      "reward": 1.2500000149011612,
      "reward_std": 0.27950849011540413,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9000000208616257,
      "step": 691
    },
    {
      "completion_length": 182.22500228881836,
      "epoch": 0.7174701918092276,
      "grad_norm": 24.62877655029297,
      "kl": 7.6640625,
      "learning_rate": 2.82675694722016e-06,
      "loss": 0.6286,
      "reward": 1.3250000029802322,
      "reward_std": 0.20019175857305527,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.9500000178813934,
      "step": 692
    },
    {
      "completion_length": 183.7250099182129,
      "epoch": 0.71850699844479,
      "grad_norm": 0.7584772109985352,
      "kl": 1.05224609375,
      "learning_rate": 2.8259110620575313e-06,
      "loss": 0.1729,
      "reward": 1.3250000178813934,
      "reward_std": 0.3245587646961212,
      "rewards/accuracy_reward": 0.37499999813735485,
      "rewards/format_reward": 0.9500000178813934,
      "step": 693
    },
    {
      "completion_length": 123.92500305175781,
      "epoch": 0.7195438050803525,
      "grad_norm": 0.2979358434677124,
      "kl": 0.254150390625,
      "learning_rate": 2.8250632440388927e-06,
      "loss": -0.0232,
      "reward": 1.7000000178813934,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 1.0,
      "step": 694
    },
    {
      "completion_length": 223.82501220703125,
      "epoch": 0.720580611715915,
      "grad_norm": 1.2255027294158936,
      "kl": 1.716552734375,
      "learning_rate": 2.8242134944001558e-06,
      "loss": 0.2352,
      "reward": 1.625,
      "reward_std": 0.2795085050165653,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.9500000178813934,
      "step": 695
    },
    {
      "completion_length": 152.20000076293945,
      "epoch": 0.7216174183514774,
      "grad_norm": 0.9140526652336121,
      "kl": 0.26806640625,
      "learning_rate": 2.82336181438005e-06,
      "loss": 0.0913,
      "reward": 1.675000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 696
    },
    {
      "completion_length": 134.87500381469727,
      "epoch": 0.7226542249870399,
      "grad_norm": 0.4066285192966461,
      "kl": 0.310791015625,
      "learning_rate": 2.822508205220117e-06,
      "loss": -0.0062,
      "reward": 1.5750000476837158,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 1.0,
      "step": 697
    },
    {
      "completion_length": 226.60000038146973,
      "epoch": 0.7236910316226024,
      "grad_norm": 0.5627696514129639,
      "kl": 1.79150390625,
      "learning_rate": 2.8216526681647118e-06,
      "loss": 0.0035,
      "reward": 1.8250000178813934,
      "reward_std": 0.2487340308725834,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9500000029802322,
      "step": 698
    },
    {
      "completion_length": 133.60000038146973,
      "epoch": 0.7247278382581649,
      "grad_norm": 0.7156115174293518,
      "kl": 0.302001953125,
      "learning_rate": 2.820795204460999e-06,
      "loss": 0.0222,
      "reward": 1.4749999940395355,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 699
    },
    {
      "completion_length": 123.40000343322754,
      "epoch": 0.7257646448937273,
      "grad_norm": 0.44991451501846313,
      "kl": 0.252685546875,
      "learning_rate": 2.819935815358953e-06,
      "loss": -0.0084,
      "reward": 1.425000011920929,
      "reward_std": 0.18026870489120483,
      "rewards/accuracy_reward": 0.42499999701976776,
      "rewards/format_reward": 1.0,
      "step": 700
    },
    {
      "completion_length": 330.52499771118164,
      "epoch": 0.7268014515292898,
      "grad_norm": 3.5027003288269043,
      "kl": 1.857177734375,
      "learning_rate": 2.819074502111354e-06,
      "loss": 0.2354,
      "reward": 1.5000000149011612,
      "reward_std": 0.41643912717700005,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.875,
      "step": 701
    },
    {
      "completion_length": 294.82500076293945,
      "epoch": 0.7278382581648523,
      "grad_norm": 34.1741828918457,
      "kl": 10.48681640625,
      "learning_rate": 2.818211265973787e-06,
      "loss": 0.4339,
      "reward": 1.2000000178813934,
      "reward_std": 0.32818929105997086,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.8750000149011612,
      "step": 702
    },
    {
      "completion_length": 178.725004196167,
      "epoch": 0.7288750648004148,
      "grad_norm": 0.9051024913787842,
      "kl": 0.39794921875,
      "learning_rate": 2.8173461082046407e-06,
      "loss": 0.0751,
      "reward": 1.3000000417232513,
      "reward_std": 0.24151314049959183,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 703
    },
    {
      "completion_length": 159.60000610351562,
      "epoch": 0.7299118714359772,
      "grad_norm": 1.2098740339279175,
      "kl": 0.4189453125,
      "learning_rate": 2.8164790300651042e-06,
      "loss": 0.0657,
      "reward": 1.5000000298023224,
      "reward_std": 0.41643914952874184,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.9750000089406967,
      "step": 704
    },
    {
      "completion_length": 143.62500190734863,
      "epoch": 0.7309486780715396,
      "grad_norm": 0.858209490776062,
      "kl": 1.26513671875,
      "learning_rate": 2.815610032819168e-06,
      "loss": 0.1172,
      "reward": 1.475000023841858,
      "reward_std": 0.31199515610933304,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 705
    },
    {
      "completion_length": 216.0250072479248,
      "epoch": 0.7319854847071021,
      "grad_norm": 0.6318058967590332,
      "kl": 0.768798828125,
      "learning_rate": 2.8147391177336176e-06,
      "loss": 0.0993,
      "reward": 1.7000000029802322,
      "reward_std": 0.1604842059314251,
      "rewards/accuracy_reward": 0.7500000093132257,
      "rewards/format_reward": 0.9500000029802322,
      "step": 706
    },
    {
      "completion_length": 224.87500381469727,
      "epoch": 0.7330222913426646,
      "grad_norm": 518.7442016601562,
      "kl": 119.13671875,
      "learning_rate": 2.8138662860780356e-06,
      "loss": 4.7663,
      "reward": 1.2500000149011612,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.925000011920929,
      "step": 707
    },
    {
      "completion_length": 115.35000038146973,
      "epoch": 0.7340590979782271,
      "grad_norm": 0.6044578552246094,
      "kl": 0.365234375,
      "learning_rate": 2.8129915391247988e-06,
      "loss": 0.0126,
      "reward": 1.4749999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 1.0,
      "step": 708
    },
    {
      "completion_length": 186.9250087738037,
      "epoch": 0.7350959046137895,
      "grad_norm": 2.9357755184173584,
      "kl": 1.806884765625,
      "learning_rate": 2.8121148781490754e-06,
      "loss": 0.1944,
      "reward": 1.4250000268220901,
      "reward_std": 0.22894952073693275,
      "rewards/accuracy_reward": 0.4749999940395355,
      "rewards/format_reward": 0.9500000178813934,
      "step": 709
    },
    {
      "completion_length": 138.32500076293945,
      "epoch": 0.736132711249352,
      "grad_norm": 0.2896471321582794,
      "kl": 0.189453125,
      "learning_rate": 2.8112363044288243e-06,
      "loss": 0.0259,
      "reward": 1.5250000059604645,
      "reward_std": 0.12436702102422714,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.9750000089406967,
      "step": 710
    },
    {
      "completion_length": 258.32500648498535,
      "epoch": 0.7371695178849145,
      "grad_norm": 18.005449295043945,
      "kl": 4.0224609375,
      "learning_rate": 2.810355819244792e-06,
      "loss": 0.4517,
      "reward": 1.350000023841858,
      "reward_std": 0.3407529331743717,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.925000011920929,
      "step": 711
    },
    {
      "completion_length": 114.47500228881836,
      "epoch": 0.738206324520477,
      "grad_norm": 0.36921098828315735,
      "kl": 0.2666015625,
      "learning_rate": 2.809473423880512e-06,
      "loss": 0.0209,
      "reward": 1.675000011920929,
      "reward_std": 0.12436702102422714,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 1.0,
      "step": 712
    },
    {
      "completion_length": 199.05000686645508,
      "epoch": 0.7392431311560393,
      "grad_norm": 0.9610119462013245,
      "kl": 1.69140625,
      "learning_rate": 2.8085891196223025e-06,
      "loss": 0.1894,
      "reward": 1.1500000208616257,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.20000000670552254,
      "rewards/format_reward": 0.9500000178813934,
      "step": 713
    },
    {
      "completion_length": 154.62500381469727,
      "epoch": 0.7402799377916018,
      "grad_norm": 1.1751526594161987,
      "kl": 0.3271484375,
      "learning_rate": 2.8077029077592633e-06,
      "loss": 0.0186,
      "reward": 1.225000023841858,
      "reward_std": 0.30463574081659317,
      "rewards/accuracy_reward": 0.27500000409781933,
      "rewards/format_reward": 0.9500000178813934,
      "step": 714
    },
    {
      "completion_length": 209.70000839233398,
      "epoch": 0.7413167444271643,
      "grad_norm": 1.0980820655822754,
      "kl": 0.487548828125,
      "learning_rate": 2.806814789583277e-06,
      "loss": 0.1487,
      "reward": 1.5250000059604645,
      "reward_std": 0.2487340308725834,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 715
    },
    {
      "completion_length": 227.0000114440918,
      "epoch": 0.7423535510627268,
      "grad_norm": 0.5780226588249207,
      "kl": 1.57666015625,
      "learning_rate": 2.8059247663890032e-06,
      "loss": 0.2161,
      "reward": 1.5250000059604645,
      "reward_std": 0.18026871606707573,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9249999970197678,
      "step": 716
    },
    {
      "completion_length": 280.20000076293945,
      "epoch": 0.7433903576982893,
      "grad_norm": 1.4871811866760254,
      "kl": 1.12109375,
      "learning_rate": 2.8050328394738803e-06,
      "loss": 0.2866,
      "reward": 1.4250000417232513,
      "reward_std": 0.4729737937450409,
      "rewards/accuracy_reward": 0.5500000081956387,
      "rewards/format_reward": 0.8750000149011612,
      "step": 717
    },
    {
      "completion_length": 310.9500045776367,
      "epoch": 0.7444271643338517,
      "grad_norm": 1.3770198822021484,
      "kl": 0.87255859375,
      "learning_rate": 2.8041390101381195e-06,
      "loss": 0.229,
      "reward": 1.4000000059604645,
      "reward_std": 0.6291945092380047,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.8750000149011612,
      "step": 718
    },
    {
      "completion_length": 225.77500343322754,
      "epoch": 0.7454639709694142,
      "grad_norm": 0.8252406120300293,
      "kl": 1.41943359375,
      "learning_rate": 2.8032432796847076e-06,
      "loss": 0.2017,
      "reward": 1.4249999970197678,
      "reward_std": 0.21275536343455315,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 719
    },
    {
      "completion_length": 256.90000343322754,
      "epoch": 0.7465007776049767,
      "grad_norm": 0.9165952205657959,
      "kl": 1.5322265625,
      "learning_rate": 2.802345649419402e-06,
      "loss": 0.2268,
      "reward": 1.375,
      "reward_std": 0.329762976616621,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.8750000149011612,
      "step": 720
    },
    {
      "completion_length": 112.55000305175781,
      "epoch": 0.7475375842405392,
      "grad_norm": 0.8553647398948669,
      "kl": 0.322265625,
      "learning_rate": 2.8014461206507292e-06,
      "loss": -0.0239,
      "reward": 1.5250000357627869,
      "reward_std": 0.30463573709130287,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 721
    },
    {
      "completion_length": 308.075008392334,
      "epoch": 0.7485743908761016,
      "grad_norm": 2.7667524814605713,
      "kl": 3.8203125,
      "learning_rate": 2.800544694689983e-06,
      "loss": 0.3563,
      "reward": 1.2750000208616257,
      "reward_std": 0.4604102000594139,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.8750000149011612,
      "step": 722
    },
    {
      "completion_length": 280.6499996185303,
      "epoch": 0.749611197511664,
      "grad_norm": 6.384510517120361,
      "kl": 3.0322265625,
      "learning_rate": 2.7996413728512236e-06,
      "loss": 0.3557,
      "reward": 1.4500000029802322,
      "reward_std": 0.26129764690995216,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9000000059604645,
      "step": 723
    },
    {
      "completion_length": 279.8000068664551,
      "epoch": 0.7506480041472265,
      "grad_norm": 7.1993584632873535,
      "kl": 2.51171875,
      "learning_rate": 2.7987361564512757e-06,
      "loss": 0.354,
      "reward": 1.3250000178813934,
      "reward_std": 0.38409097865223885,
      "rewards/accuracy_reward": 0.4249999988824129,
      "rewards/format_reward": 0.9000000357627869,
      "step": 724
    },
    {
      "completion_length": 149.2750015258789,
      "epoch": 0.751684810782789,
      "grad_norm": 3.0475800037384033,
      "kl": 2.4169921875,
      "learning_rate": 2.797829046809723e-06,
      "loss": 0.1504,
      "reward": 1.300000011920929,
      "reward_std": 0.36053744331002235,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.8750000149011612,
      "step": 725
    },
    {
      "completion_length": 177.97500228881836,
      "epoch": 0.7527216174183515,
      "grad_norm": 0.7727159857749939,
      "kl": 0.75732421875,
      "learning_rate": 2.7969200452489117e-06,
      "loss": 0.1334,
      "reward": 1.574999988079071,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9750000089406967,
      "step": 726
    },
    {
      "completion_length": 162.05000495910645,
      "epoch": 0.7537584240539139,
      "grad_norm": 1.6836689710617065,
      "kl": 1.5498046875,
      "learning_rate": 2.7960091530939453e-06,
      "loss": 0.1498,
      "reward": 1.375,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 727
    },
    {
      "completion_length": 108.67500305175781,
      "epoch": 0.7547952306894764,
      "grad_norm": 0.8589237332344055,
      "kl": 0.76123046875,
      "learning_rate": 2.7950963716726833e-06,
      "loss": 0.0888,
      "reward": 1.3750000298023224,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9500000178813934,
      "step": 728
    },
    {
      "completion_length": 119.17500114440918,
      "epoch": 0.7558320373250389,
      "grad_norm": 0.9010735750198364,
      "kl": 0.44287109375,
      "learning_rate": 2.7941817023157385e-06,
      "loss": 0.0313,
      "reward": 1.600000023841858,
      "reward_std": 0.1928323470056057,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 729
    },
    {
      "completion_length": 121.82500076293945,
      "epoch": 0.7568688439606014,
      "grad_norm": 2.7615885734558105,
      "kl": 1.18115234375,
      "learning_rate": 2.7932651463564763e-06,
      "loss": 0.1171,
      "reward": 1.300000011920929,
      "reward_std": 0.1604841984808445,
      "rewards/accuracy_reward": 0.3249999973922968,
      "rewards/format_reward": 0.9750000089406967,
      "step": 730
    },
    {
      "completion_length": 102.30000305175781,
      "epoch": 0.7579056505961638,
      "grad_norm": 0.4340980350971222,
      "kl": 0.39697265625,
      "learning_rate": 2.792346705131013e-06,
      "loss": 0.0185,
      "reward": 1.2000000178813934,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.2000000048428774,
      "rewards/format_reward": 1.0,
      "step": 731
    },
    {
      "completion_length": 118.50000381469727,
      "epoch": 0.7589424572317263,
      "grad_norm": 0.6814145445823669,
      "kl": 0.67236328125,
      "learning_rate": 2.791426379978213e-06,
      "loss": 0.0499,
      "reward": 1.5250000357627869,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 732
    },
    {
      "completion_length": 139.77500534057617,
      "epoch": 0.7599792638672888,
      "grad_norm": 3.444451093673706,
      "kl": 2.74609375,
      "learning_rate": 2.7905041722396854e-06,
      "loss": 0.1425,
      "reward": 1.4000000059604645,
      "reward_std": 0.18026872351765633,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 733
    },
    {
      "completion_length": 429.12499809265137,
      "epoch": 0.7610160705028512,
      "grad_norm": 10.602200508117676,
      "kl": 9.31494140625,
      "learning_rate": 2.7895800832597865e-06,
      "loss": 0.7934,
      "reward": 1.3249999955296516,
      "reward_std": 0.3731010667979717,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.8250000104308128,
      "step": 734
    },
    {
      "completion_length": 126.12500381469727,
      "epoch": 0.7620528771384136,
      "grad_norm": 0.5543495416641235,
      "kl": 0.70166015625,
      "learning_rate": 2.788654114385612e-06,
      "loss": -0.0408,
      "reward": 1.4750000536441803,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 735
    },
    {
      "completion_length": 106.92499923706055,
      "epoch": 0.7630896837739761,
      "grad_norm": 0.33099156618118286,
      "kl": 0.531982421875,
      "learning_rate": 2.787726266967e-06,
      "loss": -0.0037,
      "reward": 1.675000011920929,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 736
    },
    {
      "completion_length": 266.87500381469727,
      "epoch": 0.7641264904095386,
      "grad_norm": 1.9715381860733032,
      "kl": 1.3359375,
      "learning_rate": 2.7867965423565273e-06,
      "loss": 0.2879,
      "reward": 1.3000000268220901,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.4000000096857548,
      "rewards/format_reward": 0.9000000357627869,
      "step": 737
    },
    {
      "completion_length": 204.62499809265137,
      "epoch": 0.7651632970451011,
      "grad_norm": 0.3641264736652374,
      "kl": 0.6787109375,
      "learning_rate": 2.785864941909505e-06,
      "loss": 0.0829,
      "reward": 1.4750000089406967,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9500000029802322,
      "step": 738
    },
    {
      "completion_length": 233.37500381469727,
      "epoch": 0.7662001036806636,
      "grad_norm": 1.9150782823562622,
      "kl": 1.84033203125,
      "learning_rate": 2.784931466983981e-06,
      "loss": 0.2618,
      "reward": 1.2750000208616257,
      "reward_std": 0.4038755036890507,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9000000208616257,
      "step": 739
    },
    {
      "completion_length": 168.02499961853027,
      "epoch": 0.767236910316226,
      "grad_norm": 23.784427642822266,
      "kl": 9.75,
      "learning_rate": 2.783996118940734e-06,
      "loss": 0.6841,
      "reward": 1.225000038743019,
      "reward_std": 0.3245587609708309,
      "rewards/accuracy_reward": 0.30000001192092896,
      "rewards/format_reward": 0.9250000268220901,
      "step": 740
    },
    {
      "completion_length": 413.52500915527344,
      "epoch": 0.7682737169517885,
      "grad_norm": 28.01406478881836,
      "kl": 10.80078125,
      "learning_rate": 2.7830588991432745e-06,
      "loss": 0.8084,
      "reward": 1.200000062584877,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.8500000238418579,
      "step": 741
    },
    {
      "completion_length": 134.600004196167,
      "epoch": 0.769310523587351,
      "grad_norm": 1.1930382251739502,
      "kl": 1.346923828125,
      "learning_rate": 2.7821198089578417e-06,
      "loss": 0.1408,
      "reward": 1.375,
      "reward_std": 0.1604842059314251,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 742
    },
    {
      "completion_length": 269.3500061035156,
      "epoch": 0.7703473302229135,
      "grad_norm": 3.615288496017456,
      "kl": 4.735107421875,
      "learning_rate": 2.7811788497533998e-06,
      "loss": 0.4038,
      "reward": 1.2500000298023224,
      "reward_std": 0.36053742095828056,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 0.8500000089406967,
      "step": 743
    },
    {
      "completion_length": 355.05001068115234,
      "epoch": 0.7713841368584758,
      "grad_norm": 23.248554229736328,
      "kl": 9.390625,
      "learning_rate": 2.780236022901639e-06,
      "loss": 0.8676,
      "reward": 1.225000038743019,
      "reward_std": 0.632825031876564,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.800000011920929,
      "step": 744
    },
    {
      "completion_length": 301.225004196167,
      "epoch": 0.7724209434940383,
      "grad_norm": 1.500929594039917,
      "kl": 1.21484375,
      "learning_rate": 2.7792913297769723e-06,
      "loss": 0.1665,
      "reward": 1.2750000208616257,
      "reward_std": 0.4604101926088333,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.875,
      "step": 745
    },
    {
      "completion_length": 242.7750015258789,
      "epoch": 0.7734577501296008,
      "grad_norm": 0.9008773565292358,
      "kl": 1.644775390625,
      "learning_rate": 2.7783447717565313e-06,
      "loss": 0.2006,
      "reward": 1.600000023841858,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9000000059604645,
      "step": 746
    },
    {
      "completion_length": 321.52500915527344,
      "epoch": 0.7744945567651633,
      "grad_norm": 4.788297176361084,
      "kl": 4.009765625,
      "learning_rate": 2.777396350220169e-06,
      "loss": 0.436,
      "reward": 1.1500000059604645,
      "reward_std": 0.3534550555050373,
      "rewards/accuracy_reward": 0.25000000186264515,
      "rewards/format_reward": 0.9000000357627869,
      "step": 747
    },
    {
      "completion_length": 274.0750045776367,
      "epoch": 0.7755313634007258,
      "grad_norm": 2.297362804412842,
      "kl": 1.6220703125,
      "learning_rate": 2.7764460665504522e-06,
      "loss": 0.245,
      "reward": 1.4250000268220901,
      "reward_std": 0.3245587684214115,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.925000011920929,
      "step": 748
    },
    {
      "completion_length": 185.92500495910645,
      "epoch": 0.7765681700362882,
      "grad_norm": 1.0662373304367065,
      "kl": 1.704345703125,
      "learning_rate": 2.775493922132664e-06,
      "loss": 0.16,
      "reward": 1.3750000298023224,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9500000029802322,
      "step": 749
    },
    {
      "completion_length": 183.27500343322754,
      "epoch": 0.7776049766718507,
      "grad_norm": 0.667378306388855,
      "kl": 0.505615234375,
      "learning_rate": 2.7745399183548e-06,
      "loss": 0.0602,
      "reward": 1.5499999970197678,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 750
    },
    {
      "completion_length": 165.1500015258789,
      "epoch": 0.7786417833074132,
      "grad_norm": 0.5699002146720886,
      "kl": 0.941162109375,
      "learning_rate": 2.7735840566075646e-06,
      "loss": 0.0966,
      "reward": 1.4500000029802322,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 751
    },
    {
      "completion_length": 170.72500228881836,
      "epoch": 0.7796785899429757,
      "grad_norm": 3.4902076721191406,
      "kl": 1.697265625,
      "learning_rate": 2.7726263382843726e-06,
      "loss": 0.1494,
      "reward": 1.5999999940395355,
      "reward_std": 0.25609344989061356,
      "rewards/accuracy_reward": 0.6250000186264515,
      "rewards/format_reward": 0.9750000089406967,
      "step": 752
    },
    {
      "completion_length": 129.47500228881836,
      "epoch": 0.7807153965785381,
      "grad_norm": 0.3129628598690033,
      "kl": 0.272705078125,
      "learning_rate": 2.771666764781345e-06,
      "loss": 0.0141,
      "reward": 1.425000011920929,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 1.0,
      "step": 753
    },
    {
      "completion_length": 171.75000381469727,
      "epoch": 0.7817522032141005,
      "grad_norm": 0.5923978090286255,
      "kl": 0.968505859375,
      "learning_rate": 2.7707053374973055e-06,
      "loss": 0.0773,
      "reward": 1.4000000059604645,
      "reward_std": 0.21638592705130577,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9249999970197678,
      "step": 754
    },
    {
      "completion_length": 132.92500114440918,
      "epoch": 0.782789009849663,
      "grad_norm": 0.7645208835601807,
      "kl": 0.856201171875,
      "learning_rate": 2.7697420578337823e-06,
      "loss": 0.0276,
      "reward": 1.5750000178813934,
      "reward_std": 0.21275537833571434,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 755
    },
    {
      "completion_length": 174.9250030517578,
      "epoch": 0.7838258164852255,
      "grad_norm": 2.4929704666137695,
      "kl": 0.317626953125,
      "learning_rate": 2.768776927195003e-06,
      "loss": 0.0762,
      "reward": 1.600000023841858,
      "reward_std": 0.281220693141222,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 756
    },
    {
      "completion_length": 185.57500076293945,
      "epoch": 0.784862623120788,
      "grad_norm": 1.6370223760604858,
      "kl": 0.39794921875,
      "learning_rate": 2.7678099469878927e-06,
      "loss": 0.0027,
      "reward": 1.5250000655651093,
      "reward_std": 0.36053744703531265,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 757
    },
    {
      "completion_length": 108.42500114440918,
      "epoch": 0.7858994297563504,
      "grad_norm": 0.4571923613548279,
      "kl": 0.2900390625,
      "learning_rate": 2.766841118622074e-06,
      "loss": -0.0458,
      "reward": 1.3750000298023224,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 1.0,
      "step": 758
    },
    {
      "completion_length": 200.82500648498535,
      "epoch": 0.7869362363919129,
      "grad_norm": 2.9244139194488525,
      "kl": 3.072265625,
      "learning_rate": 2.765870443509863e-06,
      "loss": 0.3235,
      "reward": 1.4000000059604645,
      "reward_std": 0.4472135864198208,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9000000059604645,
      "step": 759
    },
    {
      "completion_length": 134.95000457763672,
      "epoch": 0.7879730430274754,
      "grad_norm": 1.5931257009506226,
      "kl": 0.43212890625,
      "learning_rate": 2.7648979230662682e-06,
      "loss": 0.0302,
      "reward": 1.5,
      "reward_std": 0.31199514865875244,
      "rewards/accuracy_reward": 0.5500000212341547,
      "rewards/format_reward": 0.9500000178813934,
      "step": 760
    },
    {
      "completion_length": 267.575008392334,
      "epoch": 0.7890098496630379,
      "grad_norm": 2.634326934814453,
      "kl": 3.15771484375,
      "learning_rate": 2.763923558708988e-06,
      "loss": 0.3409,
      "reward": 1.2999999970197678,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9250000268220901,
      "step": 761
    },
    {
      "completion_length": 128.4250030517578,
      "epoch": 0.7900466562986003,
      "grad_norm": 0.6787009835243225,
      "kl": 0.467041015625,
      "learning_rate": 2.7629473518584088e-06,
      "loss": 0.0092,
      "reward": 1.350000023841858,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 762
    },
    {
      "completion_length": 264.8250026702881,
      "epoch": 0.7910834629341628,
      "grad_norm": 0.48455703258514404,
      "kl": 1.8251953125,
      "learning_rate": 2.7619693039376025e-06,
      "loss": 0.2773,
      "reward": 1.2000000029802322,
      "reward_std": 0.15685366839170456,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 763
    },
    {
      "completion_length": 204.0750026702881,
      "epoch": 0.7921202695697253,
      "grad_norm": 0.7012991309165955,
      "kl": 0.8583984375,
      "learning_rate": 2.7609894163723257e-06,
      "loss": 0.1481,
      "reward": 1.600000023841858,
      "reward_std": 0.42178187519311905,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.925000011920929,
      "step": 764
    },
    {
      "completion_length": 263.72500228881836,
      "epoch": 0.7931570762052877,
      "grad_norm": 1.760319709777832,
      "kl": 1.53173828125,
      "learning_rate": 2.760007690591016e-06,
      "loss": 0.1753,
      "reward": 1.2000000327825546,
      "reward_std": 0.3371223732829094,
      "rewards/accuracy_reward": 0.32500000670552254,
      "rewards/format_reward": 0.8750000149011612,
      "step": 765
    },
    {
      "completion_length": 131.42500495910645,
      "epoch": 0.7941938828408501,
      "grad_norm": 0.4826253056526184,
      "kl": 0.946044921875,
      "learning_rate": 2.7590241280247906e-06,
      "loss": 0.1077,
      "reward": 1.5,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.5250000022351742,
      "rewards/format_reward": 0.9750000089406967,
      "step": 766
    },
    {
      "completion_length": 117.82500076293945,
      "epoch": 0.7952306894764126,
      "grad_norm": 0.8102556467056274,
      "kl": 0.491455078125,
      "learning_rate": 2.758038730107446e-06,
      "loss": 0.0176,
      "reward": 1.550000011920929,
      "reward_std": 0.21275536715984344,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 767
    },
    {
      "completion_length": 131.2750015258789,
      "epoch": 0.7962674961119751,
      "grad_norm": 0.7947399020195007,
      "kl": 1.2119140625,
      "learning_rate": 2.7570514982754507e-06,
      "loss": 0.061,
      "reward": 1.1250000149011612,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.17500000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 768
    },
    {
      "completion_length": 215.7250099182129,
      "epoch": 0.7973043027475376,
      "grad_norm": 3.324897527694702,
      "kl": 1.559814453125,
      "learning_rate": 2.7560624339679505e-06,
      "loss": 0.2325,
      "reward": 1.475000023841858,
      "reward_std": 0.3245587795972824,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 769
    },
    {
      "completion_length": 254.65000915527344,
      "epoch": 0.7983411093831001,
      "grad_norm": 2.051250696182251,
      "kl": 3.601806640625,
      "learning_rate": 2.75507153862676e-06,
      "loss": 0.2329,
      "reward": 1.300000011920929,
      "reward_std": 0.27228758856654167,
      "rewards/accuracy_reward": 0.42499999701976776,
      "rewards/format_reward": 0.8750000149011612,
      "step": 770
    },
    {
      "completion_length": 113.37500381469727,
      "epoch": 0.7993779160186625,
      "grad_norm": 0.8960034847259521,
      "kl": 1.621826171875,
      "learning_rate": 2.7540788136963637e-06,
      "loss": 0.1835,
      "reward": 1.350000023841858,
      "reward_std": 0.40387552231550217,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 0.925000011920929,
      "step": 771
    },
    {
      "completion_length": 134.89999961853027,
      "epoch": 0.800414722654225,
      "grad_norm": 0.7624129056930542,
      "kl": 0.291259765625,
      "learning_rate": 2.7530842606239135e-06,
      "loss": 0.088,
      "reward": 1.2750000059604645,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.35000000707805157,
      "rewards/format_reward": 0.9250000268220901,
      "step": 772
    },
    {
      "completion_length": 127.10000419616699,
      "epoch": 0.8014515292897875,
      "grad_norm": 0.4995887279510498,
      "kl": 0.360107421875,
      "learning_rate": 2.7520878808592255e-06,
      "loss": 0.0129,
      "reward": 1.6500000357627869,
      "reward_std": 0.3479738347232342,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 773
    },
    {
      "completion_length": 192.8000030517578,
      "epoch": 0.80248833592535,
      "grad_norm": 1.0516695976257324,
      "kl": 1.542236328125,
      "learning_rate": 2.7510896758547796e-06,
      "loss": 0.1974,
      "reward": 1.3250000029802322,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.4499999973922968,
      "rewards/format_reward": 0.8750000149011612,
      "step": 774
    },
    {
      "completion_length": 286.1500053405762,
      "epoch": 0.8035251425609123,
      "grad_norm": 1.4060876369476318,
      "kl": 2.7529296875,
      "learning_rate": 2.7500896470657166e-06,
      "loss": 0.3356,
      "reward": 1.200000062584877,
      "reward_std": 0.5842827521264553,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.8500000238418579,
      "step": 775
    },
    {
      "completion_length": 212.77499961853027,
      "epoch": 0.8045619491964748,
      "grad_norm": 1.4797383546829224,
      "kl": 1.840576171875,
      "learning_rate": 2.749087795949834e-06,
      "loss": 0.2744,
      "reward": 1.1000000536441803,
      "reward_std": 0.3913118988275528,
      "rewards/accuracy_reward": 0.20000000298023224,
      "rewards/format_reward": 0.9000000357627869,
      "step": 776
    },
    {
      "completion_length": 171.00000190734863,
      "epoch": 0.8055987558320373,
      "grad_norm": 8.790672302246094,
      "kl": 3.203125,
      "learning_rate": 2.7480841239675883e-06,
      "loss": 0.2483,
      "reward": 1.1250000298023224,
      "reward_std": 0.4814124368131161,
      "rewards/accuracy_reward": 0.25000000186264515,
      "rewards/format_reward": 0.8750000298023224,
      "step": 777
    },
    {
      "completion_length": 208.57500648498535,
      "epoch": 0.8066355624675998,
      "grad_norm": 0.8135401606559753,
      "kl": 1.341064453125,
      "learning_rate": 2.747078632582089e-06,
      "loss": 0.0399,
      "reward": 1.3750000596046448,
      "reward_std": 0.26129765436053276,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9500000029802322,
      "step": 778
    },
    {
      "completion_length": 237.90000343322754,
      "epoch": 0.8076723691031623,
      "grad_norm": 1.9624338150024414,
      "kl": 1.96484375,
      "learning_rate": 2.7460713232590973e-06,
      "loss": 0.3619,
      "reward": 1.4000000357627869,
      "reward_std": 0.45977723225951195,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.8750000149011612,
      "step": 779
    },
    {
      "completion_length": 91.25,
      "epoch": 0.8087091757387247,
      "grad_norm": 2.1631221771240234,
      "kl": 5.6884765625,
      "learning_rate": 2.745062197467027e-06,
      "loss": -0.0101,
      "reward": 1.3250000029802322,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9500000178813934,
      "step": 780
    },
    {
      "completion_length": 204.02500915527344,
      "epoch": 0.8097459823742872,
      "grad_norm": 2.164458990097046,
      "kl": 2.19189453125,
      "learning_rate": 2.7440512566769368e-06,
      "loss": 0.2447,
      "reward": 1.8500000536441803,
      "reward_std": 0.292072094976902,
      "rewards/accuracy_reward": 0.9000000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 781
    },
    {
      "completion_length": 107.25000190734863,
      "epoch": 0.8107827890098497,
      "grad_norm": 0.8715801239013672,
      "kl": 0.63037109375,
      "learning_rate": 2.7430385023625334e-06,
      "loss": 0.073,
      "reward": 1.2500000298023224,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.2750000022351742,
      "rewards/format_reward": 0.9750000089406967,
      "step": 782
    },
    {
      "completion_length": 149.97500228881836,
      "epoch": 0.8118195956454122,
      "grad_norm": 2.8529815673828125,
      "kl": 2.78369140625,
      "learning_rate": 2.7420239360001654e-06,
      "loss": 0.258,
      "reward": 1.8250000178813934,
      "reward_std": 0.3354101926088333,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.9500000178813934,
      "step": 783
    },
    {
      "completion_length": 98.1500015258789,
      "epoch": 0.8128564022809746,
      "grad_norm": 1.0965875387191772,
      "kl": 0.5849609375,
      "learning_rate": 2.741007559068825e-06,
      "loss": 0.0432,
      "reward": 1.375,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 784
    },
    {
      "completion_length": 200.9000072479248,
      "epoch": 0.813893208916537,
      "grad_norm": 0.8102331757545471,
      "kl": 1.67041015625,
      "learning_rate": 2.739989373050142e-06,
      "loss": 0.196,
      "reward": 1.1749999970197678,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.22500000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 785
    },
    {
      "completion_length": 275.60000228881836,
      "epoch": 0.8149300155520995,
      "grad_norm": 10.27814769744873,
      "kl": 7.51904296875,
      "learning_rate": 2.7389693794283847e-06,
      "loss": 0.7619,
      "reward": 1.3750000298023224,
      "reward_std": 0.45977722853422165,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.9000000208616257,
      "step": 786
    },
    {
      "completion_length": 141.5000057220459,
      "epoch": 0.815966822187662,
      "grad_norm": 1.1177572011947632,
      "kl": 1.534423828125,
      "learning_rate": 2.7379475796904555e-06,
      "loss": 0.1921,
      "reward": 1.4500000327825546,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 787
    },
    {
      "completion_length": 159.07500076293945,
      "epoch": 0.8170036288232244,
      "grad_norm": 0.926374614238739,
      "kl": 0.93505859375,
      "learning_rate": 2.73692397532589e-06,
      "loss": 0.1362,
      "reward": 1.449999988079071,
      "reward_std": 0.36053742468357086,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 788
    },
    {
      "completion_length": 151.77500343322754,
      "epoch": 0.8180404354587869,
      "grad_norm": 0.6936078071594238,
      "kl": 0.8388671875,
      "learning_rate": 2.7358985678268544e-06,
      "loss": 0.0432,
      "reward": 1.3499999940395355,
      "reward_std": 0.28485121950507164,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 789
    },
    {
      "completion_length": 166.9000072479248,
      "epoch": 0.8190772420943494,
      "grad_norm": 0.9093707799911499,
      "kl": 0.895263671875,
      "learning_rate": 2.734871358688144e-06,
      "loss": 0.1818,
      "reward": 1.5750000178813934,
      "reward_std": 0.33712238073349,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000029802322,
      "step": 790
    },
    {
      "completion_length": 286.17500495910645,
      "epoch": 0.8201140487299119,
      "grad_norm": 11.58954906463623,
      "kl": 5.29541015625,
      "learning_rate": 2.7338423494071795e-06,
      "loss": 0.6152,
      "reward": 1.4249999970197678,
      "reward_std": 0.29270510002970695,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 791
    },
    {
      "completion_length": 293.9500026702881,
      "epoch": 0.8211508553654744,
      "grad_norm": 2.5993452072143555,
      "kl": 2.10888671875,
      "learning_rate": 2.7328115414840066e-06,
      "loss": 0.3068,
      "reward": 1.2249999940395355,
      "reward_std": 0.3047742620110512,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.9000000208616257,
      "step": 792
    },
    {
      "completion_length": 179.12500381469727,
      "epoch": 0.8221876620010368,
      "grad_norm": 43.453060150146484,
      "kl": 9.94287109375,
      "learning_rate": 2.7317789364212927e-06,
      "loss": 0.847,
      "reward": 1.300000011920929,
      "reward_std": 0.30114371702075005,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.925000011920929,
      "step": 793
    },
    {
      "completion_length": 198.8250026702881,
      "epoch": 0.8232244686365993,
      "grad_norm": 1.3790888786315918,
      "kl": 1.91845703125,
      "learning_rate": 2.730744535724325e-06,
      "loss": 0.1889,
      "reward": 1.3250000178813934,
      "reward_std": 0.4652584567666054,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 0.925000011920929,
      "step": 794
    },
    {
      "completion_length": 166.50000190734863,
      "epoch": 0.8242612752721618,
      "grad_norm": 13.034732818603516,
      "kl": 8.81005859375,
      "learning_rate": 2.729708340901008e-06,
      "loss": 0.7422,
      "reward": 1.550000011920929,
      "reward_std": 0.25609344989061356,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 795
    },
    {
      "completion_length": 134.90000534057617,
      "epoch": 0.8252980819077242,
      "grad_norm": 1.2326252460479736,
      "kl": 0.88623046875,
      "learning_rate": 2.7286703534618622e-06,
      "loss": 0.0926,
      "reward": 1.6000000536441803,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 796
    },
    {
      "completion_length": 133.35000228881836,
      "epoch": 0.8263348885432866,
      "grad_norm": 1.0299975872039795,
      "kl": 1.529296875,
      "learning_rate": 2.7276305749200206e-06,
      "loss": 0.1607,
      "reward": 1.475000023841858,
      "reward_std": 0.3966546133160591,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 797
    },
    {
      "completion_length": 201.5750026702881,
      "epoch": 0.8273716951788491,
      "grad_norm": 1.292431354522705,
      "kl": 0.6396484375,
      "learning_rate": 2.7265890067912275e-06,
      "loss": 0.0736,
      "reward": 1.600000023841858,
      "reward_std": 0.324558787047863,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9500000178813934,
      "step": 798
    },
    {
      "completion_length": 197.82500839233398,
      "epoch": 0.8284085018144116,
      "grad_norm": 1.056061863899231,
      "kl": 0.56640625,
      "learning_rate": 2.725545650593837e-06,
      "loss": 0.0988,
      "reward": 1.4749999940395355,
      "reward_std": 0.25609343126416206,
      "rewards/accuracy_reward": 0.5250000152736902,
      "rewards/format_reward": 0.9500000178813934,
      "step": 799
    },
    {
      "completion_length": 151.5750026702881,
      "epoch": 0.8294453084499741,
      "grad_norm": 2.8053700923919678,
      "kl": 2.60009765625,
      "learning_rate": 2.7245005078488077e-06,
      "loss": 0.2199,
      "reward": 1.324999988079071,
      "reward_std": 0.46511994674801826,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9500000178813934,
      "step": 800
    },
    {
      "completion_length": 137.5250015258789,
      "epoch": 0.8304821150855366,
      "grad_norm": 0.9269716739654541,
      "kl": 1.56103515625,
      "learning_rate": 2.7234535800797037e-06,
      "loss": 0.1335,
      "reward": 1.0999999940395355,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.9750000089406967,
      "step": 801
    },
    {
      "completion_length": 195.5750026702881,
      "epoch": 0.831518921721099,
      "grad_norm": 5.310616970062256,
      "kl": 3.5379638671875,
      "learning_rate": 2.7224048688126915e-06,
      "loss": 0.3573,
      "reward": 1.7500000596046448,
      "reward_std": 0.41643913835287094,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 802
    },
    {
      "completion_length": 243.20000839233398,
      "epoch": 0.8325557283566615,
      "grad_norm": 11.0709810256958,
      "kl": 6.59619140625,
      "learning_rate": 2.7213543755765377e-06,
      "loss": 0.6267,
      "reward": 1.225000023841858,
      "reward_std": 0.40387551486492157,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.9250000268220901,
      "step": 803
    },
    {
      "completion_length": 199.4000072479248,
      "epoch": 0.833592534992224,
      "grad_norm": 1.7671570777893066,
      "kl": 2.40234375,
      "learning_rate": 2.720302101902605e-06,
      "loss": 0.2168,
      "reward": 1.5250000059604645,
      "reward_std": 0.4363621585071087,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 804
    },
    {
      "completion_length": 160.1750030517578,
      "epoch": 0.8346293416277865,
      "grad_norm": 0.5831762552261353,
      "kl": 1.02880859375,
      "learning_rate": 2.719248049324853e-06,
      "loss": 0.1246,
      "reward": 1.7750000357627869,
      "reward_std": 0.3407529145479202,
      "rewards/accuracy_reward": 0.7999999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 805
    },
    {
      "completion_length": 110.62500190734863,
      "epoch": 0.8356661482633488,
      "grad_norm": 0.35738223791122437,
      "kl": 0.25,
      "learning_rate": 2.718192219379835e-06,
      "loss": -0.0206,
      "reward": 1.350000023841858,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.9750000089406967,
      "step": 806
    },
    {
      "completion_length": 156.70000457763672,
      "epoch": 0.8367029548989113,
      "grad_norm": 1.1867387294769287,
      "kl": 0.779296875,
      "learning_rate": 2.7171346136066936e-06,
      "loss": 0.0683,
      "reward": 1.5750000178813934,
      "reward_std": 0.373101070523262,
      "rewards/accuracy_reward": 0.6000000052154064,
      "rewards/format_reward": 0.9750000089406967,
      "step": 807
    },
    {
      "completion_length": 196.1250057220459,
      "epoch": 0.8377397615344738,
      "grad_norm": 0.2853090763092041,
      "kl": 0.412841796875,
      "learning_rate": 2.716075233547161e-06,
      "loss": 0.0884,
      "reward": 1.2999999970197678,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 808
    },
    {
      "completion_length": 203.225004196167,
      "epoch": 0.8387765681700363,
      "grad_norm": 1.5570887327194214,
      "kl": 1.113525390625,
      "learning_rate": 2.7150140807455555e-06,
      "loss": 0.1719,
      "reward": 1.449999988079071,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000178813934,
      "step": 809
    },
    {
      "completion_length": 269.4750099182129,
      "epoch": 0.8398133748055988,
      "grad_norm": 1.8617844581604004,
      "kl": 2.31640625,
      "learning_rate": 2.7139511567487806e-06,
      "loss": 0.3669,
      "reward": 1.4500000476837158,
      "reward_std": 0.3913119025528431,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9000000357627869,
      "step": 810
    },
    {
      "completion_length": 152.3750057220459,
      "epoch": 0.8408501814411612,
      "grad_norm": 0.6547606587409973,
      "kl": 0.7275390625,
      "learning_rate": 2.712886463106321e-06,
      "loss": 0.0751,
      "reward": 1.5250000059604645,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 811
    },
    {
      "completion_length": 165.82500648498535,
      "epoch": 0.8418869880767237,
      "grad_norm": 0.41356778144836426,
      "kl": 0.31201171875,
      "learning_rate": 2.7118200013702407e-06,
      "loss": 0.0653,
      "reward": 1.675000011920929,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 812
    },
    {
      "completion_length": 276.02500915527344,
      "epoch": 0.8429237947122862,
      "grad_norm": 0.9744496941566467,
      "kl": 2.72900390625,
      "learning_rate": 2.7107517730951818e-06,
      "loss": 0.3531,
      "reward": 1.400000050663948,
      "reward_std": 0.4723408408463001,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9250000268220901,
      "step": 813
    },
    {
      "completion_length": 187.32500457763672,
      "epoch": 0.8439606013478487,
      "grad_norm": 1.1766014099121094,
      "kl": 0.46630859375,
      "learning_rate": 2.709681779838362e-06,
      "loss": 0.0853,
      "reward": 1.225000023841858,
      "reward_std": 0.3046357333660126,
      "rewards/accuracy_reward": 0.2750000096857548,
      "rewards/format_reward": 0.9500000178813934,
      "step": 814
    },
    {
      "completion_length": 262.5499973297119,
      "epoch": 0.844997407983411,
      "grad_norm": 0.8816485404968262,
      "kl": 1.880126953125,
      "learning_rate": 2.7086100231595707e-06,
      "loss": 0.2727,
      "reward": 1.1250000298023224,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.20000000298023224,
      "rewards/format_reward": 0.925000011920929,
      "step": 815
    },
    {
      "completion_length": 181.67500114440918,
      "epoch": 0.8460342146189735,
      "grad_norm": 0.4687977731227875,
      "kl": 1.150634765625,
      "learning_rate": 2.7075365046211684e-06,
      "loss": 0.0978,
      "reward": 1.2500000298023224,
      "reward_std": 0.2487340308725834,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 816
    },
    {
      "completion_length": 228.07500839233398,
      "epoch": 0.847071021254536,
      "grad_norm": 2.606320381164551,
      "kl": 3.34521484375,
      "learning_rate": 2.706461225788085e-06,
      "loss": 0.3432,
      "reward": 1.5250000059604645,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9500000029802322,
      "step": 817
    },
    {
      "completion_length": 156.0500030517578,
      "epoch": 0.8481078278900985,
      "grad_norm": 2.9354310035705566,
      "kl": 2.510498046875,
      "learning_rate": 2.7053841882278156e-06,
      "loss": 0.239,
      "reward": 1.1000000089406967,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.9750000089406967,
      "step": 818
    },
    {
      "completion_length": 190.42499923706055,
      "epoch": 0.8491446345256609,
      "grad_norm": 0.6598667502403259,
      "kl": 0.934814453125,
      "learning_rate": 2.7043053935104183e-06,
      "loss": 0.0775,
      "reward": 1.375,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 819
    },
    {
      "completion_length": 182.6250057220459,
      "epoch": 0.8501814411612234,
      "grad_norm": 0.46700766682624817,
      "kl": 1.113525390625,
      "learning_rate": 2.7032248432085143e-06,
      "loss": 0.141,
      "reward": 1.3250000327825546,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 820
    },
    {
      "completion_length": 221.8000087738037,
      "epoch": 0.8512182477967859,
      "grad_norm": 3.7838664054870605,
      "kl": 0.30810546875,
      "learning_rate": 2.702142538897283e-06,
      "loss": 0.1173,
      "reward": 1.3250000178813934,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 821
    },
    {
      "completion_length": 353.32501220703125,
      "epoch": 0.8522550544323484,
      "grad_norm": 1.2203960418701172,
      "kl": 2.563232421875,
      "learning_rate": 2.701058482154461e-06,
      "loss": 0.3616,
      "reward": 1.3750000298023224,
      "reward_std": 0.48141248524188995,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.8750000149011612,
      "step": 822
    },
    {
      "completion_length": 303.4500045776367,
      "epoch": 0.8532918610679109,
      "grad_norm": 6.787600994110107,
      "kl": 6.1015625,
      "learning_rate": 2.6999726745603396e-06,
      "loss": 0.6308,
      "reward": 1.475000038743019,
      "reward_std": 0.4237985461950302,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 823
    },
    {
      "completion_length": 325.97500801086426,
      "epoch": 0.8543286677034733,
      "grad_norm": 22.804424285888672,
      "kl": 9.36083984375,
      "learning_rate": 2.698885117697762e-06,
      "loss": 0.8731,
      "reward": 1.2750000208616257,
      "reward_std": 0.3840910233557224,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9000000208616257,
      "step": 824
    },
    {
      "completion_length": 269.49999618530273,
      "epoch": 0.8553654743390358,
      "grad_norm": 1.0976574420928955,
      "kl": 1.403564453125,
      "learning_rate": 2.697795813152122e-06,
      "loss": 0.0994,
      "reward": 1.475000023841858,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.925000011920929,
      "step": 825
    },
    {
      "completion_length": 226.20000457763672,
      "epoch": 0.8564022809745983,
      "grad_norm": 1.3571267127990723,
      "kl": 1.186279296875,
      "learning_rate": 2.6967047625113606e-06,
      "loss": 0.1751,
      "reward": 1.3250000327825546,
      "reward_std": 0.31199514865875244,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.9500000178813934,
      "step": 826
    },
    {
      "completion_length": 199.25000381469727,
      "epoch": 0.8574390876101607,
      "grad_norm": 1.5436522960662842,
      "kl": 0.467529296875,
      "learning_rate": 2.6956119673659652e-06,
      "loss": 0.0601,
      "reward": 1.7249999940395355,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 827
    },
    {
      "completion_length": 180.52500915527344,
      "epoch": 0.8584758942457231,
      "grad_norm": 2.2522006034851074,
      "kl": 0.473388671875,
      "learning_rate": 2.6945174293089644e-06,
      "loss": 0.0962,
      "reward": 1.4000000357627869,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9750000089406967,
      "step": 828
    },
    {
      "completion_length": 210.47500610351562,
      "epoch": 0.8595127008812856,
      "grad_norm": 0.6633756756782532,
      "kl": 1.012451171875,
      "learning_rate": 2.6934211499359293e-06,
      "loss": 0.1798,
      "reward": 1.1500000208616257,
      "reward_std": 0.1928323470056057,
      "rewards/accuracy_reward": 0.20000000670552254,
      "rewards/format_reward": 0.9500000029802322,
      "step": 829
    },
    {
      "completion_length": 273.35000228881836,
      "epoch": 0.8605495075168481,
      "grad_norm": 0.7028055191040039,
      "kl": 0.66015625,
      "learning_rate": 2.6923231308449686e-06,
      "loss": 0.159,
      "reward": 1.425000011920929,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.925000011920929,
      "step": 830
    },
    {
      "completion_length": 220.07500839233398,
      "epoch": 0.8615863141524106,
      "grad_norm": 1.8734394311904907,
      "kl": 1.123291015625,
      "learning_rate": 2.6912233736367267e-06,
      "loss": 0.1689,
      "reward": 1.4500000327825546,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000178813934,
      "step": 831
    },
    {
      "completion_length": 316.2000026702881,
      "epoch": 0.8626231207879731,
      "grad_norm": 1.5336312055587769,
      "kl": 1.52294921875,
      "learning_rate": 2.6901218799143837e-06,
      "loss": 0.1906,
      "reward": 1.3250000178813934,
      "reward_std": 0.36053743585944176,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9000000059604645,
      "step": 832
    },
    {
      "completion_length": 132.975004196167,
      "epoch": 0.8636599274235355,
      "grad_norm": 0.27653202414512634,
      "kl": 0.23974609375,
      "learning_rate": 2.689018651283648e-06,
      "loss": 0.0261,
      "reward": 1.6500000059604645,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 1.0,
      "step": 833
    },
    {
      "completion_length": 267.3249988555908,
      "epoch": 0.864696734059098,
      "grad_norm": 1.0966119766235352,
      "kl": 1.613037109375,
      "learning_rate": 2.6879136893527593e-06,
      "loss": 0.2441,
      "reward": 1.4500000476837158,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.925000011920929,
      "step": 834
    },
    {
      "completion_length": 174.17500114440918,
      "epoch": 0.8657335406946605,
      "grad_norm": 0.47363966703414917,
      "kl": 0.385986328125,
      "learning_rate": 2.686806995732484e-06,
      "loss": 0.0541,
      "reward": 1.425000011920929,
      "reward_std": 0.26129766926169395,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 835
    },
    {
      "completion_length": 234.4000015258789,
      "epoch": 0.866770347330223,
      "grad_norm": 2.3432672023773193,
      "kl": 3.037841796875,
      "learning_rate": 2.6856985720361114e-06,
      "loss": 0.3162,
      "reward": 1.4000000357627869,
      "reward_std": 0.40387555956840515,
      "rewards/accuracy_reward": 0.4500000048428774,
      "rewards/format_reward": 0.9500000178813934,
      "step": 836
    },
    {
      "completion_length": 188.27500343322754,
      "epoch": 0.8678071539657853,
      "grad_norm": 0.5459396243095398,
      "kl": 0.52294921875,
      "learning_rate": 2.684588419879455e-06,
      "loss": 0.0788,
      "reward": 1.5250000059604645,
      "reward_std": 0.3966546431183815,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.9750000089406967,
      "step": 837
    },
    {
      "completion_length": 330.22500801086426,
      "epoch": 0.8688439606013478,
      "grad_norm": 3.8449041843414307,
      "kl": 3.85595703125,
      "learning_rate": 2.683476540880846e-06,
      "loss": 0.4513,
      "reward": 1.4750000536441803,
      "reward_std": 0.46511993557214737,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 0.9000000208616257,
      "step": 838
    },
    {
      "completion_length": 366.2750053405762,
      "epoch": 0.8698807672369103,
      "grad_norm": 6.319846153259277,
      "kl": 7.76025390625,
      "learning_rate": 2.682362936661134e-06,
      "loss": 0.6544,
      "reward": 1.5,
      "reward_std": 0.40450848266482353,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8500000089406967,
      "step": 839
    },
    {
      "completion_length": 122.9000015258789,
      "epoch": 0.8709175738724728,
      "grad_norm": 0.4149312376976013,
      "kl": 0.285400390625,
      "learning_rate": 2.6812476088436827e-06,
      "loss": 0.0377,
      "reward": 1.4000000059604645,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 1.0,
      "step": 840
    },
    {
      "completion_length": 126.32500076293945,
      "epoch": 0.8719543805080352,
      "grad_norm": 0.42162078619003296,
      "kl": 0.322998046875,
      "learning_rate": 2.6801305590543697e-06,
      "loss": 0.0095,
      "reward": 1.3750000298023224,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.37499999813735485,
      "rewards/format_reward": 1.0,
      "step": 841
    },
    {
      "completion_length": 267.50000381469727,
      "epoch": 0.8729911871435977,
      "grad_norm": 2.6304068565368652,
      "kl": 1.82666015625,
      "learning_rate": 2.6790117889215812e-06,
      "loss": 0.3007,
      "reward": 1.125,
      "reward_std": 0.27950849011540413,
      "rewards/accuracy_reward": 0.22500000149011612,
      "rewards/format_reward": 0.9000000208616257,
      "step": 842
    },
    {
      "completion_length": 169.42500686645508,
      "epoch": 0.8740279937791602,
      "grad_norm": 1.3741490840911865,
      "kl": 0.783203125,
      "learning_rate": 2.677891300076213e-06,
      "loss": 0.0916,
      "reward": 1.375,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 843
    },
    {
      "completion_length": 173.32500457763672,
      "epoch": 0.8750648004147227,
      "grad_norm": 1.253536343574524,
      "kl": 0.6121826171875,
      "learning_rate": 2.6767690941516653e-06,
      "loss": 0.0845,
      "reward": 1.7249999940395355,
      "reward_std": 0.3479738272726536,
      "rewards/accuracy_reward": 0.7500000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 844
    },
    {
      "completion_length": 263.67500495910645,
      "epoch": 0.8761016070502852,
      "grad_norm": 1.1694815158843994,
      "kl": 1.185546875,
      "learning_rate": 2.6756451727838407e-06,
      "loss": 0.1564,
      "reward": 1.3750000149011612,
      "reward_std": 0.3299015015363693,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.875,
      "step": 845
    },
    {
      "completion_length": 181.50000190734863,
      "epoch": 0.8771384136858476,
      "grad_norm": 0.7774064540863037,
      "kl": 0.954833984375,
      "learning_rate": 2.674519537611144e-06,
      "loss": 0.0741,
      "reward": 1.5750000476837158,
      "reward_std": 0.36053744703531265,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 846
    },
    {
      "completion_length": 183.3000030517578,
      "epoch": 0.87817522032141,
      "grad_norm": 0.5851902961730957,
      "kl": 1.0263671875,
      "learning_rate": 2.673392190274478e-06,
      "loss": 0.1204,
      "reward": 1.550000011920929,
      "reward_std": 0.21275537833571434,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 847
    },
    {
      "completion_length": 326.75000381469727,
      "epoch": 0.8792120269569725,
      "grad_norm": 6.058239459991455,
      "kl": 1.504150390625,
      "learning_rate": 2.6722631324172407e-06,
      "loss": 0.2304,
      "reward": 1.3250000327825546,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.8750000149011612,
      "step": 848
    },
    {
      "completion_length": 246.55000114440918,
      "epoch": 0.880248833592535,
      "grad_norm": 1.6069053411483765,
      "kl": 0.45849609375,
      "learning_rate": 2.6711323656853235e-06,
      "loss": 0.0939,
      "reward": 1.6500000357627869,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.7250000033527613,
      "rewards/format_reward": 0.925000011920929,
      "step": 849
    },
    {
      "completion_length": 203.32500076293945,
      "epoch": 0.8812856402280974,
      "grad_norm": 0.8683740496635437,
      "kl": 0.555908203125,
      "learning_rate": 2.6699998917271096e-06,
      "loss": 0.0142,
      "reward": 1.4749999940395355,
      "reward_std": 0.3611704334616661,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 850
    },
    {
      "completion_length": 176.20000076293945,
      "epoch": 0.8823224468636599,
      "grad_norm": 2.5231356620788574,
      "kl": 2.1685791015625,
      "learning_rate": 2.6688657121934713e-06,
      "loss": 0.2105,
      "reward": 1.6250000298023224,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 851
    },
    {
      "completion_length": 313.75000381469727,
      "epoch": 0.8833592534992224,
      "grad_norm": 26.584848403930664,
      "kl": 10.794189453125,
      "learning_rate": 2.6677298287377666e-06,
      "loss": 0.9613,
      "reward": 1.350000023841858,
      "reward_std": 0.535601943731308,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9000000208616257,
      "step": 852
    },
    {
      "completion_length": 376.1000118255615,
      "epoch": 0.8843960601347849,
      "grad_norm": 12.684377670288086,
      "kl": 10.703125,
      "learning_rate": 2.666592243015837e-06,
      "loss": 0.9707,
      "reward": 1.0750000327825546,
      "reward_std": 0.542072132229805,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.8250000178813934,
      "step": 853
    },
    {
      "completion_length": 372.6500129699707,
      "epoch": 0.8854328667703474,
      "grad_norm": 1.3346270322799683,
      "kl": 3.55712890625,
      "learning_rate": 2.665452956686006e-06,
      "loss": 0.3998,
      "reward": 1.4500000178813934,
      "reward_std": 0.38409099727869034,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.8750000149011612,
      "step": 854
    },
    {
      "completion_length": 209.50000762939453,
      "epoch": 0.8864696734059098,
      "grad_norm": 6.27949333190918,
      "kl": 3.934814453125,
      "learning_rate": 2.664311971409076e-06,
      "loss": 0.3364,
      "reward": 1.3500000536441803,
      "reward_std": 0.46511994302272797,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.925000011920929,
      "step": 855
    },
    {
      "completion_length": 294.7000026702881,
      "epoch": 0.8875064800414723,
      "grad_norm": 1.2558188438415527,
      "kl": 2.42041015625,
      "learning_rate": 2.663169288848327e-06,
      "loss": 0.2752,
      "reward": 1.375,
      "reward_std": 0.49746809154748917,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 856
    },
    {
      "completion_length": 430.9500026702881,
      "epoch": 0.8885432866770347,
      "grad_norm": 1.8420369625091553,
      "kl": 3.364501953125,
      "learning_rate": 2.662024910669511e-06,
      "loss": 0.3637,
      "reward": 1.4250000268220901,
      "reward_std": 0.582266092300415,
      "rewards/accuracy_reward": 0.5999999977648258,
      "rewards/format_reward": 0.8250000029802322,
      "step": 857
    },
    {
      "completion_length": 131.52500534057617,
      "epoch": 0.8895800933125972,
      "grad_norm": 0.38641226291656494,
      "kl": 0.2255859375,
      "learning_rate": 2.6608788385408546e-06,
      "loss": -0.0141,
      "reward": 1.3499999940395355,
      "reward_std": 0.1928323470056057,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 1.0,
      "step": 858
    },
    {
      "completion_length": 321.15000915527344,
      "epoch": 0.8906168999481596,
      "grad_norm": 1.2833192348480225,
      "kl": 1.921875,
      "learning_rate": 2.6597310741330515e-06,
      "loss": 0.2401,
      "reward": 1.2750000059604645,
      "reward_std": 0.4976065903902054,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 0.8500000238418579,
      "step": 859
    },
    {
      "completion_length": 118.60000419616699,
      "epoch": 0.8916537065837221,
      "grad_norm": 0.39673030376434326,
      "kl": 0.205078125,
      "learning_rate": 2.6585816191192637e-06,
      "loss": 0.0051,
      "reward": 1.5,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 1.0,
      "step": 860
    },
    {
      "completion_length": 176.8750057220459,
      "epoch": 0.8926905132192846,
      "grad_norm": 2.3243377208709717,
      "kl": 0.39306640625,
      "learning_rate": 2.657430475175117e-06,
      "loss": 0.0705,
      "reward": 1.5000000298023224,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9500000178813934,
      "step": 861
    },
    {
      "completion_length": 292.87500381469727,
      "epoch": 0.8937273198548471,
      "grad_norm": 3.260263442993164,
      "kl": 1.212890625,
      "learning_rate": 2.656277643978699e-06,
      "loss": 0.1044,
      "reward": 1.550000011920929,
      "reward_std": 0.34448180347681046,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.9250000268220901,
      "step": 862
    },
    {
      "completion_length": 292.2000141143799,
      "epoch": 0.8947641264904095,
      "grad_norm": 1.0654324293136597,
      "kl": 0.982666015625,
      "learning_rate": 2.655123127210558e-06,
      "loss": 0.0851,
      "reward": 1.5500000268220901,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 863
    },
    {
      "completion_length": 318.0000114440918,
      "epoch": 0.895800933125972,
      "grad_norm": 146.21029663085938,
      "kl": 1.025390625,
      "learning_rate": 2.6539669265536987e-06,
      "loss": 0.1239,
      "reward": 1.2000000178813934,
      "reward_std": 0.46511993184685707,
      "rewards/accuracy_reward": 0.3000000100582838,
      "rewards/format_reward": 0.9000000208616257,
      "step": 864
    },
    {
      "completion_length": 179.82500076293945,
      "epoch": 0.8968377397615345,
      "grad_norm": 3.667255163192749,
      "kl": 0.69873046875,
      "learning_rate": 2.6528090436935803e-06,
      "loss": 0.0528,
      "reward": 1.425000011920929,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.4499999973922968,
      "rewards/format_reward": 0.9750000089406967,
      "step": 865
    },
    {
      "completion_length": 241.0500030517578,
      "epoch": 0.897874546397097,
      "grad_norm": 0.824424684047699,
      "kl": 0.658203125,
      "learning_rate": 2.6516494803181147e-06,
      "loss": 0.0801,
      "reward": 1.550000011920929,
      "reward_std": 0.4849044717848301,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 866
    },
    {
      "completion_length": 235.8499984741211,
      "epoch": 0.8989113530326595,
      "grad_norm": 0.9227349162101746,
      "kl": 0.5498046875,
      "learning_rate": 2.650488238117663e-06,
      "loss": 0.1104,
      "reward": 1.5250000059604645,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9500000029802322,
      "step": 867
    },
    {
      "completion_length": 333.700008392334,
      "epoch": 0.8999481596682218,
      "grad_norm": 1.584023356437683,
      "kl": 1.18994140625,
      "learning_rate": 2.6493253187850347e-06,
      "loss": 0.1872,
      "reward": 1.2249999940395355,
      "reward_std": 0.45318927988409996,
      "rewards/accuracy_reward": 0.35000001080334187,
      "rewards/format_reward": 0.8750000149011612,
      "step": 868
    },
    {
      "completion_length": 330.27500343322754,
      "epoch": 0.9009849663037843,
      "grad_norm": 3.394610643386841,
      "kl": 3.2109375,
      "learning_rate": 2.6481607240154826e-06,
      "loss": 0.3783,
      "reward": 1.3750000298023224,
      "reward_std": 0.3533165380358696,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.9000000208616257,
      "step": 869
    },
    {
      "completion_length": 196.25,
      "epoch": 0.9020217729393468,
      "grad_norm": 0.48655158281326294,
      "kl": 0.21875,
      "learning_rate": 2.646994455506704e-06,
      "loss": 0.0491,
      "reward": 1.625,
      "reward_std": 0.39302411302924156,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9750000089406967,
      "step": 870
    },
    {
      "completion_length": 240.25000190734863,
      "epoch": 0.9030585795749093,
      "grad_norm": 2.112935781478882,
      "kl": 1.31494140625,
      "learning_rate": 2.6458265149588345e-06,
      "loss": 0.1163,
      "reward": 1.3250000327825546,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.4000000096857548,
      "rewards/format_reward": 0.9250000268220901,
      "step": 871
    },
    {
      "completion_length": 249.52500915527344,
      "epoch": 0.9040953862104717,
      "grad_norm": 1.491579294204712,
      "kl": 1.40673828125,
      "learning_rate": 2.6446569040744464e-06,
      "loss": 0.1719,
      "reward": 1.5250000357627869,
      "reward_std": 0.3299014940857887,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 872
    },
    {
      "completion_length": 239.32500076293945,
      "epoch": 0.9051321928460342,
      "grad_norm": 0.9075613021850586,
      "kl": 0.900390625,
      "learning_rate": 2.6434856245585493e-06,
      "loss": 0.1707,
      "reward": 1.6500000357627869,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 873
    },
    {
      "completion_length": 213.05000495910645,
      "epoch": 0.9061689994815967,
      "grad_norm": 1.415359616279602,
      "kl": 1.92919921875,
      "learning_rate": 2.642312678118584e-06,
      "loss": 0.217,
      "reward": 1.3750000298023224,
      "reward_std": 0.27228760346770287,
      "rewards/accuracy_reward": 0.42500000819563866,
      "rewards/format_reward": 0.9500000178813934,
      "step": 874
    },
    {
      "completion_length": 283.8000068664551,
      "epoch": 0.9072058061171592,
      "grad_norm": 40.05736541748047,
      "kl": 9.90576171875,
      "learning_rate": 2.6411380664644208e-06,
      "loss": 0.8472,
      "reward": 1.2500000149011612,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.925000011920929,
      "step": 875
    },
    {
      "completion_length": 227.15000534057617,
      "epoch": 0.9082426127527217,
      "grad_norm": 1.3082458972930908,
      "kl": 2.181884765625,
      "learning_rate": 2.6399617913083574e-06,
      "loss": 0.0988,
      "reward": 1.225000023841858,
      "reward_std": 0.41643914207816124,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.9000000208616257,
      "step": 876
    },
    {
      "completion_length": 118.55000114440918,
      "epoch": 0.909279419388284,
      "grad_norm": 0.44255509972572327,
      "kl": 0.2578125,
      "learning_rate": 2.6387838543651183e-06,
      "loss": 0.0397,
      "reward": 1.5499999821186066,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 877
    },
    {
      "completion_length": 211.2750015258789,
      "epoch": 0.9103162260238465,
      "grad_norm": 3.5066232681274414,
      "kl": 2.357421875,
      "learning_rate": 2.6376042573518482e-06,
      "loss": 0.2314,
      "reward": 1.5000000298023224,
      "reward_std": 0.2253189980983734,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 878
    },
    {
      "completion_length": 240.5250015258789,
      "epoch": 0.911353032659409,
      "grad_norm": 2.016002655029297,
      "kl": 2.30078125,
      "learning_rate": 2.6364230019881132e-06,
      "loss": 0.354,
      "reward": 1.4250000417232513,
      "reward_std": 0.4723408557474613,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.9000000208616257,
      "step": 879
    },
    {
      "completion_length": 308.575008392334,
      "epoch": 0.9123898392949715,
      "grad_norm": 3.611039876937866,
      "kl": 2.6015625,
      "learning_rate": 2.6352400899958967e-06,
      "loss": 0.3464,
      "reward": 1.4249999970197678,
      "reward_std": 0.36053743585944176,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 880
    },
    {
      "completion_length": 282.22500228881836,
      "epoch": 0.9134266459305339,
      "grad_norm": 2.582071542739868,
      "kl": 2.25634765625,
      "learning_rate": 2.6340555230995963e-06,
      "loss": 0.152,
      "reward": 1.300000011920929,
      "reward_std": 0.4092182442545891,
      "rewards/accuracy_reward": 0.3750000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 881
    },
    {
      "completion_length": 237.97501182556152,
      "epoch": 0.9144634525660964,
      "grad_norm": 1.9722750186920166,
      "kl": 1.4404296875,
      "learning_rate": 2.6328693030260233e-06,
      "loss": 0.1863,
      "reward": 1.4249999970197678,
      "reward_std": 0.27950849011540413,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 882
    },
    {
      "completion_length": 262.6999988555908,
      "epoch": 0.9155002592016589,
      "grad_norm": 3.5364909172058105,
      "kl": 1.622314453125,
      "learning_rate": 2.631681431504397e-06,
      "loss": 0.2271,
      "reward": 1.4000000357627869,
      "reward_std": 0.3533165268599987,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.925000011920929,
      "step": 883
    },
    {
      "completion_length": 308.45000076293945,
      "epoch": 0.9165370658372214,
      "grad_norm": 2.106987237930298,
      "kl": 1.511474609375,
      "learning_rate": 2.6304919102663467e-06,
      "loss": 0.2561,
      "reward": 1.2750000208616257,
      "reward_std": 0.31719936802983284,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9000000208616257,
      "step": 884
    },
    {
      "completion_length": 103.62500190734863,
      "epoch": 0.9175738724727839,
      "grad_norm": 0.7540234923362732,
      "kl": 0.2861328125,
      "learning_rate": 2.629300741045904e-06,
      "loss": 0.0162,
      "reward": 1.5500000417232513,
      "reward_std": 0.31719936057925224,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 1.0,
      "step": 885
    },
    {
      "completion_length": 272.4250011444092,
      "epoch": 0.9186106791083463,
      "grad_norm": 5.264670372009277,
      "kl": 3.08740234375,
      "learning_rate": 2.628107925579505e-06,
      "loss": 0.3337,
      "reward": 1.5750000178813934,
      "reward_std": 0.31199517846107483,
      "rewards/accuracy_reward": 0.6500000022351742,
      "rewards/format_reward": 0.925000011920929,
      "step": 886
    },
    {
      "completion_length": 229.125,
      "epoch": 0.9196474857439088,
      "grad_norm": 1.4002385139465332,
      "kl": 0.61328125,
      "learning_rate": 2.626913465605985e-06,
      "loss": 0.0328,
      "reward": 1.4250000417232513,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9500000029802322,
      "step": 887
    },
    {
      "completion_length": 177.32500457763672,
      "epoch": 0.9206842923794712,
      "grad_norm": 8.91387939453125,
      "kl": 3.0546875,
      "learning_rate": 2.6257173628665746e-06,
      "loss": 0.237,
      "reward": 1.2000000178813934,
      "reward_std": 0.22894951701164246,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.9500000029802322,
      "step": 888
    },
    {
      "completion_length": 123.5,
      "epoch": 0.9217210990150337,
      "grad_norm": 0.1785755157470703,
      "kl": 0.228515625,
      "learning_rate": 2.6245196191049022e-06,
      "loss": -0.0007,
      "reward": 1.300000011920929,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.30000000074505806,
      "rewards/format_reward": 1.0,
      "step": 889
    },
    {
      "completion_length": 254.17500686645508,
      "epoch": 0.9227579056505961,
      "grad_norm": 3.5954160690307617,
      "kl": 2.435546875,
      "learning_rate": 2.6233202360669864e-06,
      "loss": 0.334,
      "reward": 1.275000050663948,
      "reward_std": 0.4472136162221432,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9000000357627869,
      "step": 890
    },
    {
      "completion_length": 260.72500228881836,
      "epoch": 0.9237947122861586,
      "grad_norm": 14.043997764587402,
      "kl": 6.834228515625,
      "learning_rate": 2.6221192155012366e-06,
      "loss": 0.6215,
      "reward": 1.6250000298023224,
      "reward_std": 0.3245587609708309,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 891
    },
    {
      "completion_length": 350.0750045776367,
      "epoch": 0.9248315189217211,
      "grad_norm": 92.7138442993164,
      "kl": 18.8623046875,
      "learning_rate": 2.6209165591584487e-06,
      "loss": 1.5923,
      "reward": 1.0750000178813934,
      "reward_std": 0.35394952073693275,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.8500000089406967,
      "step": 892
    },
    {
      "completion_length": 241.80001068115234,
      "epoch": 0.9258683255572836,
      "grad_norm": 42.686927795410156,
      "kl": 10.860107421875,
      "learning_rate": 2.619712268791803e-06,
      "loss": 0.9162,
      "reward": 1.4000000357627869,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 0.9500000178813934,
      "step": 893
    },
    {
      "completion_length": 269.225004196167,
      "epoch": 0.926905132192846,
      "grad_norm": 125.65151977539062,
      "kl": 40.829833984375,
      "learning_rate": 2.6185063461568624e-06,
      "loss": 3.204,
      "reward": 1.3250000476837158,
      "reward_std": 0.3804604783654213,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 894
    },
    {
      "completion_length": 165.17500495910645,
      "epoch": 0.9279419388284085,
      "grad_norm": 0.945558488368988,
      "kl": 0.5986328125,
      "learning_rate": 2.617298793011569e-06,
      "loss": 0.0603,
      "reward": 1.5000000298023224,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.5250000040978193,
      "rewards/format_reward": 0.9750000089406967,
      "step": 895
    },
    {
      "completion_length": 142.74999809265137,
      "epoch": 0.928978745463971,
      "grad_norm": 0.27745455503463745,
      "kl": 0.215576171875,
      "learning_rate": 2.6160896111162412e-06,
      "loss": 0.0277,
      "reward": 1.550000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 1.0,
      "step": 896
    },
    {
      "completion_length": 269.32501220703125,
      "epoch": 0.9300155520995335,
      "grad_norm": 3.364424705505371,
      "kl": 1.71484375,
      "learning_rate": 2.614878802233573e-06,
      "loss": 0.2486,
      "reward": 1.3750000298023224,
      "reward_std": 0.5173911079764366,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9250000268220901,
      "step": 897
    },
    {
      "completion_length": 307.37500381469727,
      "epoch": 0.931052358735096,
      "grad_norm": 8.091796875,
      "kl": 0.7021484375,
      "learning_rate": 2.613666368128629e-06,
      "loss": 0.1984,
      "reward": 1.5,
      "reward_std": 0.5715806074440479,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9000000208616257,
      "step": 898
    },
    {
      "completion_length": 216.17500114440918,
      "epoch": 0.9320891653706583,
      "grad_norm": 3.7539572715759277,
      "kl": 0.489990234375,
      "learning_rate": 2.612452310568843e-06,
      "loss": 0.166,
      "reward": 1.4500000476837158,
      "reward_std": 0.347973819822073,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 899
    },
    {
      "completion_length": 306.25000381469727,
      "epoch": 0.9331259720062208,
      "grad_norm": 19.283233642578125,
      "kl": 12.60498046875,
      "learning_rate": 2.611236631324017e-06,
      "loss": 0.5753,
      "reward": 1.3000000268220901,
      "reward_std": 0.4038755036890507,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.9000000208616257,
      "step": 900
    },
    {
      "completion_length": 301.6750068664551,
      "epoch": 0.9341627786417833,
      "grad_norm": 3.5665500164031982,
      "kl": 3.050048828125,
      "learning_rate": 2.6100193321663147e-06,
      "loss": 0.3779,
      "reward": 1.0750000327825546,
      "reward_std": 0.4019973427057266,
      "rewards/accuracy_reward": 0.17500000447034836,
      "rewards/format_reward": 0.9000000208616257,
      "step": 901
    },
    {
      "completion_length": 171.35000801086426,
      "epoch": 0.9351995852773458,
      "grad_norm": 1.0143321752548218,
      "kl": 0.2734375,
      "learning_rate": 2.608800414870263e-06,
      "loss": 0.0389,
      "reward": 1.5250000357627869,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.5500000081956387,
      "rewards/format_reward": 0.9750000089406967,
      "step": 902
    },
    {
      "completion_length": 174.27500343322754,
      "epoch": 0.9362363919129082,
      "grad_norm": 1.870166301727295,
      "kl": 0.5362548828125,
      "learning_rate": 2.6075798812127463e-06,
      "loss": 0.0717,
      "reward": 1.3250000327825546,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.9750000089406967,
      "step": 903
    },
    {
      "completion_length": 270.29999923706055,
      "epoch": 0.9372731985484707,
      "grad_norm": 1.5089102983474731,
      "kl": 1.107421875,
      "learning_rate": 2.6063577329730072e-06,
      "loss": 0.2227,
      "reward": 1.4000000357627869,
      "reward_std": 0.373101070523262,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.9000000059604645,
      "step": 904
    },
    {
      "completion_length": 157.52499771118164,
      "epoch": 0.9383100051840332,
      "grad_norm": 1.0162603855133057,
      "kl": 0.883056640625,
      "learning_rate": 2.605133971932639e-06,
      "loss": 0.1052,
      "reward": 1.8000000417232513,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 905
    },
    {
      "completion_length": 117.125,
      "epoch": 0.9393468118195957,
      "grad_norm": 0.2638833224773407,
      "kl": 0.188232421875,
      "learning_rate": 2.6039085998755895e-06,
      "loss": -0.0455,
      "reward": 1.5,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 1.0,
      "step": 906
    },
    {
      "completion_length": 220.10000228881836,
      "epoch": 0.9403836184551582,
      "grad_norm": 0.5877365469932556,
      "kl": 2.642822265625,
      "learning_rate": 2.602681618588152e-06,
      "loss": 0.2276,
      "reward": 1.3250000327825546,
      "reward_std": 0.31199513375759125,
      "rewards/accuracy_reward": 0.4000000152736902,
      "rewards/format_reward": 0.9250000268220901,
      "step": 907
    },
    {
      "completion_length": 260.37500762939453,
      "epoch": 0.9414204250907205,
      "grad_norm": 1.3724843263626099,
      "kl": 3.5703125,
      "learning_rate": 2.6014530298589674e-06,
      "loss": 0.3882,
      "reward": 1.3000000268220901,
      "reward_std": 0.3479737974703312,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.925000011920929,
      "step": 908
    },
    {
      "completion_length": 233.57500457763672,
      "epoch": 0.942457231726283,
      "grad_norm": 1.3091270923614502,
      "kl": 1.9521484375,
      "learning_rate": 2.60022283547902e-06,
      "loss": 0.1282,
      "reward": 1.375,
      "reward_std": 0.4415663853287697,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 909
    },
    {
      "completion_length": 172.9249973297119,
      "epoch": 0.9434940383618455,
      "grad_norm": 0.9247521758079529,
      "kl": 0.97021484375,
      "learning_rate": 2.5989910372416335e-06,
      "loss": 0.114,
      "reward": 1.350000023841858,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 910
    },
    {
      "completion_length": 195.77500343322754,
      "epoch": 0.944530844997408,
      "grad_norm": 1.2640548944473267,
      "kl": 1.4638671875,
      "learning_rate": 2.59775763694247e-06,
      "loss": 0.1241,
      "reward": 1.225000023841858,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 0.9500000029802322,
      "step": 911
    },
    {
      "completion_length": 269.72500228881836,
      "epoch": 0.9455676516329704,
      "grad_norm": 0.822454035282135,
      "kl": 0.965576171875,
      "learning_rate": 2.596522636379528e-06,
      "loss": 0.2064,
      "reward": 1.5000000298023224,
      "reward_std": 0.30463575199246407,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.925000011920929,
      "step": 912
    },
    {
      "completion_length": 119.125,
      "epoch": 0.9466044582685329,
      "grad_norm": 0.47956565022468567,
      "kl": 0.22607421875,
      "learning_rate": 2.595286037353138e-06,
      "loss": 0.018,
      "reward": 1.675000011920929,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 1.0,
      "step": 913
    },
    {
      "completion_length": 257.4500026702881,
      "epoch": 0.9476412649040954,
      "grad_norm": 1.073453664779663,
      "kl": 2.339599609375,
      "learning_rate": 2.5940478416659605e-06,
      "loss": 0.2984,
      "reward": 1.4000000059604645,
      "reward_std": 0.3407529331743717,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.925000011920929,
      "step": 914
    },
    {
      "completion_length": 169.75000190734863,
      "epoch": 0.9486780715396579,
      "grad_norm": 0.3584200441837311,
      "kl": 0.40869140625,
      "learning_rate": 2.5928080511229837e-06,
      "loss": 0.059,
      "reward": 1.699999988079071,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 915
    },
    {
      "completion_length": 268.6500053405762,
      "epoch": 0.9497148781752203,
      "grad_norm": 0.72416752576828,
      "kl": 1.800537109375,
      "learning_rate": 2.591566667531522e-06,
      "loss": 0.2042,
      "reward": 1.5250000059604645,
      "reward_std": 0.34860679507255554,
      "rewards/accuracy_reward": 0.6250000186264515,
      "rewards/format_reward": 0.9000000208616257,
      "step": 916
    },
    {
      "completion_length": 141.25,
      "epoch": 0.9507516848107828,
      "grad_norm": 0.3592819273471832,
      "kl": 0.32470703125,
      "learning_rate": 2.590323692701209e-06,
      "loss": 0.0149,
      "reward": 1.4000000059604645,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 1.0,
      "step": 917
    },
    {
      "completion_length": 224.8500099182129,
      "epoch": 0.9517884914463453,
      "grad_norm": 1.1734187602996826,
      "kl": 0.6982421875,
      "learning_rate": 2.589079128444001e-06,
      "loss": 0.1334,
      "reward": 1.225000023841858,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 918
    },
    {
      "completion_length": 230.95000457763672,
      "epoch": 0.9528252980819077,
      "grad_norm": 1.34982430934906,
      "kl": 0.759033203125,
      "learning_rate": 2.58783297657417e-06,
      "loss": 0.1402,
      "reward": 1.5500000417232513,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 919
    },
    {
      "completion_length": 417.5250015258789,
      "epoch": 0.9538621047174702,
      "grad_norm": 1.6819778680801392,
      "kl": 1.3720703125,
      "learning_rate": 2.586585238908303e-06,
      "loss": 0.2049,
      "reward": 1.4500000476837158,
      "reward_std": 0.5282425507903099,
      "rewards/accuracy_reward": 0.6250000223517418,
      "rewards/format_reward": 0.8250000178813934,
      "step": 920
    },
    {
      "completion_length": 179.52500343322754,
      "epoch": 0.9548989113530326,
      "grad_norm": 0.4821496307849884,
      "kl": 0.600830078125,
      "learning_rate": 2.585335917265297e-06,
      "loss": 0.0838,
      "reward": 1.4000000208616257,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.42499999701976776,
      "rewards/format_reward": 0.9750000089406967,
      "step": 921
    },
    {
      "completion_length": 266.1250057220459,
      "epoch": 0.9559357179885951,
      "grad_norm": 1.7604442834854126,
      "kl": 1.4638671875,
      "learning_rate": 2.5840850134663604e-06,
      "loss": 0.1657,
      "reward": 1.2500000298023224,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.325000012293458,
      "rewards/format_reward": 0.925000011920929,
      "step": 922
    },
    {
      "completion_length": 157.10000038146973,
      "epoch": 0.9569725246241576,
      "grad_norm": 0.7505897879600525,
      "kl": 0.798583984375,
      "learning_rate": 2.5828325293350064e-06,
      "loss": 0.1055,
      "reward": 1.600000023841858,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 923
    },
    {
      "completion_length": 172.25000190734863,
      "epoch": 0.9580093312597201,
      "grad_norm": 0.8031712174415588,
      "kl": 1.382080078125,
      "learning_rate": 2.581578466697052e-06,
      "loss": 0.1606,
      "reward": 1.3250000327825546,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 924
    },
    {
      "completion_length": 115.32500267028809,
      "epoch": 0.9590461378952825,
      "grad_norm": 0.28106948733329773,
      "kl": 0.29638671875,
      "learning_rate": 2.580322827380616e-06,
      "loss": 0.0158,
      "reward": 1.6500000357627869,
      "reward_std": 0.1369306445121765,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 1.0,
      "step": 925
    },
    {
      "completion_length": 245.8000030517578,
      "epoch": 0.960082944530845,
      "grad_norm": 1.201001763343811,
      "kl": 1.44873046875,
      "learning_rate": 2.5790656132161155e-06,
      "loss": 0.284,
      "reward": 1.425000011920929,
      "reward_std": 0.3678968548774719,
      "rewards/accuracy_reward": 0.5000000167638063,
      "rewards/format_reward": 0.925000011920929,
      "step": 926
    },
    {
      "completion_length": 117.50000190734863,
      "epoch": 0.9611197511664075,
      "grad_norm": 0.5040531158447266,
      "kl": 0.27978515625,
      "learning_rate": 2.577806826036263e-06,
      "loss": -0.001,
      "reward": 1.7250000536441803,
      "reward_std": 0.20539596676826477,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 1.0,
      "step": 927
    },
    {
      "completion_length": 106.87500381469727,
      "epoch": 0.96215655780197,
      "grad_norm": 0.43804383277893066,
      "kl": 0.2802734375,
      "learning_rate": 2.576546467676064e-06,
      "loss": 0.0252,
      "reward": 1.5999999940395355,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 1.0,
      "step": 928
    },
    {
      "completion_length": 106.35000038146973,
      "epoch": 0.9631933644375325,
      "grad_norm": 0.4537351429462433,
      "kl": 0.33935546875,
      "learning_rate": 2.575284539972815e-06,
      "loss": 0.0106,
      "reward": 1.5750000178813934,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 1.0,
      "step": 929
    },
    {
      "completion_length": 242.62500190734863,
      "epoch": 0.9642301710730948,
      "grad_norm": 47.74652099609375,
      "kl": 13.046875,
      "learning_rate": 2.5740210447660993e-06,
      "loss": 1.1571,
      "reward": 1.5,
      "reward_std": 0.3913119062781334,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.9250000268220901,
      "step": 930
    },
    {
      "completion_length": 146.2500057220459,
      "epoch": 0.9652669777086573,
      "grad_norm": 24.46285057067871,
      "kl": 5.71484375,
      "learning_rate": 2.572755983897786e-06,
      "loss": 0.4407,
      "reward": 1.625,
      "reward_std": 0.36053745076060295,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 931
    },
    {
      "completion_length": 257.9250144958496,
      "epoch": 0.9663037843442198,
      "grad_norm": 3.5134782791137695,
      "kl": 3.75927734375,
      "learning_rate": 2.5714893592120267e-06,
      "loss": 0.1903,
      "reward": 1.550000011920929,
      "reward_std": 0.17304782569408417,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 932
    },
    {
      "completion_length": 159.45000457763672,
      "epoch": 0.9673405909797823,
      "grad_norm": 2.399360418319702,
      "kl": 2.16796875,
      "learning_rate": 2.570221172555252e-06,
      "loss": 0.1898,
      "reward": 1.3500000089406967,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9750000089406967,
      "step": 933
    },
    {
      "completion_length": 180.92499923706055,
      "epoch": 0.9683773976153447,
      "grad_norm": 0.9570638537406921,
      "kl": 1.00927734375,
      "learning_rate": 2.5689514257761695e-06,
      "loss": 0.1518,
      "reward": 1.600000023841858,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 934
    },
    {
      "completion_length": 147.29999923706055,
      "epoch": 0.9694142042509072,
      "grad_norm": 10.051190376281738,
      "kl": 3.88720703125,
      "learning_rate": 2.5676801207257614e-06,
      "loss": 0.2523,
      "reward": 1.350000023841858,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 935
    },
    {
      "completion_length": 183.37499809265137,
      "epoch": 0.9704510108864697,
      "grad_norm": 1.349984884262085,
      "kl": 2.6650390625,
      "learning_rate": 2.5664072592572823e-06,
      "loss": 0.2796,
      "reward": 1.550000011920929,
      "reward_std": 0.5247505605220795,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 936
    },
    {
      "completion_length": 158.97500228881836,
      "epoch": 0.9714878175220322,
      "grad_norm": 0.5514269471168518,
      "kl": 0.19921875,
      "learning_rate": 2.565132843226253e-06,
      "loss": 0.0189,
      "reward": 1.4500000178813934,
      "reward_std": 0.3371223993599415,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 937
    },
    {
      "completion_length": 192.82500076293945,
      "epoch": 0.9725246241575947,
      "grad_norm": 22.37244987487793,
      "kl": 6.8203125,
      "learning_rate": 2.563856874490465e-06,
      "loss": 0.634,
      "reward": 1.5000000149011612,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9500000029802322,
      "step": 938
    },
    {
      "completion_length": 110.50000190734863,
      "epoch": 0.973561430793157,
      "grad_norm": 0.673575222492218,
      "kl": 0.44140625,
      "learning_rate": 2.5625793549099676e-06,
      "loss": 0.003,
      "reward": 1.5250000059604645,
      "reward_std": 0.49746808037161827,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9750000089406967,
      "step": 939
    },
    {
      "completion_length": 97.17500114440918,
      "epoch": 0.9745982374287195,
      "grad_norm": 0.3349624276161194,
      "kl": 0.352783203125,
      "learning_rate": 2.561300286347075e-06,
      "loss": 0.0131,
      "reward": 1.699999988079071,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 1.0,
      "step": 940
    },
    {
      "completion_length": 245.49999809265137,
      "epoch": 0.975635044064282,
      "grad_norm": 2.73366641998291,
      "kl": 1.84619140625,
      "learning_rate": 2.560019670666359e-06,
      "loss": 0.2749,
      "reward": 1.4500000178813934,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9000000059604645,
      "step": 941
    },
    {
      "completion_length": 138.72500610351562,
      "epoch": 0.9766718506998445,
      "grad_norm": 0.5809561610221863,
      "kl": 0.8857421875,
      "learning_rate": 2.558737509734645e-06,
      "loss": 0.0642,
      "reward": 1.4500000178813934,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 942
    },
    {
      "completion_length": 132.2750015258789,
      "epoch": 0.9777086573354069,
      "grad_norm": 0.9825214147567749,
      "kl": 0.572998046875,
      "learning_rate": 2.5574538054210123e-06,
      "loss": 0.0908,
      "reward": 1.5750000178813934,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 943
    },
    {
      "completion_length": 172.37500381469727,
      "epoch": 0.9787454639709694,
      "grad_norm": 1.4355342388153076,
      "kl": 0.579833984375,
      "learning_rate": 2.5561685595967895e-06,
      "loss": 0.1356,
      "reward": 1.4500000327825546,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.5250000208616257,
      "rewards/format_reward": 0.9250000268220901,
      "step": 944
    },
    {
      "completion_length": 162.45000076293945,
      "epoch": 0.9797822706065319,
      "grad_norm": 0.6745811104774475,
      "kl": 0.871826171875,
      "learning_rate": 2.5548817741355533e-06,
      "loss": 0.1104,
      "reward": 1.324999988079071,
      "reward_std": 0.20019176974892616,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9500000178813934,
      "step": 945
    },
    {
      "completion_length": 151.02500534057617,
      "epoch": 0.9808190772420944,
      "grad_norm": 0.9925585985183716,
      "kl": 0.726318359375,
      "learning_rate": 2.5535934509131233e-06,
      "loss": 0.075,
      "reward": 1.5999999940395355,
      "reward_std": 0.3533165454864502,
      "rewards/accuracy_reward": 0.6250000186264515,
      "rewards/format_reward": 0.9750000089406967,
      "step": 946
    },
    {
      "completion_length": 163.75000381469727,
      "epoch": 0.9818558838776568,
      "grad_norm": 0.7139672040939331,
      "kl": 0.64013671875,
      "learning_rate": 2.5523035918075613e-06,
      "loss": 0.064,
      "reward": 1.25,
      "reward_std": 0.20019175857305527,
      "rewards/accuracy_reward": 0.2750000096857548,
      "rewards/format_reward": 0.9750000089406967,
      "step": 947
    },
    {
      "completion_length": 152.00000381469727,
      "epoch": 0.9828926905132193,
      "grad_norm": 2.130598306655884,
      "kl": 0.46142578125,
      "learning_rate": 2.551012198699169e-06,
      "loss": 0.0743,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 948
    },
    {
      "completion_length": 196.75000190734863,
      "epoch": 0.9839294971487818,
      "grad_norm": 1.8375300168991089,
      "kl": 0.7880859375,
      "learning_rate": 2.549719273470484e-06,
      "loss": 0.1423,
      "reward": 1.725000023841858,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 949
    },
    {
      "completion_length": 222.3250026702881,
      "epoch": 0.9849663037843442,
      "grad_norm": 5.316202640533447,
      "kl": 4.0380859375,
      "learning_rate": 2.548424818006276e-06,
      "loss": 0.4561,
      "reward": 1.675000011920929,
      "reward_std": 0.3479738123714924,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 950
    },
    {
      "completion_length": 212.09999656677246,
      "epoch": 0.9860031104199067,
      "grad_norm": 0.7121568918228149,
      "kl": 0.4951171875,
      "learning_rate": 2.547128834193547e-06,
      "loss": 0.0836,
      "reward": 1.3500000089406967,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.42499999701976776,
      "rewards/format_reward": 0.9250000268220901,
      "step": 951
    },
    {
      "completion_length": 155.22500038146973,
      "epoch": 0.9870399170554691,
      "grad_norm": 2.3164560794830322,
      "kl": 0.57177734375,
      "learning_rate": 2.5458313239215266e-06,
      "loss": 0.0801,
      "reward": 1.5250000059604645,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 952
    },
    {
      "completion_length": 197.80000495910645,
      "epoch": 0.9880767236910316,
      "grad_norm": 1.507447361946106,
      "kl": 2.1240234375,
      "learning_rate": 2.5445322890816685e-06,
      "loss": 0.2326,
      "reward": 1.6500000357627869,
      "reward_std": 0.4723408557474613,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.925000011920929,
      "step": 953
    },
    {
      "completion_length": 174.45000076293945,
      "epoch": 0.9891135303265941,
      "grad_norm": 8.216100692749023,
      "kl": 0.30615234375,
      "learning_rate": 2.54323173156765e-06,
      "loss": 0.0657,
      "reward": 1.5250000059604645,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.9750000089406967,
      "step": 954
    },
    {
      "completion_length": 121.99999809265137,
      "epoch": 0.9901503369621566,
      "grad_norm": 0.4198542833328247,
      "kl": 0.197021484375,
      "learning_rate": 2.5419296532753684e-06,
      "loss": 0.0393,
      "reward": 1.5999999940395355,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.6000000108033419,
      "rewards/format_reward": 1.0,
      "step": 955
    },
    {
      "completion_length": 268.40000915527344,
      "epoch": 0.991187143597719,
      "grad_norm": 14.27199935913086,
      "kl": 7.554931640625,
      "learning_rate": 2.5406260561029365e-06,
      "loss": 0.7576,
      "reward": 1.425000011920929,
      "reward_std": 0.4363621771335602,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.925000011920929,
      "step": 956
    },
    {
      "completion_length": 219.4249973297119,
      "epoch": 0.9922239502332815,
      "grad_norm": 8.010781288146973,
      "kl": 5.810546875,
      "learning_rate": 2.539320941950683e-06,
      "loss": 0.423,
      "reward": 1.5000000298023224,
      "reward_std": 0.4472135826945305,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.925000011920929,
      "step": 957
    },
    {
      "completion_length": 205.37499809265137,
      "epoch": 0.993260756868844,
      "grad_norm": 6.571266174316406,
      "kl": 5.18212890625,
      "learning_rate": 2.5380143127211464e-06,
      "loss": 0.4746,
      "reward": 1.6250000298023224,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9500000178813934,
      "step": 958
    },
    {
      "completion_length": 168.30000686645508,
      "epoch": 0.9942975635044065,
      "grad_norm": 8.582088470458984,
      "kl": 3.159423828125,
      "learning_rate": 2.5367061703190743e-06,
      "loss": 0.239,
      "reward": 1.475000023841858,
      "reward_std": 0.30463573709130287,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 959
    },
    {
      "completion_length": 225.42501068115234,
      "epoch": 0.995334370139969,
      "grad_norm": 1.5330493450164795,
      "kl": 2.22802734375,
      "learning_rate": 2.535396516651421e-06,
      "loss": 0.2924,
      "reward": 1.4500000178813934,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 960
    },
    {
      "completion_length": 221.0999984741211,
      "epoch": 0.9963711767755313,
      "grad_norm": 3.0697624683380127,
      "kl": 2.22119140625,
      "learning_rate": 2.534085353627343e-06,
      "loss": 0.2233,
      "reward": 1.2750000208616257,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 961
    },
    {
      "completion_length": 362.57499504089355,
      "epoch": 0.9974079834110938,
      "grad_norm": 3.0965917110443115,
      "kl": 0.933349609375,
      "learning_rate": 2.5327726831581968e-06,
      "loss": 0.2787,
      "reward": 1.6500000059604645,
      "reward_std": 0.41643914580345154,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.8750000149011612,
      "step": 962
    },
    {
      "completion_length": 171.12499809265137,
      "epoch": 0.9984447900466563,
      "grad_norm": 1.9162588119506836,
      "kl": 0.46533203125,
      "learning_rate": 2.5314585071575373e-06,
      "loss": 0.031,
      "reward": 1.6500000357627869,
      "reward_std": 0.3731010779738426,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 963
    },
    {
      "completion_length": 260.2000102996826,
      "epoch": 0.9994815966822188,
      "grad_norm": 3.7568962574005127,
      "kl": 2.827392578125,
      "learning_rate": 2.530142827541113e-06,
      "loss": 0.3048,
      "reward": 1.4000000059604645,
      "reward_std": 0.25609344989061356,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.925000011920929,
      "step": 964
    },
    {
      "completion_length": 416.1000213623047,
      "epoch": 1.0,
      "grad_norm": 3.7568962574005127,
      "kl": 3.8359375,
      "learning_rate": 2.5288256462268662e-06,
      "loss": 0.1403,
      "reward": 1.5000000596046448,
      "reward_std": 0.5590169876813889,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.8500000238418579,
      "step": 965
    },
    {
      "completion_length": 213.12500953674316,
      "epoch": 1.0010368066355624,
      "grad_norm": 3.9320881366729736,
      "kl": 4.014892578125,
      "learning_rate": 2.5275069651349257e-06,
      "loss": 0.3534,
      "reward": 1.6250000298023224,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.6750000268220901,
      "rewards/format_reward": 0.9500000178813934,
      "step": 966
    },
    {
      "completion_length": 305.67501068115234,
      "epoch": 1.002073613271125,
      "grad_norm": 7.879758358001709,
      "kl": 0.9921875,
      "learning_rate": 2.526186786187609e-06,
      "loss": 0.2681,
      "reward": 1.5,
      "reward_std": 0.5210216678678989,
      "rewards/accuracy_reward": 0.5999999940395355,
      "rewards/format_reward": 0.9000000357627869,
      "step": 967
    },
    {
      "completion_length": 266.3250026702881,
      "epoch": 1.0031104199066874,
      "grad_norm": 4.553547382354736,
      "kl": 0.9305419921875,
      "learning_rate": 2.524865111309415e-06,
      "loss": 0.2502,
      "reward": 1.4500000029802322,
      "reward_std": 0.3354101814329624,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9000000059604645,
      "step": 968
    },
    {
      "completion_length": 224.07500648498535,
      "epoch": 1.00414722654225,
      "grad_norm": 4.466383934020996,
      "kl": 0.649169921875,
      "learning_rate": 2.523541942427026e-06,
      "loss": 0.1327,
      "reward": 1.6500000357627869,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 969
    },
    {
      "completion_length": 271.1000118255615,
      "epoch": 1.0051840331778124,
      "grad_norm": 1.9565415382385254,
      "kl": 1.4482421875,
      "learning_rate": 2.522217281469299e-06,
      "loss": 0.1794,
      "reward": 1.5250000357627869,
      "reward_std": 0.5048274882137775,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.8750000149011612,
      "step": 970
    },
    {
      "completion_length": 273.3500099182129,
      "epoch": 1.0062208398133747,
      "grad_norm": 2.1400437355041504,
      "kl": 3.146484375,
      "learning_rate": 2.5208911303672685e-06,
      "loss": 0.343,
      "reward": 1.425000011920929,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.5250000022351742,
      "rewards/format_reward": 0.9000000208616257,
      "step": 971
    },
    {
      "completion_length": 232.50000762939453,
      "epoch": 1.0072576464489373,
      "grad_norm": 4.056196689605713,
      "kl": 0.76904296875,
      "learning_rate": 2.5195634910541404e-06,
      "loss": 0.1956,
      "reward": 1.7750000357627869,
      "reward_std": 0.40387551859021187,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 972
    },
    {
      "completion_length": 279.9250087738037,
      "epoch": 1.0082944530844997,
      "grad_norm": 1.1603353023529053,
      "kl": 0.69775390625,
      "learning_rate": 2.5182343654652895e-06,
      "loss": 0.0603,
      "reward": 1.5500000417232513,
      "reward_std": 0.5408061668276787,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9000000208616257,
      "step": 973
    },
    {
      "completion_length": 361.125,
      "epoch": 1.009331259720062,
      "grad_norm": 0.9591207504272461,
      "kl": 2.771484375,
      "learning_rate": 2.5169037555382583e-06,
      "loss": 0.3757,
      "reward": 1.4500000178813934,
      "reward_std": 0.36053745076060295,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.875,
      "step": 974
    },
    {
      "completion_length": 314.2750015258789,
      "epoch": 1.0103680663556247,
      "grad_norm": 3.0426323413848877,
      "kl": 3.90234375,
      "learning_rate": 2.515571663212753e-06,
      "loss": 0.4432,
      "reward": 1.3750000298023224,
      "reward_std": 0.45977723225951195,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9000000357627869,
      "step": 975
    },
    {
      "completion_length": 282.5999984741211,
      "epoch": 1.011404872991187,
      "grad_norm": 2.4559974670410156,
      "kl": 1.0947265625,
      "learning_rate": 2.514238090430639e-06,
      "loss": 0.1671,
      "reward": 1.5999999940395355,
      "reward_std": 0.29270510002970695,
      "rewards/accuracy_reward": 0.6750000212341547,
      "rewards/format_reward": 0.9249999970197678,
      "step": 976
    },
    {
      "completion_length": 169.12500190734863,
      "epoch": 1.0124416796267497,
      "grad_norm": 0.7503567934036255,
      "kl": 1.152587890625,
      "learning_rate": 2.5129030391359427e-06,
      "loss": 0.1039,
      "reward": 1.8250000178813934,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 977
    },
    {
      "completion_length": 222.95000648498535,
      "epoch": 1.013478486262312,
      "grad_norm": 0.9791756868362427,
      "kl": 0.797607421875,
      "learning_rate": 2.5115665112748435e-06,
      "loss": 0.1764,
      "reward": 1.4250000417232513,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 978
    },
    {
      "completion_length": 191.25000381469727,
      "epoch": 1.0145152928978745,
      "grad_norm": 0.6412131786346436,
      "kl": 0.7979736328125,
      "learning_rate": 2.510228508795675e-06,
      "loss": 0.078,
      "reward": 1.475000023841858,
      "reward_std": 0.3913118988275528,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 979
    },
    {
      "completion_length": 298.9750099182129,
      "epoch": 1.015552099533437,
      "grad_norm": 0.37992602586746216,
      "kl": 1.24951171875,
      "learning_rate": 2.5088890336489185e-06,
      "loss": 0.1824,
      "reward": 1.5250000357627869,
      "reward_std": 0.4415663778781891,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9000000059604645,
      "step": 980
    },
    {
      "completion_length": 235.97500610351562,
      "epoch": 1.0165889061689994,
      "grad_norm": 0.8093326091766357,
      "kl": 1.566162109375,
      "learning_rate": 2.5075480877872034e-06,
      "loss": 0.2168,
      "reward": 1.5250000357627869,
      "reward_std": 0.4525563232600689,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9250000268220901,
      "step": 981
    },
    {
      "completion_length": 243.54999923706055,
      "epoch": 1.017625712804562,
      "grad_norm": 3.8579392433166504,
      "kl": 2.362548828125,
      "learning_rate": 2.5062056731653035e-06,
      "loss": 0.2871,
      "reward": 1.550000011920929,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.6000000108033419,
      "rewards/format_reward": 0.9500000029802322,
      "step": 982
    },
    {
      "completion_length": 270.3000087738037,
      "epoch": 1.0186625194401244,
      "grad_norm": 0.3653315007686615,
      "kl": 1.9560546875,
      "learning_rate": 2.5048617917401322e-06,
      "loss": 0.1349,
      "reward": 1.2750000655651093,
      "reward_std": 0.373101070523262,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 983
    },
    {
      "completion_length": 185.5499973297119,
      "epoch": 1.0196993260756868,
      "grad_norm": 0.9105724096298218,
      "kl": 1.087158203125,
      "learning_rate": 2.5035164454707425e-06,
      "loss": 0.1579,
      "reward": 1.5750000476837158,
      "reward_std": 0.3804604485630989,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 984
    },
    {
      "completion_length": 339.25000762939453,
      "epoch": 1.0207361327112494,
      "grad_norm": 3.0632193088531494,
      "kl": 5.13818359375,
      "learning_rate": 2.502169636318322e-06,
      "loss": 0.3792,
      "reward": 1.225000038743019,
      "reward_std": 0.3856646940112114,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.8750000149011612,
      "step": 985
    },
    {
      "completion_length": 172.12500190734863,
      "epoch": 1.0217729393468118,
      "grad_norm": 4.367365837097168,
      "kl": 1.84619140625,
      "learning_rate": 2.5008213662461906e-06,
      "loss": 0.1867,
      "reward": 1.1500000059604645,
      "reward_std": 0.3047742396593094,
      "rewards/accuracy_reward": 0.22500000707805157,
      "rewards/format_reward": 0.925000011920929,
      "step": 986
    },
    {
      "completion_length": 369.50000381469727,
      "epoch": 1.0228097459823744,
      "grad_norm": 0.5143448710441589,
      "kl": 4.598876953125,
      "learning_rate": 2.4994716372197986e-06,
      "loss": 0.4032,
      "reward": 1.3750000447034836,
      "reward_std": 0.3966546170413494,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.8750000149011612,
      "step": 987
    },
    {
      "completion_length": 241.3250026702881,
      "epoch": 1.0238465526179368,
      "grad_norm": 6.717979431152344,
      "kl": 2.7890625,
      "learning_rate": 2.4981204512067235e-06,
      "loss": 0.2272,
      "reward": 1.2750000357627869,
      "reward_std": 0.46162791550159454,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.8750000298023224,
      "step": 988
    },
    {
      "completion_length": 271.475004196167,
      "epoch": 1.0248833592534992,
      "grad_norm": 2.0469918251037598,
      "kl": 2.169921875,
      "learning_rate": 2.4967678101766655e-06,
      "loss": 0.2312,
      "reward": 1.350000023841858,
      "reward_std": 0.44533543661236763,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9000000059604645,
      "step": 989
    },
    {
      "completion_length": 235.65000534057617,
      "epoch": 1.0259201658890618,
      "grad_norm": 2.4610843658447266,
      "kl": 0.60595703125,
      "learning_rate": 2.495413716101446e-06,
      "loss": 0.1438,
      "reward": 1.5250000357627869,
      "reward_std": 0.2686570733785629,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 990
    },
    {
      "completion_length": 177.72500038146973,
      "epoch": 1.0269569725246241,
      "grad_norm": 1.2966387271881104,
      "kl": 0.51416015625,
      "learning_rate": 2.494058170955005e-06,
      "loss": 0.068,
      "reward": 1.699999988079071,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 991
    },
    {
      "completion_length": 297.5000057220459,
      "epoch": 1.0279937791601865,
      "grad_norm": 5.457717418670654,
      "kl": 9.78857421875,
      "learning_rate": 2.4927011767133986e-06,
      "loss": 0.8263,
      "reward": 1.7000000178813934,
      "reward_std": 0.4604101702570915,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.8750000149011612,
      "step": 992
    },
    {
      "completion_length": 134.30000495910645,
      "epoch": 1.0290305857957491,
      "grad_norm": 26.234420776367188,
      "kl": 0.30419921875,
      "learning_rate": 2.4913427353547945e-06,
      "loss": -0.0281,
      "reward": 1.5,
      "reward_std": 0.30463575199246407,
      "rewards/accuracy_reward": 0.5000000167638063,
      "rewards/format_reward": 1.0,
      "step": 993
    },
    {
      "completion_length": 119.55000305175781,
      "epoch": 1.0300673924313115,
      "grad_norm": 0.37444260716438293,
      "kl": 0.25048828125,
      "learning_rate": 2.489982848859469e-06,
      "loss": 0.0221,
      "reward": 1.675000011920929,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.6750000007450581,
      "rewards/format_reward": 1.0,
      "step": 994
    },
    {
      "completion_length": 256.825008392334,
      "epoch": 1.0311041990668741,
      "grad_norm": 1.2725344896316528,
      "kl": 3.269775390625,
      "learning_rate": 2.4886215192098067e-06,
      "loss": 0.4284,
      "reward": 1.5250000059604645,
      "reward_std": 0.492263849824667,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.8750000149011612,
      "step": 995
    },
    {
      "completion_length": 197.6500072479248,
      "epoch": 1.0321410057024365,
      "grad_norm": 2.2583558559417725,
      "kl": 1.155517578125,
      "learning_rate": 2.487258748390295e-06,
      "loss": 0.1296,
      "reward": 1.425000011920929,
      "reward_std": 0.32818929105997086,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.9000000357627869,
      "step": 996
    },
    {
      "completion_length": 436.72500228881836,
      "epoch": 1.0331778123379989,
      "grad_norm": 1.6040884256362915,
      "kl": 2.791015625,
      "learning_rate": 2.485894538387523e-06,
      "loss": 0.3845,
      "reward": 1.475000023841858,
      "reward_std": 0.6526095606386662,
      "rewards/accuracy_reward": 0.6750000268220901,
      "rewards/format_reward": 0.7999999970197678,
      "step": 997
    },
    {
      "completion_length": 271.0500068664551,
      "epoch": 1.0342146189735615,
      "grad_norm": 2.792837142944336,
      "kl": 1.82421875,
      "learning_rate": 2.4845288911901767e-06,
      "loss": 0.2269,
      "reward": 1.300000011920929,
      "reward_std": 0.31719936802983284,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 0.9000000059604645,
      "step": 998
    },
    {
      "completion_length": 281.075008392334,
      "epoch": 1.0352514256091239,
      "grad_norm": 1.2233655452728271,
      "kl": 1.7705078125,
      "learning_rate": 2.483161808789038e-06,
      "loss": 0.2829,
      "reward": 1.550000011920929,
      "reward_std": 0.32818929851055145,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 999
    },
    {
      "completion_length": 294.1000061035156,
      "epoch": 1.0362882322446865,
      "grad_norm": 3.1707940101623535,
      "kl": 1.6826171875,
      "learning_rate": 2.481793293176981e-06,
      "loss": 0.2716,
      "reward": 1.2000000178813934,
      "reward_std": 0.5084580183029175,
      "rewards/accuracy_reward": 0.3000000026077032,
      "rewards/format_reward": 0.9000000357627869,
      "step": 1000
    },
    {
      "completion_length": 174.37500381469727,
      "epoch": 1.0373250388802489,
      "grad_norm": 2.2555198669433594,
      "kl": 0.905029296875,
      "learning_rate": 2.480423346348968e-06,
      "loss": 0.1053,
      "reward": 1.7750000357627869,
      "reward_std": 0.3479738011956215,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 1001
    },
    {
      "completion_length": 344.47500801086426,
      "epoch": 1.0383618455158112,
      "grad_norm": 1.215617299079895,
      "kl": 2.71875,
      "learning_rate": 2.479051970302049e-06,
      "loss": 0.3774,
      "reward": 1.2000000327825546,
      "reward_std": 0.2487340271472931,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1002
    },
    {
      "completion_length": 194.17500495910645,
      "epoch": 1.0393986521513738,
      "grad_norm": 4.0535383224487305,
      "kl": 3.351318359375,
      "learning_rate": 2.4776791670353576e-06,
      "loss": 0.1693,
      "reward": 1.5250000059604645,
      "reward_std": 0.36053745448589325,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1003
    },
    {
      "completion_length": 387.32501220703125,
      "epoch": 1.0404354587869362,
      "grad_norm": 2.0389881134033203,
      "kl": 3.008056640625,
      "learning_rate": 2.476304938550106e-06,
      "loss": 0.3585,
      "reward": 1.5250000357627869,
      "reward_std": 0.4614894278347492,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1004
    },
    {
      "completion_length": 176.37500381469727,
      "epoch": 1.0414722654224986,
      "grad_norm": 5.974527835845947,
      "kl": 0.7509765625,
      "learning_rate": 2.4749292868495854e-06,
      "loss": 0.092,
      "reward": 1.875,
      "reward_std": 0.1604842059314251,
      "rewards/accuracy_reward": 0.9000000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1005
    },
    {
      "completion_length": 245.35000038146973,
      "epoch": 1.0425090720580612,
      "grad_norm": 2.9274890422821045,
      "kl": 1.8115234375,
      "learning_rate": 2.4735522139391623e-06,
      "loss": 0.2333,
      "reward": 1.7500000298023224,
      "reward_std": 0.27950847521424294,
      "rewards/accuracy_reward": 0.8250000104308128,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1006
    },
    {
      "completion_length": 184.90000343322754,
      "epoch": 1.0435458786936236,
      "grad_norm": 0.9111028909683228,
      "kl": 1.2724609375,
      "learning_rate": 2.4721737218262735e-06,
      "loss": 0.0843,
      "reward": 1.5250000059604645,
      "reward_std": 0.3047742620110512,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1007
    },
    {
      "completion_length": 184.0250072479248,
      "epoch": 1.0445826853291862,
      "grad_norm": 28.965343475341797,
      "kl": 0.28125,
      "learning_rate": 2.4707938125204265e-06,
      "loss": 0.0264,
      "reward": 1.5750000178813934,
      "reward_std": 0.40387551113963127,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1008
    },
    {
      "completion_length": 173.77500534057617,
      "epoch": 1.0456194919647486,
      "grad_norm": 0.49043765664100647,
      "kl": 0.2275390625,
      "learning_rate": 2.4694124880331923e-06,
      "loss": 0.0912,
      "reward": 1.625,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.650000024586916,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1009
    },
    {
      "completion_length": 174.9250030517578,
      "epoch": 1.046656298600311,
      "grad_norm": 0.7365868091583252,
      "kl": 0.93701171875,
      "learning_rate": 2.468029750378206e-06,
      "loss": 0.1141,
      "reward": 1.5750000178813934,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.6000000108033419,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1010
    },
    {
      "completion_length": 385.7500114440918,
      "epoch": 1.0476931052358736,
      "grad_norm": 16.921222686767578,
      "kl": 8.4140625,
      "learning_rate": 2.4666456015711653e-06,
      "loss": 0.8527,
      "reward": 1.3000000417232513,
      "reward_std": 0.5373543128371239,
      "rewards/accuracy_reward": 0.44999999925494194,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1011
    },
    {
      "completion_length": 270.9250144958496,
      "epoch": 1.048729911871436,
      "grad_norm": 5.205394744873047,
      "kl": 3.22900390625,
      "learning_rate": 2.4652600436298206e-06,
      "loss": 0.3294,
      "reward": 1.5250000357627869,
      "reward_std": 0.3731010667979717,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.925000011920929,
      "step": 1012
    },
    {
      "completion_length": 210.22500610351562,
      "epoch": 1.0497667185069985,
      "grad_norm": 5.2738037109375,
      "kl": 0.7490234375,
      "learning_rate": 2.4638730785739785e-06,
      "loss": 0.1077,
      "reward": 1.675000011920929,
      "reward_std": 0.3479738235473633,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1013
    },
    {
      "completion_length": 175.07500648498535,
      "epoch": 1.050803525142561,
      "grad_norm": 0.3870754837989807,
      "kl": 1.836669921875,
      "learning_rate": 2.4624847084254983e-06,
      "loss": 0.1884,
      "reward": 1.3750000298023224,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1014
    },
    {
      "completion_length": 139.60000228881836,
      "epoch": 1.0518403317781233,
      "grad_norm": 6.656829357147217,
      "kl": 0.237548828125,
      "learning_rate": 2.461094935208285e-06,
      "loss": -0.0181,
      "reward": 1.2750000059604645,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.27500000409781933,
      "rewards/format_reward": 1.0,
      "step": 1015
    },
    {
      "completion_length": 228.225004196167,
      "epoch": 1.052877138413686,
      "grad_norm": 0.4164898991584778,
      "kl": 0.665283203125,
      "learning_rate": 2.4597037609482902e-06,
      "loss": 0.1427,
      "reward": 1.375,
      "reward_std": 0.3354101963341236,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1016
    },
    {
      "completion_length": 218.77499771118164,
      "epoch": 1.0539139450492483,
      "grad_norm": 3.3085732460021973,
      "kl": 1.551025390625,
      "learning_rate": 2.4583111876735088e-06,
      "loss": 0.1379,
      "reward": 1.2750000059604645,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1017
    },
    {
      "completion_length": 140.6750030517578,
      "epoch": 1.0549507516848107,
      "grad_norm": 0.5721271634101868,
      "kl": 0.28271484375,
      "learning_rate": 2.456917217413973e-06,
      "loss": -0.0236,
      "reward": 1.6500000059604645,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.6500000096857548,
      "rewards/format_reward": 1.0,
      "step": 1018
    },
    {
      "completion_length": 271.0500030517578,
      "epoch": 1.0559875583203733,
      "grad_norm": 1.24607253074646,
      "kl": 0.8017578125,
      "learning_rate": 2.455521852201753e-06,
      "loss": 0.161,
      "reward": 1.1750000268220901,
      "reward_std": 0.3354101963341236,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1019
    },
    {
      "completion_length": 275.5000057220459,
      "epoch": 1.0570243649559357,
      "grad_norm": 1.0766854286193848,
      "kl": 2.4658203125,
      "learning_rate": 2.4541250940709523e-06,
      "loss": 0.2699,
      "reward": 1.3500000089406967,
      "reward_std": 0.4472135938704014,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1020
    },
    {
      "completion_length": 129.25000190734863,
      "epoch": 1.0580611715914983,
      "grad_norm": 1.8902692794799805,
      "kl": 1.03955078125,
      "learning_rate": 2.452726945057704e-06,
      "loss": 0.0865,
      "reward": 1.6000000089406967,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1021
    },
    {
      "completion_length": 120.00000190734863,
      "epoch": 1.0590979782270606,
      "grad_norm": 0.7355002164840698,
      "kl": 0.20166015625,
      "learning_rate": 2.4513274072001697e-06,
      "loss": -0.0098,
      "reward": 1.675000011920929,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 1.0,
      "step": 1022
    },
    {
      "completion_length": 170.75,
      "epoch": 1.060134784862623,
      "grad_norm": 1.266566514968872,
      "kl": 1.27099609375,
      "learning_rate": 2.4499264825385365e-06,
      "loss": 0.0977,
      "reward": 1.3250000178813934,
      "reward_std": 0.3966546505689621,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1023
    },
    {
      "completion_length": 144.475004196167,
      "epoch": 1.0611715914981856,
      "grad_norm": 1.2526934146881104,
      "kl": 0.265625,
      "learning_rate": 2.4485241731150104e-06,
      "loss": 0.028,
      "reward": 1.5250000357627869,
      "reward_std": 0.2253189980983734,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1024
    },
    {
      "completion_length": 165.47500801086426,
      "epoch": 1.062208398133748,
      "grad_norm": 3.019862651824951,
      "kl": 1.36767578125,
      "learning_rate": 2.447120480973819e-06,
      "loss": 0.1348,
      "reward": 1.5250000357627869,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1025
    },
    {
      "completion_length": 129.42500495910645,
      "epoch": 1.0632452047693106,
      "grad_norm": 1.9610161781311035,
      "kl": 0.24755859375,
      "learning_rate": 2.4457154081612032e-06,
      "loss": 0.0116,
      "reward": 1.7750000059604645,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 1.0,
      "step": 1026
    },
    {
      "completion_length": 135.72500038146973,
      "epoch": 1.064282011404873,
      "grad_norm": 3.3690032958984375,
      "kl": 2.0792236328125,
      "learning_rate": 2.444308956725418e-06,
      "loss": 0.077,
      "reward": 1.3250000178813934,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1027
    },
    {
      "completion_length": 142.1250057220459,
      "epoch": 1.0653188180404354,
      "grad_norm": 0.37863561511039734,
      "kl": 0.21630859375,
      "learning_rate": 2.4429011287167277e-06,
      "loss": -0.0329,
      "reward": 1.4750000536441803,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.47500000707805157,
      "rewards/format_reward": 1.0,
      "step": 1028
    },
    {
      "completion_length": 162.57500457763672,
      "epoch": 1.066355624675998,
      "grad_norm": 2.589862585067749,
      "kl": 0.984130859375,
      "learning_rate": 2.441491926187403e-06,
      "loss": 0.1099,
      "reward": 1.4500000178813934,
      "reward_std": 0.24524202942848206,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1029
    },
    {
      "completion_length": 175.65000534057617,
      "epoch": 1.0673924313115604,
      "grad_norm": 0.28419941663742065,
      "kl": 1.521728515625,
      "learning_rate": 2.4400813511917195e-06,
      "loss": 0.1385,
      "reward": 1.625,
      "reward_std": 0.22360678762197495,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1030
    },
    {
      "completion_length": 239.27499771118164,
      "epoch": 1.068429237947123,
      "grad_norm": 5.503000259399414,
      "kl": 1.56494140625,
      "learning_rate": 2.4386694057859516e-06,
      "loss": 0.2001,
      "reward": 1.4250000417232513,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1031
    },
    {
      "completion_length": 217.3750057220459,
      "epoch": 1.0694660445826853,
      "grad_norm": 7.886969566345215,
      "kl": 2.300048828125,
      "learning_rate": 2.4372560920283734e-06,
      "loss": 0.2426,
      "reward": 1.7500000298023224,
      "reward_std": 0.22360678017139435,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1032
    },
    {
      "completion_length": 219.3750057220459,
      "epoch": 1.0705028512182477,
      "grad_norm": 9.453092575073242,
      "kl": 2.71875,
      "learning_rate": 2.435841411979252e-06,
      "loss": 0.234,
      "reward": 1.5,
      "reward_std": 0.3407529331743717,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 1033
    },
    {
      "completion_length": 231.45000076293945,
      "epoch": 1.0715396578538103,
      "grad_norm": 1.6650032997131348,
      "kl": 0.6064453125,
      "learning_rate": 2.434425367700847e-06,
      "loss": 0.1066,
      "reward": 1.5750000178813934,
      "reward_std": 0.21638590842485428,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1034
    },
    {
      "completion_length": 231.65000534057617,
      "epoch": 1.0725764644893727,
      "grad_norm": 3.1586148738861084,
      "kl": 0.628662109375,
      "learning_rate": 2.4330079612574076e-06,
      "loss": 0.1026,
      "reward": 1.4000000208616257,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1035
    },
    {
      "completion_length": 231.07500648498535,
      "epoch": 1.073613271124935,
      "grad_norm": 3.5182924270629883,
      "kl": 2.221923828125,
      "learning_rate": 2.431589194715167e-06,
      "loss": 0.2519,
      "reward": 1.600000023841858,
      "reward_std": 0.34448178485035896,
      "rewards/accuracy_reward": 0.6750000156462193,
      "rewards/format_reward": 0.925000011920929,
      "step": 1036
    },
    {
      "completion_length": 135.10000228881836,
      "epoch": 1.0746500777604977,
      "grad_norm": 2.1421992778778076,
      "kl": 0.193359375,
      "learning_rate": 2.4301690701423425e-06,
      "loss": -0.0087,
      "reward": 1.5750000476837158,
      "reward_std": 0.2487340271472931,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 1.0,
      "step": 1037
    },
    {
      "completion_length": 225.4749984741211,
      "epoch": 1.07568688439606,
      "grad_norm": 0.7044557929039001,
      "kl": 1.645263671875,
      "learning_rate": 2.4287475896091296e-06,
      "loss": 0.2239,
      "reward": 1.475000038743019,
      "reward_std": 0.42900275811553,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1038
    },
    {
      "completion_length": 179.27500534057617,
      "epoch": 1.0767236910316227,
      "grad_norm": 0.8215576410293579,
      "kl": 2.30810546875,
      "learning_rate": 2.4273247551877027e-06,
      "loss": 0.2102,
      "reward": 1.5000000298023224,
      "reward_std": 0.3479738272726536,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1039
    },
    {
      "completion_length": 173.82500076293945,
      "epoch": 1.077760497667185,
      "grad_norm": 7.556845664978027,
      "kl": 2.841796875,
      "learning_rate": 2.4259005689522076e-06,
      "loss": 0.2651,
      "reward": 1.6250000149011612,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 1040
    },
    {
      "completion_length": 179.82500076293945,
      "epoch": 1.0787973043027475,
      "grad_norm": 15.503252983093262,
      "kl": 2.53857421875,
      "learning_rate": 2.4244750329787617e-06,
      "loss": 0.2299,
      "reward": 1.5250000059604645,
      "reward_std": 0.27950849011540413,
      "rewards/accuracy_reward": 0.5500000137835741,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1041
    },
    {
      "completion_length": 164.42499923706055,
      "epoch": 1.07983411093831,
      "grad_norm": 0.4848686456680298,
      "kl": 0.856689453125,
      "learning_rate": 2.4230481493454505e-06,
      "loss": 0.0877,
      "reward": 1.4749999940395355,
      "reward_std": 0.25609345361590385,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1042
    },
    {
      "completion_length": 171.25000190734863,
      "epoch": 1.0808709175738724,
      "grad_norm": 0.8640466928482056,
      "kl": 1.219482421875,
      "learning_rate": 2.4216199201323223e-06,
      "loss": 0.1252,
      "reward": 1.675000011920929,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1043
    },
    {
      "completion_length": 258.52499771118164,
      "epoch": 1.081907724209435,
      "grad_norm": 1.0331501960754395,
      "kl": 2.021484375,
      "learning_rate": 2.4201903474213887e-06,
      "loss": 0.2789,
      "reward": 1.5250000655651093,
      "reward_std": 0.4472136050462723,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9000000357627869,
      "step": 1044
    },
    {
      "completion_length": 217.97500801086426,
      "epoch": 1.0829445308449974,
      "grad_norm": 4.569601058959961,
      "kl": 1.484130859375,
      "learning_rate": 2.4187594332966205e-06,
      "loss": 0.2315,
      "reward": 1.675000011920929,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1045
    },
    {
      "completion_length": 284.87500381469727,
      "epoch": 1.0839813374805598,
      "grad_norm": 3.5887093544006348,
      "kl": 1.0830078125,
      "learning_rate": 2.4173271798439406e-06,
      "loss": 0.2006,
      "reward": 1.4500000476837158,
      "reward_std": 0.5408061631023884,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1046
    },
    {
      "completion_length": 229.8000030517578,
      "epoch": 1.0850181441161224,
      "grad_norm": 1.2264657020568848,
      "kl": 3.2431640625,
      "learning_rate": 2.415893589151227e-06,
      "loss": 0.3404,
      "reward": 1.4750000089406967,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1047
    },
    {
      "completion_length": 237.8000030517578,
      "epoch": 1.0860549507516848,
      "grad_norm": 10.850797653198242,
      "kl": 1.75244140625,
      "learning_rate": 2.414458663308307e-06,
      "loss": 0.175,
      "reward": 1.425000011920929,
      "reward_std": 0.29270510748028755,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1048
    },
    {
      "completion_length": 126.72500038146973,
      "epoch": 1.0870917573872472,
      "grad_norm": 2.8704419136047363,
      "kl": 0.30712890625,
      "learning_rate": 2.4130224044069536e-06,
      "loss": -0.014,
      "reward": 1.600000023841858,
      "reward_std": 0.27228759601712227,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1049
    },
    {
      "completion_length": 211.30001068115234,
      "epoch": 1.0881285640228098,
      "grad_norm": 1.047118067741394,
      "kl": 0.555419921875,
      "learning_rate": 2.4115848145408823e-06,
      "loss": 0.1497,
      "reward": 1.425000011920929,
      "reward_std": 0.3245587982237339,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1050
    },
    {
      "completion_length": 194.82499885559082,
      "epoch": 1.0891653706583722,
      "grad_norm": 2.3370983600616455,
      "kl": 1.25830078125,
      "learning_rate": 2.4101458958057504e-06,
      "loss": 0.2197,
      "reward": 1.6250000298023224,
      "reward_std": 0.3407529182732105,
      "rewards/accuracy_reward": 0.7000000067055225,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1051
    },
    {
      "completion_length": 247.9250030517578,
      "epoch": 1.0902021772939348,
      "grad_norm": 1.1062910556793213,
      "kl": 4.09375,
      "learning_rate": 2.408705650299152e-06,
      "loss": 0.3605,
      "reward": 1.5750000178813934,
      "reward_std": 0.42178187519311905,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1052
    },
    {
      "completion_length": 178.850004196167,
      "epoch": 1.0912389839294971,
      "grad_norm": 3.615429639816284,
      "kl": 1.240478515625,
      "learning_rate": 2.4072640801206136e-06,
      "loss": 0.1906,
      "reward": 1.2000000327825546,
      "reward_std": 0.26129766181111336,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 1053
    },
    {
      "completion_length": 136.82500457763672,
      "epoch": 1.0922757905650595,
      "grad_norm": 1.0081489086151123,
      "kl": 1.263671875,
      "learning_rate": 2.4058211873715953e-06,
      "loss": 0.0754,
      "reward": 1.2500000298023224,
      "reward_std": 0.30463573709130287,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.925000011920929,
      "step": 1054
    },
    {
      "completion_length": 243.55000495910645,
      "epoch": 1.0933125972006221,
      "grad_norm": 0.662563681602478,
      "kl": 3.501220703125,
      "learning_rate": 2.4043769741554834e-06,
      "loss": 0.3914,
      "reward": 1.6250000298023224,
      "reward_std": 0.4525563269853592,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1055
    },
    {
      "completion_length": 130.2750015258789,
      "epoch": 1.0943494038361845,
      "grad_norm": 3.082343816757202,
      "kl": 1.136962890625,
      "learning_rate": 2.40293144257759e-06,
      "loss": 0.0961,
      "reward": 1.7000000178813934,
      "reward_std": 0.30463574081659317,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1056
    },
    {
      "completion_length": 430.7750129699707,
      "epoch": 1.0953862104717471,
      "grad_norm": 10.758148193359375,
      "kl": 5.751953125,
      "learning_rate": 2.4014845947451485e-06,
      "loss": 0.6337,
      "reward": 1.3250000029802322,
      "reward_std": 0.4290027618408203,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.8000000268220901,
      "step": 1057
    },
    {
      "completion_length": 181.2750015258789,
      "epoch": 1.0964230171073095,
      "grad_norm": 1.5399192571640015,
      "kl": 1.5126953125,
      "learning_rate": 2.4000364327673114e-06,
      "loss": 0.1621,
      "reward": 1.6500000059604645,
      "reward_std": 0.4472136050462723,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1058
    },
    {
      "completion_length": 208.05000686645508,
      "epoch": 1.0974598237428719,
      "grad_norm": 1.2210434675216675,
      "kl": 1.3642578125,
      "learning_rate": 2.3985869587551465e-06,
      "loss": 0.1137,
      "reward": 1.475000023841858,
      "reward_std": 0.3804604709148407,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1059
    },
    {
      "completion_length": 127.70000457763672,
      "epoch": 1.0984966303784345,
      "grad_norm": 1.9639644622802734,
      "kl": 1.4990234375,
      "learning_rate": 2.3971361748216347e-06,
      "loss": 0.0412,
      "reward": 1.5750000178813934,
      "reward_std": 0.36588017642498016,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.925000011920929,
      "step": 1060
    },
    {
      "completion_length": 206.90000343322754,
      "epoch": 1.0995334370139969,
      "grad_norm": 1.0273288488388062,
      "kl": 1.095703125,
      "learning_rate": 2.395684083081667e-06,
      "loss": 0.1791,
      "reward": 1.6000000536441803,
      "reward_std": 0.40387552604079247,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1061
    },
    {
      "completion_length": 175.100004196167,
      "epoch": 1.1005702436495595,
      "grad_norm": 1.2841686010360718,
      "kl": 0.399658203125,
      "learning_rate": 2.3942306856520393e-06,
      "loss": 0.092,
      "reward": 1.675000011920929,
      "reward_std": 0.18026872351765633,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1062
    },
    {
      "completion_length": 292.4250030517578,
      "epoch": 1.1016070502851218,
      "grad_norm": 6.283333778381348,
      "kl": 3.5751953125,
      "learning_rate": 2.3927759846514525e-06,
      "loss": 0.4198,
      "reward": 1.425000011920929,
      "reward_std": 0.46511995792388916,
      "rewards/accuracy_reward": 0.5249999910593033,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1063
    },
    {
      "completion_length": 161.7000026702881,
      "epoch": 1.1026438569206842,
      "grad_norm": 0.9103049635887146,
      "kl": 2.02099609375,
      "learning_rate": 2.391319982200507e-06,
      "loss": 0.1582,
      "reward": 1.5500000417232513,
      "reward_std": 0.3046357333660126,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1064
    },
    {
      "completion_length": 203.75000190734863,
      "epoch": 1.1036806635562468,
      "grad_norm": 2.299445152282715,
      "kl": 1.322021484375,
      "learning_rate": 2.389862680421699e-06,
      "loss": 0.2158,
      "reward": 1.5500000268220901,
      "reward_std": 0.28485123440623283,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1065
    },
    {
      "completion_length": 305.350004196167,
      "epoch": 1.1047174701918092,
      "grad_norm": 2.727337121963501,
      "kl": 9.0703125,
      "learning_rate": 2.388404081439422e-06,
      "loss": 0.8764,
      "reward": 1.600000023841858,
      "reward_std": 0.4958943910896778,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1066
    },
    {
      "completion_length": 212.52500534057617,
      "epoch": 1.1057542768273716,
      "grad_norm": 26.0318660736084,
      "kl": 1.458984375,
      "learning_rate": 2.3869441873799577e-06,
      "loss": 0.143,
      "reward": 1.350000023841858,
      "reward_std": 0.4706011861562729,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1067
    },
    {
      "completion_length": 149.62500190734863,
      "epoch": 1.1067910834629342,
      "grad_norm": 1.5431420803070068,
      "kl": 1.5517578125,
      "learning_rate": 2.3854830003714773e-06,
      "loss": 0.0428,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360678762197495,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1068
    },
    {
      "completion_length": 111.67500305175781,
      "epoch": 1.1078278900984966,
      "grad_norm": 1.1982154846191406,
      "kl": 0.256591796875,
      "learning_rate": 2.384020522544036e-06,
      "loss": 0.0384,
      "reward": 1.6250000298023224,
      "reward_std": 0.13693063706159592,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 1.0,
      "step": 1069
    },
    {
      "completion_length": 161.32500076293945,
      "epoch": 1.1088646967340592,
      "grad_norm": 5.154061317443848,
      "kl": 3.11767578125,
      "learning_rate": 2.382556756029571e-06,
      "loss": 0.2601,
      "reward": 1.550000011920929,
      "reward_std": 0.40387552231550217,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1070
    },
    {
      "completion_length": 109.50000190734863,
      "epoch": 1.1099015033696216,
      "grad_norm": 1.1652772426605225,
      "kl": 0.291015625,
      "learning_rate": 2.3810917029618975e-06,
      "loss": 0.018,
      "reward": 1.3000000417232513,
      "reward_std": 0.248734038323164,
      "rewards/accuracy_reward": 0.3000000063329935,
      "rewards/format_reward": 1.0,
      "step": 1071
    },
    {
      "completion_length": 156.07499885559082,
      "epoch": 1.110938310005184,
      "grad_norm": 0.9779496788978577,
      "kl": 0.75146484375,
      "learning_rate": 2.379625365476708e-06,
      "loss": 0.0488,
      "reward": 1.375,
      "reward_std": 0.27228761091828346,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1072
    },
    {
      "completion_length": 154.07500076293945,
      "epoch": 1.1119751166407466,
      "grad_norm": 14.546243667602539,
      "kl": 1.85107421875,
      "learning_rate": 2.378157745711565e-06,
      "loss": 0.1481,
      "reward": 1.4500000178813934,
      "reward_std": 0.1928323432803154,
      "rewards/accuracy_reward": 0.4749999977648258,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1073
    },
    {
      "completion_length": 252.77500343322754,
      "epoch": 1.113011923276309,
      "grad_norm": 0.485859751701355,
      "kl": 1.296875,
      "learning_rate": 2.3766888458059024e-06,
      "loss": 0.2356,
      "reward": 1.5000000596046448,
      "reward_std": 0.4290027692914009,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1074
    },
    {
      "completion_length": 206.87499809265137,
      "epoch": 1.1140487299118715,
      "grad_norm": 1.4165109395980835,
      "kl": 10.2841796875,
      "learning_rate": 2.375218667901018e-06,
      "loss": 0.6785,
      "reward": 1.4249999970197678,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.925000011920929,
      "step": 1075
    },
    {
      "completion_length": 155.850004196167,
      "epoch": 1.115085536547434,
      "grad_norm": 33.7458381652832,
      "kl": 2.77001953125,
      "learning_rate": 2.373747214140074e-06,
      "loss": 0.2534,
      "reward": 1.3750000447034836,
      "reward_std": 0.347973819822073,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 0.925000011920929,
      "step": 1076
    },
    {
      "completion_length": 113.37500190734863,
      "epoch": 1.1161223431829963,
      "grad_norm": 1.9779350757598877,
      "kl": 0.39306640625,
      "learning_rate": 2.3722744866680927e-06,
      "loss": 0.0276,
      "reward": 1.5250000059604645,
      "reward_std": 0.21275536343455315,
      "rewards/accuracy_reward": 0.5750000011175871,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1077
    },
    {
      "completion_length": 96.97500228881836,
      "epoch": 1.117159149818559,
      "grad_norm": 0.3221043050289154,
      "kl": 0.2822265625,
      "learning_rate": 2.370800487631953e-06,
      "loss": 0.0201,
      "reward": 1.800000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.8000000044703484,
      "rewards/format_reward": 1.0,
      "step": 1078
    },
    {
      "completion_length": 168.4500026702881,
      "epoch": 1.1181959564541213,
      "grad_norm": 1.1919106245040894,
      "kl": 0.35498046875,
      "learning_rate": 2.3693252191803867e-06,
      "loss": 0.0311,
      "reward": 1.6250000298023224,
      "reward_std": 0.3479738049209118,
      "rewards/accuracy_reward": 0.7000000216066837,
      "rewards/format_reward": 0.925000011920929,
      "step": 1079
    },
    {
      "completion_length": 105.07500076293945,
      "epoch": 1.1192327630896837,
      "grad_norm": 0.6131631731987,
      "kl": 0.6875,
      "learning_rate": 2.3678486834639764e-06,
      "loss": -0.0015,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1080
    },
    {
      "completion_length": 163.0,
      "epoch": 1.1202695697252463,
      "grad_norm": 0.5884708166122437,
      "kl": 6.12939453125,
      "learning_rate": 2.3663708826351527e-06,
      "loss": 0.5169,
      "reward": 1.1500000208616257,
      "reward_std": 0.4562852270901203,
      "rewards/accuracy_reward": 0.2750000078231096,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1081
    },
    {
      "completion_length": 144.90000534057617,
      "epoch": 1.1213063763608087,
      "grad_norm": 16.477420806884766,
      "kl": 2.0087890625,
      "learning_rate": 2.3648918188481898e-06,
      "loss": 0.1819,
      "reward": 1.225000023841858,
      "reward_std": 0.4525563269853592,
      "rewards/accuracy_reward": 0.32500000670552254,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1082
    },
    {
      "completion_length": 154.07500076293945,
      "epoch": 1.1223431829963713,
      "grad_norm": 2.045259952545166,
      "kl": 1.214111328125,
      "learning_rate": 2.363411494259203e-06,
      "loss": 0.1302,
      "reward": 1.7250000536441803,
      "reward_std": 0.3533165380358696,
      "rewards/accuracy_reward": 0.7750000134110451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1083
    },
    {
      "completion_length": 104.54999923706055,
      "epoch": 1.1233799896319336,
      "grad_norm": 1.0909665822982788,
      "kl": 0.363525390625,
      "learning_rate": 2.361929911026146e-06,
      "loss": 0.0156,
      "reward": 1.6500000059604645,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1084
    },
    {
      "completion_length": 104.95000076293945,
      "epoch": 1.124416796267496,
      "grad_norm": 0.852575957775116,
      "kl": 0.97265625,
      "learning_rate": 2.3604470713088075e-06,
      "loss": 0.0517,
      "reward": 1.6250000298023224,
      "reward_std": 0.1369306445121765,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1085
    },
    {
      "completion_length": 148.725004196167,
      "epoch": 1.1254536029030586,
      "grad_norm": 0.9597511291503906,
      "kl": 0.377685546875,
      "learning_rate": 2.3589629772688055e-06,
      "loss": 0.0358,
      "reward": 1.5999999940395355,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1086
    },
    {
      "completion_length": 170.22500038146973,
      "epoch": 1.126490409538621,
      "grad_norm": 0.8441897034645081,
      "kl": 2.5185546875,
      "learning_rate": 2.3574776310695903e-06,
      "loss": 0.2241,
      "reward": 1.5,
      "reward_std": 0.28858011215925217,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 0.925000011920929,
      "step": 1087
    },
    {
      "completion_length": 212.350004196167,
      "epoch": 1.1275272161741836,
      "grad_norm": 1.9597944021224976,
      "kl": 3.5634765625,
      "learning_rate": 2.3559910348764346e-06,
      "loss": 0.3404,
      "reward": 1.4250000417232513,
      "reward_std": 0.5049660131335258,
      "rewards/accuracy_reward": 0.5500000212341547,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1088
    },
    {
      "completion_length": 152.79999923706055,
      "epoch": 1.128564022809746,
      "grad_norm": 1.717069387435913,
      "kl": 1.9423828125,
      "learning_rate": 2.354503190856434e-06,
      "loss": 0.1411,
      "reward": 1.4000000357627869,
      "reward_std": 0.49226388335227966,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.925000011920929,
      "step": 1089
    },
    {
      "completion_length": 206.45000648498535,
      "epoch": 1.1296008294453084,
      "grad_norm": 1.356911301612854,
      "kl": 1.68505859375,
      "learning_rate": 2.3530141011785045e-06,
      "loss": 0.2599,
      "reward": 1.2500000298023224,
      "reward_std": 0.5049659982323647,
      "rewards/accuracy_reward": 0.4000000096857548,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1090
    },
    {
      "completion_length": 152.62500190734863,
      "epoch": 1.130637636080871,
      "grad_norm": 4.397830486297607,
      "kl": 1.978515625,
      "learning_rate": 2.3515237680133764e-06,
      "loss": 0.2519,
      "reward": 1.3750000149011612,
      "reward_std": 0.43277180567383766,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1091
    },
    {
      "completion_length": 267.5249996185303,
      "epoch": 1.1316744427164334,
      "grad_norm": 1.6721715927124023,
      "kl": 1.498291015625,
      "learning_rate": 2.3500321935335924e-06,
      "loss": 0.2189,
      "reward": 1.3500000089406967,
      "reward_std": 0.40450848639011383,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1092
    },
    {
      "completion_length": 179.25000190734863,
      "epoch": 1.132711249351996,
      "grad_norm": 5.060147762298584,
      "kl": 5.09765625,
      "learning_rate": 2.3485393799135067e-06,
      "loss": 0.4689,
      "reward": 1.5000000149011612,
      "reward_std": 0.3479738272726536,
      "rewards/accuracy_reward": 0.6000000033527613,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1093
    },
    {
      "completion_length": 179.57499885559082,
      "epoch": 1.1337480559875583,
      "grad_norm": 8.466435432434082,
      "kl": 2.309814453125,
      "learning_rate": 2.3470453293292774e-06,
      "loss": 0.2717,
      "reward": 1.5500000417232513,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1094
    },
    {
      "completion_length": 311.2500057220459,
      "epoch": 1.1347848626231207,
      "grad_norm": 7.621610164642334,
      "kl": 5.3955078125,
      "learning_rate": 2.345550043958868e-06,
      "loss": 0.4766,
      "reward": 1.2250000089406967,
      "reward_std": 0.49088017642498016,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.7999999970197678,
      "step": 1095
    },
    {
      "completion_length": 263.1749973297119,
      "epoch": 1.1358216692586833,
      "grad_norm": 1.0027433633804321,
      "kl": 0.97802734375,
      "learning_rate": 2.344053525982041e-06,
      "loss": 0.1937,
      "reward": 1.574999988079071,
      "reward_std": 0.4525563083589077,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1096
    },
    {
      "completion_length": 341.5750026702881,
      "epoch": 1.1368584758942457,
      "grad_norm": 2.4278464317321777,
      "kl": 2.37744140625,
      "learning_rate": 2.3425557775803554e-06,
      "loss": 0.2962,
      "reward": 1.2750000208616257,
      "reward_std": 0.5533697977662086,
      "rewards/accuracy_reward": 0.45000001788139343,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1097
    },
    {
      "completion_length": 266.7000045776367,
      "epoch": 1.137895282529808,
      "grad_norm": 18.119483947753906,
      "kl": 7.452392578125,
      "learning_rate": 2.3410568009371644e-06,
      "loss": 0.7244,
      "reward": 1.675000011920929,
      "reward_std": 0.36588015779852867,
      "rewards/accuracy_reward": 0.8000000193715096,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1098
    },
    {
      "completion_length": 217.1250057220459,
      "epoch": 1.1389320891653707,
      "grad_norm": 1.204911231994629,
      "kl": 1.53857421875,
      "learning_rate": 2.3395565982376117e-06,
      "loss": 0.167,
      "reward": 1.425000011920929,
      "reward_std": 0.3371223956346512,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1099
    },
    {
      "completion_length": 203.77500534057617,
      "epoch": 1.139968895800933,
      "grad_norm": 0.6387304663658142,
      "kl": 1.47216796875,
      "learning_rate": 2.3380551716686286e-06,
      "loss": 0.235,
      "reward": 1.6500000059604645,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1100
    },
    {
      "completion_length": 252.6750087738037,
      "epoch": 1.1410057024364957,
      "grad_norm": 2.940162420272827,
      "kl": 1.165283203125,
      "learning_rate": 2.3365525234189294e-06,
      "loss": 0.1856,
      "reward": 1.1250000447034836,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.2000000048428774,
      "rewards/format_reward": 0.925000011920929,
      "step": 1101
    },
    {
      "completion_length": 341.80001068115234,
      "epoch": 1.142042509072058,
      "grad_norm": 0.830411434173584,
      "kl": 2.815673828125,
      "learning_rate": 2.3350486556790096e-06,
      "loss": 0.445,
      "reward": 1.2999999970197678,
      "reward_std": 0.44219936802983284,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1102
    },
    {
      "completion_length": 150.82500076293945,
      "epoch": 1.1430793157076204,
      "grad_norm": 38.25456237792969,
      "kl": 0.4033203125,
      "learning_rate": 2.333543570641144e-06,
      "loss": 0.036,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1103
    },
    {
      "completion_length": 160.10000610351562,
      "epoch": 1.144116122343183,
      "grad_norm": 1.2388648986816406,
      "kl": 1.23974609375,
      "learning_rate": 2.33203727049938e-06,
      "loss": 0.1099,
      "reward": 1.6500000357627869,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1104
    },
    {
      "completion_length": 128.52500343322754,
      "epoch": 1.1451529289787454,
      "grad_norm": 0.07461313158273697,
      "kl": 0.1943359375,
      "learning_rate": 2.3305297574495363e-06,
      "loss": 0.0078,
      "reward": 1.375,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 1.0,
      "step": 1105
    },
    {
      "completion_length": 266.99999618530273,
      "epoch": 1.1461897356143078,
      "grad_norm": 0.3935621678829193,
      "kl": 2.285888671875,
      "learning_rate": 2.329021033689202e-06,
      "loss": 0.1073,
      "reward": 1.3250000029802322,
      "reward_std": 0.32519175484776497,
      "rewards/accuracy_reward": 0.4250000026077032,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1106
    },
    {
      "completion_length": 140.55000495910645,
      "epoch": 1.1472265422498704,
      "grad_norm": 2.322416067123413,
      "kl": 0.85546875,
      "learning_rate": 2.327511101417729e-06,
      "loss": 0.0957,
      "reward": 1.7750000357627869,
      "reward_std": 0.13693063706159592,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1107
    },
    {
      "completion_length": 182.12500762939453,
      "epoch": 1.1482633488854328,
      "grad_norm": 1.7401773929595947,
      "kl": 1.87109375,
      "learning_rate": 2.3259999628362302e-06,
      "loss": 0.228,
      "reward": 1.5250000059604645,
      "reward_std": 0.34448183327913284,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1108
    },
    {
      "completion_length": 159.72500610351562,
      "epoch": 1.1493001555209954,
      "grad_norm": 1.2455532550811768,
      "kl": 0.74462890625,
      "learning_rate": 2.32448762014758e-06,
      "loss": 0.094,
      "reward": 1.725000023841858,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1109
    },
    {
      "completion_length": 208.05000495910645,
      "epoch": 1.1503369621565578,
      "grad_norm": 28.125736236572266,
      "kl": 7.759521484375,
      "learning_rate": 2.3229740755564063e-06,
      "loss": 0.6062,
      "reward": 1.425000011920929,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1110
    },
    {
      "completion_length": 111.25000190734863,
      "epoch": 1.1513737687921202,
      "grad_norm": 0.134598970413208,
      "kl": 0.265869140625,
      "learning_rate": 2.3214593312690885e-06,
      "loss": 0.0104,
      "reward": 1.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 1.0,
      "step": 1111
    },
    {
      "completion_length": 316.625,
      "epoch": 1.1524105754276828,
      "grad_norm": 1.7135318517684937,
      "kl": 1.057861328125,
      "learning_rate": 2.3199433894937556e-06,
      "loss": 0.2495,
      "reward": 1.3750000298023224,
      "reward_std": 0.45977722480893135,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1112
    },
    {
      "completion_length": 100.82500076293945,
      "epoch": 1.1534473820632452,
      "grad_norm": 1.3429611921310425,
      "kl": 0.369873046875,
      "learning_rate": 2.3184262524402823e-06,
      "loss": 0.0181,
      "reward": 1.675000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 1.0,
      "step": 1113
    },
    {
      "completion_length": 112.92500114440918,
      "epoch": 1.1544841886988078,
      "grad_norm": 0.9468095302581787,
      "kl": 0.24267578125,
      "learning_rate": 2.316907922320286e-06,
      "loss": -0.0086,
      "reward": 1.4000000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 1.0,
      "step": 1114
    },
    {
      "completion_length": 125.9749984741211,
      "epoch": 1.1555209953343701,
      "grad_norm": 0.4861811399459839,
      "kl": 0.239501953125,
      "learning_rate": 2.315388401347122e-06,
      "loss": 0.004,
      "reward": 1.3250000178813934,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.32500001043081284,
      "rewards/format_reward": 1.0,
      "step": 1115
    },
    {
      "completion_length": 163.12500190734863,
      "epoch": 1.1565578019699325,
      "grad_norm": 1.0500177145004272,
      "kl": 1.072509765625,
      "learning_rate": 2.313867691735884e-06,
      "loss": 0.1148,
      "reward": 1.300000011920929,
      "reward_std": 0.15685366839170456,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1116
    },
    {
      "completion_length": 163.45000457763672,
      "epoch": 1.1575946086054951,
      "grad_norm": 0.3982069194316864,
      "kl": 0.38818359375,
      "learning_rate": 2.312345795703395e-06,
      "loss": 0.0074,
      "reward": 1.3750000298023224,
      "reward_std": 0.3371223844587803,
      "rewards/accuracy_reward": 0.42500001564621925,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1117
    },
    {
      "completion_length": 279.8500061035156,
      "epoch": 1.1586314152410575,
      "grad_norm": 6.1592888832092285,
      "kl": 6.94091796875,
      "learning_rate": 2.310822715468211e-06,
      "loss": 0.6748,
      "reward": 1.225000023841858,
      "reward_std": 0.4092182442545891,
      "rewards/accuracy_reward": 0.3500000052154064,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1118
    },
    {
      "completion_length": 139.72500228881836,
      "epoch": 1.15966822187662,
      "grad_norm": 74.81144714355469,
      "kl": 19.76904296875,
      "learning_rate": 2.309298453250612e-06,
      "loss": 1.0662,
      "reward": 1.5000000298023224,
      "reward_std": 0.304635725915432,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1119
    },
    {
      "completion_length": 122.17499923706055,
      "epoch": 1.1607050285121825,
      "grad_norm": 0.7593364715576172,
      "kl": 0.323974609375,
      "learning_rate": 2.307773011272602e-06,
      "loss": -0.0027,
      "reward": 1.3499999940395355,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 1.0,
      "step": 1120
    },
    {
      "completion_length": 264.75000381469727,
      "epoch": 1.1617418351477449,
      "grad_norm": 0.7510677576065063,
      "kl": 2.908203125,
      "learning_rate": 2.3062463917579046e-06,
      "loss": 0.3516,
      "reward": 1.2500000149011612,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.32500001043081284,
      "rewards/format_reward": 0.925000011920929,
      "step": 1121
    },
    {
      "completion_length": 206.10000038146973,
      "epoch": 1.1627786417833075,
      "grad_norm": 3.3034536838531494,
      "kl": 0.33837890625,
      "learning_rate": 2.3047185969319605e-06,
      "loss": 0.1074,
      "reward": 1.625,
      "reward_std": 0.22360678762197495,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1122
    },
    {
      "completion_length": 227.97500228881836,
      "epoch": 1.1638154484188699,
      "grad_norm": 1.0775312185287476,
      "kl": 1.544921875,
      "learning_rate": 2.303189629021922e-06,
      "loss": 0.207,
      "reward": 1.4000000208616257,
      "reward_std": 0.32818931341171265,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1123
    },
    {
      "completion_length": 166.17500114440918,
      "epoch": 1.1648522550544325,
      "grad_norm": 0.43752148747444153,
      "kl": 0.21826171875,
      "learning_rate": 2.3016594902566532e-06,
      "loss": 0.0541,
      "reward": 1.5,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1124
    },
    {
      "completion_length": 135.1750030517578,
      "epoch": 1.1658890616899948,
      "grad_norm": 0.3818998634815216,
      "kl": 0.21142578125,
      "learning_rate": 2.300128182866725e-06,
      "loss": -0.0078,
      "reward": 1.6500000059604645,
      "reward_std": 0.30463575199246407,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 1.0,
      "step": 1125
    },
    {
      "completion_length": 214.30000686645508,
      "epoch": 1.1669258683255572,
      "grad_norm": 1.9454764127731323,
      "kl": 0.822998046875,
      "learning_rate": 2.29859570908441e-06,
      "loss": 0.1374,
      "reward": 1.600000023841858,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1126
    },
    {
      "completion_length": 265.1500053405762,
      "epoch": 1.1679626749611198,
      "grad_norm": 3.658550500869751,
      "kl": 1.494140625,
      "learning_rate": 2.2970620711436836e-06,
      "loss": 0.2519,
      "reward": 1.3250000178813934,
      "reward_std": 0.4290027394890785,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1127
    },
    {
      "completion_length": 141.5250015258789,
      "epoch": 1.1689994815966822,
      "grad_norm": 1.1074241399765015,
      "kl": 0.46044921875,
      "learning_rate": 2.295527271280217e-06,
      "loss": 0.0418,
      "reward": 1.350000023841858,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1128
    },
    {
      "completion_length": 285.9250068664551,
      "epoch": 1.1700362882322446,
      "grad_norm": 0.3166654706001282,
      "kl": 2.0283203125,
      "learning_rate": 2.293991311731375e-06,
      "loss": 0.2676,
      "reward": 1.425000011920929,
      "reward_std": 0.25609345734119415,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 1129
    },
    {
      "completion_length": 207.27500343322754,
      "epoch": 1.1710730948678072,
      "grad_norm": 3.2603349685668945,
      "kl": 2.12890625,
      "learning_rate": 2.2924541947362132e-06,
      "loss": 0.2093,
      "reward": 1.5250000357627869,
      "reward_std": 0.40387551859021187,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1130
    },
    {
      "completion_length": 257.24999809265137,
      "epoch": 1.1721099015033696,
      "grad_norm": 1.6190482378005981,
      "kl": 1.931396484375,
      "learning_rate": 2.2909159225354742e-06,
      "loss": 0.2589,
      "reward": 1.4000000357627869,
      "reward_std": 0.43999266624450684,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 1131
    },
    {
      "completion_length": 309.35000801086426,
      "epoch": 1.1731467081389322,
      "grad_norm": 1.865369439125061,
      "kl": 1.885009765625,
      "learning_rate": 2.2893764973715864e-06,
      "loss": 0.3008,
      "reward": 1.4749999940395355,
      "reward_std": 0.30477428063750267,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1132
    },
    {
      "completion_length": 263.5000057220459,
      "epoch": 1.1741835147744946,
      "grad_norm": 3.0509140491485596,
      "kl": 2.080078125,
      "learning_rate": 2.2878359214886555e-06,
      "loss": 0.1884,
      "reward": 1.3500000089406967,
      "reward_std": 0.3768700994551182,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1133
    },
    {
      "completion_length": 460.1749973297119,
      "epoch": 1.175220321410057,
      "grad_norm": 1.640738606452942,
      "kl": 5.125,
      "learning_rate": 2.2862941971324675e-06,
      "loss": 0.6654,
      "reward": 1.25,
      "reward_std": 0.373101070523262,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1134
    },
    {
      "completion_length": 222.4750099182129,
      "epoch": 1.1762571280456195,
      "grad_norm": 20.201419830322266,
      "kl": 7.2978515625,
      "learning_rate": 2.2847513265504824e-06,
      "loss": 0.7116,
      "reward": 1.600000023841858,
      "reward_std": 0.4525563381612301,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1135
    },
    {
      "completion_length": 154.95000457763672,
      "epoch": 1.177293934681182,
      "grad_norm": 1.1167094707489014,
      "kl": 0.981689453125,
      "learning_rate": 2.2832073119918294e-06,
      "loss": 0.0693,
      "reward": 1.4500000178813934,
      "reward_std": 0.15685367584228516,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1136
    },
    {
      "completion_length": 162.9500026702881,
      "epoch": 1.1783307413167443,
      "grad_norm": 6.627492427825928,
      "kl": 3.576904296875,
      "learning_rate": 2.281662155707307e-06,
      "loss": 0.2332,
      "reward": 1.475000023841858,
      "reward_std": 0.4723408557474613,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.925000011920929,
      "step": 1137
    },
    {
      "completion_length": 171.6500015258789,
      "epoch": 1.179367547952307,
      "grad_norm": 1.2426679134368896,
      "kl": 1.3828125,
      "learning_rate": 2.280115859949377e-06,
      "loss": 0.1562,
      "reward": 1.5000000298023224,
      "reward_std": 0.32455874979496,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1138
    },
    {
      "completion_length": 156.2000026702881,
      "epoch": 1.1804043545878693,
      "grad_norm": 0.8162038326263428,
      "kl": 1.317138671875,
      "learning_rate": 2.2785684269721627e-06,
      "loss": 0.1348,
      "reward": 1.6500000059604645,
      "reward_std": 0.22360678762197495,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1139
    },
    {
      "completion_length": 100.6500015258789,
      "epoch": 1.181441161223432,
      "grad_norm": 0.6516849994659424,
      "kl": 1.010498046875,
      "learning_rate": 2.2770198590314462e-06,
      "loss": -0.038,
      "reward": 1.600000023841858,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.6250000186264515,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1140
    },
    {
      "completion_length": 265.700008392334,
      "epoch": 1.1824779678589943,
      "grad_norm": 8.137269020080566,
      "kl": 3.6572265625,
      "learning_rate": 2.2754701583846622e-06,
      "loss": 0.4887,
      "reward": 1.4500000178813934,
      "reward_std": 0.6761630997061729,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1141
    },
    {
      "completion_length": 164.90000534057617,
      "epoch": 1.1835147744945567,
      "grad_norm": 3.577707529067993,
      "kl": 2.4912109375,
      "learning_rate": 2.2739193272908977e-06,
      "loss": 0.1773,
      "reward": 1.3250000178813934,
      "reward_std": 0.20019173622131348,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1142
    },
    {
      "completion_length": 198.350004196167,
      "epoch": 1.1845515811301193,
      "grad_norm": 3.3634819984436035,
      "kl": 1.25439453125,
      "learning_rate": 2.2723673680108876e-06,
      "loss": 0.1872,
      "reward": 1.7000000178813934,
      "reward_std": 0.33353203535079956,
      "rewards/accuracy_reward": 0.7749999910593033,
      "rewards/format_reward": 0.925000011920929,
      "step": 1143
    },
    {
      "completion_length": 262.3000068664551,
      "epoch": 1.1855883877656817,
      "grad_norm": 1.0182909965515137,
      "kl": 2.65087890625,
      "learning_rate": 2.2708142828070105e-06,
      "loss": 0.2845,
      "reward": 1.5250000357627869,
      "reward_std": 0.3913118988275528,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1144
    },
    {
      "completion_length": 362.57501220703125,
      "epoch": 1.1866251944012443,
      "grad_norm": 7.052261829376221,
      "kl": 6.25390625,
      "learning_rate": 2.2692600739432884e-06,
      "loss": 0.7407,
      "reward": 1.3000000417232513,
      "reward_std": 0.6345372498035431,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1145
    },
    {
      "completion_length": 296.1500053405762,
      "epoch": 1.1876620010368066,
      "grad_norm": 11.791178703308105,
      "kl": 5.19189453125,
      "learning_rate": 2.267704743685379e-06,
      "loss": 0.4376,
      "reward": 1.4500000476837158,
      "reward_std": 0.4291412830352783,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1146
    },
    {
      "completion_length": 250.85000610351562,
      "epoch": 1.188698807672369,
      "grad_norm": 5.4525885581970215,
      "kl": 2.895263671875,
      "learning_rate": 2.2661482943005763e-06,
      "loss": 0.2821,
      "reward": 1.4000000208616257,
      "reward_std": 0.40558773651719093,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1147
    },
    {
      "completion_length": 203.62499809265137,
      "epoch": 1.1897356143079316,
      "grad_norm": 0.6782277822494507,
      "kl": 1.380859375,
      "learning_rate": 2.2645907280578046e-06,
      "loss": 0.171,
      "reward": 1.350000038743019,
      "reward_std": 0.37310105189681053,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1148
    },
    {
      "completion_length": 499.77501678466797,
      "epoch": 1.190772420943494,
      "grad_norm": 2.9671342372894287,
      "kl": 2.22021484375,
      "learning_rate": 2.2630320472276173e-06,
      "loss": 0.3943,
      "reward": 1.225000038743019,
      "reward_std": 0.5967078506946564,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.7750000208616257,
      "step": 1149
    },
    {
      "completion_length": 508.0750045776367,
      "epoch": 1.1918092275790566,
      "grad_norm": 2.0540173053741455,
      "kl": 4.39453125,
      "learning_rate": 2.261472254082192e-06,
      "loss": 0.3224,
      "reward": 1.2000000029802322,
      "reward_std": 0.4987340494990349,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.7500000149011612,
      "step": 1150
    },
    {
      "completion_length": 269.67500495910645,
      "epoch": 1.192846034214619,
      "grad_norm": 2.222440242767334,
      "kl": 2.26806640625,
      "learning_rate": 2.259911350895328e-06,
      "loss": 0.2526,
      "reward": 1.4749999940395355,
      "reward_std": 0.45977720618247986,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1151
    },
    {
      "completion_length": 279.2750053405762,
      "epoch": 1.1938828408501814,
      "grad_norm": 1.225080132484436,
      "kl": 3.06201171875,
      "learning_rate": 2.2583493399424427e-06,
      "loss": 0.4461,
      "reward": 1.2000000178813934,
      "reward_std": 0.478316567838192,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1152
    },
    {
      "completion_length": 504.17501068115234,
      "epoch": 1.194919647485744,
      "grad_norm": 2.603325366973877,
      "kl": 3.29296875,
      "learning_rate": 2.256786223500568e-06,
      "loss": 0.4565,
      "reward": 1.2000000178813934,
      "reward_std": 0.8058728724718094,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.7250000089406967,
      "step": 1153
    },
    {
      "completion_length": 355.2750015258789,
      "epoch": 1.1959564541213064,
      "grad_norm": 4.560014247894287,
      "kl": 3.18359375,
      "learning_rate": 2.2552220038483483e-06,
      "loss": 0.4786,
      "reward": 1.425000011920929,
      "reward_std": 0.7879664748907089,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1154
    },
    {
      "completion_length": 498.1250228881836,
      "epoch": 1.196993260756869,
      "grad_norm": 2.23262882232666,
      "kl": 5.962890625,
      "learning_rate": 2.253656683266035e-06,
      "loss": 0.6254,
      "reward": 1.050000011920929,
      "reward_std": 0.6675183773040771,
      "rewards/accuracy_reward": 0.3750000111758709,
      "rewards/format_reward": 0.6750000268220901,
      "step": 1155
    },
    {
      "completion_length": 533.9000091552734,
      "epoch": 1.1980300673924313,
      "grad_norm": 8.015848159790039,
      "kl": 3.220703125,
      "learning_rate": 2.2520902640354854e-06,
      "loss": 0.4371,
      "reward": 1.2000000029802322,
      "reward_std": 0.5498777888715267,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.7250000238418579,
      "step": 1156
    },
    {
      "completion_length": 398.05001068115234,
      "epoch": 1.1990668740279937,
      "grad_norm": 2.3568358421325684,
      "kl": 3.130859375,
      "learning_rate": 2.2505227484401575e-06,
      "loss": 0.4058,
      "reward": 1.4500000178813934,
      "reward_std": 0.6388007551431656,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.7750000059604645,
      "step": 1157
    },
    {
      "completion_length": 509.17501068115234,
      "epoch": 1.2001036806635563,
      "grad_norm": 2.682429075241089,
      "kl": 3.63671875,
      "learning_rate": 2.248954138765109e-06,
      "loss": 0.5159,
      "reward": 0.8500000238418579,
      "reward_std": 0.4651199206709862,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.7250000089406967,
      "step": 1158
    },
    {
      "completion_length": 699.5250091552734,
      "epoch": 1.2011404872991187,
      "grad_norm": 4.163193225860596,
      "kl": 7.796875,
      "learning_rate": 2.2473844372969906e-06,
      "loss": 0.7881,
      "reward": 1.2000000476837158,
      "reward_std": 0.7290672734379768,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.6500000208616257,
      "step": 1159
    },
    {
      "completion_length": 333.4000053405762,
      "epoch": 1.202177293934681,
      "grad_norm": 1.7668286561965942,
      "kl": 1.573974609375,
      "learning_rate": 2.2458136463240455e-06,
      "loss": 0.3326,
      "reward": 1.3750000298023224,
      "reward_std": 0.5335852652788162,
      "rewards/accuracy_reward": 0.5000000111758709,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1160
    },
    {
      "completion_length": 240.45000839233398,
      "epoch": 1.2032141005702437,
      "grad_norm": 1.1698496341705322,
      "kl": 2.04541015625,
      "learning_rate": 2.244241768136105e-06,
      "loss": 0.2676,
      "reward": 1.3750000149011612,
      "reward_std": 0.3856646865606308,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1161
    },
    {
      "completion_length": 489.1750068664551,
      "epoch": 1.204250907205806,
      "grad_norm": 1.149490237236023,
      "kl": 2.4130859375,
      "learning_rate": 2.242668805024586e-06,
      "loss": 0.4075,
      "reward": 1.2000000178813934,
      "reward_std": 0.5356019288301468,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.75,
      "step": 1162
    },
    {
      "completion_length": 261.7750053405762,
      "epoch": 1.2052877138413687,
      "grad_norm": 6.335738658905029,
      "kl": 2.89794921875,
      "learning_rate": 2.241094759282486e-06,
      "loss": 0.3619,
      "reward": 1.600000023841858,
      "reward_std": 0.5288755036890507,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1163
    },
    {
      "completion_length": 449.8250160217285,
      "epoch": 1.206324520476931,
      "grad_norm": 2.405794620513916,
      "kl": 3.873046875,
      "learning_rate": 2.239519633204381e-06,
      "loss": 0.4872,
      "reward": 1.100000023841858,
      "reward_std": 0.6381677836179733,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.7250000089406967,
      "step": 1164
    },
    {
      "completion_length": 300.4500102996826,
      "epoch": 1.2073613271124934,
      "grad_norm": 1.931716799736023,
      "kl": 1.939208984375,
      "learning_rate": 2.2379434290864216e-06,
      "loss": 0.3559,
      "reward": 1.5000000298023224,
      "reward_std": 0.5156788975000381,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1165
    },
    {
      "completion_length": 504.9000072479248,
      "epoch": 1.208398133748056,
      "grad_norm": 2.2452728748321533,
      "kl": 2.67333984375,
      "learning_rate": 2.2363661492263312e-06,
      "loss": 0.3062,
      "reward": 1.2500000298023224,
      "reward_std": 0.4415663778781891,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.7250000089406967,
      "step": 1166
    },
    {
      "completion_length": 245.92500114440918,
      "epoch": 1.2094349403836184,
      "grad_norm": 3.8794145584106445,
      "kl": 2.95263671875,
      "learning_rate": 2.2347877959234e-06,
      "loss": 0.2499,
      "reward": 1.3750000298023224,
      "reward_std": 0.46511995792388916,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1167
    },
    {
      "completion_length": 343.25000762939453,
      "epoch": 1.2104717470191808,
      "grad_norm": 6.801175117492676,
      "kl": 3.704345703125,
      "learning_rate": 2.233208371478482e-06,
      "loss": 0.4683,
      "reward": 1.350000038743019,
      "reward_std": 0.3354101814329624,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1168
    },
    {
      "completion_length": 398.4750061035156,
      "epoch": 1.2115085536547434,
      "grad_norm": 2.9643471240997314,
      "kl": 5.6796875,
      "learning_rate": 2.231627878193997e-06,
      "loss": 0.5255,
      "reward": 1.350000023841858,
      "reward_std": 0.4958943761885166,
      "rewards/accuracy_reward": 0.5000000111758709,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1169
    },
    {
      "completion_length": 163.37500190734863,
      "epoch": 1.2125453602903058,
      "grad_norm": 11.943863868713379,
      "kl": 2.26806640625,
      "learning_rate": 2.230046318373917e-06,
      "loss": 0.097,
      "reward": 1.425000011920929,
      "reward_std": 0.42963574826717377,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 1170
    },
    {
      "completion_length": 203.0,
      "epoch": 1.2135821669258684,
      "grad_norm": 1.9657776355743408,
      "kl": 1.68310546875,
      "learning_rate": 2.228463694323774e-06,
      "loss": 0.1682,
      "reward": 1.4500000178813934,
      "reward_std": 0.2974148504436016,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 1171
    },
    {
      "completion_length": 242.9749984741211,
      "epoch": 1.2146189735614308,
      "grad_norm": 1.2443591356277466,
      "kl": 2.547607421875,
      "learning_rate": 2.2268800083506476e-06,
      "loss": 0.1818,
      "reward": 1.475000023841858,
      "reward_std": 0.4849044680595398,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.875,
      "step": 1172
    },
    {
      "completion_length": 162.95000076293945,
      "epoch": 1.2156557801969932,
      "grad_norm": 1.3411865234375,
      "kl": 1.81201171875,
      "learning_rate": 2.225295262763168e-06,
      "loss": 0.167,
      "reward": 1.4000000357627869,
      "reward_std": 0.45977720245718956,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1173
    },
    {
      "completion_length": 223.49999809265137,
      "epoch": 1.2166925868325558,
      "grad_norm": 0.7742236852645874,
      "kl": 2.1669921875,
      "learning_rate": 2.2237094598715085e-06,
      "loss": 0.2562,
      "reward": 1.5750000178813934,
      "reward_std": 0.5408061631023884,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.875,
      "step": 1174
    },
    {
      "completion_length": 250.45000457763672,
      "epoch": 1.2177293934681181,
      "grad_norm": 2.619572639465332,
      "kl": 1.431640625,
      "learning_rate": 2.222122601987384e-06,
      "loss": 0.2507,
      "reward": 1.5000000298023224,
      "reward_std": 0.423798531293869,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1175
    },
    {
      "completion_length": 248.00001335144043,
      "epoch": 1.2187662001036808,
      "grad_norm": 1.728539228439331,
      "kl": 2.7685546875,
      "learning_rate": 2.2205346914240487e-06,
      "loss": 0.3616,
      "reward": 1.5250000208616257,
      "reward_std": 0.4472136050462723,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1176
    },
    {
      "completion_length": 254.57500839233398,
      "epoch": 1.2198030067392431,
      "grad_norm": 1.1736640930175781,
      "kl": 2.730224609375,
      "learning_rate": 2.2189457304962886e-06,
      "loss": 0.3847,
      "reward": 1.5,
      "reward_std": 0.4797002375125885,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1177
    },
    {
      "completion_length": 165.75000190734863,
      "epoch": 1.2208398133748055,
      "grad_norm": 0.7769097685813904,
      "kl": 0.267822265625,
      "learning_rate": 2.2173557215204233e-06,
      "loss": 0.0461,
      "reward": 1.6500000059604645,
      "reward_std": 0.21275537833571434,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1178
    },
    {
      "completion_length": 153.87500381469727,
      "epoch": 1.2218766200103681,
      "grad_norm": 0.6941525936126709,
      "kl": 3.181640625,
      "learning_rate": 2.2157646668142998e-06,
      "loss": 0.2306,
      "reward": 1.649999976158142,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1179
    },
    {
      "completion_length": 207.65000915527344,
      "epoch": 1.2229134266459305,
      "grad_norm": 12.102463722229004,
      "kl": 3.49609375,
      "learning_rate": 2.214172568697289e-06,
      "loss": 0.3298,
      "reward": 1.5750000178813934,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1180
    },
    {
      "completion_length": 355.87500381469727,
      "epoch": 1.223950233281493,
      "grad_norm": 2.0645124912261963,
      "kl": 4.75,
      "learning_rate": 2.2125794294902818e-06,
      "loss": 0.5486,
      "reward": 1.100000023841858,
      "reward_std": 0.5769233554601669,
      "rewards/accuracy_reward": 0.27500000409781933,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1181
    },
    {
      "completion_length": 188.42500495910645,
      "epoch": 1.2249870399170555,
      "grad_norm": 4.254176139831543,
      "kl": 2.12890625,
      "learning_rate": 2.2109852515156885e-06,
      "loss": 0.3345,
      "reward": 1.475000023841858,
      "reward_std": 0.40387551859021187,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1182
    },
    {
      "completion_length": 107.5250015258789,
      "epoch": 1.2260238465526179,
      "grad_norm": 0.7304279208183289,
      "kl": 0.62451171875,
      "learning_rate": 2.209390037097434e-06,
      "loss": 0.0252,
      "reward": 1.8250000178813934,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1183
    },
    {
      "completion_length": 200.14999961853027,
      "epoch": 1.2270606531881805,
      "grad_norm": 0.655267596244812,
      "kl": 1.7373046875,
      "learning_rate": 2.207793788560952e-06,
      "loss": 0.1922,
      "reward": 1.6250000298023224,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 1184
    },
    {
      "completion_length": 148.1750030517578,
      "epoch": 1.2280974598237429,
      "grad_norm": 0.9158124923706055,
      "kl": 0.294921875,
      "learning_rate": 2.206196508233185e-06,
      "loss": 0.0711,
      "reward": 1.8750000298023224,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.9000000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1185
    },
    {
      "completion_length": 247.2000026702881,
      "epoch": 1.2291342664593055,
      "grad_norm": 1.4075510501861572,
      "kl": 1.9541015625,
      "learning_rate": 2.20459819844258e-06,
      "loss": 0.13,
      "reward": 1.4000000059604645,
      "reward_std": 0.3840910233557224,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1186
    },
    {
      "completion_length": 171.60000228881836,
      "epoch": 1.2301710730948678,
      "grad_norm": 0.9474394917488098,
      "kl": 0.662841796875,
      "learning_rate": 2.202998861519083e-06,
      "loss": 0.1273,
      "reward": 1.4000000059604645,
      "reward_std": 0.28485123440623283,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1187
    },
    {
      "completion_length": 190.0250015258789,
      "epoch": 1.2312078797304302,
      "grad_norm": 2.5202767848968506,
      "kl": 0.23974609375,
      "learning_rate": 2.2013984997941395e-06,
      "loss": 0.1328,
      "reward": 1.6500000357627869,
      "reward_std": 0.3354102037847042,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1188
    },
    {
      "completion_length": 269.8500061035156,
      "epoch": 1.2322446863659928,
      "grad_norm": 2.44858455657959,
      "kl": 1.68701171875,
      "learning_rate": 2.1997971156006877e-06,
      "loss": 0.2688,
      "reward": 1.2000000327825546,
      "reward_std": 0.335410188883543,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1189
    },
    {
      "completion_length": 348.62500762939453,
      "epoch": 1.2332814930015552,
      "grad_norm": 1.9599698781967163,
      "kl": 3.48486328125,
      "learning_rate": 2.198194711273156e-06,
      "loss": 0.4033,
      "reward": 1.2250000089406967,
      "reward_std": 0.4688488282263279,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1190
    },
    {
      "completion_length": 258.1250057220459,
      "epoch": 1.2343182996371176,
      "grad_norm": 4.182277679443359,
      "kl": 1.91015625,
      "learning_rate": 2.196591289147461e-06,
      "loss": 0.1094,
      "reward": 1.4500000476837158,
      "reward_std": 0.4290027618408203,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.925000011920929,
      "step": 1191
    },
    {
      "completion_length": 348.00000762939453,
      "epoch": 1.2353551062726802,
      "grad_norm": 5.709355354309082,
      "kl": 4.673828125,
      "learning_rate": 2.1949868515610018e-06,
      "loss": 0.5129,
      "reward": 1.5000000298023224,
      "reward_std": 0.6381677538156509,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1192
    },
    {
      "completion_length": 251.725004196167,
      "epoch": 1.2363919129082426,
      "grad_norm": 4.844602584838867,
      "kl": 0.80517578125,
      "learning_rate": 2.1933814008526593e-06,
      "loss": 0.1768,
      "reward": 1.5750000476837158,
      "reward_std": 0.2937843129038811,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 1193
    },
    {
      "completion_length": 297.0750102996826,
      "epoch": 1.237428719543805,
      "grad_norm": 0.785921573638916,
      "kl": 2.92822265625,
      "learning_rate": 2.191774939362789e-06,
      "loss": 0.408,
      "reward": 1.350000023841858,
      "reward_std": 0.6149187088012695,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1194
    },
    {
      "completion_length": 202.75,
      "epoch": 1.2384655261793676,
      "grad_norm": 2.0941972732543945,
      "kl": 1.6484375,
      "learning_rate": 2.1901674694332225e-06,
      "loss": 0.2486,
      "reward": 1.6500000357627869,
      "reward_std": 0.31199512630701065,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1195
    },
    {
      "completion_length": 256.6250057220459,
      "epoch": 1.23950233281493,
      "grad_norm": 3.958770990371704,
      "kl": 1.618896484375,
      "learning_rate": 2.1885589934072593e-06,
      "loss": 0.1736,
      "reward": 1.4250000417232513,
      "reward_std": 0.4415663927793503,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1196
    },
    {
      "completion_length": 257.82500648498535,
      "epoch": 1.2405391394504925,
      "grad_norm": 0.5299058556556702,
      "kl": 0.7626953125,
      "learning_rate": 2.1869495136296666e-06,
      "loss": 0.0838,
      "reward": 1.550000011920929,
      "reward_std": 0.36053745821118355,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1197
    },
    {
      "completion_length": 193.32500457763672,
      "epoch": 1.241575946086055,
      "grad_norm": 1.3431684970855713,
      "kl": 1.329345703125,
      "learning_rate": 2.185339032446675e-06,
      "loss": 0.1783,
      "reward": 1.5500000417232513,
      "reward_std": 0.28485122323036194,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1198
    },
    {
      "completion_length": 199.8250026702881,
      "epoch": 1.2426127527216173,
      "grad_norm": 1.8247193098068237,
      "kl": 1.277587890625,
      "learning_rate": 2.1837275522059746e-06,
      "loss": 0.1962,
      "reward": 1.5500000417232513,
      "reward_std": 0.3479738011956215,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1199
    },
    {
      "completion_length": 335.6750068664551,
      "epoch": 1.24364955935718,
      "grad_norm": 2.906721830368042,
      "kl": 1.59765625,
      "learning_rate": 2.1821150752567115e-06,
      "loss": 0.2166,
      "reward": 1.4500000327825546,
      "reward_std": 0.4038755036890507,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1200
    },
    {
      "completion_length": 151.32500076293945,
      "epoch": 1.2446863659927423,
      "grad_norm": 0.7558800578117371,
      "kl": 0.267822265625,
      "learning_rate": 2.180501603949485e-06,
      "loss": 0.0495,
      "reward": 1.4500000327825546,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1201
    },
    {
      "completion_length": 167.72500610351562,
      "epoch": 1.245723172628305,
      "grad_norm": 0.3539299964904785,
      "kl": 3.014892578125,
      "learning_rate": 2.1788871406363445e-06,
      "loss": 0.2357,
      "reward": 1.5750000178813934,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 1202
    },
    {
      "completion_length": 254.00000762939453,
      "epoch": 1.2467599792638673,
      "grad_norm": 5.103694438934326,
      "kl": 1.91455078125,
      "learning_rate": 2.1772716876707855e-06,
      "loss": 0.2527,
      "reward": 1.3750000298023224,
      "reward_std": 0.3245587609708309,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1203
    },
    {
      "completion_length": 210.57500648498535,
      "epoch": 1.2477967858994297,
      "grad_norm": 0.8210768699645996,
      "kl": 0.304443359375,
      "learning_rate": 2.1756552474077445e-06,
      "loss": 0.1541,
      "reward": 1.7500000298023224,
      "reward_std": 0.3354102149605751,
      "rewards/accuracy_reward": 0.8000000268220901,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1204
    },
    {
      "completion_length": 103.35000228881836,
      "epoch": 1.2488335925349923,
      "grad_norm": 1.1031365394592285,
      "kl": 0.89453125,
      "learning_rate": 2.174037822203599e-06,
      "loss": 0.0896,
      "reward": 1.6500000357627869,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1205
    },
    {
      "completion_length": 192.2999973297119,
      "epoch": 1.2498703991705546,
      "grad_norm": 0.5509751439094543,
      "kl": 3.65380859375,
      "learning_rate": 2.172419414416162e-06,
      "loss": 0.3186,
      "reward": 1.475000023841858,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1206
    },
    {
      "completion_length": 156.47500228881836,
      "epoch": 1.2509072058061173,
      "grad_norm": 15.85297966003418,
      "kl": 1.678466796875,
      "learning_rate": 2.1708000264046784e-06,
      "loss": 0.1949,
      "reward": 1.6500000059604645,
      "reward_std": 0.32818929851055145,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1207
    },
    {
      "completion_length": 170.90000343322754,
      "epoch": 1.2519440124416796,
      "grad_norm": 0.9234507083892822,
      "kl": 1.376220703125,
      "learning_rate": 2.1691796605298223e-06,
      "loss": 0.1696,
      "reward": 1.5000000298023224,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1208
    },
    {
      "completion_length": 107.37500190734863,
      "epoch": 1.252980819077242,
      "grad_norm": 0.49394702911376953,
      "kl": 0.27685546875,
      "learning_rate": 2.1675583191536936e-06,
      "loss": 0.0603,
      "reward": 1.425000011920929,
      "reward_std": 0.18026871606707573,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 1.0,
      "step": 1209
    },
    {
      "completion_length": 118.07500076293945,
      "epoch": 1.2540176257128046,
      "grad_norm": 0.31334245204925537,
      "kl": 0.250244140625,
      "learning_rate": 2.1659360046398127e-06,
      "loss": 0.0077,
      "reward": 1.600000023841858,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.6000000033527613,
      "rewards/format_reward": 1.0,
      "step": 1210
    },
    {
      "completion_length": 263.7000045776367,
      "epoch": 1.255054432348367,
      "grad_norm": 2.3956122398376465,
      "kl": 2.690673828125,
      "learning_rate": 2.164312719353121e-06,
      "loss": 0.3025,
      "reward": 1.6250000298023224,
      "reward_std": 0.4290027767419815,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1211
    },
    {
      "completion_length": 104.75000190734863,
      "epoch": 1.2560912389839296,
      "grad_norm": 1.3687725067138672,
      "kl": 0.489501953125,
      "learning_rate": 2.1626884656599734e-06,
      "loss": 0.0095,
      "reward": 1.675000011920929,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1212
    },
    {
      "completion_length": 207.15000534057617,
      "epoch": 1.257128045619492,
      "grad_norm": 0.5240602493286133,
      "kl": 2.193359375,
      "learning_rate": 2.1610632459281375e-06,
      "loss": 0.2094,
      "reward": 1.375,
      "reward_std": 0.3354102037847042,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 0.925000011920929,
      "step": 1213
    },
    {
      "completion_length": 101.80000305175781,
      "epoch": 1.2581648522550544,
      "grad_norm": 2.601689577102661,
      "kl": 0.537109375,
      "learning_rate": 2.159437062526788e-06,
      "loss": 0.0006,
      "reward": 1.7750000059604645,
      "reward_std": 0.10458250343799591,
      "rewards/accuracy_reward": 0.7999999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1214
    },
    {
      "completion_length": 133.20000076293945,
      "epoch": 1.259201658890617,
      "grad_norm": 0.0533587783575058,
      "kl": 0.259521484375,
      "learning_rate": 2.1578099178265045e-06,
      "loss": 0.0064,
      "reward": 1.4749999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 1.0,
      "step": 1215
    },
    {
      "completion_length": 160.20000076293945,
      "epoch": 1.2602384655261794,
      "grad_norm": 0.5499846935272217,
      "kl": 1.1591796875,
      "learning_rate": 2.1561818141992697e-06,
      "loss": 0.1508,
      "reward": 1.3250000327825546,
      "reward_std": 0.31719937175512314,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1216
    },
    {
      "completion_length": 104.64999961853027,
      "epoch": 1.261275272161742,
      "grad_norm": 0.9835723042488098,
      "kl": 0.248291015625,
      "learning_rate": 2.1545527540184615e-06,
      "loss": 0.0182,
      "reward": 1.7249999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 1.0,
      "step": 1217
    },
    {
      "completion_length": 165.10000610351562,
      "epoch": 1.2623120787973043,
      "grad_norm": 0.298230916261673,
      "kl": 0.4833984375,
      "learning_rate": 2.1529227396588546e-06,
      "loss": 0.0823,
      "reward": 1.7000000178813934,
      "reward_std": 0.22360678017139435,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1218
    },
    {
      "completion_length": 167.60000038146973,
      "epoch": 1.2633488854328667,
      "grad_norm": 3.0810084342956543,
      "kl": 1.11328125,
      "learning_rate": 2.151291773496614e-06,
      "loss": 0.1332,
      "reward": 1.4000000059604645,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1219
    },
    {
      "completion_length": 127.27500343322754,
      "epoch": 1.264385692068429,
      "grad_norm": 0.4667617976665497,
      "kl": 0.2464599609375,
      "learning_rate": 2.149659857909291e-06,
      "loss": 0.0454,
      "reward": 1.350000023841858,
      "reward_std": 0.36053744703531265,
      "rewards/accuracy_reward": 0.3750000111758709,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1220
    },
    {
      "completion_length": 100.97500038146973,
      "epoch": 1.2654224987039917,
      "grad_norm": 0.531950831413269,
      "kl": 0.354248046875,
      "learning_rate": 2.148026995275822e-06,
      "loss": 0.0324,
      "reward": 1.6250000298023224,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.6500000096857548,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1221
    },
    {
      "completion_length": 110.25000190734863,
      "epoch": 1.2664593053395543,
      "grad_norm": 0.46680694818496704,
      "kl": 0.443359375,
      "learning_rate": 2.1463931879765247e-06,
      "loss": 0.0523,
      "reward": 1.675000011920929,
      "reward_std": 0.22894952073693275,
      "rewards/accuracy_reward": 0.6999999955296516,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1222
    },
    {
      "completion_length": 115.62500190734863,
      "epoch": 1.2674961119751167,
      "grad_norm": 0.4408380687236786,
      "kl": 0.2998046875,
      "learning_rate": 2.1447584383930915e-06,
      "loss": -0.0039,
      "reward": 1.5999999940395355,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1223
    },
    {
      "completion_length": 116.375,
      "epoch": 1.268532918610679,
      "grad_norm": 0.40410295128822327,
      "kl": 0.280517578125,
      "learning_rate": 2.143122748908591e-06,
      "loss": 0.0256,
      "reward": 1.475000023841858,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1224
    },
    {
      "completion_length": 94.60000228881836,
      "epoch": 1.2695697252462415,
      "grad_norm": 0.5488646030426025,
      "kl": 0.27587890625,
      "learning_rate": 2.14148612190746e-06,
      "loss": -0.0308,
      "reward": 1.75,
      "reward_std": 0.31719936802983284,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 1.0,
      "step": 1225
    },
    {
      "completion_length": 204.25000190734863,
      "epoch": 1.270606531881804,
      "grad_norm": 1.7064461708068848,
      "kl": 1.0234375,
      "learning_rate": 2.1398485597755035e-06,
      "loss": 0.1092,
      "reward": 1.4749999940395355,
      "reward_std": 0.27950847521424294,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.925000011920929,
      "step": 1226
    },
    {
      "completion_length": 171.100004196167,
      "epoch": 1.2716433385173664,
      "grad_norm": 3.355642080307007,
      "kl": 0.697509765625,
      "learning_rate": 2.138210064899889e-06,
      "loss": 0.068,
      "reward": 1.800000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1227
    },
    {
      "completion_length": 140.57500076293945,
      "epoch": 1.272680145152929,
      "grad_norm": 0.8206340670585632,
      "kl": 0.77099609375,
      "learning_rate": 2.1365706396691432e-06,
      "loss": 0.0585,
      "reward": 1.699999988079071,
      "reward_std": 0.20019177719950676,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1228
    },
    {
      "completion_length": 211.6500015258789,
      "epoch": 1.2737169517884914,
      "grad_norm": 0.5457781553268433,
      "kl": 0.89453125,
      "learning_rate": 2.13493028647315e-06,
      "loss": 0.0865,
      "reward": 1.675000011920929,
      "reward_std": 0.25609346106648445,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1229
    },
    {
      "completion_length": 258.8000030517578,
      "epoch": 1.2747537584240538,
      "grad_norm": 0.5949947237968445,
      "kl": 4.61669921875,
      "learning_rate": 2.1332890077031454e-06,
      "loss": 0.4629,
      "reward": 1.5,
      "reward_std": 0.32818930223584175,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 1230
    },
    {
      "completion_length": 118.4000015258789,
      "epoch": 1.2757905650596164,
      "grad_norm": 12.690463066101074,
      "kl": 0.2841796875,
      "learning_rate": 2.1316468057517154e-06,
      "loss": 0.0061,
      "reward": 1.75,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 1.0,
      "step": 1231
    },
    {
      "completion_length": 165.52500343322754,
      "epoch": 1.2768273716951788,
      "grad_norm": 1.1037925481796265,
      "kl": 1.39501953125,
      "learning_rate": 2.130003683012791e-06,
      "loss": 0.1823,
      "reward": 1.574999988079071,
      "reward_std": 0.27228759229183197,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1232
    },
    {
      "completion_length": 150.10000038146973,
      "epoch": 1.2778641783307414,
      "grad_norm": 0.7952080965042114,
      "kl": 0.2578125,
      "learning_rate": 2.1283596418816447e-06,
      "loss": 0.0616,
      "reward": 1.600000023841858,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1233
    },
    {
      "completion_length": 225.67499923706055,
      "epoch": 1.2789009849663038,
      "grad_norm": 2.4153053760528564,
      "kl": 2.948974609375,
      "learning_rate": 2.1267146847548895e-06,
      "loss": 0.3239,
      "reward": 1.2500000298023224,
      "reward_std": 0.3407529219985008,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1234
    },
    {
      "completion_length": 307.8000087738037,
      "epoch": 1.2799377916018662,
      "grad_norm": 4.414966583251953,
      "kl": 3.880859375,
      "learning_rate": 2.1250688140304736e-06,
      "loss": 0.4783,
      "reward": 1.3000000417232513,
      "reward_std": 0.3046357333660126,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1235
    },
    {
      "completion_length": 210.35000610351562,
      "epoch": 1.2809745982374288,
      "grad_norm": 2.01659893989563,
      "kl": 1.3486328125,
      "learning_rate": 2.1234220321076763e-06,
      "loss": 0.2044,
      "reward": 1.5250000357627869,
      "reward_std": 0.49746809899806976,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1236
    },
    {
      "completion_length": 358.0000057220459,
      "epoch": 1.2820114048729911,
      "grad_norm": 4.004034042358398,
      "kl": 3.3154296875,
      "learning_rate": 2.1217743413871044e-06,
      "loss": 0.3045,
      "reward": 1.2750000357627869,
      "reward_std": 0.4850429967045784,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1237
    },
    {
      "completion_length": 276.2250003814697,
      "epoch": 1.2830482115085537,
      "grad_norm": 0.434862345457077,
      "kl": 6.9912109375,
      "learning_rate": 2.1201257442706914e-06,
      "loss": 0.5656,
      "reward": 1.4750000536441803,
      "reward_std": 0.20539596676826477,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1238
    },
    {
      "completion_length": 161.89999961853027,
      "epoch": 1.2840850181441161,
      "grad_norm": 16.989086151123047,
      "kl": 0.656005859375,
      "learning_rate": 2.1184762431616906e-06,
      "loss": 0.0833,
      "reward": 1.6250000298023224,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1239
    },
    {
      "completion_length": 247.2500057220459,
      "epoch": 1.2851218247796785,
      "grad_norm": 1.4616671800613403,
      "kl": 5.37646484375,
      "learning_rate": 2.1168258404646737e-06,
      "loss": 0.493,
      "reward": 1.800000011920929,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.9000000208616257,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1240
    },
    {
      "completion_length": 257.8250026702881,
      "epoch": 1.2861586314152411,
      "grad_norm": 12.225122451782227,
      "kl": 0.751220703125,
      "learning_rate": 2.1151745385855266e-06,
      "loss": 0.1749,
      "reward": 1.5500000417232513,
      "reward_std": 0.36789683997631073,
      "rewards/accuracy_reward": 0.6250000111758709,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1241
    },
    {
      "completion_length": 136.85000038146973,
      "epoch": 1.2871954380508035,
      "grad_norm": 2.7704896926879883,
      "kl": 1.5244140625,
      "learning_rate": 2.113522339931447e-06,
      "loss": 0.1441,
      "reward": 1.3750000149011612,
      "reward_std": 0.43999267742037773,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1242
    },
    {
      "completion_length": 151.17500114440918,
      "epoch": 1.288232244686366,
      "grad_norm": 0.6818279027938843,
      "kl": 1.033935546875,
      "learning_rate": 2.111869246910937e-06,
      "loss": 0.0704,
      "reward": 1.5250000059604645,
      "reward_std": 0.3047742433845997,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1243
    },
    {
      "completion_length": 309.1500053405762,
      "epoch": 1.2892690513219285,
      "grad_norm": 1.6926321983337402,
      "kl": 2.467529296875,
      "learning_rate": 2.110215261933806e-06,
      "loss": 0.2825,
      "reward": 1.5500000417232513,
      "reward_std": 0.41643911972641945,
      "rewards/accuracy_reward": 0.6500000152736902,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1244
    },
    {
      "completion_length": 262.52500915527344,
      "epoch": 1.2903058579574909,
      "grad_norm": 4.2472052574157715,
      "kl": 1.349609375,
      "learning_rate": 2.1085603874111612e-06,
      "loss": 0.1518,
      "reward": 1.5000000298023224,
      "reward_std": 0.3354101926088333,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.875,
      "step": 1245
    },
    {
      "completion_length": 214.975004196167,
      "epoch": 1.2913426645930535,
      "grad_norm": 2.5802834033966064,
      "kl": 0.95361328125,
      "learning_rate": 2.106904625755408e-06,
      "loss": 0.1651,
      "reward": 1.300000011920929,
      "reward_std": 0.22894951701164246,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1246
    },
    {
      "completion_length": 234.22500610351562,
      "epoch": 1.2923794712286159,
      "grad_norm": 1.8437985181808472,
      "kl": 1.1201171875,
      "learning_rate": 2.1052479793802435e-06,
      "loss": 0.2049,
      "reward": 1.350000023841858,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 1247
    },
    {
      "completion_length": 357.10001373291016,
      "epoch": 1.2934162778641785,
      "grad_norm": 0.7537873983383179,
      "kl": 3.578857421875,
      "learning_rate": 2.103590450700656e-06,
      "loss": 0.4093,
      "reward": 1.0750000327825546,
      "reward_std": 0.5049660094082355,
      "rewards/accuracy_reward": 0.22500000707805157,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1248
    },
    {
      "completion_length": 180.5500030517578,
      "epoch": 1.2944530844997408,
      "grad_norm": 1.090117335319519,
      "kl": 2.805419921875,
      "learning_rate": 2.1019320421329195e-06,
      "loss": 0.197,
      "reward": 1.6000000536441803,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1249
    },
    {
      "completion_length": 309.7250061035156,
      "epoch": 1.2954898911353032,
      "grad_norm": 1.707935094833374,
      "kl": 9.330078125,
      "learning_rate": 2.1002727560945897e-06,
      "loss": 0.7876,
      "reward": 1.2000000029802322,
      "reward_std": 0.3479738011956215,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.875,
      "step": 1250
    },
    {
      "completion_length": 259.02501106262207,
      "epoch": 1.2965266977708656,
      "grad_norm": 26.302583694458008,
      "kl": 1.01171875,
      "learning_rate": 2.098612595004504e-06,
      "loss": 0.1072,
      "reward": 1.4500000327825546,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1251
    },
    {
      "completion_length": 339.9750061035156,
      "epoch": 1.2975635044064282,
      "grad_norm": 2.3541882038116455,
      "kl": 4.7470703125,
      "learning_rate": 2.0969515612827717e-06,
      "loss": 0.5587,
      "reward": 1.5750000327825546,
      "reward_std": 0.5163118913769722,
      "rewards/accuracy_reward": 0.725000012665987,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1252
    },
    {
      "completion_length": 308.075008392334,
      "epoch": 1.2986003110419908,
      "grad_norm": 5.078311443328857,
      "kl": 1.603515625,
      "learning_rate": 2.0952896573507772e-06,
      "loss": 0.2998,
      "reward": 1.5250000208616257,
      "reward_std": 0.3297629877924919,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1253
    },
    {
      "completion_length": 181.25000762939453,
      "epoch": 1.2996371176775532,
      "grad_norm": 1.3796796798706055,
      "kl": 1.64599609375,
      "learning_rate": 2.0936268856311727e-06,
      "loss": 0.1874,
      "reward": 1.5,
      "reward_std": 0.3479738049209118,
      "rewards/accuracy_reward": 0.5999999940395355,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1254
    },
    {
      "completion_length": 295.6750087738037,
      "epoch": 1.3006739243131156,
      "grad_norm": 2.7654173374176025,
      "kl": 9.52099609375,
      "learning_rate": 2.0919632485478746e-06,
      "loss": 0.8503,
      "reward": 1.5000000298023224,
      "reward_std": 0.32818928360939026,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1255
    },
    {
      "completion_length": 267.07500648498535,
      "epoch": 1.301710730948678,
      "grad_norm": 37.04022216796875,
      "kl": 1.9921875,
      "learning_rate": 2.090298748526061e-06,
      "loss": 0.1794,
      "reward": 1.300000011920929,
      "reward_std": 0.45977720245718956,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.8250000327825546,
      "step": 1256
    },
    {
      "completion_length": 344.2500057220459,
      "epoch": 1.3027475375842406,
      "grad_norm": 9.935186386108398,
      "kl": 13.21484375,
      "learning_rate": 2.08863338799217e-06,
      "loss": 1.1489,
      "reward": 1.225000038743019,
      "reward_std": 0.4472135864198208,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1257
    },
    {
      "completion_length": 280.99999713897705,
      "epoch": 1.303784344219803,
      "grad_norm": 26.916645050048828,
      "kl": 6.21875,
      "learning_rate": 2.0869671693738904e-06,
      "loss": 0.3927,
      "reward": 1.5250000059604645,
      "reward_std": 0.40387552976608276,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1258
    },
    {
      "completion_length": 246.9000015258789,
      "epoch": 1.3048211508553655,
      "grad_norm": 8.15865421295166,
      "kl": 4.12744140625,
      "learning_rate": 2.0853000951001652e-06,
      "loss": 0.348,
      "reward": 1.1000000089406967,
      "reward_std": 0.3768701143562794,
      "rewards/accuracy_reward": 0.25000000186264515,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1259
    },
    {
      "completion_length": 472.5750045776367,
      "epoch": 1.305857957490928,
      "grad_norm": 6.539979934692383,
      "kl": 7.273193359375,
      "learning_rate": 2.083632167601183e-06,
      "loss": 0.5544,
      "reward": 1.3500000536441803,
      "reward_std": 0.4723408482968807,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.7500000149011612,
      "step": 1260
    },
    {
      "completion_length": 333.7750129699707,
      "epoch": 1.3068947641264903,
      "grad_norm": 12.980135917663574,
      "kl": 3.122314453125,
      "learning_rate": 2.081963389308376e-06,
      "loss": 0.3177,
      "reward": 1.475000023841858,
      "reward_std": 0.5210216455161572,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1261
    },
    {
      "completion_length": 358.12501525878906,
      "epoch": 1.307931570762053,
      "grad_norm": 5.874717712402344,
      "kl": 4.38330078125,
      "learning_rate": 2.080293762654419e-06,
      "loss": 0.4428,
      "reward": 1.4250000268220901,
      "reward_std": 0.4290027692914009,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1262
    },
    {
      "completion_length": 479.4750061035156,
      "epoch": 1.3089683773976153,
      "grad_norm": 48.87544250488281,
      "kl": 14.296875,
      "learning_rate": 2.07862329007322e-06,
      "loss": 1.1627,
      "reward": 1.100000038743019,
      "reward_std": 0.4415663853287697,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.75,
      "step": 1263
    },
    {
      "completion_length": 372.8250045776367,
      "epoch": 1.310005184033178,
      "grad_norm": 28.217016220092773,
      "kl": 19.96484375,
      "learning_rate": 2.076951973999923e-06,
      "loss": 0.7583,
      "reward": 1.2000000476837158,
      "reward_std": 0.5535083115100861,
      "rewards/accuracy_reward": 0.4000000115483999,
      "rewards/format_reward": 0.800000011920929,
      "step": 1264
    },
    {
      "completion_length": 453.80001640319824,
      "epoch": 1.3110419906687403,
      "grad_norm": 5.039831638336182,
      "kl": 2.38525390625,
      "learning_rate": 2.0752798168709e-06,
      "loss": 0.3843,
      "reward": 1.2000000476837158,
      "reward_std": 0.40387551859021187,
      "rewards/accuracy_reward": 0.3750000111758709,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1265
    },
    {
      "completion_length": 349.4000053405762,
      "epoch": 1.3120787973043027,
      "grad_norm": 9.782047271728516,
      "kl": 4.740234375,
      "learning_rate": 2.07360682112375e-06,
      "loss": 0.5457,
      "reward": 1.3250000178813934,
      "reward_std": 0.6726711243391037,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1266
    },
    {
      "completion_length": 303.5500087738037,
      "epoch": 1.3131156039398653,
      "grad_norm": 3.975912570953369,
      "kl": 1.66748046875,
      "learning_rate": 2.071932989197294e-06,
      "loss": 0.2409,
      "reward": 1.4500000178813934,
      "reward_std": 0.42963578924536705,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1267
    },
    {
      "completion_length": 201.92500686645508,
      "epoch": 1.3141524105754276,
      "grad_norm": 2.8953194618225098,
      "kl": 0.730224609375,
      "learning_rate": 2.0702583235315716e-06,
      "loss": 0.1217,
      "reward": 1.4749999940395355,
      "reward_std": 0.27950849011540413,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1268
    },
    {
      "completion_length": 375.975004196167,
      "epoch": 1.3151892172109902,
      "grad_norm": 1.7686223983764648,
      "kl": 1.0029296875,
      "learning_rate": 2.0685828265678388e-06,
      "loss": 0.2452,
      "reward": 1.2750000208616257,
      "reward_std": 0.5288755223155022,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.8000000193715096,
      "step": 1269
    },
    {
      "completion_length": 157.02499771118164,
      "epoch": 1.3162260238465526,
      "grad_norm": 1.5042879581451416,
      "kl": 1.02978515625,
      "learning_rate": 2.066906500748563e-06,
      "loss": 0.1275,
      "reward": 1.800000011920929,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.8500000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1270
    },
    {
      "completion_length": 257.100004196167,
      "epoch": 1.317262830482115,
      "grad_norm": 2.236104965209961,
      "kl": 0.393798828125,
      "learning_rate": 2.065229348517419e-06,
      "loss": 0.0451,
      "reward": 1.5250000059604645,
      "reward_std": 0.21638590842485428,
      "rewards/accuracy_reward": 0.6000000033527613,
      "rewards/format_reward": 0.925000011920929,
      "step": 1271
    },
    {
      "completion_length": 295.95000648498535,
      "epoch": 1.3182996371176776,
      "grad_norm": 0.46052783727645874,
      "kl": 2.16015625,
      "learning_rate": 2.0635513723192876e-06,
      "loss": 0.1901,
      "reward": 1.475000023841858,
      "reward_std": 0.41643913462758064,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1272
    },
    {
      "completion_length": 356.8250045776367,
      "epoch": 1.31933644375324,
      "grad_norm": 6.156704425811768,
      "kl": 2.18310546875,
      "learning_rate": 2.06187257460025e-06,
      "loss": 0.4086,
      "reward": 1.2750000357627869,
      "reward_std": 0.6291945055127144,
      "rewards/accuracy_reward": 0.42500000819563866,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1273
    },
    {
      "completion_length": 139.95000457763672,
      "epoch": 1.3203732503888026,
      "grad_norm": 6.010787487030029,
      "kl": 2.75927734375,
      "learning_rate": 2.0601929578075857e-06,
      "loss": 0.2277,
      "reward": 1.600000023841858,
      "reward_std": 0.30477424710989,
      "rewards/accuracy_reward": 0.6750000156462193,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1274
    },
    {
      "completion_length": 162.87500190734863,
      "epoch": 1.321410057024365,
      "grad_norm": 1.2043706178665161,
      "kl": 0.456298828125,
      "learning_rate": 2.0585125243897673e-06,
      "loss": 0.0403,
      "reward": 1.5250000357627869,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.5500000137835741,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1275
    },
    {
      "completion_length": 306.22500228881836,
      "epoch": 1.3224468636599274,
      "grad_norm": 4.1956706047058105,
      "kl": 0.492919921875,
      "learning_rate": 2.0568312767964573e-06,
      "loss": 0.1374,
      "reward": 1.125,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1276
    },
    {
      "completion_length": 276.97501373291016,
      "epoch": 1.32348367029549,
      "grad_norm": 0.5680062770843506,
      "kl": 1.7470703125,
      "learning_rate": 2.0551492174785077e-06,
      "loss": 0.1708,
      "reward": 1.475000023841858,
      "reward_std": 0.5769233480095863,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1277
    },
    {
      "completion_length": 200.02500343322754,
      "epoch": 1.3245204769310523,
      "grad_norm": 6.632391929626465,
      "kl": 0.780029296875,
      "learning_rate": 2.053466348887951e-06,
      "loss": 0.1845,
      "reward": 1.800000011920929,
      "reward_std": 0.40387552604079247,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 1278
    },
    {
      "completion_length": 244.70000457763672,
      "epoch": 1.325557283566615,
      "grad_norm": 3.6601107120513916,
      "kl": 0.83251953125,
      "learning_rate": 2.0517826734780003e-06,
      "loss": 0.1642,
      "reward": 1.4500000178813934,
      "reward_std": 0.3046357333660126,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1279
    },
    {
      "completion_length": 99.9000015258789,
      "epoch": 1.3265940902021773,
      "grad_norm": 0.35116100311279297,
      "kl": 0.29541015625,
      "learning_rate": 2.050098193703046e-06,
      "loss": 0.0298,
      "reward": 1.800000011920929,
      "reward_std": 0.16770510375499725,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1280
    },
    {
      "completion_length": 151.14999961853027,
      "epoch": 1.3276308968377397,
      "grad_norm": 2.063858985900879,
      "kl": 1.90966796875,
      "learning_rate": 2.04841291201865e-06,
      "loss": 0.2199,
      "reward": 1.4249999821186066,
      "reward_std": 0.21638590842485428,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 1281
    },
    {
      "completion_length": 202.80000686645508,
      "epoch": 1.328667703473302,
      "grad_norm": 1.1473495960235596,
      "kl": 0.9140625,
      "learning_rate": 2.0467268308815423e-06,
      "loss": 0.1356,
      "reward": 1.550000011920929,
      "reward_std": 0.4723408445715904,
      "rewards/accuracy_reward": 0.6250000037252903,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1282
    },
    {
      "completion_length": 108.72500419616699,
      "epoch": 1.3297045101088647,
      "grad_norm": 0.49558553099632263,
      "kl": 0.28662109375,
      "learning_rate": 2.0450399527496203e-06,
      "loss": 0.0247,
      "reward": 1.5750000178813934,
      "reward_std": 0.20539596676826477,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 1.0,
      "step": 1283
    },
    {
      "completion_length": 221.8250026702881,
      "epoch": 1.330741316744427,
      "grad_norm": 0.9170733094215393,
      "kl": 1.209716796875,
      "learning_rate": 2.043352280081942e-06,
      "loss": 0.2026,
      "reward": 1.5500000417232513,
      "reward_std": 0.5659333989024162,
      "rewards/accuracy_reward": 0.6249999925494194,
      "rewards/format_reward": 0.925000011920929,
      "step": 1284
    },
    {
      "completion_length": 179.37500381469727,
      "epoch": 1.3317781233799897,
      "grad_norm": 0.7182348966598511,
      "kl": 0.850341796875,
      "learning_rate": 2.041663815338724e-06,
      "loss": 0.1016,
      "reward": 1.5000000149011612,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1285
    },
    {
      "completion_length": 164.39999961853027,
      "epoch": 1.332814930015552,
      "grad_norm": 2.3556573390960693,
      "kl": 2.061279296875,
      "learning_rate": 2.039974560981337e-06,
      "loss": 0.1464,
      "reward": 1.6500000059604645,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1286
    },
    {
      "completion_length": 309.7999954223633,
      "epoch": 1.3338517366511145,
      "grad_norm": 15.727317810058594,
      "kl": 16.7392578125,
      "learning_rate": 2.0382845194723033e-06,
      "loss": 1.4665,
      "reward": 1.1250000298023224,
      "reward_std": 0.6401844471693039,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.800000011920929,
      "step": 1287
    },
    {
      "completion_length": 123.27500343322754,
      "epoch": 1.334888543286677,
      "grad_norm": 30.45476722717285,
      "kl": 0.635986328125,
      "learning_rate": 2.036593693275293e-06,
      "loss": 0.019,
      "reward": 1.5250000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1288
    },
    {
      "completion_length": 248.42500686645508,
      "epoch": 1.3359253499222394,
      "grad_norm": 0.7280150055885315,
      "kl": 1.158203125,
      "learning_rate": 2.0349020848551197e-06,
      "loss": 0.1204,
      "reward": 1.275000050663948,
      "reward_std": 0.4849044606089592,
      "rewards/accuracy_reward": 0.42500001564621925,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1289
    },
    {
      "completion_length": 212.7750072479248,
      "epoch": 1.336962156557802,
      "grad_norm": 1.562294840812683,
      "kl": 0.937255859375,
      "learning_rate": 2.0332096966777368e-06,
      "loss": 0.131,
      "reward": 1.2750000357627869,
      "reward_std": 0.3804604709148407,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1290
    },
    {
      "completion_length": 317.0250053405762,
      "epoch": 1.3379989631933644,
      "grad_norm": 6.326425075531006,
      "kl": 5.015625,
      "learning_rate": 2.0315165312102347e-06,
      "loss": 0.4916,
      "reward": 1.449999988079071,
      "reward_std": 0.4723408445715904,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1291
    },
    {
      "completion_length": 169.32500076293945,
      "epoch": 1.3390357698289268,
      "grad_norm": 39.2647819519043,
      "kl": 11.615234375,
      "learning_rate": 2.0298225909208375e-06,
      "loss": 0.8409,
      "reward": 1.4500000476837158,
      "reward_std": 0.20539595931768417,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1292
    },
    {
      "completion_length": 164.32500457763672,
      "epoch": 1.3400725764644894,
      "grad_norm": 1.0471571683883667,
      "kl": 0.7060546875,
      "learning_rate": 2.0281278782788985e-06,
      "loss": 0.0951,
      "reward": 1.4749999940395355,
      "reward_std": 0.17304782569408417,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1293
    },
    {
      "completion_length": 199.62500381469727,
      "epoch": 1.3411093831000518,
      "grad_norm": 0.6253998875617981,
      "kl": 1.9755859375,
      "learning_rate": 2.0264323957548966e-06,
      "loss": 0.2257,
      "reward": 1.475000023841858,
      "reward_std": 0.3047742620110512,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.925000011920929,
      "step": 1294
    },
    {
      "completion_length": 267.0750026702881,
      "epoch": 1.3421461897356144,
      "grad_norm": 1.7353161573410034,
      "kl": 2.8291015625,
      "learning_rate": 2.0247361458204336e-06,
      "loss": 0.3355,
      "reward": 1.1250000298023224,
      "reward_std": 0.36053743213415146,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1295
    },
    {
      "completion_length": 255.30000495910645,
      "epoch": 1.3431829963711768,
      "grad_norm": 1.32989501953125,
      "kl": 1.211181640625,
      "learning_rate": 2.0230391309482278e-06,
      "loss": 0.2027,
      "reward": 1.425000011920929,
      "reward_std": 0.47247935086488724,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.925000011920929,
      "step": 1296
    },
    {
      "completion_length": 353.1750030517578,
      "epoch": 1.3442198030067392,
      "grad_norm": 0.5341919660568237,
      "kl": 4.642578125,
      "learning_rate": 2.0213413536121164e-06,
      "loss": 0.5698,
      "reward": 1.3750000298023224,
      "reward_std": 0.7769765928387642,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1297
    },
    {
      "completion_length": 345.5750141143799,
      "epoch": 1.3452566096423018,
      "grad_norm": 3.6098403930664062,
      "kl": 2.868896484375,
      "learning_rate": 2.0196428162870455e-06,
      "loss": 0.4044,
      "reward": 1.350000023841858,
      "reward_std": 0.5732928328216076,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1298
    },
    {
      "completion_length": 258.3500061035156,
      "epoch": 1.3462934162778641,
      "grad_norm": 3.2862420082092285,
      "kl": 0.93603515625,
      "learning_rate": 2.017943521449069e-06,
      "loss": 0.1827,
      "reward": 1.225000023841858,
      "reward_std": 0.42178186774253845,
      "rewards/accuracy_reward": 0.3250000011175871,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1299
    },
    {
      "completion_length": 159.47500038146973,
      "epoch": 1.3473302229134267,
      "grad_norm": 2.2226035594940186,
      "kl": 0.3154296875,
      "learning_rate": 2.016243471575346e-06,
      "loss": 0.0745,
      "reward": 1.600000023841858,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.6250000111758709,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1300
    },
    {
      "completion_length": 241.12500190734863,
      "epoch": 1.3483670295489891,
      "grad_norm": 4.94175386428833,
      "kl": 3.16162109375,
      "learning_rate": 2.014542669144136e-06,
      "loss": 0.3438,
      "reward": 1.5750000178813934,
      "reward_std": 0.4290027543902397,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 1301
    },
    {
      "completion_length": 285.5750045776367,
      "epoch": 1.3494038361845515,
      "grad_norm": 0.8761221170425415,
      "kl": 1.416259765625,
      "learning_rate": 2.012841116634796e-06,
      "loss": 0.2923,
      "reward": 1.4250000417232513,
      "reward_std": 0.5894870012998581,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1302
    },
    {
      "completion_length": 299.7750129699707,
      "epoch": 1.3504406428201141,
      "grad_norm": 1.5156244039535522,
      "kl": 2.20556640625,
      "learning_rate": 2.011138816527775e-06,
      "loss": 0.3612,
      "reward": 1.6250000596046448,
      "reward_std": 0.40387552231550217,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1303
    },
    {
      "completion_length": 199.1250057220459,
      "epoch": 1.3514774494556765,
      "grad_norm": 0.821329653263092,
      "kl": 0.312255859375,
      "learning_rate": 2.009435771304613e-06,
      "loss": 0.0109,
      "reward": 1.5250000059604645,
      "reward_std": 0.1369306445121765,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1304
    },
    {
      "completion_length": 206.35000228881836,
      "epoch": 1.352514256091239,
      "grad_norm": 0.8979102373123169,
      "kl": 1.183837890625,
      "learning_rate": 2.0077319834479364e-06,
      "loss": 0.0987,
      "reward": 1.6000000089406967,
      "reward_std": 0.17304782569408417,
      "rewards/accuracy_reward": 0.6750000007450581,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1305
    },
    {
      "completion_length": 161.97500228881836,
      "epoch": 1.3535510627268015,
      "grad_norm": 0.4250115752220154,
      "kl": 0.3291015625,
      "learning_rate": 2.0060274554414534e-06,
      "loss": 0.0632,
      "reward": 1.4500000178813934,
      "reward_std": 0.30463573709130287,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1306
    },
    {
      "completion_length": 255.27500343322754,
      "epoch": 1.3545878693623639,
      "grad_norm": 2.771432876586914,
      "kl": 1.2548828125,
      "learning_rate": 2.0043221897699516e-06,
      "loss": 0.1758,
      "reward": 1.5999999940395355,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1307
    },
    {
      "completion_length": 153.32500076293945,
      "epoch": 1.3556246759979265,
      "grad_norm": 1.119118332862854,
      "kl": 1.18994140625,
      "learning_rate": 2.0026161889192937e-06,
      "loss": 0.1228,
      "reward": 1.5250000059604645,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.5500000063329935,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1308
    },
    {
      "completion_length": 196.87500953674316,
      "epoch": 1.3566614826334888,
      "grad_norm": 0.8415226340293884,
      "kl": 1.863037109375,
      "learning_rate": 2.000909455376414e-06,
      "loss": 0.2324,
      "reward": 1.675000011920929,
      "reward_std": 0.3407529294490814,
      "rewards/accuracy_reward": 0.7249999940395355,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1309
    },
    {
      "completion_length": 159.50000381469727,
      "epoch": 1.3576982892690515,
      "grad_norm": 1.2588375806808472,
      "kl": 1.39892578125,
      "learning_rate": 1.999201991629316e-06,
      "loss": 0.1529,
      "reward": 1.6500000357627869,
      "reward_std": 0.335410188883543,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1310
    },
    {
      "completion_length": 238.82500839233398,
      "epoch": 1.3587350959046138,
      "grad_norm": 2.7150652408599854,
      "kl": 1.916748046875,
      "learning_rate": 1.9974938001670653e-06,
      "loss": 0.2961,
      "reward": 1.7000000476837158,
      "reward_std": 0.3913118839263916,
      "rewards/accuracy_reward": 0.7750000134110451,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1311
    },
    {
      "completion_length": 343.2000045776367,
      "epoch": 1.3597719025401762,
      "grad_norm": 1.2668253183364868,
      "kl": 4.1181640625,
      "learning_rate": 1.9957848834797907e-06,
      "loss": 0.3664,
      "reward": 1.1250000149011612,
      "reward_std": 0.34860678389668465,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1312
    },
    {
      "completion_length": 220.475004196167,
      "epoch": 1.3608087091757386,
      "grad_norm": 4.202035427093506,
      "kl": 1.3388671875,
      "learning_rate": 1.994075244058678e-06,
      "loss": 0.1765,
      "reward": 1.5499999821186066,
      "reward_std": 0.36053744703531265,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1313
    },
    {
      "completion_length": 244.44999885559082,
      "epoch": 1.3618455158113012,
      "grad_norm": 0.495758980512619,
      "kl": 1.9638671875,
      "learning_rate": 1.992364884395964e-06,
      "loss": 0.2726,
      "reward": 1.3750000149011612,
      "reward_std": 0.3678968548774719,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1314
    },
    {
      "completion_length": 220.75000190734863,
      "epoch": 1.3628823224468636,
      "grad_norm": 7.6908745765686035,
      "kl": 3.709228515625,
      "learning_rate": 1.9906538069849394e-06,
      "loss": 0.4358,
      "reward": 1.600000023841858,
      "reward_std": 0.3407529145479202,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1315
    },
    {
      "completion_length": 149.04999923706055,
      "epoch": 1.3639191290824262,
      "grad_norm": 0.523462176322937,
      "kl": 1.835693359375,
      "learning_rate": 1.9889420143199378e-06,
      "loss": 0.1679,
      "reward": 1.5000000298023224,
      "reward_std": 0.3354101814329624,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 1316
    },
    {
      "completion_length": 117.12499809265137,
      "epoch": 1.3649559357179886,
      "grad_norm": 1.952399492263794,
      "kl": 4.02685546875,
      "learning_rate": 1.987229508896337e-06,
      "loss": 0.1088,
      "reward": 1.5250000059604645,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1317
    },
    {
      "completion_length": 201.97500228881836,
      "epoch": 1.365992742353551,
      "grad_norm": 5.1983842849731445,
      "kl": 0.33984375,
      "learning_rate": 1.9855162932105538e-06,
      "loss": 0.1112,
      "reward": 1.5000000149011612,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1318
    },
    {
      "completion_length": 192.45000648498535,
      "epoch": 1.3670295489891136,
      "grad_norm": 0.7994645833969116,
      "kl": 2.2470703125,
      "learning_rate": 1.98380236976004e-06,
      "loss": 0.2463,
      "reward": 1.725000023841858,
      "reward_std": 0.3354102075099945,
      "rewards/accuracy_reward": 0.8000000268220901,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1319
    },
    {
      "completion_length": 297.1500015258789,
      "epoch": 1.368066355624676,
      "grad_norm": 3.1475718021392822,
      "kl": 2.51220703125,
      "learning_rate": 1.9820877410432804e-06,
      "loss": 0.3617,
      "reward": 1.4249999970197678,
      "reward_std": 0.3479738123714924,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1320
    },
    {
      "completion_length": 104.80000114440918,
      "epoch": 1.3691031622602385,
      "grad_norm": 0.5522794723510742,
      "kl": 0.20751953125,
      "learning_rate": 1.9803724095597845e-06,
      "loss": 0.0092,
      "reward": 1.7000000178813934,
      "reward_std": 0.12436700984835625,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 1.0,
      "step": 1321
    },
    {
      "completion_length": 345.65000915527344,
      "epoch": 1.370139968895801,
      "grad_norm": 2.9535956382751465,
      "kl": 4.50244140625,
      "learning_rate": 1.978656377810091e-06,
      "loss": 0.3934,
      "reward": 1.425000011920929,
      "reward_std": 0.497468076646328,
      "rewards/accuracy_reward": 0.5749999992549419,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1322
    },
    {
      "completion_length": 343.65000343322754,
      "epoch": 1.3711767755313633,
      "grad_norm": 1.0855612754821777,
      "kl": 1.342041015625,
      "learning_rate": 1.976939648295756e-06,
      "loss": 0.3052,
      "reward": 1.1750000268220901,
      "reward_std": 0.44892581552267075,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1323
    },
    {
      "completion_length": 297.0500030517578,
      "epoch": 1.372213582166926,
      "grad_norm": 1.9501731395721436,
      "kl": 4.21484375,
      "learning_rate": 1.9752222235193536e-06,
      "loss": 0.3417,
      "reward": 1.275000050663948,
      "reward_std": 0.5048274956643581,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1324
    },
    {
      "completion_length": 353.6750030517578,
      "epoch": 1.3732503888024883,
      "grad_norm": 3.1614885330200195,
      "kl": 2.147705078125,
      "learning_rate": 1.9735041059844727e-06,
      "loss": 0.4237,
      "reward": 1.0750000178813934,
      "reward_std": 0.45977721363306046,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1325
    },
    {
      "completion_length": 254.95001029968262,
      "epoch": 1.374287195438051,
      "grad_norm": 1.7063990831375122,
      "kl": 3.00390625,
      "learning_rate": 1.9717852981957092e-06,
      "loss": 0.3488,
      "reward": 1.25,
      "reward_std": 0.5643596909940243,
      "rewards/accuracy_reward": 0.3999999985098839,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1326
    },
    {
      "completion_length": 395.6250057220459,
      "epoch": 1.3753240020736133,
      "grad_norm": 3.218597650527954,
      "kl": 4.737060546875,
      "learning_rate": 1.9700658026586687e-06,
      "loss": 0.4662,
      "reward": 1.0250000208616257,
      "reward_std": 0.36588016897439957,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.7999999970197678,
      "step": 1327
    },
    {
      "completion_length": 307.10000228881836,
      "epoch": 1.3763608087091757,
      "grad_norm": 1.8125945329666138,
      "kl": 3.138671875,
      "learning_rate": 1.968345621879957e-06,
      "loss": 0.4781,
      "reward": 1.5250000059604645,
      "reward_std": 0.5715806037187576,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1328
    },
    {
      "completion_length": 217.62500381469727,
      "epoch": 1.3773976153447383,
      "grad_norm": 1.349875807762146,
      "kl": 1.9482421875,
      "learning_rate": 1.9666247583671796e-06,
      "loss": 0.2255,
      "reward": 1.574999988079071,
      "reward_std": 0.22360678762197495,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1329
    },
    {
      "completion_length": 163.37500381469727,
      "epoch": 1.3784344219803006,
      "grad_norm": 2.4353957176208496,
      "kl": 0.28515625,
      "learning_rate": 1.9649032146289377e-06,
      "loss": 0.0256,
      "reward": 1.3499999940395355,
      "reward_std": 0.16770510375499725,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1330
    },
    {
      "completion_length": 176.20000648498535,
      "epoch": 1.3794712286158632,
      "grad_norm": 0.4093508720397949,
      "kl": 3.2236328125,
      "learning_rate": 1.9631809931748218e-06,
      "loss": 0.3603,
      "reward": 1.4749999940395355,
      "reward_std": 0.4525563195347786,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.925000011920929,
      "step": 1331
    },
    {
      "completion_length": 354.90000343322754,
      "epoch": 1.3805080352514256,
      "grad_norm": 120.4134750366211,
      "kl": 50.94775390625,
      "learning_rate": 1.9614580965154143e-06,
      "loss": 2.1455,
      "reward": 1.5000000298023224,
      "reward_std": 0.5461488850414753,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1332
    },
    {
      "completion_length": 248.62500381469727,
      "epoch": 1.381544841886988,
      "grad_norm": 1.1852550506591797,
      "kl": 2.482666015625,
      "learning_rate": 1.9597345271622784e-06,
      "loss": 0.2442,
      "reward": 1.5,
      "reward_std": 0.2854842245578766,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 0.925000011920929,
      "step": 1333
    },
    {
      "completion_length": 407.375,
      "epoch": 1.3825816485225506,
      "grad_norm": 5.533868312835693,
      "kl": 4.03466796875,
      "learning_rate": 1.9580102876279596e-06,
      "loss": 0.5496,
      "reward": 1.350000023841858,
      "reward_std": 0.5210216753184795,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8000000268220901,
      "step": 1334
    },
    {
      "completion_length": 191.85000038146973,
      "epoch": 1.383618455158113,
      "grad_norm": 3.50870418548584,
      "kl": 0.885498046875,
      "learning_rate": 1.95628538042598e-06,
      "loss": 0.1257,
      "reward": 1.5500000417232513,
      "reward_std": 0.40387553349137306,
      "rewards/accuracy_reward": 0.6250000223517418,
      "rewards/format_reward": 0.925000011920929,
      "step": 1335
    },
    {
      "completion_length": 148.4249973297119,
      "epoch": 1.3846552617936756,
      "grad_norm": 5.719729423522949,
      "kl": 0.66552734375,
      "learning_rate": 1.9545598080708355e-06,
      "loss": 0.093,
      "reward": 1.4249999970197678,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1336
    },
    {
      "completion_length": 190.05000495910645,
      "epoch": 1.385692068429238,
      "grad_norm": 1.0899783372879028,
      "kl": 2.8780517578125,
      "learning_rate": 1.95283357307799e-06,
      "loss": 0.2897,
      "reward": 1.7000000178813934,
      "reward_std": 0.20019173622131348,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1337
    },
    {
      "completion_length": 390.47501373291016,
      "epoch": 1.3867288750648004,
      "grad_norm": 14.337153434753418,
      "kl": 6.26953125,
      "learning_rate": 1.9511066779638754e-06,
      "loss": 0.7641,
      "reward": 1.2000000476837158,
      "reward_std": 0.4399926923215389,
      "rewards/accuracy_reward": 0.3750000111758709,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1338
    },
    {
      "completion_length": 139.92499923706055,
      "epoch": 1.387765681700363,
      "grad_norm": 3.601635694503784,
      "kl": 0.927734375,
      "learning_rate": 1.9493791252458842e-06,
      "loss": 0.109,
      "reward": 1.5750000178813934,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.6250000037252903,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1339
    },
    {
      "completion_length": 147.82500076293945,
      "epoch": 1.3888024883359253,
      "grad_norm": 2.475231647491455,
      "kl": 1.29345703125,
      "learning_rate": 1.9476509174423695e-06,
      "loss": 0.1011,
      "reward": 1.2750000059604645,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.3000000063329935,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1340
    },
    {
      "completion_length": 272.7000026702881,
      "epoch": 1.389839294971488,
      "grad_norm": 3.056259870529175,
      "kl": 2.72265625,
      "learning_rate": 1.9459220570726375e-06,
      "loss": 0.3923,
      "reward": 1.6500000357627869,
      "reward_std": 0.47768355160951614,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1341
    },
    {
      "completion_length": 379.625,
      "epoch": 1.3908761016070503,
      "grad_norm": 2.5830676555633545,
      "kl": 2.778564453125,
      "learning_rate": 1.9441925466569454e-06,
      "loss": 0.4417,
      "reward": 1.350000023841858,
      "reward_std": 0.5461488962173462,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.800000011920929,
      "step": 1342
    },
    {
      "completion_length": 274.6750030517578,
      "epoch": 1.3919129082426127,
      "grad_norm": 2.2792797088623047,
      "kl": 1.99365234375,
      "learning_rate": 1.9424623887165005e-06,
      "loss": 0.3201,
      "reward": 1.475000023841858,
      "reward_std": 0.3804604522883892,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9000000357627869,
      "step": 1343
    },
    {
      "completion_length": 327.22500228881836,
      "epoch": 1.392949714878175,
      "grad_norm": 2.4801011085510254,
      "kl": 1.701171875,
      "learning_rate": 1.9407315857734516e-06,
      "loss": 0.1142,
      "reward": 1.5500000417232513,
      "reward_std": 0.3804604634642601,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.8499999940395355,
      "step": 1344
    },
    {
      "completion_length": 249.52500343322754,
      "epoch": 1.3939865215137377,
      "grad_norm": 3.4298818111419678,
      "kl": 1.82177734375,
      "learning_rate": 1.939000140350889e-06,
      "loss": 0.2832,
      "reward": 1.7000000178813934,
      "reward_std": 0.3913118876516819,
      "rewards/accuracy_reward": 0.7750000134110451,
      "rewards/format_reward": 0.925000011920929,
      "step": 1345
    },
    {
      "completion_length": 142.95000648498535,
      "epoch": 1.3950233281493,
      "grad_norm": 3.945064067840576,
      "kl": 2.00634765625,
      "learning_rate": 1.9372680549728394e-06,
      "loss": 0.1959,
      "reward": 1.7249999940395355,
      "reward_std": 0.21638591215014458,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1346
    },
    {
      "completion_length": 241.475004196167,
      "epoch": 1.3960601347848627,
      "grad_norm": 2.005232095718384,
      "kl": 1.3388671875,
      "learning_rate": 1.9355353321642615e-06,
      "loss": 0.1454,
      "reward": 1.475000023841858,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1347
    },
    {
      "completion_length": 243.9500026702881,
      "epoch": 1.397096941420425,
      "grad_norm": 2.505420684814453,
      "kl": 3.240966796875,
      "learning_rate": 1.9338019744510444e-06,
      "loss": 0.378,
      "reward": 1.4000000357627869,
      "reward_std": 0.4237985163927078,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 1348
    },
    {
      "completion_length": 260.20000076293945,
      "epoch": 1.3981337480559874,
      "grad_norm": 5.2407546043396,
      "kl": 1.7978515625,
      "learning_rate": 1.932067984360002e-06,
      "loss": 0.2908,
      "reward": 1.7000000476837158,
      "reward_std": 0.447213601320982,
      "rewards/accuracy_reward": 0.8000000193715096,
      "rewards/format_reward": 0.9000000357627869,
      "step": 1349
    },
    {
      "completion_length": 243.350004196167,
      "epoch": 1.39917055469155,
      "grad_norm": 1.1847293376922607,
      "kl": 1.4765625,
      "learning_rate": 1.9303333644188712e-06,
      "loss": 0.2516,
      "reward": 1.6500000059604645,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1350
    },
    {
      "completion_length": 183.25000190734863,
      "epoch": 1.4002073613271124,
      "grad_norm": 2.2103288173675537,
      "kl": 2.0478515625,
      "learning_rate": 1.9285981171563054e-06,
      "loss": 0.267,
      "reward": 1.7000000476837158,
      "reward_std": 0.3804604858160019,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1351
    },
    {
      "completion_length": 312.20000648498535,
      "epoch": 1.401244167962675,
      "grad_norm": 4.538423538208008,
      "kl": 3.09814453125,
      "learning_rate": 1.9268622451018734e-06,
      "loss": 0.3925,
      "reward": 1.350000023841858,
      "reward_std": 0.42963574081659317,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.875,
      "step": 1352
    },
    {
      "completion_length": 240.82500648498535,
      "epoch": 1.4022809745982374,
      "grad_norm": 0.9791562557220459,
      "kl": 3.373046875,
      "learning_rate": 1.925125750786055e-06,
      "loss": 0.3425,
      "reward": 1.4000000357627869,
      "reward_std": 0.5715806111693382,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1353
    },
    {
      "completion_length": 279.325008392334,
      "epoch": 1.4033177812337998,
      "grad_norm": 2.6334242820739746,
      "kl": 2.3408203125,
      "learning_rate": 1.9233886367402364e-06,
      "loss": 0.3473,
      "reward": 1.1500000208616257,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1354
    },
    {
      "completion_length": 195.25,
      "epoch": 1.4043545878693624,
      "grad_norm": 2.7357373237609863,
      "kl": 0.626708984375,
      "learning_rate": 1.9216509054967076e-06,
      "loss": 0.0761,
      "reward": 1.550000011920929,
      "reward_std": 0.28485121205449104,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1355
    },
    {
      "completion_length": 150.5750026702881,
      "epoch": 1.4053913945049248,
      "grad_norm": 2.6245734691619873,
      "kl": 1.7177734375,
      "learning_rate": 1.919912559588659e-06,
      "loss": 0.1806,
      "reward": 1.5250000059604645,
      "reward_std": 0.279508501291275,
      "rewards/accuracy_reward": 0.5500000063329935,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1356
    },
    {
      "completion_length": 197.17501068115234,
      "epoch": 1.4064282011404874,
      "grad_norm": 2.298128604888916,
      "kl": 2.6669921875,
      "learning_rate": 1.9181736015501754e-06,
      "loss": 0.2947,
      "reward": 1.4250000417232513,
      "reward_std": 0.28485123068094254,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.925000011920929,
      "step": 1357
    },
    {
      "completion_length": 193.05000495910645,
      "epoch": 1.4074650077760498,
      "grad_norm": 1.1060773134231567,
      "kl": 2.9697265625,
      "learning_rate": 1.916434033916235e-06,
      "loss": 0.3492,
      "reward": 1.350000023841858,
      "reward_std": 0.45977720618247986,
      "rewards/accuracy_reward": 0.44999999925494194,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1358
    },
    {
      "completion_length": 333.15001106262207,
      "epoch": 1.4085018144116122,
      "grad_norm": 1.7391694784164429,
      "kl": 2.19970703125,
      "learning_rate": 1.914693859222705e-06,
      "loss": 0.3666,
      "reward": 1.3750000298023224,
      "reward_std": 0.3856646940112114,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1359
    },
    {
      "completion_length": 310.1750030517578,
      "epoch": 1.4095386210471748,
      "grad_norm": 5.543538570404053,
      "kl": 25.02783203125,
      "learning_rate": 1.9129530800063377e-06,
      "loss": 2.0473,
      "reward": 1.5750000178813934,
      "reward_std": 0.3533165454864502,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1360
    },
    {
      "completion_length": 303.875,
      "epoch": 1.4105754276827371,
      "grad_norm": 72.21014404296875,
      "kl": 1.54541015625,
      "learning_rate": 1.9112116988047646e-06,
      "loss": 0.2335,
      "reward": 1.3750000298023224,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1361
    },
    {
      "completion_length": 286.0500011444092,
      "epoch": 1.4116122343182997,
      "grad_norm": 65.4577407836914,
      "kl": 22.9541015625,
      "learning_rate": 1.9094697181564967e-06,
      "loss": 1.9458,
      "reward": 1.4500000178813934,
      "reward_std": 0.3479738123714924,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1362
    },
    {
      "completion_length": 199.5499973297119,
      "epoch": 1.4126490409538621,
      "grad_norm": 3.5749762058258057,
      "kl": 2.787109375,
      "learning_rate": 1.907727140600919e-06,
      "loss": 0.1877,
      "reward": 1.5250000357627869,
      "reward_std": 0.4290027320384979,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 1363
    },
    {
      "completion_length": 184.100004196167,
      "epoch": 1.4136858475894245,
      "grad_norm": 4.195121765136719,
      "kl": 3.4029541015625,
      "learning_rate": 1.9059839686782853e-06,
      "loss": 0.1832,
      "reward": 1.5750000178813934,
      "reward_std": 0.4415663816034794,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1364
    },
    {
      "completion_length": 201.15000915527344,
      "epoch": 1.414722654224987,
      "grad_norm": 2.0089199542999268,
      "kl": 1.5,
      "learning_rate": 1.9042402049297163e-06,
      "loss": 0.1528,
      "reward": 1.475000023841858,
      "reward_std": 0.45977721735835075,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9000000357627869,
      "step": 1365
    },
    {
      "completion_length": 242.22500610351562,
      "epoch": 1.4157594608605495,
      "grad_norm": 4.145914077758789,
      "kl": 3.361572265625,
      "learning_rate": 1.9024958518971964e-06,
      "loss": 0.3238,
      "reward": 1.475000038743019,
      "reward_std": 0.33712238073349,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1366
    },
    {
      "completion_length": 273.37500762939453,
      "epoch": 1.416796267496112,
      "grad_norm": 1.178970217704773,
      "kl": 1.828125,
      "learning_rate": 1.9007509121235676e-06,
      "loss": 0.2617,
      "reward": 1.475000023841858,
      "reward_std": 0.3479738049209118,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1367
    },
    {
      "completion_length": 146.25000190734863,
      "epoch": 1.4178330741316745,
      "grad_norm": 2.0879671573638916,
      "kl": 1.00146484375,
      "learning_rate": 1.8990053881525282e-06,
      "loss": 0.1353,
      "reward": 1.5250000357627869,
      "reward_std": 0.31719937175512314,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1368
    },
    {
      "completion_length": 292.45000648498535,
      "epoch": 1.4188698807672369,
      "grad_norm": 1.4969731569290161,
      "kl": 2.66796875,
      "learning_rate": 1.897259282528628e-06,
      "loss": 0.324,
      "reward": 1.225000038743019,
      "reward_std": 0.31719936802983284,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1369
    },
    {
      "completion_length": 196.87500381469727,
      "epoch": 1.4199066874027995,
      "grad_norm": 0.8825180530548096,
      "kl": 1.77880859375,
      "learning_rate": 1.8955125977972644e-06,
      "loss": 0.1783,
      "reward": 1.6500000357627869,
      "reward_std": 0.38409100100398064,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1370
    },
    {
      "completion_length": 238.02501106262207,
      "epoch": 1.4209434940383618,
      "grad_norm": 1.4576854705810547,
      "kl": 1.9921875,
      "learning_rate": 1.8937653365046792e-06,
      "loss": 0.2668,
      "reward": 1.3000000417232513,
      "reward_std": 0.4723408482968807,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1371
    },
    {
      "completion_length": 196.47500133514404,
      "epoch": 1.4219803006739244,
      "grad_norm": 1.6968486309051514,
      "kl": 0.517578125,
      "learning_rate": 1.892017501197954e-06,
      "loss": 0.1419,
      "reward": 1.5000000149011612,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 1372
    },
    {
      "completion_length": 389.4500045776367,
      "epoch": 1.4230171073094868,
      "grad_norm": 3.5373055934906006,
      "kl": 2.83740234375,
      "learning_rate": 1.8902690944250086e-06,
      "loss": 0.5057,
      "reward": 1.3750000298023224,
      "reward_std": 0.6149186939001083,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1373
    },
    {
      "completion_length": 185.05000495910645,
      "epoch": 1.4240539139450492,
      "grad_norm": 3.059485673904419,
      "kl": 1.74072265625,
      "learning_rate": 1.8885201187345948e-06,
      "loss": 0.2048,
      "reward": 1.625,
      "reward_std": 0.22360679134726524,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 1374
    },
    {
      "completion_length": 190.62500381469727,
      "epoch": 1.4250907205806116,
      "grad_norm": 1.2744885683059692,
      "kl": 2.24658203125,
      "learning_rate": 1.8867705766762936e-06,
      "loss": 0.1086,
      "reward": 1.1500000208616257,
      "reward_std": 0.3407529145479202,
      "rewards/accuracy_reward": 0.2500000037252903,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1375
    },
    {
      "completion_length": 94.32500267028809,
      "epoch": 1.4261275272161742,
      "grad_norm": 0.6009107232093811,
      "kl": 0.4296875,
      "learning_rate": 1.8850204708005127e-06,
      "loss": 0.1247,
      "reward": 1.6500000357627869,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1376
    },
    {
      "completion_length": 202.850004196167,
      "epoch": 1.4271643338517366,
      "grad_norm": 0.49962109327316284,
      "kl": 0.6357421875,
      "learning_rate": 1.8832698036584805e-06,
      "loss": 0.1524,
      "reward": 1.600000023841858,
      "reward_std": 0.41643913835287094,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1377
    },
    {
      "completion_length": 136.77499961853027,
      "epoch": 1.4282011404872992,
      "grad_norm": 2.90053129196167,
      "kl": 0.54052734375,
      "learning_rate": 1.8815185778022447e-06,
      "loss": 0.079,
      "reward": 1.3250000327825546,
      "reward_std": 0.24873404577374458,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1378
    },
    {
      "completion_length": 106.85000038146973,
      "epoch": 1.4292379471228616,
      "grad_norm": 0.4404795467853546,
      "kl": 1.04052734375,
      "learning_rate": 1.8797667957846663e-06,
      "loss": 0.0969,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1379
    },
    {
      "completion_length": 192.80000686645508,
      "epoch": 1.430274753758424,
      "grad_norm": 1.2581452131271362,
      "kl": 1.458984375,
      "learning_rate": 1.878014460159418e-06,
      "loss": 0.1953,
      "reward": 1.675000011920929,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1380
    },
    {
      "completion_length": 428.6000118255615,
      "epoch": 1.4313115603939865,
      "grad_norm": 14.338582038879395,
      "kl": 6.367919921875,
      "learning_rate": 1.8762615734809787e-06,
      "loss": 0.6138,
      "reward": 1.3500000089406967,
      "reward_std": 0.576923344284296,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1381
    },
    {
      "completion_length": 345.4000015258789,
      "epoch": 1.432348367029549,
      "grad_norm": 13.323136329650879,
      "kl": 4.8623046875,
      "learning_rate": 1.8745081383046317e-06,
      "loss": 0.583,
      "reward": 1.5000000298023224,
      "reward_std": 0.5445752069354057,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1382
    },
    {
      "completion_length": 221.0500087738037,
      "epoch": 1.4333851736651115,
      "grad_norm": 0.6289466023445129,
      "kl": 2.7431640625,
      "learning_rate": 1.8727541571864588e-06,
      "loss": 0.3706,
      "reward": 1.6250000596046448,
      "reward_std": 0.3913118876516819,
      "rewards/accuracy_reward": 0.700000025331974,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1383
    },
    {
      "completion_length": 168.15000343322754,
      "epoch": 1.434421980300674,
      "grad_norm": 2.4519495964050293,
      "kl": 2.47216796875,
      "learning_rate": 1.870999632683338e-06,
      "loss": 0.2879,
      "reward": 1.675000011920929,
      "reward_std": 0.16770510375499725,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1384
    },
    {
      "completion_length": 101.60000228881836,
      "epoch": 1.4354587869362363,
      "grad_norm": 3.0664381980895996,
      "kl": 0.244873046875,
      "learning_rate": 1.8692445673529398e-06,
      "loss": 0.004,
      "reward": 1.7000000178813934,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 1.0,
      "step": 1385
    },
    {
      "completion_length": 288.49999809265137,
      "epoch": 1.436495593571799,
      "grad_norm": 0.5069672465324402,
      "kl": 8.235595703125,
      "learning_rate": 1.867488963753723e-06,
      "loss": 0.8141,
      "reward": 1.3750000298023224,
      "reward_std": 0.39728759974241257,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1386
    },
    {
      "completion_length": 181.87500381469727,
      "epoch": 1.4375324002073613,
      "grad_norm": 29.51852798461914,
      "kl": 4.054443359375,
      "learning_rate": 1.8657328244449301e-06,
      "loss": 0.3738,
      "reward": 1.75,
      "reward_std": 0.1767766997218132,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1387
    },
    {
      "completion_length": 245.850004196167,
      "epoch": 1.4385692068429239,
      "grad_norm": 19.067703247070312,
      "kl": 2.6376953125,
      "learning_rate": 1.8639761519865866e-06,
      "loss": 0.2866,
      "reward": 1.6500000208616257,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.7250000182539225,
      "rewards/format_reward": 0.925000011920929,
      "step": 1388
    },
    {
      "completion_length": 167.39999961853027,
      "epoch": 1.4396060134784863,
      "grad_norm": 0.3625662326812744,
      "kl": 1.27001953125,
      "learning_rate": 1.8622189489394926e-06,
      "loss": 0.1855,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510375499725,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1389
    },
    {
      "completion_length": 330.87500762939453,
      "epoch": 1.4406428201140487,
      "grad_norm": 5.789668560028076,
      "kl": 5.293212890625,
      "learning_rate": 1.8604612178652234e-06,
      "loss": 0.4377,
      "reward": 1.2500000149011612,
      "reward_std": 0.40387551859021187,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.875,
      "step": 1390
    },
    {
      "completion_length": 145.75000381469727,
      "epoch": 1.4416796267496113,
      "grad_norm": 9.631103515625,
      "kl": 2.2275390625,
      "learning_rate": 1.8587029613261241e-06,
      "loss": 0.0912,
      "reward": 1.5500000417232513,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.600000012665987,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1391
    },
    {
      "completion_length": 237.0750026702881,
      "epoch": 1.4427164333851736,
      "grad_norm": 1.7411144971847534,
      "kl": 1.08447265625,
      "learning_rate": 1.8569441818853047e-06,
      "loss": 0.1139,
      "reward": 1.5750000178813934,
      "reward_std": 0.30463575199246407,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.925000011920929,
      "step": 1392
    },
    {
      "completion_length": 177.82500648498535,
      "epoch": 1.4437532400207362,
      "grad_norm": 1.8259129524230957,
      "kl": 2.00634765625,
      "learning_rate": 1.8551848821066383e-06,
      "loss": 0.2332,
      "reward": 1.4500000178813934,
      "reward_std": 0.41643914580345154,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1393
    },
    {
      "completion_length": 297.4250011444092,
      "epoch": 1.4447900466562986,
      "grad_norm": 0.878925621509552,
      "kl": 2.45947265625,
      "learning_rate": 1.8534250645547568e-06,
      "loss": 0.2842,
      "reward": 1.425000011920929,
      "reward_std": 0.36789683997631073,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1394
    },
    {
      "completion_length": 291.425012588501,
      "epoch": 1.445826853291861,
      "grad_norm": 2.780809164047241,
      "kl": 4.52490234375,
      "learning_rate": 1.8516647317950455e-06,
      "loss": 0.3475,
      "reward": 1.2750000208616257,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1395
    },
    {
      "completion_length": 232.52500534057617,
      "epoch": 1.4468636599274236,
      "grad_norm": 11.077188491821289,
      "kl": 1.074951171875,
      "learning_rate": 1.849903886393642e-06,
      "loss": 0.2272,
      "reward": 1.600000023841858,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 1396
    },
    {
      "completion_length": 134.60000228881836,
      "epoch": 1.447900466562986,
      "grad_norm": 0.9025780558586121,
      "kl": 0.34912109375,
      "learning_rate": 1.848142530917431e-06,
      "loss": 0.0357,
      "reward": 1.3250000178813934,
      "reward_std": 0.15685366094112396,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1397
    },
    {
      "completion_length": 238.72500228881836,
      "epoch": 1.4489372731985486,
      "grad_norm": 3.455339193344116,
      "kl": 1.527587890625,
      "learning_rate": 1.8463806679340409e-06,
      "loss": 0.2651,
      "reward": 1.7000000476837158,
      "reward_std": 0.3479738011956215,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1398
    },
    {
      "completion_length": 89.25000190734863,
      "epoch": 1.449974079834111,
      "grad_norm": 0.4030129313468933,
      "kl": 0.3681640625,
      "learning_rate": 1.844618300011839e-06,
      "loss": 0.0142,
      "reward": 1.5,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 1.0,
      "step": 1399
    },
    {
      "completion_length": 258.55001068115234,
      "epoch": 1.4510108864696734,
      "grad_norm": 2.5429656505584717,
      "kl": 1.571533203125,
      "learning_rate": 1.8428554297199293e-06,
      "loss": 0.2594,
      "reward": 1.4750000536441803,
      "reward_std": 0.3966546058654785,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9000000357627869,
      "step": 1400
    },
    {
      "completion_length": 346.0,
      "epoch": 1.4520476931052357,
      "grad_norm": 2.216012954711914,
      "kl": 1.3095703125,
      "learning_rate": 1.8410920596281483e-06,
      "loss": 0.3248,
      "reward": 1.3250000327825546,
      "reward_std": 0.3479738123714924,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1401
    },
    {
      "completion_length": 239.6750030517578,
      "epoch": 1.4530844997407983,
      "grad_norm": 3.4484033584594727,
      "kl": 2.7900390625,
      "learning_rate": 1.8393281923070614e-06,
      "loss": 0.3214,
      "reward": 1.4000000357627869,
      "reward_std": 0.4472135752439499,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1402
    },
    {
      "completion_length": 190.600004196167,
      "epoch": 1.454121306376361,
      "grad_norm": 0.9628894925117493,
      "kl": 0.29833984375,
      "learning_rate": 1.837563830327958e-06,
      "loss": 0.1015,
      "reward": 1.475000023841858,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1403
    },
    {
      "completion_length": 299.62500953674316,
      "epoch": 1.4551581130119233,
      "grad_norm": 16.452903747558594,
      "kl": 14.7822265625,
      "learning_rate": 1.8357989762628493e-06,
      "loss": 1.2684,
      "reward": 1.6250000596046448,
      "reward_std": 0.4290027618408203,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1404
    },
    {
      "completion_length": 147.2500057220459,
      "epoch": 1.4561949196474857,
      "grad_norm": 48.58760452270508,
      "kl": 3.30419921875,
      "learning_rate": 1.8340336326844627e-06,
      "loss": 0.2987,
      "reward": 1.4500000178813934,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1405
    },
    {
      "completion_length": 139.2500057220459,
      "epoch": 1.457231726283048,
      "grad_norm": 12.247923851013184,
      "kl": 1.459716796875,
      "learning_rate": 1.832267802166241e-06,
      "loss": 0.1251,
      "reward": 1.625,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1406
    },
    {
      "completion_length": 126.57500267028809,
      "epoch": 1.4582685329186107,
      "grad_norm": 0.12291527539491653,
      "kl": 1.54150390625,
      "learning_rate": 1.8305014872823346e-06,
      "loss": 0.1752,
      "reward": 1.5750000178813934,
      "reward_std": 0.17304781079292297,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1407
    },
    {
      "completion_length": 135.95000076293945,
      "epoch": 1.459305339554173,
      "grad_norm": 3.804716110229492,
      "kl": 1.363525390625,
      "learning_rate": 1.8287346906076024e-06,
      "loss": 0.1411,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1408
    },
    {
      "completion_length": 150.75000381469727,
      "epoch": 1.4603421461897357,
      "grad_norm": 2.0299649238586426,
      "kl": 0.79345703125,
      "learning_rate": 1.8269674147176044e-06,
      "loss": 0.0995,
      "reward": 1.5250000357627869,
      "reward_std": 0.403875507414341,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1409
    },
    {
      "completion_length": 245.3250026702881,
      "epoch": 1.461378952825298,
      "grad_norm": 2.6532015800476074,
      "kl": 0.55029296875,
      "learning_rate": 1.825199662188598e-06,
      "loss": 0.0766,
      "reward": 1.2750000059604645,
      "reward_std": 0.37844379246234894,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 1410
    },
    {
      "completion_length": 151.9000015258789,
      "epoch": 1.4624157594608604,
      "grad_norm": 0.45854178071022034,
      "kl": 0.419921875,
      "learning_rate": 1.823431435597538e-06,
      "loss": 0.0667,
      "reward": 1.6500000059604645,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1411
    },
    {
      "completion_length": 393.9000053405762,
      "epoch": 1.463452566096423,
      "grad_norm": 3.40095591545105,
      "kl": 4.75830078125,
      "learning_rate": 1.8216627375220684e-06,
      "loss": 0.4276,
      "reward": 1.4500000178813934,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1412
    },
    {
      "completion_length": 210.37500190734863,
      "epoch": 1.4644893727319854,
      "grad_norm": 11.09891128540039,
      "kl": 0.6650390625,
      "learning_rate": 1.819893570540521e-06,
      "loss": 0.1252,
      "reward": 1.550000011920929,
      "reward_std": 0.25609345361590385,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1413
    },
    {
      "completion_length": 99.67500114440918,
      "epoch": 1.465526179367548,
      "grad_norm": 0.5185099244117737,
      "kl": 0.38720703125,
      "learning_rate": 1.818123937231911e-06,
      "loss": 0.0038,
      "reward": 1.625,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.650000000372529,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1414
    },
    {
      "completion_length": 251.52500534057617,
      "epoch": 1.4665629860031104,
      "grad_norm": 3.0295069217681885,
      "kl": 1.4248046875,
      "learning_rate": 1.8163538401759342e-06,
      "loss": 0.2459,
      "reward": 1.5250000059604645,
      "reward_std": 0.3354102075099945,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1415
    },
    {
      "completion_length": 244.975004196167,
      "epoch": 1.4675997926386728,
      "grad_norm": 5.051865577697754,
      "kl": 3.203125,
      "learning_rate": 1.8145832819529614e-06,
      "loss": 0.3631,
      "reward": 1.5500000417232513,
      "reward_std": 0.45977726206183434,
      "rewards/accuracy_reward": 0.6250000223517418,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1416
    },
    {
      "completion_length": 122.9000015258789,
      "epoch": 1.4686365992742354,
      "grad_norm": 4.377063274383545,
      "kl": 2.48486328125,
      "learning_rate": 1.8128122651440362e-06,
      "loss": 0.1759,
      "reward": 1.3500000089406967,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1417
    },
    {
      "completion_length": 96.32500076293945,
      "epoch": 1.4696734059097978,
      "grad_norm": 6.121978759765625,
      "kl": 0.453857421875,
      "learning_rate": 1.8110407923308702e-06,
      "loss": 0.0328,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1418
    },
    {
      "completion_length": 189.47500610351562,
      "epoch": 1.4707102125453604,
      "grad_norm": 7.747506141662598,
      "kl": 3.40234375,
      "learning_rate": 1.8092688660958408e-06,
      "loss": 0.3534,
      "reward": 1.225000023841858,
      "reward_std": 0.27950849011540413,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1419
    },
    {
      "completion_length": 232.7000026702881,
      "epoch": 1.4717470191809228,
      "grad_norm": 6.417865753173828,
      "kl": 3.60107421875,
      "learning_rate": 1.807496489021985e-06,
      "loss": 0.3762,
      "reward": 1.5500000417232513,
      "reward_std": 0.3245587833225727,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.925000011920929,
      "step": 1420
    },
    {
      "completion_length": 244.12500190734863,
      "epoch": 1.4727838258164851,
      "grad_norm": 5.064550399780273,
      "kl": 2.221435546875,
      "learning_rate": 1.8057236636929988e-06,
      "loss": 0.3437,
      "reward": 1.5750000178813934,
      "reward_std": 0.3354101926088333,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1421
    },
    {
      "completion_length": 234.55000686645508,
      "epoch": 1.4738206324520478,
      "grad_norm": 3.3823163509368896,
      "kl": 2.0166015625,
      "learning_rate": 1.8039503926932294e-06,
      "loss": 0.2434,
      "reward": 1.5000000298023224,
      "reward_std": 0.3354102149605751,
      "rewards/accuracy_reward": 0.6000000182539225,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1422
    },
    {
      "completion_length": 145.2500057220459,
      "epoch": 1.4748574390876101,
      "grad_norm": 4.185642719268799,
      "kl": 0.616943359375,
      "learning_rate": 1.8021766786076754e-06,
      "loss": 0.0692,
      "reward": 1.25,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1423
    },
    {
      "completion_length": 341.6500129699707,
      "epoch": 1.4758942457231727,
      "grad_norm": 6.28878116607666,
      "kl": 2.884765625,
      "learning_rate": 1.8004025240219807e-06,
      "loss": 0.4261,
      "reward": 1.5500000417232513,
      "reward_std": 0.40387551859021187,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1424
    },
    {
      "completion_length": 166.67500114440918,
      "epoch": 1.4769310523587351,
      "grad_norm": 3.579341411590576,
      "kl": 1.2705078125,
      "learning_rate": 1.7986279315224315e-06,
      "loss": 0.1264,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1425
    },
    {
      "completion_length": 151.92500114440918,
      "epoch": 1.4779678589942975,
      "grad_norm": 0.24669739603996277,
      "kl": 0.78076171875,
      "learning_rate": 1.7968529036959518e-06,
      "loss": 0.0722,
      "reward": 1.5250000059604645,
      "reward_std": 0.10458250343799591,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1426
    },
    {
      "completion_length": 211.15000343322754,
      "epoch": 1.47900466562986,
      "grad_norm": 0.9443327784538269,
      "kl": 0.842529296875,
      "learning_rate": 1.7950774431301009e-06,
      "loss": 0.1043,
      "reward": 1.7000000178813934,
      "reward_std": 0.3407529294490814,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1427
    },
    {
      "completion_length": 150.10000038146973,
      "epoch": 1.4800414722654225,
      "grad_norm": 1.4577233791351318,
      "kl": 3.143310546875,
      "learning_rate": 1.7933015524130691e-06,
      "loss": 0.2531,
      "reward": 1.475000023841858,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1428
    },
    {
      "completion_length": 88.25000190734863,
      "epoch": 1.481078278900985,
      "grad_norm": 9.675968170166016,
      "kl": 0.2255859375,
      "learning_rate": 1.7915252341336732e-06,
      "loss": -0.0211,
      "reward": 1.5000000298023224,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 1.0,
      "step": 1429
    },
    {
      "completion_length": 197.47500228881836,
      "epoch": 1.4821150855365475,
      "grad_norm": 0.5715732574462891,
      "kl": 0.4140625,
      "learning_rate": 1.7897484908813534e-06,
      "loss": 0.0736,
      "reward": 1.5750000476837158,
      "reward_std": 0.41643913462758064,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1430
    },
    {
      "completion_length": 144.27500343322754,
      "epoch": 1.4831518921721099,
      "grad_norm": 5.203686714172363,
      "kl": 2.2109375,
      "learning_rate": 1.7879713252461689e-06,
      "loss": 0.1812,
      "reward": 1.2750000059604645,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.3000000063329935,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1431
    },
    {
      "completion_length": 202.17500495910645,
      "epoch": 1.4841886988076722,
      "grad_norm": 3.741713285446167,
      "kl": 2.13818359375,
      "learning_rate": 1.786193739818796e-06,
      "loss": 0.2348,
      "reward": 1.550000011920929,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1432
    },
    {
      "completion_length": 376.2000045776367,
      "epoch": 1.4852255054432348,
      "grad_norm": 4.261992931365967,
      "kl": 4.01025390625,
      "learning_rate": 1.7844157371905217e-06,
      "loss": 0.5322,
      "reward": 1.2000000327825546,
      "reward_std": 0.484904445707798,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1433
    },
    {
      "completion_length": 185.77500343322754,
      "epoch": 1.4862623120787974,
      "grad_norm": 1.0868406295776367,
      "kl": 3.3740234375,
      "learning_rate": 1.782637319953242e-06,
      "loss": 0.3241,
      "reward": 1.5000000298023224,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1434
    },
    {
      "completion_length": 294.6500015258789,
      "epoch": 1.4872991187143598,
      "grad_norm": 7.851423263549805,
      "kl": 1.19677734375,
      "learning_rate": 1.7808584906994563e-06,
      "loss": 0.2256,
      "reward": 1.2000000327825546,
      "reward_std": 0.403875507414341,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.9000000357627869,
      "step": 1435
    },
    {
      "completion_length": 359.1250057220459,
      "epoch": 1.4883359253499222,
      "grad_norm": 1.853022575378418,
      "kl": 1.81689453125,
      "learning_rate": 1.7790792520222657e-06,
      "loss": 0.1214,
      "reward": 1.4250000268220901,
      "reward_std": 0.41643913090229034,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.800000011920929,
      "step": 1436
    },
    {
      "completion_length": 228.52500534057617,
      "epoch": 1.4893727319854846,
      "grad_norm": 0.7809069752693176,
      "kl": 0.87060546875,
      "learning_rate": 1.7772996065153676e-06,
      "loss": 0.081,
      "reward": 1.2750000059604645,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1437
    },
    {
      "completion_length": 208.5000057220459,
      "epoch": 1.4904095386210472,
      "grad_norm": 3.053826093673706,
      "kl": 1.2216796875,
      "learning_rate": 1.775519556773052e-06,
      "loss": 0.1771,
      "reward": 1.425000011920929,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1438
    },
    {
      "completion_length": 215.40000343322754,
      "epoch": 1.4914463452566096,
      "grad_norm": 4.729836463928223,
      "kl": 1.341552734375,
      "learning_rate": 1.7737391053901998e-06,
      "loss": 0.1857,
      "reward": 1.350000023841858,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1439
    },
    {
      "completion_length": 202.62500381469727,
      "epoch": 1.4924831518921722,
      "grad_norm": 0.5005960464477539,
      "kl": 1.277099609375,
      "learning_rate": 1.7719582549622755e-06,
      "loss": 0.1389,
      "reward": 1.4000000208616257,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 1440
    },
    {
      "completion_length": 259.65000343322754,
      "epoch": 1.4935199585277346,
      "grad_norm": 1.7078673839569092,
      "kl": 1.71240234375,
      "learning_rate": 1.7701770080853261e-06,
      "loss": 0.2559,
      "reward": 1.4000000208616257,
      "reward_std": 0.27228762954473495,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 1441
    },
    {
      "completion_length": 164.92500495910645,
      "epoch": 1.494556765163297,
      "grad_norm": 0.4092227518558502,
      "kl": 1.8466796875,
      "learning_rate": 1.7683953673559771e-06,
      "loss": 0.1498,
      "reward": 1.7500000298023224,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1442
    },
    {
      "completion_length": 158.20000457763672,
      "epoch": 1.4955935717988595,
      "grad_norm": 3.3610963821411133,
      "kl": 1.3515625,
      "learning_rate": 1.7666133353714268e-06,
      "loss": 0.1636,
      "reward": 1.4000000059604645,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.42499999701976776,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1443
    },
    {
      "completion_length": 200.0750026702881,
      "epoch": 1.496630378434422,
      "grad_norm": 0.3641982972621918,
      "kl": 1.6630859375,
      "learning_rate": 1.7648309147294452e-06,
      "loss": 0.2213,
      "reward": 1.6500000059604645,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1444
    },
    {
      "completion_length": 221.5250072479248,
      "epoch": 1.4976671850699845,
      "grad_norm": 2.0081844329833984,
      "kl": 2.14794921875,
      "learning_rate": 1.7630481080283683e-06,
      "loss": 0.2462,
      "reward": 1.2250000089406967,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.2749999985098839,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1445
    },
    {
      "completion_length": 218.32500457763672,
      "epoch": 1.498703991705547,
      "grad_norm": 1.0023314952850342,
      "kl": 1.4482421875,
      "learning_rate": 1.7612649178670946e-06,
      "loss": 0.1987,
      "reward": 1.5500000417232513,
      "reward_std": 0.40558771044015884,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.925000011920929,
      "step": 1446
    },
    {
      "completion_length": 217.5500030517578,
      "epoch": 1.4997407983411093,
      "grad_norm": 1.1659680604934692,
      "kl": 1.01171875,
      "learning_rate": 1.7594813468450822e-06,
      "loss": 0.1317,
      "reward": 1.75,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1447
    },
    {
      "completion_length": 263.3000011444092,
      "epoch": 1.500777604976672,
      "grad_norm": 0.3303806781768799,
      "kl": 2.60009765625,
      "learning_rate": 1.7576973975623436e-06,
      "loss": 0.2062,
      "reward": 1.4000000357627869,
      "reward_std": 0.4237985461950302,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.925000011920929,
      "step": 1448
    },
    {
      "completion_length": 217.8250026702881,
      "epoch": 1.5018144116122343,
      "grad_norm": 1.0276261568069458,
      "kl": 1.326416015625,
      "learning_rate": 1.7559130726194434e-06,
      "loss": 0.1334,
      "reward": 1.4749999940395355,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1449
    },
    {
      "completion_length": 437.4250068664551,
      "epoch": 1.5028512182477969,
      "grad_norm": 1.156264066696167,
      "kl": 1.967529296875,
      "learning_rate": 1.7541283746174943e-06,
      "loss": 0.4732,
      "reward": 1.475000023841858,
      "reward_std": 0.5156789049506187,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1450
    },
    {
      "completion_length": 170.72500038146973,
      "epoch": 1.5038880248833593,
      "grad_norm": 0.7211430668830872,
      "kl": 0.8359375,
      "learning_rate": 1.7523433061581511e-06,
      "loss": 0.0813,
      "reward": 1.6500000059604645,
      "reward_std": 0.3245587684214115,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1451
    },
    {
      "completion_length": 233.29999923706055,
      "epoch": 1.5049248315189216,
      "grad_norm": 1.8370705842971802,
      "kl": 1.745849609375,
      "learning_rate": 1.7505578698436103e-06,
      "loss": 0.21,
      "reward": 1.3750000447034836,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1452
    },
    {
      "completion_length": 317.9750003814697,
      "epoch": 1.505961638154484,
      "grad_norm": 9.33315658569336,
      "kl": 6.892333984375,
      "learning_rate": 1.7487720682766035e-06,
      "loss": 0.6925,
      "reward": 1.5250000357627869,
      "reward_std": 0.41643912345170975,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1453
    },
    {
      "completion_length": 237.50000381469727,
      "epoch": 1.5069984447900466,
      "grad_norm": 7.423142910003662,
      "kl": 3.14892578125,
      "learning_rate": 1.7469859040603956e-06,
      "loss": 0.2776,
      "reward": 1.600000023841858,
      "reward_std": 0.3046357296407223,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 1454
    },
    {
      "completion_length": 331.25001335144043,
      "epoch": 1.5080352514256092,
      "grad_norm": 2.906614065170288,
      "kl": 3.95947265625,
      "learning_rate": 1.74519937979878e-06,
      "loss": 0.4799,
      "reward": 1.5750000476837158,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1455
    },
    {
      "completion_length": 140.75000190734863,
      "epoch": 1.5090720580611716,
      "grad_norm": 3.1353986263275146,
      "kl": 1.02490234375,
      "learning_rate": 1.7434124980960746e-06,
      "loss": 0.0841,
      "reward": 1.425000011920929,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1456
    },
    {
      "completion_length": 327.9750061035156,
      "epoch": 1.510108864696734,
      "grad_norm": 2.66587495803833,
      "kl": 2.343505859375,
      "learning_rate": 1.7416252615571188e-06,
      "loss": 0.3487,
      "reward": 1.1750000417232513,
      "reward_std": 0.48867350444197655,
      "rewards/accuracy_reward": 0.27500000409781933,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1457
    },
    {
      "completion_length": 328.7750129699707,
      "epoch": 1.5111456713322964,
      "grad_norm": 2.1067070960998535,
      "kl": 2.60693359375,
      "learning_rate": 1.7398376727872688e-06,
      "loss": 0.3707,
      "reward": 1.4000000059604645,
      "reward_std": 0.5210216231644154,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1458
    },
    {
      "completion_length": 226.82500076293945,
      "epoch": 1.512182477967859,
      "grad_norm": 1.5981792211532593,
      "kl": 2.7060546875,
      "learning_rate": 1.7380497343923939e-06,
      "loss": 0.2833,
      "reward": 1.475000023841858,
      "reward_std": 0.3371223956346512,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1459
    },
    {
      "completion_length": 310.27499771118164,
      "epoch": 1.5132192846034216,
      "grad_norm": 1.6134463548660278,
      "kl": 1.82861328125,
      "learning_rate": 1.7362614489788746e-06,
      "loss": 0.2917,
      "reward": 1.3750000298023224,
      "reward_std": 0.40387551486492157,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1460
    },
    {
      "completion_length": 259.0750026702881,
      "epoch": 1.514256091238984,
      "grad_norm": 2.272364616394043,
      "kl": 2.445068359375,
      "learning_rate": 1.734472819153596e-06,
      "loss": 0.3094,
      "reward": 1.4000000059604645,
      "reward_std": 0.3407529182732105,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1461
    },
    {
      "completion_length": 174.95000457763672,
      "epoch": 1.5152928978745464,
      "grad_norm": 0.6720190048217773,
      "kl": 0.226318359375,
      "learning_rate": 1.7326838475239455e-06,
      "loss": 0.0733,
      "reward": 1.75,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1462
    },
    {
      "completion_length": 485.5500030517578,
      "epoch": 1.5163297045101087,
      "grad_norm": 3.0803070068359375,
      "kl": 4.446044921875,
      "learning_rate": 1.7308945366978087e-06,
      "loss": 0.6859,
      "reward": 1.1000000089406967,
      "reward_std": 0.6020505987107754,
      "rewards/accuracy_reward": 0.32500000670552254,
      "rewards/format_reward": 0.7750000059604645,
      "step": 1463
    },
    {
      "completion_length": 275.40000343322754,
      "epoch": 1.5173665111456713,
      "grad_norm": 0.49723660945892334,
      "kl": 2.191650390625,
      "learning_rate": 1.7291048892835658e-06,
      "loss": 0.32,
      "reward": 1.475000038743019,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 1464
    },
    {
      "completion_length": 190.39999961853027,
      "epoch": 1.518403317781234,
      "grad_norm": 0.2684459984302521,
      "kl": 0.29150390625,
      "learning_rate": 1.727314907890088e-06,
      "loss": 0.0613,
      "reward": 1.4749999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1465
    },
    {
      "completion_length": 127.55000114440918,
      "epoch": 1.5194401244167963,
      "grad_norm": 0.3302256464958191,
      "kl": 0.25146484375,
      "learning_rate": 1.725524595126733e-06,
      "loss": 0.0094,
      "reward": 1.4500000476837158,
      "reward_std": 0.2487340271472931,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 1.0,
      "step": 1466
    },
    {
      "completion_length": 195.600004196167,
      "epoch": 1.5204769310523587,
      "grad_norm": 0.4501000642776489,
      "kl": 1.028076171875,
      "learning_rate": 1.7237339536033417e-06,
      "loss": 0.1607,
      "reward": 1.4750000089406967,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1467
    },
    {
      "completion_length": 174.77500343322754,
      "epoch": 1.521513737687921,
      "grad_norm": 0.3627508580684662,
      "kl": 0.228271484375,
      "learning_rate": 1.7219429859302336e-06,
      "loss": 0.0539,
      "reward": 1.725000023841858,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1468
    },
    {
      "completion_length": 184.475004196167,
      "epoch": 1.5225505443234837,
      "grad_norm": 0.448893666267395,
      "kl": 0.669921875,
      "learning_rate": 1.7201516947182045e-06,
      "loss": 0.1209,
      "reward": 1.6750000417232513,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1469
    },
    {
      "completion_length": 267.6750030517578,
      "epoch": 1.5235873509590463,
      "grad_norm": 1.081580638885498,
      "kl": 2.113525390625,
      "learning_rate": 1.7183600825785218e-06,
      "loss": 0.2725,
      "reward": 1.7000000178813934,
      "reward_std": 0.27228760346770287,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1470
    },
    {
      "completion_length": 362.82501220703125,
      "epoch": 1.5246241575946087,
      "grad_norm": 3.3258347511291504,
      "kl": 2.7412109375,
      "learning_rate": 1.71656815212292e-06,
      "loss": 0.4794,
      "reward": 1.4500000178813934,
      "reward_std": 0.6149186752736568,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1471
    },
    {
      "completion_length": 126.37500190734863,
      "epoch": 1.525660964230171,
      "grad_norm": 0.5189715027809143,
      "kl": 0.267333984375,
      "learning_rate": 1.7147759059635987e-06,
      "loss": 0.0503,
      "reward": 1.5500000417232513,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1472
    },
    {
      "completion_length": 112.82500076293945,
      "epoch": 1.5266977708657334,
      "grad_norm": 0.3307163715362549,
      "kl": 0.283447265625,
      "learning_rate": 1.7129833467132167e-06,
      "loss": -0.003,
      "reward": 1.4000000059604645,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1473
    },
    {
      "completion_length": 179.07500076293945,
      "epoch": 1.527734577501296,
      "grad_norm": 0.42351534962654114,
      "kl": 1.09619140625,
      "learning_rate": 1.7111904769848892e-06,
      "loss": 0.0423,
      "reward": 1.5,
      "reward_std": 0.1369306445121765,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1474
    },
    {
      "completion_length": 222.62500381469727,
      "epoch": 1.5287713841368584,
      "grad_norm": 0.4376266598701477,
      "kl": 0.935791015625,
      "learning_rate": 1.7093972993921855e-06,
      "loss": 0.1267,
      "reward": 1.625,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1475
    },
    {
      "completion_length": 411.6250114440918,
      "epoch": 1.529808190772421,
      "grad_norm": 3.8951869010925293,
      "kl": 3.93310546875,
      "learning_rate": 1.7076038165491215e-06,
      "loss": 0.5593,
      "reward": 1.4500000178813934,
      "reward_std": 0.5590169690549374,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1476
    },
    {
      "completion_length": 133.95000076293945,
      "epoch": 1.5308449974079834,
      "grad_norm": 4.108523845672607,
      "kl": 0.268310546875,
      "learning_rate": 1.7058100310701605e-06,
      "loss": -0.03,
      "reward": 1.425000011920929,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 1.0,
      "step": 1477
    },
    {
      "completion_length": 160.47500228881836,
      "epoch": 1.5318818040435458,
      "grad_norm": 0.4052504897117615,
      "kl": 1.052978515625,
      "learning_rate": 1.7040159455702044e-06,
      "loss": 0.1561,
      "reward": 1.4500000178813934,
      "reward_std": 0.3046357296407223,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1478
    },
    {
      "completion_length": 221.30000114440918,
      "epoch": 1.5329186106791084,
      "grad_norm": 0.8664781451225281,
      "kl": 2.45751953125,
      "learning_rate": 1.7022215626645942e-06,
      "loss": 0.2669,
      "reward": 1.7500000298023224,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1479
    },
    {
      "completion_length": 115.20000267028809,
      "epoch": 1.5339554173146708,
      "grad_norm": 2.1453802585601807,
      "kl": 0.21435546875,
      "learning_rate": 1.7004268849691045e-06,
      "loss": 0.0208,
      "reward": 1.7249999940395355,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 1.0,
      "step": 1480
    },
    {
      "completion_length": 97.4000015258789,
      "epoch": 1.5349922239502334,
      "grad_norm": 0.1985270082950592,
      "kl": 0.3603515625,
      "learning_rate": 1.6986319150999382e-06,
      "loss": 0.0396,
      "reward": 1.7249999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 1.0,
      "step": 1481
    },
    {
      "completion_length": 162.35000038146973,
      "epoch": 1.5360290305857958,
      "grad_norm": 0.38188207149505615,
      "kl": 1.603271484375,
      "learning_rate": 1.696836655673726e-06,
      "loss": 0.1492,
      "reward": 1.600000023841858,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1482
    },
    {
      "completion_length": 313.3750057220459,
      "epoch": 1.5370658372213581,
      "grad_norm": 3.4883224964141846,
      "kl": 7.850341796875,
      "learning_rate": 1.695041109307519e-06,
      "loss": 0.7717,
      "reward": 1.0750000178813934,
      "reward_std": 0.5533697828650475,
      "rewards/accuracy_reward": 0.25000000931322575,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1483
    },
    {
      "completion_length": 365.2750053405762,
      "epoch": 1.5381026438569205,
      "grad_norm": 12.899331092834473,
      "kl": 2.89892578125,
      "learning_rate": 1.6932452786187875e-06,
      "loss": 0.4664,
      "reward": 1.225000023841858,
      "reward_std": 0.3354101814329624,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1484
    },
    {
      "completion_length": 254.17500114440918,
      "epoch": 1.5391394504924831,
      "grad_norm": 1.319108247756958,
      "kl": 3.1611328125,
      "learning_rate": 1.6914491662254159e-06,
      "loss": 0.3845,
      "reward": 1.6500000059604645,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1485
    },
    {
      "completion_length": 169.95000457763672,
      "epoch": 1.5401762571280457,
      "grad_norm": 1.1633234024047852,
      "kl": 0.43603515625,
      "learning_rate": 1.6896527747456997e-06,
      "loss": 0.0378,
      "reward": 1.5250000208616257,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1486
    },
    {
      "completion_length": 106.10000228881836,
      "epoch": 1.5412130637636081,
      "grad_norm": 0.48456689715385437,
      "kl": 0.24267578125,
      "learning_rate": 1.6878561067983412e-06,
      "loss": -0.0568,
      "reward": 1.8250000178813934,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 1.0,
      "step": 1487
    },
    {
      "completion_length": 240.45000457763672,
      "epoch": 1.5422498703991705,
      "grad_norm": 0.9362939596176147,
      "kl": 2.4794921875,
      "learning_rate": 1.6860591650024454e-06,
      "loss": 0.3434,
      "reward": 1.475000038743019,
      "reward_std": 0.2920721061527729,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 1488
    },
    {
      "completion_length": 284.25000762939453,
      "epoch": 1.5432866770347329,
      "grad_norm": 2.9841935634613037,
      "kl": 1.6005859375,
      "learning_rate": 1.684261951977516e-06,
      "loss": 0.2635,
      "reward": 1.3250000178813934,
      "reward_std": 0.3354101814329624,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1489
    },
    {
      "completion_length": 156.02500534057617,
      "epoch": 1.5443234836702955,
      "grad_norm": 0.9497570395469666,
      "kl": 1.114501953125,
      "learning_rate": 1.6824644703434537e-06,
      "loss": 0.1514,
      "reward": 1.5750000178813934,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.6000000182539225,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1490
    },
    {
      "completion_length": 209.2750072479248,
      "epoch": 1.545360290305858,
      "grad_norm": 0.2899838387966156,
      "kl": 2.5693359375,
      "learning_rate": 1.6806667227205487e-06,
      "loss": 0.2662,
      "reward": 1.4500000178813934,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 1491
    },
    {
      "completion_length": 160.82500457763672,
      "epoch": 1.5463970969414205,
      "grad_norm": 2.7640109062194824,
      "kl": 1.458984375,
      "learning_rate": 1.6788687117294811e-06,
      "loss": 0.1608,
      "reward": 1.4000000357627869,
      "reward_std": 0.37310105934739113,
      "rewards/accuracy_reward": 0.45000001415610313,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1492
    },
    {
      "completion_length": 157.67500114440918,
      "epoch": 1.5474339035769828,
      "grad_norm": 0.7922787070274353,
      "kl": 1.201416015625,
      "learning_rate": 1.677070439991313e-06,
      "loss": 0.1548,
      "reward": 1.6500000059604645,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1493
    },
    {
      "completion_length": 162.29999923706055,
      "epoch": 1.5484707102125452,
      "grad_norm": 4.747604846954346,
      "kl": 2.978515625,
      "learning_rate": 1.6752719101274876e-06,
      "loss": 0.2523,
      "reward": 1.5750000476837158,
      "reward_std": 0.3479738123714924,
      "rewards/accuracy_reward": 0.6250000223517418,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1494
    },
    {
      "completion_length": 156.52500534057617,
      "epoch": 1.5495075168481078,
      "grad_norm": 1.032403826713562,
      "kl": 0.344970703125,
      "learning_rate": 1.673473124759825e-06,
      "loss": 0.0548,
      "reward": 1.800000011920929,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1495
    },
    {
      "completion_length": 162.87500190734863,
      "epoch": 1.5505443234836704,
      "grad_norm": 0.3170575499534607,
      "kl": 1.254638671875,
      "learning_rate": 1.671674086510516e-06,
      "loss": 0.1184,
      "reward": 1.6500000059604645,
      "reward_std": 0.32818929478526115,
      "rewards/accuracy_reward": 0.6750000268220901,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1496
    },
    {
      "completion_length": 209.50000190734863,
      "epoch": 1.5515811301192328,
      "grad_norm": 1.0937821865081787,
      "kl": 1.545654296875,
      "learning_rate": 1.6698747980021213e-06,
      "loss": 0.133,
      "reward": 1.5750000178813934,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1497
    },
    {
      "completion_length": 104.4000015258789,
      "epoch": 1.5526179367547952,
      "grad_norm": 1.5526025295257568,
      "kl": 0.243896484375,
      "learning_rate": 1.6680752618575665e-06,
      "loss": -0.0018,
      "reward": 1.9000000059604645,
      "reward_std": 0.18026872351765633,
      "rewards/accuracy_reward": 0.9000000059604645,
      "rewards/format_reward": 1.0,
      "step": 1498
    },
    {
      "completion_length": 203.17499923706055,
      "epoch": 1.5536547433903576,
      "grad_norm": 0.3623444736003876,
      "kl": 2.51123046875,
      "learning_rate": 1.6662754807001376e-06,
      "loss": 0.1818,
      "reward": 1.550000011920929,
      "reward_std": 0.4290027692914009,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1499
    },
    {
      "completion_length": 203.7500057220459,
      "epoch": 1.5546915500259202,
      "grad_norm": 1.1907838582992554,
      "kl": 1.669921875,
      "learning_rate": 1.664475457153478e-06,
      "loss": 0.2063,
      "reward": 1.6500000059604645,
      "reward_std": 0.24873405322432518,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1500
    },
    {
      "completion_length": 168.37500381469727,
      "epoch": 1.5557283566614828,
      "grad_norm": 1.3985354900360107,
      "kl": 0.7236328125,
      "learning_rate": 1.6626751938415847e-06,
      "loss": 0.1602,
      "reward": 1.5500000268220901,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1501
    },
    {
      "completion_length": 253.4250087738037,
      "epoch": 1.5567651632970452,
      "grad_norm": 2.0021469593048096,
      "kl": 1.75341796875,
      "learning_rate": 1.6608746933888036e-06,
      "loss": 0.1339,
      "reward": 1.7750000059604645,
      "reward_std": 0.2974148504436016,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 1502
    },
    {
      "completion_length": 243.32500457763672,
      "epoch": 1.5578019699326076,
      "grad_norm": 0.421642541885376,
      "kl": 1.3193359375,
      "learning_rate": 1.6590739584198273e-06,
      "loss": 0.2241,
      "reward": 1.4000000208616257,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1503
    },
    {
      "completion_length": 194.5750026702881,
      "epoch": 1.55883877656817,
      "grad_norm": 0.661420464515686,
      "kl": 1.241455078125,
      "learning_rate": 1.6572729915596884e-06,
      "loss": 0.0457,
      "reward": 1.425000011920929,
      "reward_std": 0.3479738049209118,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1504
    },
    {
      "completion_length": 349.5250053405762,
      "epoch": 1.5598755832037325,
      "grad_norm": 8.147127151489258,
      "kl": 6.46875,
      "learning_rate": 1.6554717954337604e-06,
      "loss": 0.4454,
      "reward": 1.2000000178813934,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1505
    },
    {
      "completion_length": 212.5500087738037,
      "epoch": 1.560912389839295,
      "grad_norm": 0.6585958003997803,
      "kl": 3.0361328125,
      "learning_rate": 1.6536703726677483e-06,
      "loss": 0.3097,
      "reward": 1.4000000059604645,
      "reward_std": 0.4237985350191593,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.9000000357627869,
      "step": 1506
    },
    {
      "completion_length": 208.1250057220459,
      "epoch": 1.5619491964748575,
      "grad_norm": 1.088865876197815,
      "kl": 0.254638671875,
      "learning_rate": 1.6518687258876884e-06,
      "loss": 0.0666,
      "reward": 1.1250000149011612,
      "reward_std": 0.30463573709130287,
      "rewards/accuracy_reward": 0.17500000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1507
    },
    {
      "completion_length": 252.65000534057617,
      "epoch": 1.56298600311042,
      "grad_norm": 0.4406292736530304,
      "kl": 0.35107421875,
      "learning_rate": 1.6500668577199448e-06,
      "loss": 0.18,
      "reward": 1.600000038743019,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1508
    },
    {
      "completion_length": 252.37499809265137,
      "epoch": 1.5640228097459823,
      "grad_norm": 0.7803511023521423,
      "kl": 2.29052734375,
      "learning_rate": 1.648264770791202e-06,
      "loss": 0.3081,
      "reward": 1.475000023841858,
      "reward_std": 0.27950849011540413,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1509
    },
    {
      "completion_length": 150.12500190734863,
      "epoch": 1.565059616381545,
      "grad_norm": 0.6692258715629578,
      "kl": 0.607666015625,
      "learning_rate": 1.6464624677284663e-06,
      "loss": 0.0313,
      "reward": 1.5500000417232513,
      "reward_std": 0.41643911972641945,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1510
    },
    {
      "completion_length": 272.35000228881836,
      "epoch": 1.5660964230171073,
      "grad_norm": 0.6051040291786194,
      "kl": 2.58056640625,
      "learning_rate": 1.6446599511590558e-06,
      "loss": 0.3326,
      "reward": 1.2750000208616257,
      "reward_std": 0.4092182256281376,
      "rewards/accuracy_reward": 0.42499999701976776,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1511
    },
    {
      "completion_length": 297.75000190734863,
      "epoch": 1.5671332296526699,
      "grad_norm": 2.3226115703582764,
      "kl": 0.70849609375,
      "learning_rate": 1.6428572237106023e-06,
      "loss": 0.1106,
      "reward": 1.550000011920929,
      "reward_std": 0.284851249307394,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1512
    },
    {
      "completion_length": 287.9000053405762,
      "epoch": 1.5681700362882323,
      "grad_norm": 3.223996162414551,
      "kl": 2.470458984375,
      "learning_rate": 1.641054288011044e-06,
      "loss": 0.3918,
      "reward": 1.4750000089406967,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1513
    },
    {
      "completion_length": 201.95000076293945,
      "epoch": 1.5692068429237946,
      "grad_norm": 6.340351581573486,
      "kl": 0.7060546875,
      "learning_rate": 1.6392511466886231e-06,
      "loss": 0.134,
      "reward": 1.2000000327825546,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 0.925000011920929,
      "step": 1514
    },
    {
      "completion_length": 175.0250072479248,
      "epoch": 1.570243649559357,
      "grad_norm": 2.9794259071350098,
      "kl": 0.84814453125,
      "learning_rate": 1.637447802371881e-06,
      "loss": 0.1037,
      "reward": 1.425000011920929,
      "reward_std": 0.12436702102422714,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1515
    },
    {
      "completion_length": 256.1250057220459,
      "epoch": 1.5712804561949196,
      "grad_norm": 1.7978872060775757,
      "kl": 1.8125,
      "learning_rate": 1.635644257689656e-06,
      "loss": 0.2891,
      "reward": 1.550000011920929,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1516
    },
    {
      "completion_length": 358.0750026702881,
      "epoch": 1.5723172628304822,
      "grad_norm": 1.119584560394287,
      "kl": 3.794189453125,
      "learning_rate": 1.6338405152710764e-06,
      "loss": 0.5007,
      "reward": 1.600000023841858,
      "reward_std": 0.4723408445715904,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.875,
      "step": 1517
    },
    {
      "completion_length": 263.29999923706055,
      "epoch": 1.5733540694660446,
      "grad_norm": 0.628728985786438,
      "kl": 1.5361328125,
      "learning_rate": 1.6320365777455616e-06,
      "loss": 0.2288,
      "reward": 1.3500000089406967,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1518
    },
    {
      "completion_length": 221.97500228881836,
      "epoch": 1.574390876101607,
      "grad_norm": 4.136834621429443,
      "kl": 5.1005859375,
      "learning_rate": 1.6302324477428133e-06,
      "loss": 0.4654,
      "reward": 1.5250000357627869,
      "reward_std": 0.5533697605133057,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1519
    },
    {
      "completion_length": 345.6000099182129,
      "epoch": 1.5754276827371694,
      "grad_norm": 17.222375869750977,
      "kl": 6.73828125,
      "learning_rate": 1.6284281278928149e-06,
      "loss": 0.7314,
      "reward": 1.6500000059604645,
      "reward_std": 0.6635995022952557,
      "rewards/accuracy_reward": 0.8000000268220901,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1520
    },
    {
      "completion_length": 163.07500648498535,
      "epoch": 1.576464489372732,
      "grad_norm": 4.385284900665283,
      "kl": 1.533935546875,
      "learning_rate": 1.6266236208258257e-06,
      "loss": 0.1571,
      "reward": 1.7249999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1521
    },
    {
      "completion_length": 264.12500190734863,
      "epoch": 1.5775012960082946,
      "grad_norm": 1.2917828559875488,
      "kl": 2.667236328125,
      "learning_rate": 1.6248189291723782e-06,
      "loss": 0.3353,
      "reward": 1.4000000059604645,
      "reward_std": 0.3354101777076721,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.925000011920929,
      "step": 1522
    },
    {
      "completion_length": 322.75000953674316,
      "epoch": 1.578538102643857,
      "grad_norm": 1.4020047187805176,
      "kl": 0.292236328125,
      "learning_rate": 1.6230140555632748e-06,
      "loss": 0.0598,
      "reward": 1.4000000059604645,
      "reward_std": 0.34860679507255554,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1523
    },
    {
      "completion_length": 355.3250102996826,
      "epoch": 1.5795749092794193,
      "grad_norm": 1.2283453941345215,
      "kl": 3.8046875,
      "learning_rate": 1.6212090026295819e-06,
      "loss": 0.4933,
      "reward": 1.5500000268220901,
      "reward_std": 0.403875507414341,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1524
    },
    {
      "completion_length": 208.0250072479248,
      "epoch": 1.5806117159149817,
      "grad_norm": 4.229063987731934,
      "kl": 2.224609375,
      "learning_rate": 1.6194037730026283e-06,
      "loss": 0.2554,
      "reward": 1.5750000178813934,
      "reward_std": 0.36053745821118355,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1525
    },
    {
      "completion_length": 317.55000495910645,
      "epoch": 1.5816485225505443,
      "grad_norm": 1.123126745223999,
      "kl": 1.52490234375,
      "learning_rate": 1.6175983693139994e-06,
      "loss": 0.2079,
      "reward": 1.475000023841858,
      "reward_std": 0.4092182330787182,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1526
    },
    {
      "completion_length": 107.67500114440918,
      "epoch": 1.582685329186107,
      "grad_norm": 1.740889310836792,
      "kl": 0.271484375,
      "learning_rate": 1.6157927941955354e-06,
      "loss": 0.0049,
      "reward": 1.425000011920929,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 1.0,
      "step": 1527
    },
    {
      "completion_length": 151.25000190734863,
      "epoch": 1.5837221358216693,
      "grad_norm": 1.7476391792297363,
      "kl": 1.57470703125,
      "learning_rate": 1.613987050279326e-06,
      "loss": 0.1968,
      "reward": 1.4000000208616257,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.45000001788139343,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1528
    },
    {
      "completion_length": 170.39999771118164,
      "epoch": 1.5847589424572317,
      "grad_norm": 0.23554155230522156,
      "kl": 1.2509765625,
      "learning_rate": 1.6121811401977062e-06,
      "loss": 0.1608,
      "reward": 1.550000011920929,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.925000011920929,
      "step": 1529
    },
    {
      "completion_length": 241.22500610351562,
      "epoch": 1.585795749092794,
      "grad_norm": 2.0842747688293457,
      "kl": 0.82373046875,
      "learning_rate": 1.6103750665832547e-06,
      "loss": 0.1447,
      "reward": 1.5999999940395355,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.650000000372529,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1530
    },
    {
      "completion_length": 215.600004196167,
      "epoch": 1.5868325557283567,
      "grad_norm": 0.22583654522895813,
      "kl": 1.046630859375,
      "learning_rate": 1.608568832068787e-06,
      "loss": 0.1359,
      "reward": 1.5250000208616257,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1531
    },
    {
      "completion_length": 217.25000381469727,
      "epoch": 1.5878693623639193,
      "grad_norm": 7.7239089012146,
      "kl": 1.028076171875,
      "learning_rate": 1.6067624392873545e-06,
      "loss": 0.1174,
      "reward": 1.600000023841858,
      "reward_std": 0.45413000881671906,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1532
    },
    {
      "completion_length": 201.600004196167,
      "epoch": 1.5889061689994817,
      "grad_norm": 2.654834032058716,
      "kl": 1.376953125,
      "learning_rate": 1.6049558908722385e-06,
      "loss": 0.1816,
      "reward": 1.6750000417232513,
      "reward_std": 0.4849044531583786,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1533
    },
    {
      "completion_length": 245.22500228881836,
      "epoch": 1.589942975635044,
      "grad_norm": 1.0713359117507935,
      "kl": 2.8271484375,
      "learning_rate": 1.603149189456948e-06,
      "loss": 0.3116,
      "reward": 1.725000023841858,
      "reward_std": 0.3354102075099945,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1534
    },
    {
      "completion_length": 148.20000457763672,
      "epoch": 1.5909797822706064,
      "grad_norm": 0.886328399181366,
      "kl": 1.534423828125,
      "learning_rate": 1.6013423376752131e-06,
      "loss": 0.098,
      "reward": 1.5250000357627869,
      "reward_std": 0.13693063706159592,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1535
    },
    {
      "completion_length": 347.8999996185303,
      "epoch": 1.592016588906169,
      "grad_norm": 2.5690722465515137,
      "kl": 2.233642578125,
      "learning_rate": 1.5995353381609864e-06,
      "loss": 0.2542,
      "reward": 1.350000038743019,
      "reward_std": 0.3731010667979717,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1536
    },
    {
      "completion_length": 158.79999923706055,
      "epoch": 1.5930533955417314,
      "grad_norm": 0.9137997627258301,
      "kl": 0.298828125,
      "learning_rate": 1.597728193548432e-06,
      "loss": 0.0629,
      "reward": 1.6500000059604645,
      "reward_std": 0.18026873469352722,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1537
    },
    {
      "completion_length": 174.25000190734863,
      "epoch": 1.594090202177294,
      "grad_norm": 0.7593628764152527,
      "kl": 1.142333984375,
      "learning_rate": 1.5959209064719291e-06,
      "loss": 0.1649,
      "reward": 1.6250000298023224,
      "reward_std": 0.3479738011956215,
      "rewards/accuracy_reward": 0.6750000063329935,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1538
    },
    {
      "completion_length": 327.800012588501,
      "epoch": 1.5951270088128564,
      "grad_norm": 2.3554060459136963,
      "kl": 2.83251953125,
      "learning_rate": 1.594113479566062e-06,
      "loss": 0.2768,
      "reward": 1.475000023841858,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1539
    },
    {
      "completion_length": 216.55000686645508,
      "epoch": 1.5961638154484188,
      "grad_norm": 1.6953859329223633,
      "kl": 0.8837890625,
      "learning_rate": 1.5923059154656195e-06,
      "loss": 0.1239,
      "reward": 1.675000011920929,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.7250000033527613,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1540
    },
    {
      "completion_length": 132.6500015258789,
      "epoch": 1.5972006220839814,
      "grad_norm": 0.8555593490600586,
      "kl": 0.274658203125,
      "learning_rate": 1.5904982168055912e-06,
      "loss": 0.0857,
      "reward": 1.4749999940395355,
      "reward_std": 0.3046357296407223,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 1.0,
      "step": 1541
    },
    {
      "completion_length": 159.29999923706055,
      "epoch": 1.5982374287195438,
      "grad_norm": 0.48500317335128784,
      "kl": 0.344970703125,
      "learning_rate": 1.5886903862211622e-06,
      "loss": 0.0902,
      "reward": 1.5500000417232513,
      "reward_std": 0.373101070523262,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1542
    },
    {
      "completion_length": 303.6000003814697,
      "epoch": 1.5992742353551064,
      "grad_norm": 1.3442599773406982,
      "kl": 3.2978515625,
      "learning_rate": 1.58688242634771e-06,
      "loss": 0.3328,
      "reward": 1.4000000059604645,
      "reward_std": 0.31199514865875244,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1543
    },
    {
      "completion_length": 387.8250160217285,
      "epoch": 1.6003110419906688,
      "grad_norm": 1.801302433013916,
      "kl": 5.2314453125,
      "learning_rate": 1.5850743398208008e-06,
      "loss": 0.513,
      "reward": 1.425000011920929,
      "reward_std": 0.5841442123055458,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1544
    },
    {
      "completion_length": 244.67499923706055,
      "epoch": 1.6013478486262311,
      "grad_norm": 2.613616943359375,
      "kl": 1.364501953125,
      "learning_rate": 1.5832661292761848e-06,
      "loss": 0.1866,
      "reward": 1.425000011920929,
      "reward_std": 0.40387551486492157,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1545
    },
    {
      "completion_length": 295.6000003814697,
      "epoch": 1.6023846552617935,
      "grad_norm": 4.084285736083984,
      "kl": 3.672119140625,
      "learning_rate": 1.5814577973497938e-06,
      "loss": 0.4096,
      "reward": 1.7249999940395355,
      "reward_std": 0.2650667019188404,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1546
    },
    {
      "completion_length": 166.1500015258789,
      "epoch": 1.6034214618973561,
      "grad_norm": 2.872854232788086,
      "kl": 0.380859375,
      "learning_rate": 1.5796493466777356e-06,
      "loss": 0.0488,
      "reward": 1.6000000089406967,
      "reward_std": 0.1442900411784649,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1547
    },
    {
      "completion_length": 196.4250030517578,
      "epoch": 1.6044582685329187,
      "grad_norm": 2.0663416385650635,
      "kl": 1.898193359375,
      "learning_rate": 1.5778407798962925e-06,
      "loss": 0.1581,
      "reward": 1.5750000178813934,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1548
    },
    {
      "completion_length": 305.82501220703125,
      "epoch": 1.6054950751684811,
      "grad_norm": 1.5803462266921997,
      "kl": 0.78271484375,
      "learning_rate": 1.576032099641915e-06,
      "loss": 0.2732,
      "reward": 1.4000000059604645,
      "reward_std": 0.6040672399103642,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1549
    },
    {
      "completion_length": 174.47500038146973,
      "epoch": 1.6065318818040435,
      "grad_norm": 1.148290753364563,
      "kl": 1.66748046875,
      "learning_rate": 1.5742233085512183e-06,
      "loss": 0.0676,
      "reward": 1.5750000178813934,
      "reward_std": 0.40387552976608276,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1550
    },
    {
      "completion_length": 211.50000953674316,
      "epoch": 1.6075686884396059,
      "grad_norm": 1.313335657119751,
      "kl": 1.8564453125,
      "learning_rate": 1.572414409260981e-06,
      "loss": 0.2061,
      "reward": 1.600000023841858,
      "reward_std": 0.38409099727869034,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1551
    },
    {
      "completion_length": 98.52500343322754,
      "epoch": 1.6086054950751685,
      "grad_norm": 1.1569651365280151,
      "kl": 0.256103515625,
      "learning_rate": 1.5706054044081385e-06,
      "loss": 0.0219,
      "reward": 1.7249999940395355,
      "reward_std": 0.1442900449037552,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1552
    },
    {
      "completion_length": 393.100004196167,
      "epoch": 1.609642301710731,
      "grad_norm": 0.44840702414512634,
      "kl": 1.95068359375,
      "learning_rate": 1.56879629662978e-06,
      "loss": 0.4238,
      "reward": 1.475000023841858,
      "reward_std": 0.5282425284385681,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1553
    },
    {
      "completion_length": 203.4250030517578,
      "epoch": 1.6106791083462935,
      "grad_norm": 9.99445915222168,
      "kl": 1.92578125,
      "learning_rate": 1.566987088563145e-06,
      "loss": 0.2378,
      "reward": 1.600000023841858,
      "reward_std": 0.21638590842485428,
      "rewards/accuracy_reward": 0.6500000022351742,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1554
    },
    {
      "completion_length": 396.0250015258789,
      "epoch": 1.6117159149818558,
      "grad_norm": 12.977739334106445,
      "kl": 6.28515625,
      "learning_rate": 1.5651777828456183e-06,
      "loss": 0.6928,
      "reward": 1.6500000357627869,
      "reward_std": 0.6959476321935654,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1555
    },
    {
      "completion_length": 399.100004196167,
      "epoch": 1.6127527216174182,
      "grad_norm": 5.44423770904541,
      "kl": 3.6025390625,
      "learning_rate": 1.5633683821147291e-06,
      "loss": 0.534,
      "reward": 1.4000000208616257,
      "reward_std": 0.5517960898578167,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1556
    },
    {
      "completion_length": 339.4250011444092,
      "epoch": 1.6137895282529808,
      "grad_norm": 2.6319031715393066,
      "kl": 4.25048828125,
      "learning_rate": 1.5615588890081424e-06,
      "loss": 0.5455,
      "reward": 1.2500000298023224,
      "reward_std": 0.540806170552969,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1557
    },
    {
      "completion_length": 691.7500228881836,
      "epoch": 1.6148263348885434,
      "grad_norm": 2.091921329498291,
      "kl": 4.77490234375,
      "learning_rate": 1.5597493061636607e-06,
      "loss": 0.6976,
      "reward": 1.0750000178813934,
      "reward_std": 0.7314124554395676,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.6749999970197678,
      "step": 1558
    },
    {
      "completion_length": 453.77501106262207,
      "epoch": 1.6158631415241058,
      "grad_norm": 2.37636399269104,
      "kl": 3.71484375,
      "learning_rate": 1.5579396362192148e-06,
      "loss": 0.3806,
      "reward": 1.0750000029802322,
      "reward_std": 0.5100317224860191,
      "rewards/accuracy_reward": 0.29999999701976776,
      "rewards/format_reward": 0.7750000059604645,
      "step": 1559
    },
    {
      "completion_length": 495.950008392334,
      "epoch": 1.6168999481596682,
      "grad_norm": 4.216372489929199,
      "kl": 4.93701171875,
      "learning_rate": 1.556129881812864e-06,
      "loss": 0.6983,
      "reward": 1.225000038743019,
      "reward_std": 0.6092714965343475,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.7500000149011612,
      "step": 1560
    },
    {
      "completion_length": 203.35000228881836,
      "epoch": 1.6179367547952306,
      "grad_norm": 5.362270355224609,
      "kl": 1.5576171875,
      "learning_rate": 1.55432004558279e-06,
      "loss": 0.2287,
      "reward": 1.5250000059604645,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1561
    },
    {
      "completion_length": 442.7000141143799,
      "epoch": 1.6189735614307932,
      "grad_norm": 3.0055434703826904,
      "kl": 1.2265625,
      "learning_rate": 1.5525101301672946e-06,
      "loss": 0.3656,
      "reward": 1.3750000149011612,
      "reward_std": 0.6281152926385403,
      "rewards/accuracy_reward": 0.5500000081956387,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1562
    },
    {
      "completion_length": 301.3250045776367,
      "epoch": 1.6200103680663558,
      "grad_norm": 2.8841614723205566,
      "kl": 1.67919921875,
      "learning_rate": 1.550700138204794e-06,
      "loss": 0.2526,
      "reward": 1.550000011920929,
      "reward_std": 0.4045085087418556,
      "rewards/accuracy_reward": 0.6500000096857548,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1563
    },
    {
      "completion_length": 256.25000190734863,
      "epoch": 1.6210471747019182,
      "grad_norm": 1.5231977701187134,
      "kl": 3.0517578125,
      "learning_rate": 1.5488900723338173e-06,
      "loss": 0.284,
      "reward": 1.25,
      "reward_std": 0.4291412979364395,
      "rewards/accuracy_reward": 0.350000012665987,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1564
    },
    {
      "completion_length": 308.4500045776367,
      "epoch": 1.6220839813374806,
      "grad_norm": 0.8757073879241943,
      "kl": 4.7578125,
      "learning_rate": 1.547079935192999e-06,
      "loss": 0.568,
      "reward": 1.449999988079071,
      "reward_std": 0.50311528891325,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1565
    },
    {
      "completion_length": 253.64999961853027,
      "epoch": 1.623120787973043,
      "grad_norm": 5.401318073272705,
      "kl": 4.2685546875,
      "learning_rate": 1.5452697294210807e-06,
      "loss": 0.4422,
      "reward": 1.2750000059604645,
      "reward_std": 0.2974148392677307,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1566
    },
    {
      "completion_length": 255.12500190734863,
      "epoch": 1.6241575946086055,
      "grad_norm": 10.566208839416504,
      "kl": 3.754638671875,
      "learning_rate": 1.5434594576569009e-06,
      "loss": 0.421,
      "reward": 1.2750000357627869,
      "reward_std": 0.4958943948149681,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1567
    },
    {
      "completion_length": 104.60000228881836,
      "epoch": 1.625194401244168,
      "grad_norm": 0.4463522434234619,
      "kl": 0.328125,
      "learning_rate": 1.541649122539397e-06,
      "loss": 0.0442,
      "reward": 1.300000011920929,
      "reward_std": 0.30463575199246407,
      "rewards/accuracy_reward": 0.30000000819563866,
      "rewards/format_reward": 1.0,
      "step": 1568
    },
    {
      "completion_length": 107.47500228881836,
      "epoch": 1.6262312078797305,
      "grad_norm": 0.45555442571640015,
      "kl": 0.220947265625,
      "learning_rate": 1.5398387267075966e-06,
      "loss": -0.0156,
      "reward": 1.5250000059604645,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 1.0,
      "step": 1569
    },
    {
      "completion_length": 153.92500114440918,
      "epoch": 1.627268014515293,
      "grad_norm": 0.24410176277160645,
      "kl": 0.6796875,
      "learning_rate": 1.5380282728006174e-06,
      "loss": 0.0791,
      "reward": 1.600000023841858,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1570
    },
    {
      "completion_length": 181.37500190734863,
      "epoch": 1.6283048211508553,
      "grad_norm": 4.3899312019348145,
      "kl": 1.115966796875,
      "learning_rate": 1.5362177634576608e-06,
      "loss": 0.1201,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1571
    },
    {
      "completion_length": 298.2000026702881,
      "epoch": 1.6293416277864177,
      "grad_norm": 3.581695079803467,
      "kl": 3.4921875,
      "learning_rate": 1.5344072013180093e-06,
      "loss": 0.3671,
      "reward": 1.3250000327825546,
      "reward_std": 0.6047002524137497,
      "rewards/accuracy_reward": 0.47500000707805157,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1572
    },
    {
      "completion_length": 156.0,
      "epoch": 1.6303784344219803,
      "grad_norm": 1.390693187713623,
      "kl": 0.886474609375,
      "learning_rate": 1.5325965890210223e-06,
      "loss": 0.0974,
      "reward": 1.6500000059604645,
      "reward_std": 0.18026871606707573,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1573
    },
    {
      "completion_length": 359.3500099182129,
      "epoch": 1.6314152410575429,
      "grad_norm": 2.581305742263794,
      "kl": 3.68212890625,
      "learning_rate": 1.5307859292061326e-06,
      "loss": 0.4285,
      "reward": 1.375,
      "reward_std": 0.46511995792388916,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1574
    },
    {
      "completion_length": 210.4500026702881,
      "epoch": 1.6324520476931053,
      "grad_norm": 6.038915157318115,
      "kl": 0.876220703125,
      "learning_rate": 1.5289752245128416e-06,
      "loss": 0.1413,
      "reward": 1.475000023841858,
      "reward_std": 0.28485120832920074,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1575
    },
    {
      "completion_length": 207.3249969482422,
      "epoch": 1.6334888543286676,
      "grad_norm": 1.6099038124084473,
      "kl": 2.033447265625,
      "learning_rate": 1.5271644775807173e-06,
      "loss": 0.231,
      "reward": 1.5,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1576
    },
    {
      "completion_length": 122.95000267028809,
      "epoch": 1.63452566096423,
      "grad_norm": 0.775686502456665,
      "kl": 0.7314453125,
      "learning_rate": 1.525353691049388e-06,
      "loss": 0.0345,
      "reward": 1.8250000178813934,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1577
    },
    {
      "completion_length": 154.67500495910645,
      "epoch": 1.6355624675997926,
      "grad_norm": 0.4884706139564514,
      "kl": 0.86083984375,
      "learning_rate": 1.5235428675585404e-06,
      "loss": 0.1491,
      "reward": 1.6500000357627869,
      "reward_std": 0.41643914207816124,
      "rewards/accuracy_reward": 0.700000025331974,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1578
    },
    {
      "completion_length": 149.8750057220459,
      "epoch": 1.6365992742353552,
      "grad_norm": 1.7934447526931763,
      "kl": 0.4140625,
      "learning_rate": 1.5217320097479152e-06,
      "loss": 0.0524,
      "reward": 1.5000000298023224,
      "reward_std": 0.41643914952874184,
      "rewards/accuracy_reward": 0.5250000022351742,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1579
    },
    {
      "completion_length": 199.40000534057617,
      "epoch": 1.6376360808709176,
      "grad_norm": 0.4854935109615326,
      "kl": 1.685791015625,
      "learning_rate": 1.5199211202573021e-06,
      "loss": 0.1541,
      "reward": 1.600000023841858,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 1580
    },
    {
      "completion_length": 110.95000267028809,
      "epoch": 1.63867288750648,
      "grad_norm": 2.2596967220306396,
      "kl": 0.27294921875,
      "learning_rate": 1.5181102017265389e-06,
      "loss": -0.0106,
      "reward": 1.7750000059604645,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.7749999985098839,
      "rewards/format_reward": 1.0,
      "step": 1581
    },
    {
      "completion_length": 110.32500076293945,
      "epoch": 1.6397096941420424,
      "grad_norm": 0.32057473063468933,
      "kl": 0.295654296875,
      "learning_rate": 1.5162992567955044e-06,
      "loss": -0.0126,
      "reward": 1.375,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 1.0,
      "step": 1582
    },
    {
      "completion_length": 410.02501678466797,
      "epoch": 1.640746500777605,
      "grad_norm": 0.5861947536468506,
      "kl": 4.6337890625,
      "learning_rate": 1.5144882881041147e-06,
      "loss": 0.5881,
      "reward": 1.4000000059604645,
      "reward_std": 0.6274823173880577,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1583
    },
    {
      "completion_length": 155.55000495910645,
      "epoch": 1.6417833074131676,
      "grad_norm": 6.338595867156982,
      "kl": 1.751953125,
      "learning_rate": 1.5126772982923235e-06,
      "loss": 0.1598,
      "reward": 1.3750000298023224,
      "reward_std": 0.38409099727869034,
      "rewards/accuracy_reward": 0.4500000160187483,
      "rewards/format_reward": 0.925000011920929,
      "step": 1584
    },
    {
      "completion_length": 159.07500457763672,
      "epoch": 1.64282011404873,
      "grad_norm": 1.1443921327590942,
      "kl": 2.912841796875,
      "learning_rate": 1.510866290000113e-06,
      "loss": 0.2813,
      "reward": 1.6250000298023224,
      "reward_std": 0.23617041110992432,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1585
    },
    {
      "completion_length": 250.20001029968262,
      "epoch": 1.6438569206842923,
      "grad_norm": 5.0934953689575195,
      "kl": 1.823486328125,
      "learning_rate": 1.5090552658674932e-06,
      "loss": 0.256,
      "reward": 1.5500000417232513,
      "reward_std": 0.32818929851055145,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1586
    },
    {
      "completion_length": 281.1500015258789,
      "epoch": 1.6448937273198547,
      "grad_norm": 1.359196424484253,
      "kl": 6.293212890625,
      "learning_rate": 1.5072442285344971e-06,
      "loss": 0.5661,
      "reward": 1.2500000149011612,
      "reward_std": 0.46511995047330856,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1587
    },
    {
      "completion_length": 154.77499961853027,
      "epoch": 1.6459305339554173,
      "grad_norm": 11.460147857666016,
      "kl": 1.093994140625,
      "learning_rate": 1.5054331806411767e-06,
      "loss": 0.0877,
      "reward": 1.4500000178813934,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1588
    },
    {
      "completion_length": 200.77500343322754,
      "epoch": 1.64696734059098,
      "grad_norm": 0.5465800762176514,
      "kl": 2.06982421875,
      "learning_rate": 1.5036221248275996e-06,
      "loss": 0.2584,
      "reward": 1.6000000536441803,
      "reward_std": 0.3479738123714924,
      "rewards/accuracy_reward": 0.650000024586916,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1589
    },
    {
      "completion_length": 221.95000457763672,
      "epoch": 1.6480041472265423,
      "grad_norm": 0.8402000665664673,
      "kl": 2.9384765625,
      "learning_rate": 1.5018110637338452e-06,
      "loss": 0.1917,
      "reward": 1.3500000089406967,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1590
    },
    {
      "completion_length": 241.94999885559082,
      "epoch": 1.6490409538621047,
      "grad_norm": 1.211803674697876,
      "kl": 2.736083984375,
      "learning_rate": 1.5e-06,
      "loss": 0.2163,
      "reward": 1.5250000208616257,
      "reward_std": 0.22894951701164246,
      "rewards/accuracy_reward": 0.6000000033527613,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1591
    },
    {
      "completion_length": 159.375,
      "epoch": 1.650077760497667,
      "grad_norm": 2.434783935546875,
      "kl": 2.064208984375,
      "learning_rate": 1.498188936266155e-06,
      "loss": 0.1954,
      "reward": 1.7750000357627869,
      "reward_std": 0.36053744331002235,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1592
    },
    {
      "completion_length": 262.2250118255615,
      "epoch": 1.6511145671332297,
      "grad_norm": 0.5691776871681213,
      "kl": 1.02587890625,
      "learning_rate": 1.4963778751724005e-06,
      "loss": 0.2409,
      "reward": 1.725000023841858,
      "reward_std": 0.3354102075099945,
      "rewards/accuracy_reward": 0.8000000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1593
    },
    {
      "completion_length": 107.75000190734863,
      "epoch": 1.6521513737687923,
      "grad_norm": 0.8404366970062256,
      "kl": 0.28369140625,
      "learning_rate": 1.4945668193588233e-06,
      "loss": 0.0043,
      "reward": 1.550000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 1.0,
      "step": 1594
    },
    {
      "completion_length": 216.37500381469727,
      "epoch": 1.6531881804043547,
      "grad_norm": 0.4120844006538391,
      "kl": 0.34814453125,
      "learning_rate": 1.492755771465503e-06,
      "loss": 0.1214,
      "reward": 1.5250000059604645,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1595
    },
    {
      "completion_length": 199.2000026702881,
      "epoch": 1.654224987039917,
      "grad_norm": 0.5536773800849915,
      "kl": 1.704833984375,
      "learning_rate": 1.4909447341325067e-06,
      "loss": 0.1642,
      "reward": 1.550000011920929,
      "reward_std": 0.36053743213415146,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1596
    },
    {
      "completion_length": 163.5750026702881,
      "epoch": 1.6552617936754794,
      "grad_norm": 5.34987211227417,
      "kl": 0.679443359375,
      "learning_rate": 1.4891337099998875e-06,
      "loss": 0.0953,
      "reward": 1.4500000476837158,
      "reward_std": 0.3731010667979717,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1597
    },
    {
      "completion_length": 123.05000305175781,
      "epoch": 1.656298600311042,
      "grad_norm": 0.5703464150428772,
      "kl": 0.23193359375,
      "learning_rate": 1.4873227017076763e-06,
      "loss": 0.0431,
      "reward": 1.8250000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 1.0,
      "step": 1598
    },
    {
      "completion_length": 304.200008392334,
      "epoch": 1.6573354069466044,
      "grad_norm": 0.6134169697761536,
      "kl": 2.039794921875,
      "learning_rate": 1.4855117118958856e-06,
      "loss": 0.2122,
      "reward": 1.3500000089406967,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1599
    },
    {
      "completion_length": 170.00000190734863,
      "epoch": 1.658372213582167,
      "grad_norm": 0.38233059644699097,
      "kl": 0.826904296875,
      "learning_rate": 1.4837007432044961e-06,
      "loss": 0.0966,
      "reward": 1.375,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.4000000078231096,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1600
    },
    {
      "completion_length": 318.7750015258789,
      "epoch": 1.6594090202177294,
      "grad_norm": 4.728154182434082,
      "kl": 3.4892578125,
      "learning_rate": 1.4818897982734612e-06,
      "loss": 0.2733,
      "reward": 1.6500000357627869,
      "reward_std": 0.5084579959511757,
      "rewards/accuracy_reward": 0.7750000134110451,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1601
    },
    {
      "completion_length": 281.7500114440918,
      "epoch": 1.6604458268532918,
      "grad_norm": 2.4841439723968506,
      "kl": 1.1044921875,
      "learning_rate": 1.4800788797426977e-06,
      "loss": 0.2515,
      "reward": 1.175000011920929,
      "reward_std": 0.25609343871474266,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1602
    },
    {
      "completion_length": 349.22501373291016,
      "epoch": 1.6614826334888542,
      "grad_norm": 1.3508840799331665,
      "kl": 3.63525390625,
      "learning_rate": 1.478267990252085e-06,
      "loss": 0.3127,
      "reward": 1.1750000268220901,
      "reward_std": 0.44892580807209015,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1603
    },
    {
      "completion_length": 203.3750057220459,
      "epoch": 1.6625194401244168,
      "grad_norm": 2.026071310043335,
      "kl": 1.2587890625,
      "learning_rate": 1.4764571324414595e-06,
      "loss": 0.1681,
      "reward": 1.5250000357627869,
      "reward_std": 0.292072132229805,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1604
    },
    {
      "completion_length": 163.80000114440918,
      "epoch": 1.6635562467599794,
      "grad_norm": 0.2313051074743271,
      "kl": 0.25537109375,
      "learning_rate": 1.474646308950612e-06,
      "loss": 0.0659,
      "reward": 1.600000023841858,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1605
    },
    {
      "completion_length": 122.25,
      "epoch": 1.6645930533955418,
      "grad_norm": 0.35142838954925537,
      "kl": 0.445556640625,
      "learning_rate": 1.4728355224192826e-06,
      "loss": 0.0286,
      "reward": 1.5999999940395355,
      "reward_std": 0.21638590842485428,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1606
    },
    {
      "completion_length": 227.82500457763672,
      "epoch": 1.6656298600311041,
      "grad_norm": 0.5352864861488342,
      "kl": 1.51806640625,
      "learning_rate": 1.4710247754871587e-06,
      "loss": 0.205,
      "reward": 1.5000000298023224,
      "reward_std": 0.4415663778781891,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1607
    },
    {
      "completion_length": 116.69999885559082,
      "epoch": 1.6666666666666665,
      "grad_norm": 0.6908564567565918,
      "kl": 0.255859375,
      "learning_rate": 1.4692140707938675e-06,
      "loss": 0.0067,
      "reward": 1.9500000178813934,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.9500000029802322,
      "rewards/format_reward": 1.0,
      "step": 1608
    },
    {
      "completion_length": 117.20000076293945,
      "epoch": 1.6677034733022291,
      "grad_norm": 0.4036959111690521,
      "kl": 0.218505859375,
      "learning_rate": 1.467403410978978e-06,
      "loss": 0.0104,
      "reward": 1.425000011920929,
      "reward_std": 0.26129765436053276,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 1.0,
      "step": 1609
    },
    {
      "completion_length": 216.85000228881836,
      "epoch": 1.6687402799377917,
      "grad_norm": 0.5966922640800476,
      "kl": 0.321533203125,
      "learning_rate": 1.465592798681991e-06,
      "loss": 0.0942,
      "reward": 1.5250000357627869,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1610
    },
    {
      "completion_length": 110.35000610351562,
      "epoch": 1.669777086573354,
      "grad_norm": 0.4677380323410034,
      "kl": 0.333984375,
      "learning_rate": 1.4637822365423397e-06,
      "loss": 0.0127,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1611
    },
    {
      "completion_length": 100.60000038146973,
      "epoch": 1.6708138932089165,
      "grad_norm": 0.29280760884284973,
      "kl": 0.275146484375,
      "learning_rate": 1.4619717271993828e-06,
      "loss": 0.0177,
      "reward": 1.7750000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 1.0,
      "step": 1612
    },
    {
      "completion_length": 207.77500534057617,
      "epoch": 1.6718506998444789,
      "grad_norm": 1.227375864982605,
      "kl": 2.556640625,
      "learning_rate": 1.4601612732924036e-06,
      "loss": 0.2885,
      "reward": 1.550000011920929,
      "reward_std": 0.4363621845841408,
      "rewards/accuracy_reward": 0.6250000223517418,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1613
    },
    {
      "completion_length": 159.8000030517578,
      "epoch": 1.6728875064800415,
      "grad_norm": 0.9971756339073181,
      "kl": 1.3330078125,
      "learning_rate": 1.4583508774606033e-06,
      "loss": 0.0911,
      "reward": 1.6500000059604645,
      "reward_std": 0.24873404577374458,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1614
    },
    {
      "completion_length": 268.62500953674316,
      "epoch": 1.673924313115604,
      "grad_norm": 2.016587495803833,
      "kl": 1.754150390625,
      "learning_rate": 1.4565405423430994e-06,
      "loss": 0.1122,
      "reward": 1.6500000357627869,
      "reward_std": 0.3052687272429466,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1615
    },
    {
      "completion_length": 108.97500228881836,
      "epoch": 1.6749611197511665,
      "grad_norm": 0.03610905632376671,
      "kl": 0.216796875,
      "learning_rate": 1.4547302705789198e-06,
      "loss": 0.0083,
      "reward": 1.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 1.0,
      "step": 1616
    },
    {
      "completion_length": 127.0,
      "epoch": 1.6759979263867288,
      "grad_norm": 0.3895142078399658,
      "kl": 0.29296875,
      "learning_rate": 1.4529200648070016e-06,
      "loss": 0.0191,
      "reward": 1.800000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 1.0,
      "step": 1617
    },
    {
      "completion_length": 167.25,
      "epoch": 1.6770347330222912,
      "grad_norm": 0.15201283991336823,
      "kl": 0.28759765625,
      "learning_rate": 1.4511099276661832e-06,
      "loss": 0.0544,
      "reward": 1.3499999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1618
    },
    {
      "completion_length": 207.55000686645508,
      "epoch": 1.6780715396578538,
      "grad_norm": 0.21813489496707916,
      "kl": 0.4111328125,
      "learning_rate": 1.4492998617952062e-06,
      "loss": 0.1047,
      "reward": 1.5250000357627869,
      "reward_std": 0.22360678017139435,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1619
    },
    {
      "completion_length": 171.8000030517578,
      "epoch": 1.6791083462934164,
      "grad_norm": 0.6411747336387634,
      "kl": 1.269287109375,
      "learning_rate": 1.447489869832706e-06,
      "loss": 0.0934,
      "reward": 1.3250000178813934,
      "reward_std": 0.31719937548041344,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1620
    },
    {
      "completion_length": 116.4749984741211,
      "epoch": 1.6801451529289788,
      "grad_norm": 1.831243634223938,
      "kl": 0.17822265625,
      "learning_rate": 1.4456799544172104e-06,
      "loss": -0.0055,
      "reward": 1.8499999940395355,
      "reward_std": 0.23617041110992432,
      "rewards/accuracy_reward": 0.8500000238418579,
      "rewards/format_reward": 1.0,
      "step": 1621
    },
    {
      "completion_length": 184.05000686645508,
      "epoch": 1.6811819595645412,
      "grad_norm": 0.413693904876709,
      "kl": 0.224853515625,
      "learning_rate": 1.4438701181871363e-06,
      "loss": 0.0633,
      "reward": 1.5750000178813934,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.6000000201165676,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1622
    },
    {
      "completion_length": 113.0250015258789,
      "epoch": 1.6822187662001036,
      "grad_norm": 0.29346320033073425,
      "kl": 0.248046875,
      "learning_rate": 1.4420603637807857e-06,
      "loss": -0.0001,
      "reward": 1.5250000357627869,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 1.0,
      "step": 1623
    },
    {
      "completion_length": 262.75000381469727,
      "epoch": 1.6832555728356662,
      "grad_norm": 1.1153827905654907,
      "kl": 1.1875,
      "learning_rate": 1.4402506938363398e-06,
      "loss": 0.1261,
      "reward": 1.5,
      "reward_std": 0.30463573709130287,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.925000011920929,
      "step": 1624
    },
    {
      "completion_length": 229.87500762939453,
      "epoch": 1.6842923794712286,
      "grad_norm": 1.9630428552627563,
      "kl": 0.88818359375,
      "learning_rate": 1.4384411109918577e-06,
      "loss": 0.1332,
      "reward": 1.7000000178813934,
      "reward_std": 0.3479737974703312,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1625
    },
    {
      "completion_length": 214.90000534057617,
      "epoch": 1.6853291861067912,
      "grad_norm": 0.46512189507484436,
      "kl": 0.61181640625,
      "learning_rate": 1.4366316178852714e-06,
      "loss": 0.1527,
      "reward": 1.6250000298023224,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1626
    },
    {
      "completion_length": 116.64999961853027,
      "epoch": 1.6863659927423535,
      "grad_norm": 2.697237253189087,
      "kl": 0.9498291015625,
      "learning_rate": 1.4348222171543822e-06,
      "loss": 0.095,
      "reward": 1.550000011920929,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1627
    },
    {
      "completion_length": 184.7500057220459,
      "epoch": 1.687402799377916,
      "grad_norm": 0.5999835133552551,
      "kl": 0.671630859375,
      "learning_rate": 1.4330129114368554e-06,
      "loss": 0.0842,
      "reward": 1.675000011920929,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1628
    },
    {
      "completion_length": 204.97500228881836,
      "epoch": 1.6884396060134785,
      "grad_norm": 2.4206836223602295,
      "kl": 1.625,
      "learning_rate": 1.4312037033702205e-06,
      "loss": 0.1712,
      "reward": 1.75,
      "reward_std": 0.32818931341171265,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1629
    },
    {
      "completion_length": 107.70000267028809,
      "epoch": 1.689476412649041,
      "grad_norm": 0.4894731044769287,
      "kl": 0.219970703125,
      "learning_rate": 1.4293945955918617e-06,
      "loss": -0.0092,
      "reward": 1.4500000178813934,
      "reward_std": 0.06846531480550766,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 1.0,
      "step": 1630
    },
    {
      "completion_length": 176.52500343322754,
      "epoch": 1.6905132192846035,
      "grad_norm": 0.2627868354320526,
      "kl": 1.17529296875,
      "learning_rate": 1.4275855907390186e-06,
      "loss": 0.1381,
      "reward": 1.600000023841858,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1631
    },
    {
      "completion_length": 219.15000343322754,
      "epoch": 1.691550025920166,
      "grad_norm": 4.7461466789245605,
      "kl": 3.079345703125,
      "learning_rate": 1.425776691448782e-06,
      "loss": 0.3135,
      "reward": 1.5999999940395355,
      "reward_std": 0.2795085050165653,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1632
    },
    {
      "completion_length": 201.17500495910645,
      "epoch": 1.6925868325557283,
      "grad_norm": 0.37242236733436584,
      "kl": 3.163330078125,
      "learning_rate": 1.4239679003580854e-06,
      "loss": 0.3266,
      "reward": 1.3000000268220901,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1633
    },
    {
      "completion_length": 115.22500228881836,
      "epoch": 1.6936236391912907,
      "grad_norm": 4.702274799346924,
      "kl": 0.3017578125,
      "learning_rate": 1.4221592201037075e-06,
      "loss": 0.0156,
      "reward": 1.550000011920929,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.5749999973922968,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1634
    },
    {
      "completion_length": 270.3250045776367,
      "epoch": 1.6946604458268533,
      "grad_norm": 8.115036964416504,
      "kl": 5.443359375,
      "learning_rate": 1.420350653322264e-06,
      "loss": 0.4798,
      "reward": 1.475000023841858,
      "reward_std": 0.32818933203816414,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1635
    },
    {
      "completion_length": 218.47500610351562,
      "epoch": 1.6956972524624159,
      "grad_norm": 5.428323268890381,
      "kl": 3.03564453125,
      "learning_rate": 1.4185422026502062e-06,
      "loss": 0.3195,
      "reward": 1.1250000149011612,
      "reward_std": 0.2812207005918026,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1636
    },
    {
      "completion_length": 201.3250026702881,
      "epoch": 1.6967340590979783,
      "grad_norm": 0.41541728377342224,
      "kl": 2.009033203125,
      "learning_rate": 1.416733870723815e-06,
      "loss": 0.2323,
      "reward": 1.6500000059604645,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.925000011920929,
      "step": 1637
    },
    {
      "completion_length": 260.975004196167,
      "epoch": 1.6977708657335406,
      "grad_norm": 0.4679148197174072,
      "kl": 2.0595703125,
      "learning_rate": 1.4149256601791993e-06,
      "loss": 0.3123,
      "reward": 1.4500000178813934,
      "reward_std": 0.3930240832269192,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1638
    },
    {
      "completion_length": 106.625,
      "epoch": 1.698807672369103,
      "grad_norm": 3.114511489868164,
      "kl": 0.36279296875,
      "learning_rate": 1.4131175736522897e-06,
      "loss": 0.0242,
      "reward": 1.9749999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.9750000089406967,
      "rewards/format_reward": 1.0,
      "step": 1639
    },
    {
      "completion_length": 157.3499984741211,
      "epoch": 1.6998444790046656,
      "grad_norm": 0.44841909408569336,
      "kl": 1.156982421875,
      "learning_rate": 1.411309613778838e-06,
      "loss": 0.1365,
      "reward": 1.699999988079071,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1640
    },
    {
      "completion_length": 271.9000072479248,
      "epoch": 1.7008812856402282,
      "grad_norm": 0.4403691291809082,
      "kl": 0.539306640625,
      "learning_rate": 1.4095017831944087e-06,
      "loss": 0.1344,
      "reward": 1.3500000089406967,
      "reward_std": 0.2795085050165653,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1641
    },
    {
      "completion_length": 118.9000015258789,
      "epoch": 1.7019180922757906,
      "grad_norm": 0.30528485774993896,
      "kl": 0.2333984375,
      "learning_rate": 1.4076940845343808e-06,
      "loss": 0.0026,
      "reward": 1.6500000059604645,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.6500000152736902,
      "rewards/format_reward": 1.0,
      "step": 1642
    },
    {
      "completion_length": 167.05000114440918,
      "epoch": 1.702954898911353,
      "grad_norm": 1.617598295211792,
      "kl": 0.653076171875,
      "learning_rate": 1.4058865204339383e-06,
      "loss": 0.0373,
      "reward": 1.7000000178813934,
      "reward_std": 0.15685367584228516,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1643
    },
    {
      "completion_length": 242.39999771118164,
      "epoch": 1.7039917055469154,
      "grad_norm": 0.044586073607206345,
      "kl": 2.680908203125,
      "learning_rate": 1.4040790935280714e-06,
      "loss": 0.3243,
      "reward": 1.600000023841858,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1644
    },
    {
      "completion_length": 118.62500190734863,
      "epoch": 1.705028512182478,
      "grad_norm": 1.295025110244751,
      "kl": 0.28564453125,
      "learning_rate": 1.4022718064515678e-06,
      "loss": 0.0111,
      "reward": 1.625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 1.0,
      "step": 1645
    },
    {
      "completion_length": 202.25000190734863,
      "epoch": 1.7060653188180406,
      "grad_norm": 1.3782089948654175,
      "kl": 0.6884765625,
      "learning_rate": 1.4004646618390139e-06,
      "loss": 0.1291,
      "reward": 1.5,
      "reward_std": 0.27228761836886406,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1646
    },
    {
      "completion_length": 260.6250114440918,
      "epoch": 1.707102125453603,
      "grad_norm": 2.5833706855773926,
      "kl": 2.291015625,
      "learning_rate": 1.3986576623247865e-06,
      "loss": 0.2426,
      "reward": 1.4749999940395355,
      "reward_std": 0.38109345734119415,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1647
    },
    {
      "completion_length": 113.37500190734863,
      "epoch": 1.7081389320891653,
      "grad_norm": 0.9900299906730652,
      "kl": 0.28271484375,
      "learning_rate": 1.3968508105430525e-06,
      "loss": 0.0018,
      "reward": 1.625,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 1.0,
      "step": 1648
    },
    {
      "completion_length": 206.67500114440918,
      "epoch": 1.7091757387247277,
      "grad_norm": 0.37716108560562134,
      "kl": 1.8212890625,
      "learning_rate": 1.3950441091277612e-06,
      "loss": 0.2549,
      "reward": 1.4000000059604645,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.4500000011175871,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1649
    },
    {
      "completion_length": 188.55000114440918,
      "epoch": 1.7102125453602903,
      "grad_norm": 1.695388913154602,
      "kl": 1.5244140625,
      "learning_rate": 1.393237560712646e-06,
      "loss": 0.1435,
      "reward": 1.475000023841858,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1650
    },
    {
      "completion_length": 160.92500495910645,
      "epoch": 1.711249351995853,
      "grad_norm": 1.9501590728759766,
      "kl": 0.2236328125,
      "learning_rate": 1.391431167931213e-06,
      "loss": 0.0411,
      "reward": 1.75,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1651
    },
    {
      "completion_length": 202.375,
      "epoch": 1.7122861586314153,
      "grad_norm": 0.16232004761695862,
      "kl": 1.083251953125,
      "learning_rate": 1.389624933416746e-06,
      "loss": 0.1748,
      "reward": 1.6250000298023224,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1652
    },
    {
      "completion_length": 176.90000343322754,
      "epoch": 1.7133229652669777,
      "grad_norm": 0.8067182898521423,
      "kl": 1.25048828125,
      "learning_rate": 1.3878188598022941e-06,
      "loss": 0.1323,
      "reward": 1.449999988079071,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1653
    },
    {
      "completion_length": 179.80000114440918,
      "epoch": 1.71435977190254,
      "grad_norm": 0.28877392411231995,
      "kl": 0.2696533203125,
      "learning_rate": 1.3860129497206747e-06,
      "loss": 0.0804,
      "reward": 1.5999999940395355,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1654
    },
    {
      "completion_length": 260.5000057220459,
      "epoch": 1.7153965785381027,
      "grad_norm": 0.5322703123092651,
      "kl": 2.345947265625,
      "learning_rate": 1.3842072058044649e-06,
      "loss": 0.2565,
      "reward": 1.5250000357627869,
      "reward_std": 0.40387552976608276,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1655
    },
    {
      "completion_length": 130.4000015258789,
      "epoch": 1.716433385173665,
      "grad_norm": 0.6268528699874878,
      "kl": 0.223876953125,
      "learning_rate": 1.382401630686001e-06,
      "loss": 0.0517,
      "reward": 1.7750000059604645,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 1.0,
      "step": 1656
    },
    {
      "completion_length": 116.57499885559082,
      "epoch": 1.7174701918092277,
      "grad_norm": 0.3194897174835205,
      "kl": 0.26318359375,
      "learning_rate": 1.3805962269973722e-06,
      "loss": -0.0135,
      "reward": 1.600000023841858,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 1.0,
      "step": 1657
    },
    {
      "completion_length": 319.850004196167,
      "epoch": 1.71850699844479,
      "grad_norm": 1.0248870849609375,
      "kl": 4.117431640625,
      "learning_rate": 1.3787909973704186e-06,
      "loss": 0.4725,
      "reward": 1.5000000447034836,
      "reward_std": 0.40387551113963127,
      "rewards/accuracy_reward": 0.6000000182539225,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1658
    },
    {
      "completion_length": 108.22500228881836,
      "epoch": 1.7195438050803524,
      "grad_norm": 3.075923204421997,
      "kl": 0.233642578125,
      "learning_rate": 1.3769859444367255e-06,
      "loss": 0.0176,
      "reward": 1.675000011920929,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 1.0,
      "step": 1659
    },
    {
      "completion_length": 210.6250057220459,
      "epoch": 1.720580611715915,
      "grad_norm": 0.2640899419784546,
      "kl": 1.01953125,
      "learning_rate": 1.3751810708276225e-06,
      "loss": 0.1408,
      "reward": 1.475000023841858,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1660
    },
    {
      "completion_length": 117.40000343322754,
      "epoch": 1.7216174183514774,
      "grad_norm": 1.2959752082824707,
      "kl": 0.2353515625,
      "learning_rate": 1.3733763791741748e-06,
      "loss": 0.0243,
      "reward": 1.550000011920929,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 1.0,
      "step": 1661
    },
    {
      "completion_length": 197.52500343322754,
      "epoch": 1.72265422498704,
      "grad_norm": 0.5643317699432373,
      "kl": 2.107177734375,
      "learning_rate": 1.3715718721071858e-06,
      "loss": 0.2302,
      "reward": 1.725000023841858,
      "reward_std": 0.40387552976608276,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1662
    },
    {
      "completion_length": 437.15000915527344,
      "epoch": 1.7236910316226024,
      "grad_norm": 1.0964477062225342,
      "kl": 4.185546875,
      "learning_rate": 1.3697675522571872e-06,
      "loss": 0.5379,
      "reward": 1.3000000417232513,
      "reward_std": 0.6094099879264832,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1663
    },
    {
      "completion_length": 219.9500026702881,
      "epoch": 1.7247278382581648,
      "grad_norm": 1.2283393144607544,
      "kl": 1.347900390625,
      "learning_rate": 1.367963422254439e-06,
      "loss": 0.1927,
      "reward": 1.425000011920929,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1664
    },
    {
      "completion_length": 113.72500419616699,
      "epoch": 1.7257646448937272,
      "grad_norm": 0.6352521181106567,
      "kl": 0.247802734375,
      "learning_rate": 1.3661594847289239e-06,
      "loss": 0.0097,
      "reward": 1.625,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 1.0,
      "step": 1665
    },
    {
      "completion_length": 161.4500026702881,
      "epoch": 1.7268014515292898,
      "grad_norm": 0.9560681581497192,
      "kl": 0.270263671875,
      "learning_rate": 1.3643557423103448e-06,
      "loss": 0.0937,
      "reward": 1.800000011920929,
      "reward_std": 0.22360679879784584,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1666
    },
    {
      "completion_length": 165.62500190734863,
      "epoch": 1.7278382581648524,
      "grad_norm": 0.7034723162651062,
      "kl": 1.19287109375,
      "learning_rate": 1.3625521976281192e-06,
      "loss": 0.1538,
      "reward": 1.5750000327825546,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1667
    },
    {
      "completion_length": 258.9499988555908,
      "epoch": 1.7288750648004148,
      "grad_norm": 0.8078364133834839,
      "kl": 1.2313232421875,
      "learning_rate": 1.3607488533113774e-06,
      "loss": 0.2207,
      "reward": 1.2750000208616257,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1668
    },
    {
      "completion_length": 167.100004196167,
      "epoch": 1.7299118714359771,
      "grad_norm": 0.26684144139289856,
      "kl": 0.213623046875,
      "learning_rate": 1.358945711988956e-06,
      "loss": 0.0518,
      "reward": 1.699999988079071,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1669
    },
    {
      "completion_length": 162.95000076293945,
      "epoch": 1.7309486780715395,
      "grad_norm": 0.2954098880290985,
      "kl": 1.203125,
      "learning_rate": 1.3571427762893975e-06,
      "loss": 0.107,
      "reward": 1.5750000178813934,
      "reward_std": 0.18026871606707573,
      "rewards/accuracy_reward": 0.6000000033527613,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1670
    },
    {
      "completion_length": 266.37500953674316,
      "epoch": 1.7319854847071021,
      "grad_norm": 0.40566912293434143,
      "kl": 1.091796875,
      "learning_rate": 1.3553400488409442e-06,
      "loss": 0.2136,
      "reward": 1.574999988079071,
      "reward_std": 0.27228759974241257,
      "rewards/accuracy_reward": 0.6500000096857548,
      "rewards/format_reward": 0.925000011920929,
      "step": 1671
    },
    {
      "completion_length": 275.875,
      "epoch": 1.7330222913426647,
      "grad_norm": 4.127599716186523,
      "kl": 3.32080078125,
      "learning_rate": 1.3535375322715338e-06,
      "loss": 0.3764,
      "reward": 1.449999988079071,
      "reward_std": 0.5048274919390678,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1672
    },
    {
      "completion_length": 145.82500648498535,
      "epoch": 1.734059097978227,
      "grad_norm": 1.3361109495162964,
      "kl": 1.259033203125,
      "learning_rate": 1.351735229208798e-06,
      "loss": 0.1036,
      "reward": 1.425000011920929,
      "reward_std": 0.20019175112247467,
      "rewards/accuracy_reward": 0.4500000011175871,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1673
    },
    {
      "completion_length": 206.55000495910645,
      "epoch": 1.7350959046137895,
      "grad_norm": 1.2055398225784302,
      "kl": 1.6787109375,
      "learning_rate": 1.349933142280055e-06,
      "loss": 0.1976,
      "reward": 1.4750000536441803,
      "reward_std": 0.4723408371210098,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1674
    },
    {
      "completion_length": 395.5500068664551,
      "epoch": 1.7361327112493519,
      "grad_norm": 0.9115234613418579,
      "kl": 3.5263671875,
      "learning_rate": 1.3481312741123117e-06,
      "loss": 0.2621,
      "reward": 1.375,
      "reward_std": 0.4981010966002941,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1675
    },
    {
      "completion_length": 305.97501373291016,
      "epoch": 1.7371695178849145,
      "grad_norm": 2.5328431129455566,
      "kl": 2.07373046875,
      "learning_rate": 1.346329627332252e-06,
      "loss": 0.3261,
      "reward": 1.4500000178813934,
      "reward_std": 0.3731010779738426,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1676
    },
    {
      "completion_length": 309.8000087738037,
      "epoch": 1.738206324520477,
      "grad_norm": 0.9459800720214844,
      "kl": 1.197998046875,
      "learning_rate": 1.34452820456624e-06,
      "loss": 0.2595,
      "reward": 1.1500000208616257,
      "reward_std": 0.41643916070461273,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1677
    },
    {
      "completion_length": 153.8750057220459,
      "epoch": 1.7392431311560395,
      "grad_norm": 0.6364434957504272,
      "kl": 0.249267578125,
      "learning_rate": 1.3427270084403114e-06,
      "loss": 0.0669,
      "reward": 1.675000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1678
    },
    {
      "completion_length": 255.24999809265137,
      "epoch": 1.7402799377916018,
      "grad_norm": 0.2327195107936859,
      "kl": 2.372802734375,
      "learning_rate": 1.3409260415801732e-06,
      "loss": 0.3188,
      "reward": 1.4250000417232513,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.925000011920929,
      "step": 1679
    },
    {
      "completion_length": 314.7250061035156,
      "epoch": 1.7413167444271642,
      "grad_norm": 4.738442897796631,
      "kl": 1.777099609375,
      "learning_rate": 1.3391253066111962e-06,
      "loss": 0.2694,
      "reward": 1.375,
      "reward_std": 0.3606759384274483,
      "rewards/accuracy_reward": 0.47500000707805157,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1680
    },
    {
      "completion_length": 243.1250057220459,
      "epoch": 1.7423535510627268,
      "grad_norm": 0.2917191982269287,
      "kl": 2.29541015625,
      "learning_rate": 1.3373248061584154e-06,
      "loss": 0.3098,
      "reward": 1.7750000059604645,
      "reward_std": 0.25609345361590385,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1681
    },
    {
      "completion_length": 298.89999771118164,
      "epoch": 1.7433903576982894,
      "grad_norm": 1.8764444589614868,
      "kl": 2.10302734375,
      "learning_rate": 1.3355245428465218e-06,
      "loss": 0.2798,
      "reward": 1.675000011920929,
      "reward_std": 0.4415663853287697,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1682
    },
    {
      "completion_length": 393.37499809265137,
      "epoch": 1.7444271643338518,
      "grad_norm": 8.258137702941895,
      "kl": 3.956787109375,
      "learning_rate": 1.3337245192998627e-06,
      "loss": 0.5742,
      "reward": 1.3000000268220901,
      "reward_std": 0.45318930596113205,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1683
    },
    {
      "completion_length": 227.02500915527344,
      "epoch": 1.7454639709694142,
      "grad_norm": 1.1636691093444824,
      "kl": 1.012939453125,
      "learning_rate": 1.3319247381424333e-06,
      "loss": 0.1633,
      "reward": 1.5000000298023224,
      "reward_std": 0.3479738235473633,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1684
    },
    {
      "completion_length": 544.9750061035156,
      "epoch": 1.7465007776049766,
      "grad_norm": 2.0953562259674072,
      "kl": 5.66015625,
      "learning_rate": 1.330125201997879e-06,
      "loss": 0.5279,
      "reward": 1.1250000298023224,
      "reward_std": 0.5607291907072067,
      "rewards/accuracy_reward": 0.3500000052154064,
      "rewards/format_reward": 0.7750000059604645,
      "step": 1685
    },
    {
      "completion_length": 348.375,
      "epoch": 1.7475375842405392,
      "grad_norm": 5.163078784942627,
      "kl": 3.19091796875,
      "learning_rate": 1.3283259134894844e-06,
      "loss": 0.4759,
      "reward": 1.4500000476837158,
      "reward_std": 0.4604101926088333,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1686
    },
    {
      "completion_length": 400.35001373291016,
      "epoch": 1.7485743908761016,
      "grad_norm": 3.5182042121887207,
      "kl": 4.67578125,
      "learning_rate": 1.3265268752401755e-06,
      "loss": 0.3921,
      "reward": 1.1750000268220901,
      "reward_std": 0.47234081104397774,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.800000011920929,
      "step": 1687
    },
    {
      "completion_length": 489.5500183105469,
      "epoch": 1.7496111975116642,
      "grad_norm": 0.4392010271549225,
      "kl": 2.59814453125,
      "learning_rate": 1.3247280898725125e-06,
      "loss": 0.3865,
      "reward": 1.3250000476837158,
      "reward_std": 0.5649927109479904,
      "rewards/accuracy_reward": 0.5500000156462193,
      "rewards/format_reward": 0.7750000059604645,
      "step": 1688
    },
    {
      "completion_length": 481.375,
      "epoch": 1.7506480041472265,
      "grad_norm": 2.8733747005462646,
      "kl": 4.671875,
      "learning_rate": 1.3229295600086873e-06,
      "loss": 0.5259,
      "reward": 1.2250000089406967,
      "reward_std": 0.6130405366420746,
      "rewards/accuracy_reward": 0.4250000100582838,
      "rewards/format_reward": 0.800000011920929,
      "step": 1689
    },
    {
      "completion_length": 292.2749996185303,
      "epoch": 1.751684810782789,
      "grad_norm": 1.616410732269287,
      "kl": 1.1044921875,
      "learning_rate": 1.3211312882705192e-06,
      "loss": 0.2193,
      "reward": 1.3250000327825546,
      "reward_std": 0.3046357296407223,
      "rewards/accuracy_reward": 0.4250000026077032,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1690
    },
    {
      "completion_length": 357.9250144958496,
      "epoch": 1.7527216174183515,
      "grad_norm": 1.5785585641860962,
      "kl": 4.5654296875,
      "learning_rate": 1.3193332772794515e-06,
      "loss": 0.2922,
      "reward": 1.2500000149011612,
      "reward_std": 0.4415663853287697,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1691
    },
    {
      "completion_length": 393.9750061035156,
      "epoch": 1.753758424053914,
      "grad_norm": 2.245445489883423,
      "kl": 4.9765625,
      "learning_rate": 1.3175355296565466e-06,
      "loss": 0.6097,
      "reward": 1.2000000327825546,
      "reward_std": 0.41643913462758064,
      "rewards/accuracy_reward": 0.37500000186264515,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1692
    },
    {
      "completion_length": 450.95001220703125,
      "epoch": 1.7547952306894765,
      "grad_norm": 1.7147431373596191,
      "kl": 1.007568359375,
      "learning_rate": 1.3157380480224844e-06,
      "loss": 0.3057,
      "reward": 1.350000023841858,
      "reward_std": 0.6166309043765068,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1693
    },
    {
      "completion_length": 200.52500534057617,
      "epoch": 1.755832037325039,
      "grad_norm": 0.47567418217658997,
      "kl": 0.35205078125,
      "learning_rate": 1.3139408349975549e-06,
      "loss": 0.1028,
      "reward": 1.4250000417232513,
      "reward_std": 0.3046357184648514,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1694
    },
    {
      "completion_length": 199.29999923706055,
      "epoch": 1.7568688439606013,
      "grad_norm": 2.4895272254943848,
      "kl": 1.810546875,
      "learning_rate": 1.3121438932016595e-06,
      "loss": 0.2271,
      "reward": 1.6250000298023224,
      "reward_std": 0.27950847521424294,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1695
    },
    {
      "completion_length": 93.2750015258789,
      "epoch": 1.7579056505961637,
      "grad_norm": 0.4038166105747223,
      "kl": 0.654296875,
      "learning_rate": 1.3103472252543003e-06,
      "loss": 0.0834,
      "reward": 1.5250000208616257,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1696
    },
    {
      "completion_length": 195.37500953674316,
      "epoch": 1.7589424572317263,
      "grad_norm": 2.357900381088257,
      "kl": 2.333984375,
      "learning_rate": 1.3085508337745844e-06,
      "loss": 0.2501,
      "reward": 1.725000023841858,
      "reward_std": 0.3354101926088333,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1697
    },
    {
      "completion_length": 299.55000495910645,
      "epoch": 1.7599792638672889,
      "grad_norm": 4.066470146179199,
      "kl": 3.29541015625,
      "learning_rate": 1.306754721381213e-06,
      "loss": 0.3886,
      "reward": 1.5,
      "reward_std": 0.3479738272726536,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1698
    },
    {
      "completion_length": 195.05000114440918,
      "epoch": 1.7610160705028512,
      "grad_norm": 2.4610085487365723,
      "kl": 2.865234375,
      "learning_rate": 1.3049588906924814e-06,
      "loss": 0.2184,
      "reward": 1.4000000208616257,
      "reward_std": 0.45977721735835075,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1699
    },
    {
      "completion_length": 143.95000457763672,
      "epoch": 1.7620528771384136,
      "grad_norm": 1.7119483947753906,
      "kl": 1.673828125,
      "learning_rate": 1.3031633443262742e-06,
      "loss": 0.1461,
      "reward": 1.425000011920929,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1700
    },
    {
      "completion_length": 207.39999771118164,
      "epoch": 1.763089683773976,
      "grad_norm": 1.165863037109375,
      "kl": 1.869873046875,
      "learning_rate": 1.3013680849000619e-06,
      "loss": 0.2443,
      "reward": 1.625,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1701
    },
    {
      "completion_length": 139.10000228881836,
      "epoch": 1.7641264904095386,
      "grad_norm": 1.0776835680007935,
      "kl": 1.6533203125,
      "learning_rate": 1.2995731150308957e-06,
      "loss": 0.1518,
      "reward": 1.875,
      "reward_std": 0.22360678017139435,
      "rewards/accuracy_reward": 0.9249999970197678,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1702
    },
    {
      "completion_length": 206.87500190734863,
      "epoch": 1.7651632970451012,
      "grad_norm": 2.873706340789795,
      "kl": 2.02880859375,
      "learning_rate": 1.2977784373354061e-06,
      "loss": 0.2131,
      "reward": 1.375,
      "reward_std": 0.3479738235473633,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1703
    },
    {
      "completion_length": 253.75000381469727,
      "epoch": 1.7662001036806636,
      "grad_norm": 0.7443107962608337,
      "kl": 1.505859375,
      "learning_rate": 1.2959840544297957e-06,
      "loss": 0.2283,
      "reward": 1.4250000566244125,
      "reward_std": 0.4038754962384701,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1704
    },
    {
      "completion_length": 237.8499984741211,
      "epoch": 1.767236910316226,
      "grad_norm": 2.906121253967285,
      "kl": 3.813720703125,
      "learning_rate": 1.2941899689298402e-06,
      "loss": 0.4106,
      "reward": 1.7250000536441803,
      "reward_std": 0.3913118876516819,
      "rewards/accuracy_reward": 0.8000000268220901,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1705
    },
    {
      "completion_length": 139.92500114440918,
      "epoch": 1.7682737169517884,
      "grad_norm": 2.2459774017333984,
      "kl": 1.25634765625,
      "learning_rate": 1.2923961834508785e-06,
      "loss": 0.0538,
      "reward": 1.350000023841858,
      "reward_std": 0.27950847521424294,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1706
    },
    {
      "completion_length": 160.55000686645508,
      "epoch": 1.769310523587351,
      "grad_norm": 0.6946339011192322,
      "kl": 1.557861328125,
      "learning_rate": 1.2906027006078152e-06,
      "loss": 0.1377,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.8250000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1707
    },
    {
      "completion_length": 193.47500610351562,
      "epoch": 1.7703473302229136,
      "grad_norm": 2.409801483154297,
      "kl": 2.78955078125,
      "learning_rate": 1.288809523015111e-06,
      "loss": 0.3181,
      "reward": 1.3250000178813934,
      "reward_std": 0.48867350444197655,
      "rewards/accuracy_reward": 0.4249999951571226,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1708
    },
    {
      "completion_length": 150.92500114440918,
      "epoch": 1.771384136858476,
      "grad_norm": 0.6411232948303223,
      "kl": 0.862548828125,
      "learning_rate": 1.287016653286784e-06,
      "loss": 0.1047,
      "reward": 1.425000011920929,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1709
    },
    {
      "completion_length": 294.50000190734863,
      "epoch": 1.7724209434940383,
      "grad_norm": 3.621253490447998,
      "kl": 3.5322265625,
      "learning_rate": 1.2852240940364018e-06,
      "loss": 0.4518,
      "reward": 1.4000000059604645,
      "reward_std": 0.38943372294306755,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1710
    },
    {
      "completion_length": 248.05000686645508,
      "epoch": 1.7734577501296007,
      "grad_norm": 1.0850621461868286,
      "kl": 1.03369140625,
      "learning_rate": 1.28343184787708e-06,
      "loss": 0.204,
      "reward": 1.6250000298023224,
      "reward_std": 0.4525563307106495,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1711
    },
    {
      "completion_length": 188.02500343322754,
      "epoch": 1.7744945567651633,
      "grad_norm": 2.1935651302337646,
      "kl": 1.64306640625,
      "learning_rate": 1.2816399174214785e-06,
      "loss": 0.1811,
      "reward": 1.725000023841858,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1712
    },
    {
      "completion_length": 195.0000057220459,
      "epoch": 1.775531363400726,
      "grad_norm": 1.4455136060714722,
      "kl": 1.5380859375,
      "learning_rate": 1.2798483052817956e-06,
      "loss": 0.183,
      "reward": 1.5250000059604645,
      "reward_std": 0.3354101963341236,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1713
    },
    {
      "completion_length": 151.30000686645508,
      "epoch": 1.7765681700362883,
      "grad_norm": 1.1415375471115112,
      "kl": 0.890869140625,
      "learning_rate": 1.2780570140697667e-06,
      "loss": 0.0601,
      "reward": 1.675000011920929,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1714
    },
    {
      "completion_length": 289.45000076293945,
      "epoch": 1.7776049766718507,
      "grad_norm": 1.0119621753692627,
      "kl": 1.9130859375,
      "learning_rate": 1.2762660463966584e-06,
      "loss": 0.1743,
      "reward": 1.5250000059604645,
      "reward_std": 0.373101070523262,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1715
    },
    {
      "completion_length": 152.47500228881836,
      "epoch": 1.778641783307413,
      "grad_norm": 0.6372405886650085,
      "kl": 2.10205078125,
      "learning_rate": 1.2744754048732672e-06,
      "loss": 0.2332,
      "reward": 1.8000000417232513,
      "reward_std": 0.3479737974703312,
      "rewards/accuracy_reward": 0.8500000238418579,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1716
    },
    {
      "completion_length": 294.5750045776367,
      "epoch": 1.7796785899429757,
      "grad_norm": 0.4199138283729553,
      "kl": 3.05908203125,
      "learning_rate": 1.2726850921099117e-06,
      "loss": 0.4174,
      "reward": 1.4250000566244125,
      "reward_std": 0.3913118727505207,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.9000000357627869,
      "step": 1717
    },
    {
      "completion_length": 103.77499961853027,
      "epoch": 1.780715396578538,
      "grad_norm": 0.9671004414558411,
      "kl": 0.2900390625,
      "learning_rate": 1.2708951107164345e-06,
      "loss": -0.0021,
      "reward": 1.5000000298023224,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 1.0,
      "step": 1718
    },
    {
      "completion_length": 206.975004196167,
      "epoch": 1.7817522032141007,
      "grad_norm": 2.2271177768707275,
      "kl": 2.0693359375,
      "learning_rate": 1.2691054633021914e-06,
      "loss": 0.2187,
      "reward": 1.2750000208616257,
      "reward_std": 0.3371223844587803,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1719
    },
    {
      "completion_length": 155.97500610351562,
      "epoch": 1.782789009849663,
      "grad_norm": 0.91446453332901,
      "kl": 0.313720703125,
      "learning_rate": 1.2673161524760548e-06,
      "loss": 0.0516,
      "reward": 1.6500000357627869,
      "reward_std": 0.23617041110992432,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1720
    },
    {
      "completion_length": 160.30000495910645,
      "epoch": 1.7838258164852254,
      "grad_norm": 0.9384238719940186,
      "kl": 1.2392578125,
      "learning_rate": 1.2655271808464041e-06,
      "loss": 0.1296,
      "reward": 1.600000023841858,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1721
    },
    {
      "completion_length": 204.350004196167,
      "epoch": 1.784862623120788,
      "grad_norm": 0.4190017879009247,
      "kl": 2.197021484375,
      "learning_rate": 1.2637385510211255e-06,
      "loss": 0.2493,
      "reward": 1.300000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1722
    },
    {
      "completion_length": 209.25000190734863,
      "epoch": 1.7858994297563504,
      "grad_norm": 0.058677058666944504,
      "kl": 1.091796875,
      "learning_rate": 1.261950265607606e-06,
      "loss": 0.1672,
      "reward": 1.6250000298023224,
      "reward_std": 0.23617040365934372,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1723
    },
    {
      "completion_length": 241.8750057220459,
      "epoch": 1.786936236391913,
      "grad_norm": 2.870100259780884,
      "kl": 2.03955078125,
      "learning_rate": 1.2601623272127315e-06,
      "loss": 0.2512,
      "reward": 1.3000000268220901,
      "reward_std": 0.39365706965327263,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1724
    },
    {
      "completion_length": 148.17499923706055,
      "epoch": 1.7879730430274754,
      "grad_norm": 0.45264932513237,
      "kl": 0.904296875,
      "learning_rate": 1.2583747384428813e-06,
      "loss": 0.1311,
      "reward": 1.3750000298023224,
      "reward_std": 0.18026870489120483,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1725
    },
    {
      "completion_length": 153.14999771118164,
      "epoch": 1.7890098496630378,
      "grad_norm": 1.5778464078903198,
      "kl": 1.006591796875,
      "learning_rate": 1.2565875019039259e-06,
      "loss": 0.1198,
      "reward": 1.4250000417232513,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.44999999925494194,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1726
    },
    {
      "completion_length": 303.72500801086426,
      "epoch": 1.7900466562986002,
      "grad_norm": 1.394773006439209,
      "kl": 3.194580078125,
      "learning_rate": 1.2548006202012198e-06,
      "loss": 0.4276,
      "reward": 1.5500000417232513,
      "reward_std": 0.40387551486492157,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1727
    },
    {
      "completion_length": 395.25000762939453,
      "epoch": 1.7910834629341628,
      "grad_norm": 2.0842819213867188,
      "kl": 3.142578125,
      "learning_rate": 1.2530140959396049e-06,
      "loss": 0.4972,
      "reward": 1.350000023841858,
      "reward_std": 0.7085113078355789,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1728
    },
    {
      "completion_length": 93.37500190734863,
      "epoch": 1.7921202695697254,
      "grad_norm": 1.1371591091156006,
      "kl": 0.28076171875,
      "learning_rate": 1.2512279317233968e-06,
      "loss": 0.0116,
      "reward": 1.5000000298023224,
      "reward_std": 0.13693062961101532,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 1.0,
      "step": 1729
    },
    {
      "completion_length": 240.55000114440918,
      "epoch": 1.7931570762052877,
      "grad_norm": 0.3758859634399414,
      "kl": 3.274658203125,
      "learning_rate": 1.2494421301563904e-06,
      "loss": 0.3887,
      "reward": 1.5750000476837158,
      "reward_std": 0.37844377011060715,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1730
    },
    {
      "completion_length": 247.3250026702881,
      "epoch": 1.7941938828408501,
      "grad_norm": 13.461213111877441,
      "kl": 4.13427734375,
      "learning_rate": 1.2476566938418494e-06,
      "loss": 0.4357,
      "reward": 1.2500000447034836,
      "reward_std": 0.30463574454188347,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.925000011920929,
      "step": 1731
    },
    {
      "completion_length": 164.7750072479248,
      "epoch": 1.7952306894764125,
      "grad_norm": 0.4518619775772095,
      "kl": 1.48828125,
      "learning_rate": 1.2458716253825062e-06,
      "loss": 0.1191,
      "reward": 1.2000000327825546,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.22500000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1732
    },
    {
      "completion_length": 177.95000457763672,
      "epoch": 1.7962674961119751,
      "grad_norm": 1.2171616554260254,
      "kl": 0.61328125,
      "learning_rate": 1.2440869273805566e-06,
      "loss": 0.1541,
      "reward": 1.5750000178813934,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1733
    },
    {
      "completion_length": 156.32500076293945,
      "epoch": 1.7973043027475377,
      "grad_norm": 2.8522934913635254,
      "kl": 1.1611328125,
      "learning_rate": 1.242302602437657e-06,
      "loss": 0.114,
      "reward": 1.725000023841858,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1734
    },
    {
      "completion_length": 140.27499961853027,
      "epoch": 1.7983411093831,
      "grad_norm": 0.3764083683490753,
      "kl": 1.037841796875,
      "learning_rate": 1.2405186531549181e-06,
      "loss": 0.0853,
      "reward": 1.699999988079071,
      "reward_std": 0.20019175112247467,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1735
    },
    {
      "completion_length": 194.6750087738037,
      "epoch": 1.7993779160186625,
      "grad_norm": 1.3961740732192993,
      "kl": 0.92138671875,
      "learning_rate": 1.2387350821329061e-06,
      "loss": 0.1744,
      "reward": 1.725000023841858,
      "reward_std": 0.2920720987021923,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1736
    },
    {
      "completion_length": 257.575008392334,
      "epoch": 1.8004147226542249,
      "grad_norm": 1.2615939378738403,
      "kl": 1.832275390625,
      "learning_rate": 1.2369518919716318e-06,
      "loss": 0.2549,
      "reward": 1.5000000447034836,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1737
    },
    {
      "completion_length": 250.00000762939453,
      "epoch": 1.8014515292897875,
      "grad_norm": 0.34040528535842896,
      "kl": 1.0751953125,
      "learning_rate": 1.2351690852705553e-06,
      "loss": 0.2489,
      "reward": 1.375,
      "reward_std": 0.3354102037847042,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1738
    },
    {
      "completion_length": 208.95000648498535,
      "epoch": 1.80248833592535,
      "grad_norm": 0.7414734959602356,
      "kl": 2.16748046875,
      "learning_rate": 1.2333866646285734e-06,
      "loss": 0.2397,
      "reward": 1.4500000178813934,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1739
    },
    {
      "completion_length": 147.5250015258789,
      "epoch": 1.8035251425609125,
      "grad_norm": 0.5828838348388672,
      "kl": 1.392822265625,
      "learning_rate": 1.2316046326440236e-06,
      "loss": 0.1437,
      "reward": 1.6000000536441803,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1740
    },
    {
      "completion_length": 190.62500190734863,
      "epoch": 1.8045619491964748,
      "grad_norm": 1.4413193464279175,
      "kl": 2.00341796875,
      "learning_rate": 1.2298229919146742e-06,
      "loss": 0.2509,
      "reward": 1.4000000059604645,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1741
    },
    {
      "completion_length": 146.20000648498535,
      "epoch": 1.8055987558320372,
      "grad_norm": 0.4844883382320404,
      "kl": 0.25341796875,
      "learning_rate": 1.228041745037725e-06,
      "loss": 0.0718,
      "reward": 1.4000000059604645,
      "reward_std": 0.1442900486290455,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1742
    },
    {
      "completion_length": 207.55000114440918,
      "epoch": 1.8066355624675998,
      "grad_norm": 2.3084030151367188,
      "kl": 0.81298828125,
      "learning_rate": 1.2262608946098004e-06,
      "loss": 0.1324,
      "reward": 1.5750000178813934,
      "reward_std": 0.26129766181111336,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1743
    },
    {
      "completion_length": 158.89999961853027,
      "epoch": 1.8076723691031624,
      "grad_norm": 0.41830238699913025,
      "kl": 2.94482421875,
      "learning_rate": 1.224480443226948e-06,
      "loss": 0.3097,
      "reward": 1.675000011920929,
      "reward_std": 0.3913118727505207,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1744
    },
    {
      "completion_length": 293.1750030517578,
      "epoch": 1.8087091757387248,
      "grad_norm": 3.3972856998443604,
      "kl": 4.56640625,
      "learning_rate": 1.2227003934846327e-06,
      "loss": 0.4965,
      "reward": 1.625,
      "reward_std": 0.4614894241094589,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1745
    },
    {
      "completion_length": 394.25000381469727,
      "epoch": 1.8097459823742872,
      "grad_norm": 2.418309450149536,
      "kl": 1.626220703125,
      "learning_rate": 1.220920747977735e-06,
      "loss": 0.271,
      "reward": 0.925000011920929,
      "reward_std": 0.3966546207666397,
      "rewards/accuracy_reward": 0.10000000335276127,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1746
    },
    {
      "completion_length": 243.85000801086426,
      "epoch": 1.8107827890098496,
      "grad_norm": 0.8050068020820618,
      "kl": 2.1611328125,
      "learning_rate": 1.2191415093005438e-06,
      "loss": 0.2769,
      "reward": 1.5,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1747
    },
    {
      "completion_length": 212.2000026702881,
      "epoch": 1.8118195956454122,
      "grad_norm": 0.7463563680648804,
      "kl": 0.81689453125,
      "learning_rate": 1.2173626800467587e-06,
      "loss": 0.141,
      "reward": 1.3750000149011612,
      "reward_std": 0.3913118802011013,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1748
    },
    {
      "completion_length": 89.70000267028809,
      "epoch": 1.8128564022809746,
      "grad_norm": 0.7790433168411255,
      "kl": 0.46875,
      "learning_rate": 1.2155842628094786e-06,
      "loss": 0.0045,
      "reward": 1.5999999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 1.0,
      "step": 1749
    },
    {
      "completion_length": 209.25000190734863,
      "epoch": 1.8138932089165372,
      "grad_norm": 1.4553327560424805,
      "kl": 1.29931640625,
      "learning_rate": 1.2138062601812042e-06,
      "loss": 0.1713,
      "reward": 1.5,
      "reward_std": 0.27950848639011383,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 1750
    },
    {
      "completion_length": 99.82500267028809,
      "epoch": 1.8149300155520995,
      "grad_norm": 0.4023449122905731,
      "kl": 0.25341796875,
      "learning_rate": 1.2120286747538314e-06,
      "loss": -0.0057,
      "reward": 1.675000011920929,
      "reward_std": 0.06846531480550766,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 1.0,
      "step": 1751
    },
    {
      "completion_length": 215.1750030517578,
      "epoch": 1.815966822187662,
      "grad_norm": 5.118110656738281,
      "kl": 2.81103515625,
      "learning_rate": 1.210251509118647e-06,
      "loss": 0.293,
      "reward": 1.5250000059604645,
      "reward_std": 0.1369306445121765,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1752
    },
    {
      "completion_length": 104.70000267028809,
      "epoch": 1.8170036288232243,
      "grad_norm": 0.2352413386106491,
      "kl": 0.262451171875,
      "learning_rate": 1.208474765866327e-06,
      "loss": 0.0062,
      "reward": 1.7249999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 1.0,
      "step": 1753
    },
    {
      "completion_length": 170.12500190734863,
      "epoch": 1.818040435458787,
      "grad_norm": 0.32878950238227844,
      "kl": 0.37255859375,
      "learning_rate": 1.206698447586931e-06,
      "loss": 0.0954,
      "reward": 1.5999999940395355,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1754
    },
    {
      "completion_length": 154.725004196167,
      "epoch": 1.8190772420943495,
      "grad_norm": 1.2951363325119019,
      "kl": 1.0927734375,
      "learning_rate": 1.2049225568698992e-06,
      "loss": 0.1117,
      "reward": 1.4000000059604645,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.42499999701976776,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1755
    },
    {
      "completion_length": 155.22500038146973,
      "epoch": 1.820114048729912,
      "grad_norm": 0.4612466096878052,
      "kl": 1.195068359375,
      "learning_rate": 1.2031470963040483e-06,
      "loss": 0.1253,
      "reward": 1.5250000357627869,
      "reward_std": 0.4290027692914009,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1756
    },
    {
      "completion_length": 220.72500038146973,
      "epoch": 1.8211508553654743,
      "grad_norm": 1.5667455196380615,
      "kl": 2.27685546875,
      "learning_rate": 1.201372068477569e-06,
      "loss": 0.2602,
      "reward": 1.6250000298023224,
      "reward_std": 0.2812206968665123,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1757
    },
    {
      "completion_length": 201.92500495910645,
      "epoch": 1.8221876620010367,
      "grad_norm": 1.2501370906829834,
      "kl": 2.5498046875,
      "learning_rate": 1.1995974759780194e-06,
      "loss": 0.2853,
      "reward": 1.425000011920929,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1758
    },
    {
      "completion_length": 114.87500190734863,
      "epoch": 1.8232244686365993,
      "grad_norm": 0.5402330756187439,
      "kl": 0.31103515625,
      "learning_rate": 1.197823321392325e-06,
      "loss": 0.0147,
      "reward": 1.800000011920929,
      "reward_std": 0.12436702102422714,
      "rewards/accuracy_reward": 0.7999999970197678,
      "rewards/format_reward": 1.0,
      "step": 1759
    },
    {
      "completion_length": 205.850004196167,
      "epoch": 1.8242612752721619,
      "grad_norm": 0.4346194267272949,
      "kl": 1.8212890625,
      "learning_rate": 1.1960496073067707e-06,
      "loss": 0.2358,
      "reward": 1.425000011920929,
      "reward_std": 0.16770510375499725,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1760
    },
    {
      "completion_length": 156.875,
      "epoch": 1.8252980819077242,
      "grad_norm": 1.3507357835769653,
      "kl": 1.203369140625,
      "learning_rate": 1.1942763363070014e-06,
      "loss": 0.0688,
      "reward": 1.600000023841858,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1761
    },
    {
      "completion_length": 296.3500118255615,
      "epoch": 1.8263348885432866,
      "grad_norm": 4.305059432983398,
      "kl": 3.052001953125,
      "learning_rate": 1.1925035109780146e-06,
      "loss": 0.3334,
      "reward": 1.5,
      "reward_std": 0.38409097492694855,
      "rewards/accuracy_reward": 0.5999999977648258,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1762
    },
    {
      "completion_length": 155.44999885559082,
      "epoch": 1.827371695178849,
      "grad_norm": 0.37885773181915283,
      "kl": 1.04296875,
      "learning_rate": 1.1907311339041592e-06,
      "loss": 0.0823,
      "reward": 1.8250000178813934,
      "reward_std": 0.27228759601712227,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1763
    },
    {
      "completion_length": 210.77499961853027,
      "epoch": 1.8284085018144116,
      "grad_norm": 0.6505246758460999,
      "kl": 0.936767578125,
      "learning_rate": 1.1889592076691297e-06,
      "loss": 0.1182,
      "reward": 1.4500000178813934,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1764
    },
    {
      "completion_length": 102.07500267028809,
      "epoch": 1.8294453084499742,
      "grad_norm": 0.9405878782272339,
      "kl": 0.290283203125,
      "learning_rate": 1.1871877348559639e-06,
      "loss": 0.005,
      "reward": 1.5,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1765
    },
    {
      "completion_length": 155.0250015258789,
      "epoch": 1.8304821150855366,
      "grad_norm": 0.451928049325943,
      "kl": 1.29931640625,
      "learning_rate": 1.1854167180470385e-06,
      "loss": 0.1072,
      "reward": 1.6250000298023224,
      "reward_std": 0.22531898319721222,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1766
    },
    {
      "completion_length": 147.97500228881836,
      "epoch": 1.831518921721099,
      "grad_norm": 1.0946840047836304,
      "kl": 0.22412109375,
      "learning_rate": 1.183646159824066e-06,
      "loss": 0.0549,
      "reward": 1.5750000029802322,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1767
    },
    {
      "completion_length": 156.4250030517578,
      "epoch": 1.8325557283566614,
      "grad_norm": 0.3518161475658417,
      "kl": 0.310546875,
      "learning_rate": 1.181876062768089e-06,
      "loss": 0.0465,
      "reward": 1.6500000059604645,
      "reward_std": 0.36053745448589325,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1768
    },
    {
      "completion_length": 206.225004196167,
      "epoch": 1.833592534992224,
      "grad_norm": 0.8643704056739807,
      "kl": 1.40771484375,
      "learning_rate": 1.1801064294594797e-06,
      "loss": 0.2228,
      "reward": 1.6500000059604645,
      "reward_std": 0.3354101963341236,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1769
    },
    {
      "completion_length": 223.64999961853027,
      "epoch": 1.8346293416277866,
      "grad_norm": 0.5106009840965271,
      "kl": 1.833251953125,
      "learning_rate": 1.178337262477932e-06,
      "loss": 0.1917,
      "reward": 1.5250000059604645,
      "reward_std": 0.25609344616532326,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1770
    },
    {
      "completion_length": 171.9250087738037,
      "epoch": 1.835666148263349,
      "grad_norm": 1.0903598070144653,
      "kl": 1.017822265625,
      "learning_rate": 1.1765685644024624e-06,
      "loss": 0.0995,
      "reward": 1.7500000298023224,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.7749999985098839,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1771
    },
    {
      "completion_length": 269.050012588501,
      "epoch": 1.8367029548989113,
      "grad_norm": 1.58475661277771,
      "kl": 1.67724609375,
      "learning_rate": 1.1748003378114023e-06,
      "loss": 0.1046,
      "reward": 1.5000000298023224,
      "reward_std": 0.42963574454188347,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1772
    },
    {
      "completion_length": 124.97500228881836,
      "epoch": 1.8377397615344737,
      "grad_norm": 0.42571085691452026,
      "kl": 0.212646484375,
      "learning_rate": 1.1730325852823965e-06,
      "loss": 0.0375,
      "reward": 1.4749999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 1.0,
      "step": 1773
    },
    {
      "completion_length": 190.22500228881836,
      "epoch": 1.8387765681700363,
      "grad_norm": 2.391396999359131,
      "kl": 0.664306640625,
      "learning_rate": 1.1712653093923977e-06,
      "loss": 0.1056,
      "reward": 1.6250000298023224,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1774
    },
    {
      "completion_length": 178.65000534057617,
      "epoch": 1.839813374805599,
      "grad_norm": 0.3581797480583191,
      "kl": 1.29443359375,
      "learning_rate": 1.1694985127176657e-06,
      "loss": 0.1552,
      "reward": 1.6000000536441803,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1775
    },
    {
      "completion_length": 182.50000381469727,
      "epoch": 1.8408501814411613,
      "grad_norm": 1.3145884275436401,
      "kl": 0.640625,
      "learning_rate": 1.1677321978337596e-06,
      "loss": 0.003,
      "reward": 1.475000023841858,
      "reward_std": 0.21275537088513374,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1776
    },
    {
      "completion_length": 226.77501106262207,
      "epoch": 1.8418869880767237,
      "grad_norm": 0.6741175055503845,
      "kl": 1.28173828125,
      "learning_rate": 1.165966367315538e-06,
      "loss": 0.1894,
      "reward": 1.8499999940395355,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.9000000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1777
    },
    {
      "completion_length": 273.4250068664551,
      "epoch": 1.842923794712286,
      "grad_norm": 0.37939608097076416,
      "kl": 1.74267578125,
      "learning_rate": 1.164201023737151e-06,
      "loss": 0.2582,
      "reward": 1.425000011920929,
      "reward_std": 0.39302409440279007,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1778
    },
    {
      "completion_length": 263.87500381469727,
      "epoch": 1.8439606013478487,
      "grad_norm": 2.4952118396759033,
      "kl": 1.37109375,
      "learning_rate": 1.1624361696720424e-06,
      "loss": 0.1401,
      "reward": 1.4500000476837158,
      "reward_std": 0.45977719873189926,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.925000011920929,
      "step": 1779
    },
    {
      "completion_length": 228.975004196167,
      "epoch": 1.844997407983411,
      "grad_norm": 7.276212692260742,
      "kl": 1.14599609375,
      "learning_rate": 1.1606718076929387e-06,
      "loss": 0.1986,
      "reward": 1.550000011920929,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.6000000052154064,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1780
    },
    {
      "completion_length": 143.85000228881836,
      "epoch": 1.8460342146189737,
      "grad_norm": 2.0451505184173584,
      "kl": 0.3984375,
      "learning_rate": 1.158907940371852e-06,
      "loss": 0.0495,
      "reward": 1.4500000178813934,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1781
    },
    {
      "completion_length": 127.5250015258789,
      "epoch": 1.847071021254536,
      "grad_norm": 0.33295977115631104,
      "kl": 0.20556640625,
      "learning_rate": 1.1571445702800712e-06,
      "loss": 0.018,
      "reward": 1.5999999940395355,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6000000182539225,
      "rewards/format_reward": 1.0,
      "step": 1782
    },
    {
      "completion_length": 121.67500114440918,
      "epoch": 1.8481078278900984,
      "grad_norm": 0.32990562915802,
      "kl": 0.212890625,
      "learning_rate": 1.1553816999881617e-06,
      "loss": -0.0039,
      "reward": 1.7249999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 1.0,
      "step": 1783
    },
    {
      "completion_length": 253.75000381469727,
      "epoch": 1.8491446345256608,
      "grad_norm": 0.3000059127807617,
      "kl": 1.16552734375,
      "learning_rate": 1.1536193320659596e-06,
      "loss": 0.189,
      "reward": 1.5250000059604645,
      "reward_std": 0.21638590097427368,
      "rewards/accuracy_reward": 0.5749999973922968,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1784
    },
    {
      "completion_length": 271.82500076293945,
      "epoch": 1.8501814411612234,
      "grad_norm": 1.5088744163513184,
      "kl": 3.0732421875,
      "learning_rate": 1.1518574690825691e-06,
      "loss": 0.3893,
      "reward": 1.6750000417232513,
      "reward_std": 0.3913118988275528,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1785
    },
    {
      "completion_length": 172.65000343322754,
      "epoch": 1.851218247796786,
      "grad_norm": 5.147366523742676,
      "kl": 1.6201171875,
      "learning_rate": 1.150096113606358e-06,
      "loss": 0.2063,
      "reward": 1.675000011920929,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1786
    },
    {
      "completion_length": 314.1750030517578,
      "epoch": 1.8522550544323484,
      "grad_norm": 2.1646902561187744,
      "kl": 4.66015625,
      "learning_rate": 1.1483352682049552e-06,
      "loss": 0.5139,
      "reward": 1.300000011920929,
      "reward_std": 0.5409446619451046,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1787
    },
    {
      "completion_length": 228.42501068115234,
      "epoch": 1.8532918610679108,
      "grad_norm": 2.4354076385498047,
      "kl": 0.26904296875,
      "learning_rate": 1.1465749354452435e-06,
      "loss": 0.0876,
      "reward": 1.625,
      "reward_std": 0.20019174739718437,
      "rewards/accuracy_reward": 0.6750000007450581,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1788
    },
    {
      "completion_length": 123.35000228881836,
      "epoch": 1.8543286677034732,
      "grad_norm": 0.36107137799263,
      "kl": 0.206298828125,
      "learning_rate": 1.1448151178933613e-06,
      "loss": 0.0454,
      "reward": 1.5750000178813934,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.5749999992549419,
      "rewards/format_reward": 1.0,
      "step": 1789
    },
    {
      "completion_length": 175.1500072479248,
      "epoch": 1.8553654743390358,
      "grad_norm": 1.548614501953125,
      "kl": 0.971923828125,
      "learning_rate": 1.1430558181146956e-06,
      "loss": 0.1015,
      "reward": 1.2000000178813934,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.22500000707805157,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1790
    },
    {
      "completion_length": 266.82501220703125,
      "epoch": 1.8564022809745984,
      "grad_norm": 1.0396159887313843,
      "kl": 1.69189453125,
      "learning_rate": 1.1412970386738757e-06,
      "loss": 0.2626,
      "reward": 1.3000000268220901,
      "reward_std": 0.35704541578888893,
      "rewards/accuracy_reward": 0.37500001303851604,
      "rewards/format_reward": 0.925000011920929,
      "step": 1791
    },
    {
      "completion_length": 272.77500915527344,
      "epoch": 1.8574390876101607,
      "grad_norm": 2.518148183822632,
      "kl": 2.255615234375,
      "learning_rate": 1.1395387821347766e-06,
      "loss": 0.2858,
      "reward": 1.5000000298023224,
      "reward_std": 0.4525563009083271,
      "rewards/accuracy_reward": 0.5750000216066837,
      "rewards/format_reward": 0.925000011920929,
      "step": 1792
    },
    {
      "completion_length": 152.60000228881836,
      "epoch": 1.8584758942457231,
      "grad_norm": 5.471208095550537,
      "kl": 0.904052734375,
      "learning_rate": 1.1377810510605075e-06,
      "loss": 0.1344,
      "reward": 1.725000023841858,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1793
    },
    {
      "completion_length": 215.92500114440918,
      "epoch": 1.8595127008812855,
      "grad_norm": 0.036844827234745026,
      "kl": 1.365234375,
      "learning_rate": 1.136023848013414e-06,
      "loss": 0.2076,
      "reward": 1.3250000178813934,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1794
    },
    {
      "completion_length": 186.94999885559082,
      "epoch": 1.8605495075168481,
      "grad_norm": 1.577744722366333,
      "kl": 0.36279296875,
      "learning_rate": 1.1342671755550695e-06,
      "loss": 0.0738,
      "reward": 1.5,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1795
    },
    {
      "completion_length": 236.3750114440918,
      "epoch": 1.8615863141524107,
      "grad_norm": 0.6909179091453552,
      "kl": 0.349609375,
      "learning_rate": 1.1325110362462771e-06,
      "loss": 0.1053,
      "reward": 1.5250000357627869,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1796
    },
    {
      "completion_length": 249.350004196167,
      "epoch": 1.862623120787973,
      "grad_norm": 0.4353656470775604,
      "kl": 3.93359375,
      "learning_rate": 1.13075543264706e-06,
      "loss": 0.4378,
      "reward": 1.5,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 1797
    },
    {
      "completion_length": 203.45001029968262,
      "epoch": 1.8636599274235355,
      "grad_norm": 7.002437591552734,
      "kl": 2.75146484375,
      "learning_rate": 1.129000367316662e-06,
      "loss": 0.2961,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1798
    },
    {
      "completion_length": 122.72500038146973,
      "epoch": 1.8646967340590979,
      "grad_norm": 0.2963770031929016,
      "kl": 0.27001953125,
      "learning_rate": 1.1272458428135413e-06,
      "loss": 0.0223,
      "reward": 1.6500000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.650000000372529,
      "rewards/format_reward": 1.0,
      "step": 1799
    },
    {
      "completion_length": 270.20001220703125,
      "epoch": 1.8657335406946605,
      "grad_norm": 0.3787246346473694,
      "kl": 1.204833984375,
      "learning_rate": 1.1254918616953688e-06,
      "loss": 0.2406,
      "reward": 1.6500000059604645,
      "reward_std": 0.28485123440623283,
      "rewards/accuracy_reward": 0.7249999940395355,
      "rewards/format_reward": 0.925000011920929,
      "step": 1800
    },
    {
      "completion_length": 323.62500381469727,
      "epoch": 1.866770347330223,
      "grad_norm": 4.747866630554199,
      "kl": 4.41650390625,
      "learning_rate": 1.1237384265190213e-06,
      "loss": 0.4899,
      "reward": 1.4500000178813934,
      "reward_std": 0.4415663927793503,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.875,
      "step": 1801
    },
    {
      "completion_length": 162.25000190734863,
      "epoch": 1.8678071539657854,
      "grad_norm": 3.6878278255462646,
      "kl": 1.429443359375,
      "learning_rate": 1.1219855398405825e-06,
      "loss": 0.1522,
      "reward": 1.3000000268220901,
      "reward_std": 0.2487340308725834,
      "rewards/accuracy_reward": 0.32500000670552254,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1802
    },
    {
      "completion_length": 214.67500495910645,
      "epoch": 1.8688439606013478,
      "grad_norm": 0.9373306632041931,
      "kl": 1.298095703125,
      "learning_rate": 1.1202332042153338e-06,
      "loss": 0.1917,
      "reward": 1.2750000357627869,
      "reward_std": 0.28485123068094254,
      "rewards/accuracy_reward": 0.32499999925494194,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1803
    },
    {
      "completion_length": 221.52499961853027,
      "epoch": 1.8698807672369102,
      "grad_norm": 0.8081947565078735,
      "kl": 0.2734375,
      "learning_rate": 1.1184814221977554e-06,
      "loss": 0.1022,
      "reward": 1.3750000298023224,
      "reward_std": 0.31733788549900055,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1804
    },
    {
      "completion_length": 181.975004196167,
      "epoch": 1.8709175738724728,
      "grad_norm": 0.3742946982383728,
      "kl": 0.85986328125,
      "learning_rate": 1.1167301963415193e-06,
      "loss": 0.0933,
      "reward": 1.350000023841858,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1805
    },
    {
      "completion_length": 266.27499771118164,
      "epoch": 1.8719543805080352,
      "grad_norm": 5.045395851135254,
      "kl": 2.68798828125,
      "learning_rate": 1.1149795291994874e-06,
      "loss": 0.3369,
      "reward": 1.550000011920929,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1806
    },
    {
      "completion_length": 116.7750015258789,
      "epoch": 1.8729911871435978,
      "grad_norm": 0.8432880640029907,
      "kl": 0.2197265625,
      "learning_rate": 1.1132294233237064e-06,
      "loss": 0.0056,
      "reward": 1.6000000089406967,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1807
    },
    {
      "completion_length": 321.4000053405762,
      "epoch": 1.8740279937791602,
      "grad_norm": 2.597256660461426,
      "kl": 2.2613525390625,
      "learning_rate": 1.1114798812654055e-06,
      "loss": 0.3055,
      "reward": 1.550000011920929,
      "reward_std": 0.49226386845111847,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1808
    },
    {
      "completion_length": 207.87500190734863,
      "epoch": 1.8750648004147226,
      "grad_norm": 0.5160976648330688,
      "kl": 1.82470703125,
      "learning_rate": 1.1097309055749915e-06,
      "loss": 0.2081,
      "reward": 1.4000000059604645,
      "reward_std": 0.21275535970926285,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1809
    },
    {
      "completion_length": 113.17500305175781,
      "epoch": 1.8761016070502852,
      "grad_norm": 0.7997815012931824,
      "kl": 0.281982421875,
      "learning_rate": 1.1079824988020467e-06,
      "loss": 0.0544,
      "reward": 1.675000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 1.0,
      "step": 1810
    },
    {
      "completion_length": 203.90000343322754,
      "epoch": 1.8771384136858476,
      "grad_norm": 1.020841121673584,
      "kl": 2.4033203125,
      "learning_rate": 1.1062346634953213e-06,
      "loss": 0.2653,
      "reward": 1.7750000357627869,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1811
    },
    {
      "completion_length": 355.0250053405762,
      "epoch": 1.8781752203214102,
      "grad_norm": 5.2479705810546875,
      "kl": 4.98486328125,
      "learning_rate": 1.1044874022027363e-06,
      "loss": 0.5485,
      "reward": 1.4250000417232513,
      "reward_std": 0.5533697754144669,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1812
    },
    {
      "completion_length": 355.5750026702881,
      "epoch": 1.8792120269569725,
      "grad_norm": 9.089150428771973,
      "kl": 4.91796875,
      "learning_rate": 1.1027407174713724e-06,
      "loss": 0.5858,
      "reward": 1.0500000417232513,
      "reward_std": 0.49226390197873116,
      "rewards/accuracy_reward": 0.2000000048428774,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1813
    },
    {
      "completion_length": 298.5750102996826,
      "epoch": 1.880248833592535,
      "grad_norm": 5.795817852020264,
      "kl": 4.29833984375,
      "learning_rate": 1.100994611847472e-06,
      "loss": 0.4714,
      "reward": 1.6000000536441803,
      "reward_std": 0.5031153038144112,
      "rewards/accuracy_reward": 0.7000000216066837,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1814
    },
    {
      "completion_length": 179.1750030517578,
      "epoch": 1.8812856402280973,
      "grad_norm": 0.4812382459640503,
      "kl": 1.103271484375,
      "learning_rate": 1.0992490878764327e-06,
      "loss": 0.1006,
      "reward": 1.75,
      "reward_std": 0.1604842059314251,
      "rewards/accuracy_reward": 0.7999999970197678,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1815
    },
    {
      "completion_length": 413.0250015258789,
      "epoch": 1.88232244686366,
      "grad_norm": 2.979985237121582,
      "kl": 4.40380859375,
      "learning_rate": 1.097504148102804e-06,
      "loss": 0.6087,
      "reward": 1.2500000298023224,
      "reward_std": 0.45977722853422165,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1816
    },
    {
      "completion_length": 248.55000495910645,
      "epoch": 1.8833592534992225,
      "grad_norm": 1.9136056900024414,
      "kl": 2.117431640625,
      "learning_rate": 1.0957597950702837e-06,
      "loss": 0.2905,
      "reward": 1.4500000029802322,
      "reward_std": 0.27228761091828346,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1817
    },
    {
      "completion_length": 149.9000015258789,
      "epoch": 1.884396060134785,
      "grad_norm": 3.6969919204711914,
      "kl": 1.101806640625,
      "learning_rate": 1.0940160313217152e-06,
      "loss": 0.1208,
      "reward": 1.475000038743019,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1818
    },
    {
      "completion_length": 262.2250118255615,
      "epoch": 1.8854328667703473,
      "grad_norm": 0.5404413938522339,
      "kl": 1.449462890625,
      "learning_rate": 1.0922728593990813e-06,
      "loss": 0.2044,
      "reward": 1.6250000298023224,
      "reward_std": 0.5031153149902821,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1819
    },
    {
      "completion_length": 303.17499923706055,
      "epoch": 1.8864696734059097,
      "grad_norm": 2.9051778316497803,
      "kl": 2.6064453125,
      "learning_rate": 1.0905302818435038e-06,
      "loss": 0.2468,
      "reward": 1.375,
      "reward_std": 0.2686570808291435,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1820
    },
    {
      "completion_length": 217.55000495910645,
      "epoch": 1.8875064800414723,
      "grad_norm": 2.8954715728759766,
      "kl": 1.001953125,
      "learning_rate": 1.0887883011952357e-06,
      "loss": 0.1922,
      "reward": 1.574999988079071,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1821
    },
    {
      "completion_length": 306.35000228881836,
      "epoch": 1.8885432866770349,
      "grad_norm": 1.8013978004455566,
      "kl": 2.07275390625,
      "learning_rate": 1.087046919993663e-06,
      "loss": 0.3474,
      "reward": 1.7000000178813934,
      "reward_std": 0.5031152777373791,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1822
    },
    {
      "completion_length": 303.6500015258789,
      "epoch": 1.8895800933125972,
      "grad_norm": 0.41030386090278625,
      "kl": 1.762939453125,
      "learning_rate": 1.085306140777295e-06,
      "loss": 0.2851,
      "reward": 1.2000000029802322,
      "reward_std": 0.4092182330787182,
      "rewards/accuracy_reward": 0.325000012293458,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1823
    },
    {
      "completion_length": 271.50000762939453,
      "epoch": 1.8906168999481596,
      "grad_norm": 3.335235118865967,
      "kl": 2.220947265625,
      "learning_rate": 1.0835659660837653e-06,
      "loss": 0.2506,
      "reward": 1.300000011920929,
      "reward_std": 0.25609346479177475,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1824
    },
    {
      "completion_length": 262.85000228881836,
      "epoch": 1.891653706583722,
      "grad_norm": 0.33565187454223633,
      "kl": 2.37890625,
      "learning_rate": 1.0818263984498253e-06,
      "loss": 0.3014,
      "reward": 1.5750000476837158,
      "reward_std": 0.3046357333660126,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 1825
    },
    {
      "completion_length": 358.3000068664551,
      "epoch": 1.8926905132192846,
      "grad_norm": 0.7493380904197693,
      "kl": 5.809326171875,
      "learning_rate": 1.0800874404113417e-06,
      "loss": 0.6782,
      "reward": 1.475000023841858,
      "reward_std": 0.4525563269853592,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1826
    },
    {
      "completion_length": 642.1250190734863,
      "epoch": 1.8937273198548472,
      "grad_norm": 21.93817138671875,
      "kl": 8.14306640625,
      "learning_rate": 1.0783490945032925e-06,
      "loss": 0.8109,
      "reward": 1.0750000029802322,
      "reward_std": 0.5769233554601669,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.7249999940395355,
      "step": 1827
    },
    {
      "completion_length": 187.100004196167,
      "epoch": 1.8947641264904096,
      "grad_norm": 0.654755711555481,
      "kl": 1.386962890625,
      "learning_rate": 1.0766113632597636e-06,
      "loss": 0.1964,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1828
    },
    {
      "completion_length": 152.35000610351562,
      "epoch": 1.895800933125972,
      "grad_norm": 1.1775734424591064,
      "kl": 0.19580078125,
      "learning_rate": 1.074874249213945e-06,
      "loss": 0.0854,
      "reward": 1.850000023841858,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1829
    },
    {
      "completion_length": 398.40000343322754,
      "epoch": 1.8968377397615344,
      "grad_norm": 2.1487603187561035,
      "kl": 4.818359375,
      "learning_rate": 1.0731377548981265e-06,
      "loss": 0.4655,
      "reward": 1.350000038743019,
      "reward_std": 0.5967078916728497,
      "rewards/accuracy_reward": 0.5250000208616257,
      "rewards/format_reward": 0.8250000327825546,
      "step": 1830
    },
    {
      "completion_length": 494.6250057220459,
      "epoch": 1.897874546397097,
      "grad_norm": 3.8342268466949463,
      "kl": 6.6142578125,
      "learning_rate": 1.0714018828436947e-06,
      "loss": 0.6902,
      "reward": 1.4250000268220901,
      "reward_std": 0.40921827033162117,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.8000000193715096,
      "step": 1831
    },
    {
      "completion_length": 362.75000762939453,
      "epoch": 1.8989113530326596,
      "grad_norm": 6.237865447998047,
      "kl": 2.207275390625,
      "learning_rate": 1.0696666355811287e-06,
      "loss": 0.3944,
      "reward": 1.350000038743019,
      "reward_std": 0.5715806297957897,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1832
    },
    {
      "completion_length": 351.225004196167,
      "epoch": 1.899948159668222,
      "grad_norm": 3.259511947631836,
      "kl": 2.474853515625,
      "learning_rate": 1.0679320156399981e-06,
      "loss": 0.2497,
      "reward": 1.4500000178813934,
      "reward_std": 0.19346532225608826,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.875,
      "step": 1833
    },
    {
      "completion_length": 355.1499996185303,
      "epoch": 1.9009849663037843,
      "grad_norm": 3.741173028945923,
      "kl": 3.993408203125,
      "learning_rate": 1.0661980255489555e-06,
      "loss": 0.5295,
      "reward": 1.4250000417232513,
      "reward_std": 0.31719934940338135,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1834
    },
    {
      "completion_length": 151.57500457763672,
      "epoch": 1.9020217729393467,
      "grad_norm": 3.732590436935425,
      "kl": 2.19287109375,
      "learning_rate": 1.064464667835739e-06,
      "loss": 0.1009,
      "reward": 1.4500000178813934,
      "reward_std": 0.5156789049506187,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1835
    },
    {
      "completion_length": 451.87500762939453,
      "epoch": 1.9030585795749093,
      "grad_norm": 2.423400402069092,
      "kl": 2.49462890625,
      "learning_rate": 1.0627319450271611e-06,
      "loss": 0.4216,
      "reward": 1.2000000327825546,
      "reward_std": 0.690300427377224,
      "rewards/accuracy_reward": 0.4000000134110451,
      "rewards/format_reward": 0.800000011920929,
      "step": 1836
    },
    {
      "completion_length": 342.95000648498535,
      "epoch": 1.9040953862104717,
      "grad_norm": 4.658113956451416,
      "kl": 3.33251953125,
      "learning_rate": 1.060999859649111e-06,
      "loss": 0.4544,
      "reward": 1.600000023841858,
      "reward_std": 0.36053744331002235,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1837
    },
    {
      "completion_length": 433.6000061035156,
      "epoch": 1.9051321928460343,
      "grad_norm": 1.5304324626922607,
      "kl": 3.76513671875,
      "learning_rate": 1.0592684142265484e-06,
      "loss": 0.5858,
      "reward": 1.3500000536441803,
      "reward_std": 0.45977720618247986,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.8250000327825546,
      "step": 1838
    },
    {
      "completion_length": 491.9250030517578,
      "epoch": 1.9061689994815967,
      "grad_norm": 2.982764959335327,
      "kl": 1.926513671875,
      "learning_rate": 1.0575376112834998e-06,
      "loss": 0.3757,
      "reward": 1.3499999940395355,
      "reward_std": 0.4045084938406944,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.8000000268220901,
      "step": 1839
    },
    {
      "completion_length": 252.82501220703125,
      "epoch": 1.907205806117159,
      "grad_norm": 0.9904256463050842,
      "kl": 1.57568359375,
      "learning_rate": 1.0558074533430546e-06,
      "loss": 0.1511,
      "reward": 1.1250000149011612,
      "reward_std": 0.1928323432803154,
      "rewards/accuracy_reward": 0.20000000298023224,
      "rewards/format_reward": 0.925000011920929,
      "step": 1840
    },
    {
      "completion_length": 99.25000381469727,
      "epoch": 1.9082426127527217,
      "grad_norm": 1.0968436002731323,
      "kl": 0.47216796875,
      "learning_rate": 1.0540779429273632e-06,
      "loss": 0.0489,
      "reward": 1.5250000208616257,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1841
    },
    {
      "completion_length": 241.75000381469727,
      "epoch": 1.909279419388284,
      "grad_norm": 1.6434630155563354,
      "kl": 1.84716796875,
      "learning_rate": 1.0523490825576304e-06,
      "loss": 0.2819,
      "reward": 1.7750000059604645,
      "reward_std": 0.38409100845456123,
      "rewards/accuracy_reward": 0.8500000238418579,
      "rewards/format_reward": 0.925000011920929,
      "step": 1842
    },
    {
      "completion_length": 196.2000026702881,
      "epoch": 1.9103162260238467,
      "grad_norm": 4.200102806091309,
      "kl": 1.955322265625,
      "learning_rate": 1.0506208747541158e-06,
      "loss": 0.2434,
      "reward": 1.3250000029802322,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1843
    },
    {
      "completion_length": 210.4000015258789,
      "epoch": 1.911353032659409,
      "grad_norm": 0.38087666034698486,
      "kl": 0.367919921875,
      "learning_rate": 1.0488933220361246e-06,
      "loss": 0.0898,
      "reward": 1.6250000596046448,
      "reward_std": 0.41643914580345154,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1844
    },
    {
      "completion_length": 151.8250026702881,
      "epoch": 1.9123898392949714,
      "grad_norm": 0.7956162095069885,
      "kl": 1.002685546875,
      "learning_rate": 1.0471664269220103e-06,
      "loss": 0.1219,
      "reward": 1.475000023841858,
      "reward_std": 0.3297629877924919,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1845
    },
    {
      "completion_length": 274.3000087738037,
      "epoch": 1.9134266459305338,
      "grad_norm": 0.9696609973907471,
      "kl": 5.397705078125,
      "learning_rate": 1.0454401919291648e-06,
      "loss": 0.6238,
      "reward": 1.5000000149011612,
      "reward_std": 0.45977720618247986,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1846
    },
    {
      "completion_length": 331.57501220703125,
      "epoch": 1.9144634525660964,
      "grad_norm": 11.981886863708496,
      "kl": 3.90234375,
      "learning_rate": 1.04371461957402e-06,
      "loss": 0.5325,
      "reward": 1.4500000476837158,
      "reward_std": 0.436362162232399,
      "rewards/accuracy_reward": 0.6000000182539225,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1847
    },
    {
      "completion_length": 352.42501068115234,
      "epoch": 1.915500259201659,
      "grad_norm": 8.505236625671387,
      "kl": 4.437255859375,
      "learning_rate": 1.0419897123720403e-06,
      "loss": 0.5587,
      "reward": 1.6250000596046448,
      "reward_std": 0.4723408594727516,
      "rewards/accuracy_reward": 0.7500000223517418,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1848
    },
    {
      "completion_length": 158.5750026702881,
      "epoch": 1.9165370658372214,
      "grad_norm": 1.6408089399337769,
      "kl": 1.48681640625,
      "learning_rate": 1.0402654728377219e-06,
      "loss": 0.1275,
      "reward": 1.6500000059604645,
      "reward_std": 0.2920720987021923,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1849
    },
    {
      "completion_length": 100.72500228881836,
      "epoch": 1.9175738724727838,
      "grad_norm": 0.6208429336547852,
      "kl": 0.25,
      "learning_rate": 1.0385419034845858e-06,
      "loss": -0.0043,
      "reward": 1.425000011920929,
      "reward_std": 0.27228759601712227,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1850
    },
    {
      "completion_length": 154.95000648498535,
      "epoch": 1.9186106791083462,
      "grad_norm": 2.2388131618499756,
      "kl": 1.45458984375,
      "learning_rate": 1.036819006825178e-06,
      "loss": 0.1409,
      "reward": 1.5,
      "reward_std": 0.30463575199246407,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1851
    },
    {
      "completion_length": 275.7750072479248,
      "epoch": 1.9196474857439088,
      "grad_norm": 1.517973780632019,
      "kl": 0.9404296875,
      "learning_rate": 1.0350967853710628e-06,
      "loss": 0.1928,
      "reward": 1.5250000059604645,
      "reward_std": 0.5715805999934673,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1852
    },
    {
      "completion_length": 155.72500610351562,
      "epoch": 1.9206842923794714,
      "grad_norm": 0.5169199109077454,
      "kl": 1.428466796875,
      "learning_rate": 1.0333752416328206e-06,
      "loss": 0.0918,
      "reward": 1.3500000089406967,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.3750000111758709,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1853
    },
    {
      "completion_length": 210.600004196167,
      "epoch": 1.9217210990150337,
      "grad_norm": 0.7431930303573608,
      "kl": 1.59765625,
      "learning_rate": 1.0316543781200429e-06,
      "loss": 0.1882,
      "reward": 1.5750000476837158,
      "reward_std": 0.5408061742782593,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1854
    },
    {
      "completion_length": 312.65000343322754,
      "epoch": 1.9227579056505961,
      "grad_norm": 2.7001216411590576,
      "kl": 1.943359375,
      "learning_rate": 1.0299341973413316e-06,
      "loss": 0.2359,
      "reward": 1.1750000268220901,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1855
    },
    {
      "completion_length": 210.70000648498535,
      "epoch": 1.9237947122861585,
      "grad_norm": 4.911709785461426,
      "kl": 0.89599609375,
      "learning_rate": 1.0282147018042908e-06,
      "loss": 0.1635,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360678762197495,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1856
    },
    {
      "completion_length": 302.22500228881836,
      "epoch": 1.924831518921721,
      "grad_norm": 8.21536922454834,
      "kl": 1.489990234375,
      "learning_rate": 1.0264958940155282e-06,
      "loss": 0.273,
      "reward": 1.5250000357627869,
      "reward_std": 0.4922638610005379,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1857
    },
    {
      "completion_length": 204.2249984741211,
      "epoch": 1.9258683255572837,
      "grad_norm": 5.6338725090026855,
      "kl": 1.3759765625,
      "learning_rate": 1.0247777764806463e-06,
      "loss": 0.1064,
      "reward": 1.6749999821186066,
      "reward_std": 0.25609347969293594,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1858
    },
    {
      "completion_length": 208.62500762939453,
      "epoch": 1.926905132192846,
      "grad_norm": 0.6463760733604431,
      "kl": 0.350830078125,
      "learning_rate": 1.0230603517042444e-06,
      "loss": 0.0992,
      "reward": 1.5499999970197678,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9249999970197678,
      "step": 1859
    },
    {
      "completion_length": 149.17500495910645,
      "epoch": 1.9279419388284085,
      "grad_norm": 0.46324238181114197,
      "kl": 1.07275390625,
      "learning_rate": 1.0213436221899091e-06,
      "loss": 0.0835,
      "reward": 1.6500000059604645,
      "reward_std": 0.10458250343799591,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1860
    },
    {
      "completion_length": 295.1750068664551,
      "epoch": 1.9289787454639709,
      "grad_norm": 0.3605247735977173,
      "kl": 4.64794921875,
      "learning_rate": 1.0196275904402154e-06,
      "loss": 0.5267,
      "reward": 1.2000000178813934,
      "reward_std": 0.3354101702570915,
      "rewards/accuracy_reward": 0.29999999701976776,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1861
    },
    {
      "completion_length": 154.10000038146973,
      "epoch": 1.9300155520995335,
      "grad_norm": 2.2866482734680176,
      "kl": 0.276611328125,
      "learning_rate": 1.0179122589567201e-06,
      "loss": 0.0411,
      "reward": 1.4000000059604645,
      "reward_std": 0.14429005607962608,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1862
    },
    {
      "completion_length": 294.5999984741211,
      "epoch": 1.931052358735096,
      "grad_norm": 0.3986778259277344,
      "kl": 1.63330078125,
      "learning_rate": 1.0161976302399604e-06,
      "loss": 0.3185,
      "reward": 1.5750000327825546,
      "reward_std": 0.34075290337204933,
      "rewards/accuracy_reward": 0.6750000156462193,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1863
    },
    {
      "completion_length": 393.9000015258789,
      "epoch": 1.9320891653706584,
      "grad_norm": 2.3842616081237793,
      "kl": 2.63671875,
      "learning_rate": 1.014483706789446e-06,
      "loss": 0.4191,
      "reward": 1.6250000298023224,
      "reward_std": 0.5973408445715904,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1864
    },
    {
      "completion_length": 448.62500762939453,
      "epoch": 1.9331259720062208,
      "grad_norm": 8.79536247253418,
      "kl": 3.38671875,
      "learning_rate": 1.0127704911036636e-06,
      "loss": 0.4878,
      "reward": 1.3250000178813934,
      "reward_std": 0.6183430999517441,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.8000000268220901,
      "step": 1865
    },
    {
      "completion_length": 406.0999984741211,
      "epoch": 1.9341627786417832,
      "grad_norm": 2.2143993377685547,
      "kl": 5.48046875,
      "learning_rate": 1.0110579856800627e-06,
      "loss": 0.7072,
      "reward": 1.225000023841858,
      "reward_std": 0.5590169690549374,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1866
    },
    {
      "completion_length": 331.2250061035156,
      "epoch": 1.9351995852773458,
      "grad_norm": 1.509391188621521,
      "kl": 2.050537109375,
      "learning_rate": 1.0093461930150604e-06,
      "loss": 0.3489,
      "reward": 1.5000000298023224,
      "reward_std": 0.41657764464616776,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1867
    },
    {
      "completion_length": 151.22500038146973,
      "epoch": 1.9362363919129082,
      "grad_norm": 1.790948748588562,
      "kl": 0.880615234375,
      "learning_rate": 1.007635115604036e-06,
      "loss": 0.0995,
      "reward": 1.5000000298023224,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1868
    },
    {
      "completion_length": 389.9000072479248,
      "epoch": 1.9372731985484708,
      "grad_norm": 0.3396458923816681,
      "kl": 2.37158203125,
      "learning_rate": 1.0059247559413224e-06,
      "loss": 0.2399,
      "reward": 1.3750000447034836,
      "reward_std": 0.4849044792354107,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1869
    },
    {
      "completion_length": 339.0500030517578,
      "epoch": 1.9383100051840332,
      "grad_norm": 1.0203490257263184,
      "kl": 4.2451171875,
      "learning_rate": 1.0042151165202094e-06,
      "loss": 0.5323,
      "reward": 1.2750000357627869,
      "reward_std": 0.41643913835287094,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1870
    },
    {
      "completion_length": 316.4250068664551,
      "epoch": 1.9393468118195956,
      "grad_norm": 3.4523062705993652,
      "kl": 5.239501953125,
      "learning_rate": 1.0025061998329348e-06,
      "loss": 0.5682,
      "reward": 1.3750000596046448,
      "reward_std": 0.565933421254158,
      "rewards/accuracy_reward": 0.5250000208616257,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1871
    },
    {
      "completion_length": 697.7250137329102,
      "epoch": 1.9403836184551582,
      "grad_norm": 1.787365436553955,
      "kl": 6.781494140625,
      "learning_rate": 1.0007980083706845e-06,
      "loss": 0.7802,
      "reward": 1.0500000417232513,
      "reward_std": 0.6658061668276787,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.7000000029802322,
      "step": 1872
    },
    {
      "completion_length": 150.75000381469727,
      "epoch": 1.9414204250907205,
      "grad_norm": 3.339010715484619,
      "kl": 1.56787109375,
      "learning_rate": 9.99090544623586e-07,
      "loss": 0.1362,
      "reward": 1.7249999940395355,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1873
    },
    {
      "completion_length": 443.3750057220459,
      "epoch": 1.9424572317262832,
      "grad_norm": 2.877967119216919,
      "kl": 3.479736328125,
      "learning_rate": 9.973838110807066e-07,
      "loss": 0.4387,
      "reward": 1.1249999850988388,
      "reward_std": 0.4290027692914009,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.8250000029802322,
      "step": 1874
    },
    {
      "completion_length": 303.7250099182129,
      "epoch": 1.9434940383618455,
      "grad_norm": 0.6481685638427734,
      "kl": 3.72802734375,
      "learning_rate": 9.956778102300485e-07,
      "loss": 0.4148,
      "reward": 1.4000000059604645,
      "reward_std": 0.49226387962698936,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1875
    },
    {
      "completion_length": 266.450008392334,
      "epoch": 1.944530844997408,
      "grad_norm": 1.3731906414031982,
      "kl": 3.2451171875,
      "learning_rate": 9.939725445585467e-07,
      "loss": 0.3688,
      "reward": 1.5750000327825546,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.6750000268220901,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1876
    },
    {
      "completion_length": 262.5500011444092,
      "epoch": 1.9455676516329703,
      "grad_norm": 1.7140556573867798,
      "kl": 3.1298828125,
      "learning_rate": 9.922680165520633e-07,
      "loss": 0.3096,
      "reward": 1.350000023841858,
      "reward_std": 0.3966546207666397,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1877
    },
    {
      "completion_length": 102.92500305175781,
      "epoch": 1.946604458268533,
      "grad_norm": 3.247180223464966,
      "kl": 0.350830078125,
      "learning_rate": 9.90564228695387e-07,
      "loss": -0.0103,
      "reward": 1.425000011920929,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.4249999988824129,
      "rewards/format_reward": 1.0,
      "step": 1878
    },
    {
      "completion_length": 410.7500057220459,
      "epoch": 1.9476412649040955,
      "grad_norm": 0.5788431167602539,
      "kl": 1.53564453125,
      "learning_rate": 9.888611834722254e-07,
      "loss": 0.176,
      "reward": 1.3750000149011612,
      "reward_std": 0.3731010667979717,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.8499999940395355,
      "step": 1879
    },
    {
      "completion_length": 203.5500030517578,
      "epoch": 1.9486780715396579,
      "grad_norm": 0.4296620488166809,
      "kl": 2.08447265625,
      "learning_rate": 9.871588833652043e-07,
      "loss": 0.2596,
      "reward": 1.5750000029802322,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1880
    },
    {
      "completion_length": 302.9750003814697,
      "epoch": 1.9497148781752203,
      "grad_norm": 1.4566224813461304,
      "kl": 0.8955078125,
      "learning_rate": 9.854573308558638e-07,
      "loss": 0.1655,
      "reward": 1.3499999940395355,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.8499999940395355,
      "step": 1881
    },
    {
      "completion_length": 197.2750072479248,
      "epoch": 1.9507516848107826,
      "grad_norm": 1.4151586294174194,
      "kl": 2.52099609375,
      "learning_rate": 9.837565284246541e-07,
      "loss": 0.3082,
      "reward": 1.6500000059604645,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1882
    },
    {
      "completion_length": 199.32500076293945,
      "epoch": 1.9517884914463453,
      "grad_norm": 2.0754356384277344,
      "kl": 1.36572265625,
      "learning_rate": 9.820564785509312e-07,
      "loss": 0.1703,
      "reward": 1.4250000566244125,
      "reward_std": 0.45977723225951195,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1883
    },
    {
      "completion_length": 241.00000381469727,
      "epoch": 1.9528252980819079,
      "grad_norm": 0.5099223256111145,
      "kl": 1.987548828125,
      "learning_rate": 9.803571837129548e-07,
      "loss": 0.2949,
      "reward": 1.5500000268220901,
      "reward_std": 0.25609343498945236,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.925000011920929,
      "step": 1884
    },
    {
      "completion_length": 307.07500648498535,
      "epoch": 1.9538621047174702,
      "grad_norm": 0.44971713423728943,
      "kl": 3.94189453125,
      "learning_rate": 9.786586463878835e-07,
      "loss": 0.5022,
      "reward": 1.2750000059604645,
      "reward_std": 0.3407529219985008,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1885
    },
    {
      "completion_length": 312.40000343322754,
      "epoch": 1.9548989113530326,
      "grad_norm": 6.418180465698242,
      "kl": 3.01806640625,
      "learning_rate": 9.769608690517723e-07,
      "loss": 0.3368,
      "reward": 1.3250000178813934,
      "reward_std": 0.47234082594513893,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.7750000059604645,
      "step": 1886
    },
    {
      "completion_length": 197.40000534057617,
      "epoch": 1.955935717988595,
      "grad_norm": 1.2701430320739746,
      "kl": 1.67822265625,
      "learning_rate": 9.752638541795667e-07,
      "loss": 0.1824,
      "reward": 1.4500000178813934,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.925000011920929,
      "step": 1887
    },
    {
      "completion_length": 455.57501220703125,
      "epoch": 1.9569725246241576,
      "grad_norm": 3.930828094482422,
      "kl": 3.11572265625,
      "learning_rate": 9.735676042451037e-07,
      "loss": 0.4811,
      "reward": 1.3250000476837158,
      "reward_std": 0.5356019139289856,
      "rewards/accuracy_reward": 0.5000000111758709,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1888
    },
    {
      "completion_length": 197.5000057220459,
      "epoch": 1.9580093312597202,
      "grad_norm": 4.1649346351623535,
      "kl": 0.983642578125,
      "learning_rate": 9.718721217211015e-07,
      "loss": 0.1585,
      "reward": 1.5500000417232513,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 1889
    },
    {
      "completion_length": 250.600004196167,
      "epoch": 1.9590461378952826,
      "grad_norm": 0.3494587242603302,
      "kl": 1.55078125,
      "learning_rate": 9.701774090791628e-07,
      "loss": 0.217,
      "reward": 1.5,
      "reward_std": 0.1928323470056057,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 1890
    },
    {
      "completion_length": 153.9749984741211,
      "epoch": 1.960082944530845,
      "grad_norm": 0.8926106691360474,
      "kl": 0.245361328125,
      "learning_rate": 9.684834687897656e-07,
      "loss": 0.0825,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360679134726524,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1891
    },
    {
      "completion_length": 458.95001220703125,
      "epoch": 1.9611197511664074,
      "grad_norm": 1.8985650539398193,
      "kl": 3.068115234375,
      "learning_rate": 9.66790303322264e-07,
      "loss": 0.5416,
      "reward": 1.425000011920929,
      "reward_std": 0.540806170552969,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.800000011920929,
      "step": 1892
    },
    {
      "completion_length": 395.15001106262207,
      "epoch": 1.96215655780197,
      "grad_norm": 2.027270793914795,
      "kl": 2.281005859375,
      "learning_rate": 9.650979151448808e-07,
      "loss": 0.3557,
      "reward": 1.3250000327825546,
      "reward_std": 0.39728761464357376,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1893
    },
    {
      "completion_length": 156.17500495910645,
      "epoch": 1.9631933644375326,
      "grad_norm": 3.82222580909729,
      "kl": 0.25,
      "learning_rate": 9.634063067247074e-07,
      "loss": 0.0566,
      "reward": 1.5750000476837158,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.6000000052154064,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1894
    },
    {
      "completion_length": 385.0000057220459,
      "epoch": 1.964230171073095,
      "grad_norm": 0.30109697580337524,
      "kl": 2.16796875,
      "learning_rate": 9.61715480527697e-07,
      "loss": 0.2793,
      "reward": 1.4500000178813934,
      "reward_std": 0.5533697828650475,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.8499999940395355,
      "step": 1895
    },
    {
      "completion_length": 404.350004196167,
      "epoch": 1.9652669777086573,
      "grad_norm": 2.2040598392486572,
      "kl": 4.5947265625,
      "learning_rate": 9.600254390186637e-07,
      "loss": 0.566,
      "reward": 1.2750000208616257,
      "reward_std": 0.5535082593560219,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1896
    },
    {
      "completion_length": 497.7000141143799,
      "epoch": 1.9663037843442197,
      "grad_norm": 5.503700256347656,
      "kl": 3.828857421875,
      "learning_rate": 9.583361846612765e-07,
      "loss": 0.5718,
      "reward": 1.475000023841858,
      "reward_std": 0.7416308633983135,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.800000011920929,
      "step": 1897
    },
    {
      "completion_length": 211.2500057220459,
      "epoch": 1.9673405909797823,
      "grad_norm": 3.9459316730499268,
      "kl": 1.50146484375,
      "learning_rate": 9.566477199180586e-07,
      "loss": 0.1866,
      "reward": 1.5500000417232513,
      "reward_std": 0.27950849011540413,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 1898
    },
    {
      "completion_length": 304.97500228881836,
      "epoch": 1.9683773976153447,
      "grad_norm": 1.4846832752227783,
      "kl": 3.548828125,
      "learning_rate": 9.549600472503802e-07,
      "loss": 0.36,
      "reward": 1.2500000149011612,
      "reward_std": 0.4363621734082699,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1899
    },
    {
      "completion_length": 494.24999809265137,
      "epoch": 1.9694142042509073,
      "grad_norm": 5.710992336273193,
      "kl": 6.60107421875,
      "learning_rate": 9.532731691184581e-07,
      "loss": 0.8096,
      "reward": 1.1500000208616257,
      "reward_std": 0.4415663927793503,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.7750000059604645,
      "step": 1900
    },
    {
      "completion_length": 345.05000495910645,
      "epoch": 1.9704510108864697,
      "grad_norm": 1.5863420963287354,
      "kl": 2.3662109375,
      "learning_rate": 9.515870879813505e-07,
      "loss": 0.3272,
      "reward": 1.5250000357627869,
      "reward_std": 0.5847772285342216,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1901
    },
    {
      "completion_length": 181.57500839233398,
      "epoch": 1.971487817522032,
      "grad_norm": 2.184415340423584,
      "kl": 2.1875,
      "learning_rate": 9.499018062969543e-07,
      "loss": 0.2611,
      "reward": 1.4500000476837158,
      "reward_std": 0.3479738086462021,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1902
    },
    {
      "completion_length": 252.7500057220459,
      "epoch": 1.9725246241575947,
      "grad_norm": 1.3678945302963257,
      "kl": 2.06884765625,
      "learning_rate": 9.482173265219997e-07,
      "loss": 0.2767,
      "reward": 1.5750000476837158,
      "reward_std": 0.4363621771335602,
      "rewards/accuracy_reward": 0.6750000268220901,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1903
    },
    {
      "completion_length": 151.92500495910645,
      "epoch": 1.973561430793157,
      "grad_norm": 1.299222707748413,
      "kl": 1.2138671875,
      "learning_rate": 9.465336511120496e-07,
      "loss": 0.1129,
      "reward": 1.475000038743019,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1904
    },
    {
      "completion_length": 382.5499973297119,
      "epoch": 1.9745982374287196,
      "grad_norm": 6.568785190582275,
      "kl": 4.384521484375,
      "learning_rate": 9.448507825214927e-07,
      "loss": 0.5428,
      "reward": 1.350000023841858,
      "reward_std": 0.49746809154748917,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1905
    },
    {
      "completion_length": 249.35000038146973,
      "epoch": 1.975635044064282,
      "grad_norm": 0.8650876879692078,
      "kl": 3.6494140625,
      "learning_rate": 9.431687232035433e-07,
      "loss": 0.4195,
      "reward": 1.6500000357627869,
      "reward_std": 0.4604102075099945,
      "rewards/accuracy_reward": 0.7500000298023224,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1906
    },
    {
      "completion_length": 245.6500072479248,
      "epoch": 1.9766718506998444,
      "grad_norm": 4.518108367919922,
      "kl": 3.12109375,
      "learning_rate": 9.414874756102333e-07,
      "loss": 0.331,
      "reward": 1.7000000476837158,
      "reward_std": 0.4237985536456108,
      "rewards/accuracy_reward": 0.8250000327825546,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1907
    },
    {
      "completion_length": 301.8499984741211,
      "epoch": 1.9777086573354068,
      "grad_norm": 0.5200906991958618,
      "kl": 3.833984375,
      "learning_rate": 9.398070421924143e-07,
      "loss": 0.4564,
      "reward": 1.4750000536441803,
      "reward_std": 0.6149186789989471,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1908
    },
    {
      "completion_length": 346.9000015258789,
      "epoch": 1.9787454639709694,
      "grad_norm": 1.6915509700775146,
      "kl": 5.0751953125,
      "learning_rate": 9.381274253997501e-07,
      "loss": 0.2743,
      "reward": 1.425000011920929,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1909
    },
    {
      "completion_length": 305.325008392334,
      "epoch": 1.979782270606532,
      "grad_norm": 5.309218883514404,
      "kl": 3.3740234375,
      "learning_rate": 9.364486276807123e-07,
      "loss": 0.4393,
      "reward": 1.300000011920929,
      "reward_std": 0.45413000881671906,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1910
    },
    {
      "completion_length": 154.75000381469727,
      "epoch": 1.9808190772420944,
      "grad_norm": 2.066056728363037,
      "kl": 0.671142578125,
      "learning_rate": 9.347706514825813e-07,
      "loss": 0.0532,
      "reward": 1.3750000447034836,
      "reward_std": 0.353316530585289,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1911
    },
    {
      "completion_length": 387.37500762939453,
      "epoch": 1.9818558838776568,
      "grad_norm": 1.2347073554992676,
      "kl": 4.736572265625,
      "learning_rate": 9.330934992514374e-07,
      "loss": 0.6312,
      "reward": 1.6250000298023224,
      "reward_std": 0.4472135975956917,
      "rewards/accuracy_reward": 0.7750000134110451,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1912
    },
    {
      "completion_length": 406.7500114440918,
      "epoch": 1.9828926905132191,
      "grad_norm": 0.9930393099784851,
      "kl": 3.92578125,
      "learning_rate": 9.314171734321614e-07,
      "loss": 0.4137,
      "reward": 1.4750000089406967,
      "reward_std": 0.38409099355340004,
      "rewards/accuracy_reward": 0.6250000037252903,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1913
    },
    {
      "completion_length": 217.45000648498535,
      "epoch": 1.9839294971487818,
      "grad_norm": 3.541429281234741,
      "kl": 1.580810546875,
      "learning_rate": 9.297416764684285e-07,
      "loss": 0.2431,
      "reward": 1.2000000327825546,
      "reward_std": 0.32818930223584175,
      "rewards/accuracy_reward": 0.2750000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1914
    },
    {
      "completion_length": 195.02500534057617,
      "epoch": 1.9849663037843444,
      "grad_norm": 1.2164496183395386,
      "kl": 2.34326171875,
      "learning_rate": 9.280670108027065e-07,
      "loss": 0.2629,
      "reward": 1.3000000268220901,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1915
    },
    {
      "completion_length": 152.05000495910645,
      "epoch": 1.9860031104199067,
      "grad_norm": 1.7377318143844604,
      "kl": 1.78955078125,
      "learning_rate": 9.2639317887625e-07,
      "loss": 0.1915,
      "reward": 1.5500000417232513,
      "reward_std": 0.447213601320982,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1916
    },
    {
      "completion_length": 388.225004196167,
      "epoch": 1.9870399170554691,
      "grad_norm": 0.8515089154243469,
      "kl": 3.8017578125,
      "learning_rate": 9.247201831291001e-07,
      "loss": 0.4497,
      "reward": 1.4000000059604645,
      "reward_std": 0.4958944320678711,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1917
    },
    {
      "completion_length": 235.10000610351562,
      "epoch": 1.9880767236910315,
      "grad_norm": 1.671351671218872,
      "kl": 0.8701171875,
      "learning_rate": 9.230480260000769e-07,
      "loss": 0.1938,
      "reward": 1.5250000059604645,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1918
    },
    {
      "completion_length": 310.70000648498535,
      "epoch": 1.989113530326594,
      "grad_norm": 1.697324514389038,
      "kl": 2.175048828125,
      "learning_rate": 9.213767099267801e-07,
      "loss": 0.34,
      "reward": 1.5000000298023224,
      "reward_std": 0.5533697865903378,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1919
    },
    {
      "completion_length": 456.2000102996826,
      "epoch": 1.9901503369621567,
      "grad_norm": 9.253554344177246,
      "kl": 5.18310546875,
      "learning_rate": 9.19706237345581e-07,
      "loss": 0.6159,
      "reward": 1.2750000357627869,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.800000011920929,
      "step": 1920
    },
    {
      "completion_length": 253.95000457763672,
      "epoch": 1.991187143597719,
      "grad_norm": 0.7404049634933472,
      "kl": 0.40478515625,
      "learning_rate": 9.180366106916239e-07,
      "loss": 0.1588,
      "reward": 1.6250000298023224,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 1921
    },
    {
      "completion_length": 138.0000057220459,
      "epoch": 1.9922239502332815,
      "grad_norm": 0.33539578318595886,
      "kl": 0.913330078125,
      "learning_rate": 9.163678323988172e-07,
      "loss": 0.0743,
      "reward": 1.4749999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1922
    },
    {
      "completion_length": 251.1750030517578,
      "epoch": 1.9932607568688439,
      "grad_norm": 2.1586689949035645,
      "kl": 2.477783203125,
      "learning_rate": 9.146999048998349e-07,
      "loss": 0.2985,
      "reward": 1.4000000208616257,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 1923
    },
    {
      "completion_length": 294.5500030517578,
      "epoch": 1.9942975635044065,
      "grad_norm": 0.4859706163406372,
      "kl": 2.2275390625,
      "learning_rate": 9.130328306261095e-07,
      "loss": 0.3606,
      "reward": 1.3250000327825546,
      "reward_std": 0.3804604671895504,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1924
    },
    {
      "completion_length": 399.6999988555908,
      "epoch": 1.995334370139969,
      "grad_norm": 4.5429158210754395,
      "kl": 2.865966796875,
      "learning_rate": 9.113666120078303e-07,
      "loss": 0.4637,
      "reward": 1.3250000178813934,
      "reward_std": 0.5156789049506187,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1925
    },
    {
      "completion_length": 296.475004196167,
      "epoch": 1.9963711767755314,
      "grad_norm": 1.8262897729873657,
      "kl": 2.650146484375,
      "learning_rate": 9.097012514739386e-07,
      "loss": 0.3707,
      "reward": 1.5500000417232513,
      "reward_std": 0.4797002449631691,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1926
    },
    {
      "completion_length": 298.425012588501,
      "epoch": 1.9974079834110938,
      "grad_norm": 3.1941144466400146,
      "kl": 2.65869140625,
      "learning_rate": 9.080367514521257e-07,
      "loss": 0.3707,
      "reward": 1.3250000327825546,
      "reward_std": 0.4922638610005379,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1927
    },
    {
      "completion_length": 332.325008392334,
      "epoch": 1.9984447900466562,
      "grad_norm": 0.6302455067634583,
      "kl": 1.169677734375,
      "learning_rate": 9.063731143688274e-07,
      "loss": 0.1151,
      "reward": 1.550000011920929,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.6750000212341547,
      "rewards/format_reward": 0.875,
      "step": 1928
    },
    {
      "completion_length": 448.6000061035156,
      "epoch": 1.9994815966822188,
      "grad_norm": 1.1343616247177124,
      "kl": 5.01953125,
      "learning_rate": 9.047103426492231e-07,
      "loss": 0.6224,
      "reward": 1.3750000447034836,
      "reward_std": 0.6526094973087311,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.8000000268220901,
      "step": 1929
    },
    {
      "completion_length": 101.1500015258789,
      "epoch": 2.0,
      "grad_norm": 1.2842036485671997,
      "kl": 0.703125,
      "learning_rate": 9.030484387172285e-07,
      "loss": -0.0309,
      "reward": 1.5,
      "reward_std": 0.3209684118628502,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1930
    },
    {
      "completion_length": 587.1000061035156,
      "epoch": 2.0031104199066876,
      "grad_norm": 25.002439498901367,
      "kl": 10.64306640625,
      "learning_rate": 9.013874049954969e-07,
      "loss": 1.2085,
      "reward": 1.350000038743019,
      "reward_std": 0.6343987286090851,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.7500000149011612,
      "step": 1931
    },
    {
      "completion_length": 339.67499923706055,
      "epoch": 2.00414722654225,
      "grad_norm": 2.448984146118164,
      "kl": 0.6630859375,
      "learning_rate": 8.997272439054105e-07,
      "loss": 0.2627,
      "reward": 1.6500000059604645,
      "reward_std": 0.41643915325403214,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1932
    },
    {
      "completion_length": 388.02500343322754,
      "epoch": 2.0051840331778124,
      "grad_norm": 0.6137410402297974,
      "kl": 1.2470703125,
      "learning_rate": 8.980679578670812e-07,
      "loss": 0.3221,
      "reward": 1.3250000327825546,
      "reward_std": 0.4723408371210098,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1933
    },
    {
      "completion_length": 144.67500495910645,
      "epoch": 2.0062208398133747,
      "grad_norm": 0.31965622305870056,
      "kl": 0.271728515625,
      "learning_rate": 8.964095492993445e-07,
      "loss": 0.0577,
      "reward": 1.5999999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1934
    },
    {
      "completion_length": 565.1500091552734,
      "epoch": 2.007257646448937,
      "grad_norm": 13.702900886535645,
      "kl": 6.2353515625,
      "learning_rate": 8.947520206197571e-07,
      "loss": 0.7508,
      "reward": 1.225000023841858,
      "reward_std": 0.5587125048041344,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.75,
      "step": 1935
    },
    {
      "completion_length": 447.1250228881836,
      "epoch": 2.0082944530845,
      "grad_norm": 5.5287017822265625,
      "kl": 3.430908203125,
      "learning_rate": 8.930953742445925e-07,
      "loss": 0.3388,
      "reward": 1.3000000417232513,
      "reward_std": 0.34609565138816833,
      "rewards/accuracy_reward": 0.4749999977648258,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1936
    },
    {
      "completion_length": 534.4000015258789,
      "epoch": 2.0093312597200623,
      "grad_norm": 2.396836519241333,
      "kl": 5.28955078125,
      "learning_rate": 8.914396125888392e-07,
      "loss": 0.6551,
      "reward": 1.5250000357627869,
      "reward_std": 0.5031152851879597,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.7750000059604645,
      "step": 1937
    },
    {
      "completion_length": 241.77500343322754,
      "epoch": 2.0103680663556247,
      "grad_norm": 1.7937800884246826,
      "kl": 2.09326171875,
      "learning_rate": 8.897847380661942e-07,
      "loss": 0.3027,
      "reward": 1.4500000476837158,
      "reward_std": 0.4723408408463001,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1938
    },
    {
      "completion_length": 536.6000061035156,
      "epoch": 2.011404872991187,
      "grad_norm": 2.8050217628479004,
      "kl": 2.77099609375,
      "learning_rate": 8.881307530890635e-07,
      "loss": 0.3741,
      "reward": 1.3250000327825546,
      "reward_std": 0.6850961893796921,
      "rewards/accuracy_reward": 0.5500000063329935,
      "rewards/format_reward": 0.7750000059604645,
      "step": 1939
    },
    {
      "completion_length": 293.00000762939453,
      "epoch": 2.0124416796267495,
      "grad_norm": 4.146521091461182,
      "kl": 1.767333984375,
      "learning_rate": 8.864776600685534e-07,
      "loss": 0.2605,
      "reward": 1.5000000596046448,
      "reward_std": 0.5031153038144112,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9000000357627869,
      "step": 1940
    },
    {
      "completion_length": 394.87501525878906,
      "epoch": 2.0134784862623123,
      "grad_norm": 10.113649368286133,
      "kl": 3.1591796875,
      "learning_rate": 8.848254614144737e-07,
      "loss": 0.5648,
      "reward": 1.5250000357627869,
      "reward_std": 0.6400459408760071,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1941
    },
    {
      "completion_length": 234.07500839233398,
      "epoch": 2.0145152928978747,
      "grad_norm": 1.9152692556381226,
      "kl": 2.36328125,
      "learning_rate": 8.831741595353267e-07,
      "loss": 0.2941,
      "reward": 1.4500000327825546,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 1942
    },
    {
      "completion_length": 96.47500228881836,
      "epoch": 2.015552099533437,
      "grad_norm": 0.2278638631105423,
      "kl": 0.1981201171875,
      "learning_rate": 8.815237568383098e-07,
      "loss": -0.0067,
      "reward": 1.4000000059604645,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 1.0,
      "step": 1943
    },
    {
      "completion_length": 292.35000228881836,
      "epoch": 2.0165889061689994,
      "grad_norm": 1.3832695484161377,
      "kl": 2.01171875,
      "learning_rate": 8.798742557293088e-07,
      "loss": 0.3169,
      "reward": 1.4250000268220901,
      "reward_std": 0.4363621473312378,
      "rewards/accuracy_reward": 0.5250000040978193,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1944
    },
    {
      "completion_length": 144.80000114440918,
      "epoch": 2.017625712804562,
      "grad_norm": 0.638437807559967,
      "kl": 1.096435546875,
      "learning_rate": 8.782256586128959e-07,
      "loss": 0.1227,
      "reward": 1.3499999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1945
    },
    {
      "completion_length": 290.3000030517578,
      "epoch": 2.018662519440124,
      "grad_norm": 2.7349393367767334,
      "kl": 2.3310546875,
      "learning_rate": 8.765779678923241e-07,
      "loss": 0.387,
      "reward": 1.600000023841858,
      "reward_std": 0.4525563307106495,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1946
    },
    {
      "completion_length": 239.60000228881836,
      "epoch": 2.019699326075687,
      "grad_norm": 2.573413133621216,
      "kl": 3.767578125,
      "learning_rate": 8.749311859695262e-07,
      "loss": 0.4075,
      "reward": 1.475000023841858,
      "reward_std": 0.3354101963341236,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 1947
    },
    {
      "completion_length": 286.7000102996826,
      "epoch": 2.0207361327112494,
      "grad_norm": 5.202550411224365,
      "kl": 3.20263671875,
      "learning_rate": 8.732853152451103e-07,
      "loss": 0.3901,
      "reward": 1.475000023841858,
      "reward_std": 0.3533165529370308,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1948
    },
    {
      "completion_length": 405.67501068115234,
      "epoch": 2.021772939346812,
      "grad_norm": 4.781026363372803,
      "kl": 4.447265625,
      "learning_rate": 8.716403581183558e-07,
      "loss": 0.6164,
      "reward": 1.0500000417232513,
      "reward_std": 0.3966546058654785,
      "rewards/accuracy_reward": 0.22500000335276127,
      "rewards/format_reward": 0.8250000178813934,
      "step": 1949
    },
    {
      "completion_length": 294.6750030517578,
      "epoch": 2.022809745982374,
      "grad_norm": 1.7645463943481445,
      "kl": 2.8505859375,
      "learning_rate": 8.699963169872099e-07,
      "loss": 0.4148,
      "reward": 1.6000000089406967,
      "reward_std": 0.292072132229805,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1950
    },
    {
      "completion_length": 278.625,
      "epoch": 2.0238465526179366,
      "grad_norm": 1.5119249820709229,
      "kl": 2.864990234375,
      "learning_rate": 8.683531942482845e-07,
      "loss": 0.3631,
      "reward": 1.600000023841858,
      "reward_std": 0.36789682507514954,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1951
    },
    {
      "completion_length": 238.80000495910645,
      "epoch": 2.0248833592534994,
      "grad_norm": 3.432887315750122,
      "kl": 0.625244140625,
      "learning_rate": 8.667109922968544e-07,
      "loss": 0.1869,
      "reward": 1.6000000536441803,
      "reward_std": 0.36053745448589325,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 1952
    },
    {
      "completion_length": 145.8750057220459,
      "epoch": 2.0259201658890618,
      "grad_norm": 0.9231489896774292,
      "kl": 0.4404296875,
      "learning_rate": 8.6506971352685e-07,
      "loss": 0.0674,
      "reward": 1.5750000178813934,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.6000000033527613,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1953
    },
    {
      "completion_length": 261.100004196167,
      "epoch": 2.026956972524624,
      "grad_norm": 1.614526391029358,
      "kl": 2.0166015625,
      "learning_rate": 8.634293603308571e-07,
      "loss": 0.3477,
      "reward": 1.3250000327825546,
      "reward_std": 0.335410188883543,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1954
    },
    {
      "completion_length": 198.04999923706055,
      "epoch": 2.0279937791601865,
      "grad_norm": 0.6759334206581116,
      "kl": 0.40771484375,
      "learning_rate": 8.61789935100111e-07,
      "loss": 0.1074,
      "reward": 1.4500000178813934,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.49999999813735485,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1955
    },
    {
      "completion_length": 290.5249996185303,
      "epoch": 2.029030585795749,
      "grad_norm": 1.3796406984329224,
      "kl": 0.65576171875,
      "learning_rate": 8.60151440224497e-07,
      "loss": 0.185,
      "reward": 1.4249999970197678,
      "reward_std": 0.36053741723299026,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1956
    },
    {
      "completion_length": 144.10000610351562,
      "epoch": 2.0300673924313117,
      "grad_norm": 0.5158128142356873,
      "kl": 0.36279296875,
      "learning_rate": 8.585138780925402e-07,
      "loss": 0.0665,
      "reward": 1.8250000178813934,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1957
    },
    {
      "completion_length": 299.42501068115234,
      "epoch": 2.031104199066874,
      "grad_norm": 0.6663864850997925,
      "kl": 0.46142578125,
      "learning_rate": 8.568772510914095e-07,
      "loss": 0.1285,
      "reward": 1.425000011920929,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1958
    },
    {
      "completion_length": 245.27499961853027,
      "epoch": 2.0321410057024365,
      "grad_norm": 0.3739622235298157,
      "kl": 1.2841796875,
      "learning_rate": 8.552415616069089e-07,
      "loss": 0.2343,
      "reward": 1.7000000476837158,
      "reward_std": 0.347973819822073,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1959
    },
    {
      "completion_length": 189.1750030517578,
      "epoch": 2.033177812337999,
      "grad_norm": 0.4717879593372345,
      "kl": 0.224853515625,
      "learning_rate": 8.536068120234761e-07,
      "loss": 0.1196,
      "reward": 1.8250000178813934,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.8750000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1960
    },
    {
      "completion_length": 298.3250026702881,
      "epoch": 2.0342146189735613,
      "grad_norm": 3.002955198287964,
      "kl": 2.294921875,
      "learning_rate": 8.519730047241778e-07,
      "loss": 0.2449,
      "reward": 1.6250000596046448,
      "reward_std": 0.46511997282505035,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1961
    },
    {
      "completion_length": 384.875,
      "epoch": 2.035251425609124,
      "grad_norm": 0.3119354248046875,
      "kl": 1.218017578125,
      "learning_rate": 8.503401420907091e-07,
      "loss": 0.3001,
      "reward": 1.4000000357627869,
      "reward_std": 0.40387551859021187,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1962
    },
    {
      "completion_length": 296.625,
      "epoch": 2.0362882322446865,
      "grad_norm": 1.4484668970108032,
      "kl": 1.799072265625,
      "learning_rate": 8.487082265033863e-07,
      "loss": 0.2896,
      "reward": 1.7000000178813934,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1963
    },
    {
      "completion_length": 241.3500099182129,
      "epoch": 2.037325038880249,
      "grad_norm": 1.377748727798462,
      "kl": 1.66015625,
      "learning_rate": 8.470772603411454e-07,
      "loss": 0.284,
      "reward": 1.4000000357627869,
      "reward_std": 0.4525563046336174,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 1964
    },
    {
      "completion_length": 294.50000381469727,
      "epoch": 2.0383618455158112,
      "grad_norm": 0.36533716320991516,
      "kl": 0.247314453125,
      "learning_rate": 8.454472459815382e-07,
      "loss": 0.1353,
      "reward": 1.275000050663948,
      "reward_std": 0.3407529331743717,
      "rewards/accuracy_reward": 0.3750000149011612,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1965
    },
    {
      "completion_length": 237.3500099182129,
      "epoch": 2.0393986521513736,
      "grad_norm": 1.3538765907287598,
      "kl": 1.386474609375,
      "learning_rate": 8.438181858007311e-07,
      "loss": 0.2302,
      "reward": 1.4750000536441803,
      "reward_std": 0.3479737974703312,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 1966
    },
    {
      "completion_length": 233.77500534057617,
      "epoch": 2.0404354587869364,
      "grad_norm": 2.16424822807312,
      "kl": 1.57763671875,
      "learning_rate": 8.421900821734956e-07,
      "loss": 0.2031,
      "reward": 1.300000011920929,
      "reward_std": 0.16770510375499725,
      "rewards/accuracy_reward": 0.37500000186264515,
      "rewards/format_reward": 0.925000011920929,
      "step": 1967
    },
    {
      "completion_length": 324.2750015258789,
      "epoch": 2.041472265422499,
      "grad_norm": 0.43854859471321106,
      "kl": 1.393798828125,
      "learning_rate": 8.405629374732128e-07,
      "loss": 0.1856,
      "reward": 1.600000023841858,
      "reward_std": 0.40387551859021187,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.8750000298023224,
      "step": 1968
    },
    {
      "completion_length": 278.95001220703125,
      "epoch": 2.042509072058061,
      "grad_norm": 0.8608680367469788,
      "kl": 0.977294921875,
      "learning_rate": 8.389367540718632e-07,
      "loss": 0.1545,
      "reward": 1.6500000059604645,
      "reward_std": 0.27228761836886406,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1969
    },
    {
      "completion_length": 338.450008392334,
      "epoch": 2.0435458786936236,
      "grad_norm": 0.23539359867572784,
      "kl": 0.22802734375,
      "learning_rate": 8.37311534340027e-07,
      "loss": 0.1755,
      "reward": 1.425000011920929,
      "reward_std": 0.3354101702570915,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.875,
      "step": 1970
    },
    {
      "completion_length": 337.1500072479248,
      "epoch": 2.044582685329186,
      "grad_norm": 0.27423250675201416,
      "kl": 1.2412109375,
      "learning_rate": 8.356872806468792e-07,
      "loss": 0.2948,
      "reward": 1.3250000327825546,
      "reward_std": 0.39728759974241257,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1971
    },
    {
      "completion_length": 336.4750099182129,
      "epoch": 2.045619491964749,
      "grad_norm": 0.5105375051498413,
      "kl": 1.388671875,
      "learning_rate": 8.340639953601876e-07,
      "loss": 0.3461,
      "reward": 1.4000000208616257,
      "reward_std": 0.40387551113963127,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1972
    },
    {
      "completion_length": 366.6250057220459,
      "epoch": 2.046656298600311,
      "grad_norm": 0.3919083774089813,
      "kl": 0.260498046875,
      "learning_rate": 8.324416808463073e-07,
      "loss": 0.3035,
      "reward": 1.3000000268220901,
      "reward_std": 0.40387552976608276,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.8500000089406967,
      "step": 1973
    },
    {
      "completion_length": 186.20000457763672,
      "epoch": 2.0476931052358736,
      "grad_norm": 0.3490411043167114,
      "kl": 0.25439453125,
      "learning_rate": 8.308203394701784e-07,
      "loss": 0.0366,
      "reward": 1.4500000178813934,
      "reward_std": 0.17304782569408417,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000029802322,
      "step": 1974
    },
    {
      "completion_length": 391.1750068664551,
      "epoch": 2.048729911871436,
      "grad_norm": 0.3857538104057312,
      "kl": 0.23974609375,
      "learning_rate": 8.291999735953218e-07,
      "loss": 0.2549,
      "reward": 1.4500000178813934,
      "reward_std": 0.44892581924796104,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.8499999940395355,
      "step": 1975
    },
    {
      "completion_length": 236.5000057220459,
      "epoch": 2.0497667185069983,
      "grad_norm": 1.3688260316848755,
      "kl": 1.823974609375,
      "learning_rate": 8.275805855838387e-07,
      "loss": 0.2326,
      "reward": 1.600000023841858,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 1976
    },
    {
      "completion_length": 193.47500610351562,
      "epoch": 2.0508035251425607,
      "grad_norm": 0.15756937861442566,
      "kl": 0.364990234375,
      "learning_rate": 8.259621777964013e-07,
      "loss": 0.1099,
      "reward": 1.4250000268220901,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1977
    },
    {
      "completion_length": 231.5000057220459,
      "epoch": 2.0518403317781235,
      "grad_norm": 0.2756265699863434,
      "kl": 0.251953125,
      "learning_rate": 8.24344752592256e-07,
      "loss": 0.1648,
      "reward": 1.3000000268220901,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.925000011920929,
      "step": 1978
    },
    {
      "completion_length": 339.37500953674316,
      "epoch": 2.052877138413686,
      "grad_norm": 0.41065216064453125,
      "kl": 0.258544921875,
      "learning_rate": 8.227283123292151e-07,
      "loss": 0.2282,
      "reward": 1.4500000178813934,
      "reward_std": 0.3479738272726536,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.8750000149011612,
      "step": 1979
    },
    {
      "completion_length": 286.65000343322754,
      "epoch": 2.0539139450492483,
      "grad_norm": 3.3530197143554688,
      "kl": 1.588623046875,
      "learning_rate": 8.211128593636557e-07,
      "loss": 0.2137,
      "reward": 1.4249999970197678,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.5250000022351742,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1980
    },
    {
      "completion_length": 269.57501220703125,
      "epoch": 2.0549507516848107,
      "grad_norm": 3.9249496459960938,
      "kl": 2.6611328125,
      "learning_rate": 8.194983960505148e-07,
      "loss": 0.3662,
      "reward": 1.4749999940395355,
      "reward_std": 0.3913118876516819,
      "rewards/accuracy_reward": 0.5750000048428774,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1981
    },
    {
      "completion_length": 286.7500057220459,
      "epoch": 2.055987558320373,
      "grad_norm": 3.7559006214141846,
      "kl": 2.265625,
      "learning_rate": 8.178849247432889e-07,
      "loss": 0.2103,
      "reward": 1.3750000149011612,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1982
    },
    {
      "completion_length": 285.6750030517578,
      "epoch": 2.057024364955936,
      "grad_norm": 0.793855607509613,
      "kl": 1.099365234375,
      "learning_rate": 8.162724477940259e-07,
      "loss": 0.285,
      "reward": 1.7000000178813934,
      "reward_std": 0.3479738347232342,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1983
    },
    {
      "completion_length": 339.9000129699707,
      "epoch": 2.0580611715914983,
      "grad_norm": 0.9394145011901855,
      "kl": 1.03369140625,
      "learning_rate": 8.146609675533254e-07,
      "loss": 0.1925,
      "reward": 1.5000000298023224,
      "reward_std": 0.4045085161924362,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1984
    },
    {
      "completion_length": 280.62500381469727,
      "epoch": 2.0590979782270606,
      "grad_norm": 0.8912813663482666,
      "kl": 0.317138671875,
      "learning_rate": 8.130504863703335e-07,
      "loss": 0.2282,
      "reward": 1.5500000417232513,
      "reward_std": 0.4472136087715626,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1985
    },
    {
      "completion_length": 195.0999984741211,
      "epoch": 2.060134784862623,
      "grad_norm": 0.48543399572372437,
      "kl": 0.269287109375,
      "learning_rate": 8.11441006592741e-07,
      "loss": 0.0977,
      "reward": 1.5499999821186066,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1986
    },
    {
      "completion_length": 394.7750015258789,
      "epoch": 2.0611715914981854,
      "grad_norm": 2.5876479148864746,
      "kl": 2.205810546875,
      "learning_rate": 8.098325305667781e-07,
      "loss": 0.4181,
      "reward": 1.5250000357627869,
      "reward_std": 0.47247934341430664,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1987
    },
    {
      "completion_length": 241.100004196167,
      "epoch": 2.0622083981337482,
      "grad_norm": 1.0407298803329468,
      "kl": 0.297119140625,
      "learning_rate": 8.082250606372107e-07,
      "loss": 0.1688,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 1988
    },
    {
      "completion_length": 390.42501068115234,
      "epoch": 2.0632452047693106,
      "grad_norm": 2.0177807807922363,
      "kl": 0.221435546875,
      "learning_rate": 8.066185991473413e-07,
      "loss": 0.2286,
      "reward": 1.6000000536441803,
      "reward_std": 0.5156788900494576,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1989
    },
    {
      "completion_length": 185.60000801086426,
      "epoch": 2.064282011404873,
      "grad_norm": 1.0948463678359985,
      "kl": 0.40185546875,
      "learning_rate": 8.050131484389979e-07,
      "loss": 0.1128,
      "reward": 1.5250000059604645,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1990
    },
    {
      "completion_length": 280.4250030517578,
      "epoch": 2.0653188180404354,
      "grad_norm": 2.9920551776885986,
      "kl": 1.01025390625,
      "learning_rate": 8.034087108525391e-07,
      "loss": 0.2619,
      "reward": 1.5750000476837158,
      "reward_std": 0.5031152777373791,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1991
    },
    {
      "completion_length": 189.62500953674316,
      "epoch": 2.0663556246759978,
      "grad_norm": 1.8554394245147705,
      "kl": 0.278076171875,
      "learning_rate": 8.018052887268441e-07,
      "loss": 0.1142,
      "reward": 1.699999988079071,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1992
    },
    {
      "completion_length": 382.37500762939453,
      "epoch": 2.0673924313115606,
      "grad_norm": 2.798607587814331,
      "kl": 1.88818359375,
      "learning_rate": 8.002028843993127e-07,
      "loss": 0.2803,
      "reward": 1.3000000566244125,
      "reward_std": 0.5408061668276787,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 0.8500000238418579,
      "step": 1993
    },
    {
      "completion_length": 243.52499961853027,
      "epoch": 2.068429237947123,
      "grad_norm": 3.0551185607910156,
      "kl": 0.2724609375,
      "learning_rate": 7.986015002058604e-07,
      "loss": 0.1509,
      "reward": 1.2500000298023224,
      "reward_std": 0.3245587572455406,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.925000011920929,
      "step": 1994
    },
    {
      "completion_length": 235.3249969482422,
      "epoch": 2.0694660445826853,
      "grad_norm": 1.484575629234314,
      "kl": 1.033203125,
      "learning_rate": 7.970011384809171e-07,
      "loss": 0.1316,
      "reward": 1.5250000059604645,
      "reward_std": 0.2974148504436016,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9000000208616257,
      "step": 1995
    },
    {
      "completion_length": 316.5250053405762,
      "epoch": 2.0705028512182477,
      "grad_norm": 3.713959217071533,
      "kl": 1.633544921875,
      "learning_rate": 7.954018015574206e-07,
      "loss": 0.219,
      "reward": 1.5000000298023224,
      "reward_std": 0.40450848639011383,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.875,
      "step": 1996
    },
    {
      "completion_length": 281.600004196167,
      "epoch": 2.07153965785381,
      "grad_norm": 2.3042683601379395,
      "kl": 1.9228515625,
      "learning_rate": 7.938034917668154e-07,
      "loss": 0.185,
      "reward": 1.675000011920929,
      "reward_std": 0.36053742468357086,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 1997
    },
    {
      "completion_length": 126.85000038146973,
      "epoch": 2.072576464489373,
      "grad_norm": 0.43266192078590393,
      "kl": 1.15478515625,
      "learning_rate": 7.92206211439048e-07,
      "loss": 0.0888,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 1998
    },
    {
      "completion_length": 86.80000114440918,
      "epoch": 2.0736132711249353,
      "grad_norm": 0.4028664827346802,
      "kl": 0.330322265625,
      "learning_rate": 7.906099629025667e-07,
      "loss": 0.0339,
      "reward": 1.5750000178813934,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.5750000048428774,
      "rewards/format_reward": 1.0,
      "step": 1999
    },
    {
      "completion_length": 184.82500076293945,
      "epoch": 2.0746500777604977,
      "grad_norm": 1.6196669340133667,
      "kl": 1.696533203125,
      "learning_rate": 7.890147484843112e-07,
      "loss": 0.0752,
      "reward": 1.5250000357627869,
      "reward_std": 0.36053742468357086,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2000
    },
    {
      "completion_length": 188.02501106262207,
      "epoch": 2.07568688439606,
      "grad_norm": 2.214606523513794,
      "kl": 1.37060546875,
      "learning_rate": 7.874205705097185e-07,
      "loss": 0.1885,
      "reward": 1.6500000357627869,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2001
    },
    {
      "completion_length": 188.07500457763672,
      "epoch": 2.0767236910316225,
      "grad_norm": 1.1378860473632812,
      "kl": 0.546875,
      "learning_rate": 7.858274313027117e-07,
      "loss": 0.1597,
      "reward": 1.5250000208616257,
      "reward_std": 0.18026872351765633,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2002
    },
    {
      "completion_length": 196.32500648498535,
      "epoch": 2.0777604976671853,
      "grad_norm": 1.1834465265274048,
      "kl": 0.3642578125,
      "learning_rate": 7.842353331857005e-07,
      "loss": 0.1344,
      "reward": 1.8750000298023224,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.9250000268220901,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2003
    },
    {
      "completion_length": 146.02500343322754,
      "epoch": 2.0787973043027477,
      "grad_norm": 0.8654714226722717,
      "kl": 0.3408203125,
      "learning_rate": 7.826442784795763e-07,
      "loss": 0.0355,
      "reward": 1.550000011920929,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2004
    },
    {
      "completion_length": 334.4000053405762,
      "epoch": 2.07983411093831,
      "grad_norm": 0.4969034790992737,
      "kl": 0.253173828125,
      "learning_rate": 7.810542695037115e-07,
      "loss": 0.2195,
      "reward": 1.425000011920929,
      "reward_std": 0.36053742840886116,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2005
    },
    {
      "completion_length": 82.1500015258789,
      "epoch": 2.0808709175738724,
      "grad_norm": 0.3520283102989197,
      "kl": 0.35498046875,
      "learning_rate": 7.79465308575952e-07,
      "loss": 0.0146,
      "reward": 1.7000000178813934,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 1.0,
      "step": 2006
    },
    {
      "completion_length": 98.75000381469727,
      "epoch": 2.081907724209435,
      "grad_norm": 0.1894538402557373,
      "kl": 0.296875,
      "learning_rate": 7.778773980126163e-07,
      "loss": 0.0076,
      "reward": 1.7249999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 1.0,
      "step": 2007
    },
    {
      "completion_length": 153.77500534057617,
      "epoch": 2.082944530844997,
      "grad_norm": 0.2839549779891968,
      "kl": 0.324462890625,
      "learning_rate": 7.762905401284917e-07,
      "loss": 0.0537,
      "reward": 1.4000000059604645,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.42499999701976776,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2008
    },
    {
      "completion_length": 234.37500190734863,
      "epoch": 2.08398133748056,
      "grad_norm": 0.2582257390022278,
      "kl": 0.3232421875,
      "learning_rate": 7.747047372368327e-07,
      "loss": 0.1262,
      "reward": 1.5750000178813934,
      "reward_std": 0.27228761091828346,
      "rewards/accuracy_reward": 0.6500000096857548,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2009
    },
    {
      "completion_length": 178.02500534057617,
      "epoch": 2.0850181441161224,
      "grad_norm": 0.42866042256355286,
      "kl": 0.211669921875,
      "learning_rate": 7.731199916493528e-07,
      "loss": 0.1196,
      "reward": 1.4749999940395355,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2010
    },
    {
      "completion_length": 232.5500030517578,
      "epoch": 2.086054950751685,
      "grad_norm": 0.38205114006996155,
      "kl": 0.48583984375,
      "learning_rate": 7.715363056762267e-07,
      "loss": 0.1418,
      "reward": 1.2750000208616257,
      "reward_std": 0.26865706220269203,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2011
    },
    {
      "completion_length": 186.07500171661377,
      "epoch": 2.087091757387247,
      "grad_norm": 0.4646410644054413,
      "kl": 0.30517578125,
      "learning_rate": 7.699536816260835e-07,
      "loss": 0.0907,
      "reward": 1.7500000596046448,
      "reward_std": 0.3479738049209118,
      "rewards/accuracy_reward": 0.8000000268220901,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2012
    },
    {
      "completion_length": 279.3250102996826,
      "epoch": 2.0881285640228096,
      "grad_norm": 0.32570788264274597,
      "kl": 0.243408203125,
      "learning_rate": 7.683721218060041e-07,
      "loss": 0.2162,
      "reward": 1.725000023841858,
      "reward_std": 0.3354101926088333,
      "rewards/accuracy_reward": 0.8250000104308128,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2013
    },
    {
      "completion_length": 179.55000591278076,
      "epoch": 2.0891653706583724,
      "grad_norm": 0.22735489904880524,
      "kl": 0.29345703125,
      "learning_rate": 7.667916285215178e-07,
      "loss": 0.0882,
      "reward": 1.5,
      "reward_std": 0.22360678017139435,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2014
    },
    {
      "completion_length": 291.30000495910645,
      "epoch": 2.0902021772939348,
      "grad_norm": 0.36799928545951843,
      "kl": 0.21484375,
      "learning_rate": 7.652122040766007e-07,
      "loss": 0.2275,
      "reward": 1.600000023841858,
      "reward_std": 0.4723408594727516,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2015
    },
    {
      "completion_length": 200.350004196167,
      "epoch": 2.091238983929497,
      "grad_norm": 0.3728603720664978,
      "kl": 0.30908203125,
      "learning_rate": 7.636338507736689e-07,
      "loss": 0.0897,
      "reward": 1.6250000149011612,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2016
    },
    {
      "completion_length": 244.1250114440918,
      "epoch": 2.0922757905650595,
      "grad_norm": 0.29106128215789795,
      "kl": 0.41455078125,
      "learning_rate": 7.620565709135786e-07,
      "loss": 0.0648,
      "reward": 1.600000023841858,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 2017
    },
    {
      "completion_length": 199.27500343322754,
      "epoch": 2.093312597200622,
      "grad_norm": 0.5537993907928467,
      "kl": 0.296875,
      "learning_rate": 7.604803667956189e-07,
      "loss": 0.1206,
      "reward": 1.6250000298023224,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2018
    },
    {
      "completion_length": 302.37500762939453,
      "epoch": 2.0943494038361847,
      "grad_norm": 1.0727345943450928,
      "kl": 0.330078125,
      "learning_rate": 7.589052407175146e-07,
      "loss": 0.2081,
      "reward": 1.4500000327825546,
      "reward_std": 0.3913119174540043,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2019
    },
    {
      "completion_length": 289.700008392334,
      "epoch": 2.095386210471747,
      "grad_norm": 1.341968297958374,
      "kl": 0.33642578125,
      "learning_rate": 7.57331194975414e-07,
      "loss": 0.2016,
      "reward": 1.475000023841858,
      "reward_std": 0.40387551486492157,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2020
    },
    {
      "completion_length": 143.2500057220459,
      "epoch": 2.0964230171073095,
      "grad_norm": 0.7596957087516785,
      "kl": 0.261474609375,
      "learning_rate": 7.557582318638953e-07,
      "loss": 0.0524,
      "reward": 1.675000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2021
    },
    {
      "completion_length": 197.52500534057617,
      "epoch": 2.097459823742872,
      "grad_norm": 0.9973276257514954,
      "kl": 0.269287109375,
      "learning_rate": 7.541863536759551e-07,
      "loss": 0.0842,
      "reward": 1.4500000178813934,
      "reward_std": 0.20539596676826477,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2022
    },
    {
      "completion_length": 142.0500030517578,
      "epoch": 2.0984966303784343,
      "grad_norm": 0.9867051839828491,
      "kl": 0.23583984375,
      "learning_rate": 7.526155627030102e-07,
      "loss": 0.0587,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2023
    },
    {
      "completion_length": 102.0,
      "epoch": 2.099533437013997,
      "grad_norm": 0.07684917002916336,
      "kl": 0.41845703125,
      "learning_rate": 7.510458612348911e-07,
      "loss": 0.0154,
      "reward": 1.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 1.0,
      "step": 2024
    },
    {
      "completion_length": 194.67499923706055,
      "epoch": 2.1005702436495595,
      "grad_norm": 1.6170426607131958,
      "kl": 0.40966796875,
      "learning_rate": 7.494772515598424e-07,
      "loss": 0.1343,
      "reward": 1.5499999970197678,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2025
    },
    {
      "completion_length": 153.87500190734863,
      "epoch": 2.101607050285122,
      "grad_norm": 0.6535692811012268,
      "kl": 0.3037109375,
      "learning_rate": 7.47909735964515e-07,
      "loss": 0.0611,
      "reward": 1.5999999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2026
    },
    {
      "completion_length": 138.50000190734863,
      "epoch": 2.1026438569206842,
      "grad_norm": 0.5811618566513062,
      "kl": 0.297119140625,
      "learning_rate": 7.463433167339648e-07,
      "loss": 0.0692,
      "reward": 1.7500000298023224,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2027
    },
    {
      "completion_length": 338.4250068664551,
      "epoch": 2.1036806635562466,
      "grad_norm": 0.9311997890472412,
      "kl": 0.4462890625,
      "learning_rate": 7.447779961516519e-07,
      "loss": 0.1481,
      "reward": 1.2750000208616257,
      "reward_std": 0.533585250377655,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.8250000029802322,
      "step": 2028
    },
    {
      "completion_length": 148.52499961853027,
      "epoch": 2.1047174701918094,
      "grad_norm": 0.5147411227226257,
      "kl": 0.302490234375,
      "learning_rate": 7.43213776499432e-07,
      "loss": 0.0894,
      "reward": 1.550000011920929,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2029
    },
    {
      "completion_length": 152.37500190734863,
      "epoch": 2.105754276827372,
      "grad_norm": 0.5189192891120911,
      "kl": 0.37646484375,
      "learning_rate": 7.416506600575577e-07,
      "loss": 0.1113,
      "reward": 1.5250000059604645,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.5750000067055225,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2030
    },
    {
      "completion_length": 217.2500057220459,
      "epoch": 2.106791083462934,
      "grad_norm": 0.4477660357952118,
      "kl": 0.4052734375,
      "learning_rate": 7.400886491046718e-07,
      "loss": 0.1798,
      "reward": 1.5000000298023224,
      "reward_std": 0.38409100472927094,
      "rewards/accuracy_reward": 0.6000000052154064,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2031
    },
    {
      "completion_length": 245.62499809265137,
      "epoch": 2.1078278900984966,
      "grad_norm": 0.3453424274921417,
      "kl": 0.310791015625,
      "learning_rate": 7.385277459178084e-07,
      "loss": 0.1418,
      "reward": 1.5250000357627869,
      "reward_std": 0.31719937175512314,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2032
    },
    {
      "completion_length": 243.27500915527344,
      "epoch": 2.108864696734059,
      "grad_norm": 0.35364580154418945,
      "kl": 0.28466796875,
      "learning_rate": 7.369679527723827e-07,
      "loss": 0.1779,
      "reward": 1.4500000178813934,
      "reward_std": 0.3479738011956215,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 2033
    },
    {
      "completion_length": 294.40000343322754,
      "epoch": 2.1099015033696213,
      "grad_norm": 0.34067314863204956,
      "kl": 0.39599609375,
      "learning_rate": 7.354092719421955e-07,
      "loss": 0.198,
      "reward": 1.4250000417232513,
      "reward_std": 0.3768701031804085,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2034
    },
    {
      "completion_length": 192.10000801086426,
      "epoch": 2.110938310005184,
      "grad_norm": 0.7962273359298706,
      "kl": 0.23779296875,
      "learning_rate": 7.33851705699424e-07,
      "loss": 0.1156,
      "reward": 1.675000011920929,
      "reward_std": 0.3678968511521816,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2035
    },
    {
      "completion_length": 257.1750068664551,
      "epoch": 2.1119751166407466,
      "grad_norm": 0.3637199103832245,
      "kl": 0.3955078125,
      "learning_rate": 7.322952563146213e-07,
      "loss": 0.1109,
      "reward": 1.6000000536441803,
      "reward_std": 0.3966546356678009,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2036
    },
    {
      "completion_length": 240.8750057220459,
      "epoch": 2.113011923276309,
      "grad_norm": 0.21260559558868408,
      "kl": 0.284423828125,
      "learning_rate": 7.307399260567115e-07,
      "loss": 0.0722,
      "reward": 1.350000023841858,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2037
    },
    {
      "completion_length": 156.17500114440918,
      "epoch": 2.1140487299118713,
      "grad_norm": 0.4064944088459015,
      "kl": 0.28662109375,
      "learning_rate": 7.291857171929894e-07,
      "loss": 0.019,
      "reward": 1.5250000059604645,
      "reward_std": 0.21638590097427368,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2038
    },
    {
      "completion_length": 202.75000381469727,
      "epoch": 2.1150855365474337,
      "grad_norm": 0.3152235448360443,
      "kl": 0.375,
      "learning_rate": 7.276326319891128e-07,
      "loss": 0.1019,
      "reward": 1.625,
      "reward_std": 0.22894952818751335,
      "rewards/accuracy_reward": 0.7000000067055225,
      "rewards/format_reward": 0.925000011920929,
      "step": 2039
    },
    {
      "completion_length": 201.22500228881836,
      "epoch": 2.1161223431829965,
      "grad_norm": 0.163133904337883,
      "kl": 0.30908203125,
      "learning_rate": 7.260806727091027e-07,
      "loss": 0.1122,
      "reward": 1.7750000357627869,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2040
    },
    {
      "completion_length": 461.2000026702881,
      "epoch": 2.117159149818559,
      "grad_norm": 0.5306205153465271,
      "kl": 0.427001953125,
      "learning_rate": 7.245298416153376e-07,
      "loss": 0.1461,
      "reward": 1.350000023841858,
      "reward_std": 0.7144869491457939,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.8250000029802322,
      "step": 2041
    },
    {
      "completion_length": 170.0999984741211,
      "epoch": 2.1181959564541213,
      "grad_norm": 0.2855950593948364,
      "kl": 0.24365234375,
      "learning_rate": 7.229801409685544e-07,
      "loss": 0.0939,
      "reward": 1.475000023841858,
      "reward_std": 0.23617039993405342,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2042
    },
    {
      "completion_length": 115.72500419616699,
      "epoch": 2.1192327630896837,
      "grad_norm": 0.28930866718292236,
      "kl": 0.291015625,
      "learning_rate": 7.214315730278371e-07,
      "loss": 0.0156,
      "reward": 1.675000011920929,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 1.0,
      "step": 2043
    },
    {
      "completion_length": 157.82500076293945,
      "epoch": 2.120269569725246,
      "grad_norm": 0.4090651869773865,
      "kl": 0.318115234375,
      "learning_rate": 7.198841400506234e-07,
      "loss": 0.0705,
      "reward": 1.725000023841858,
      "reward_std": 0.32818930968642235,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2044
    },
    {
      "completion_length": 349.8750114440918,
      "epoch": 2.121306376360809,
      "grad_norm": 0.4122838079929352,
      "kl": 0.313720703125,
      "learning_rate": 7.183378442926936e-07,
      "loss": 0.0776,
      "reward": 1.4000000357627869,
      "reward_std": 0.4290027469396591,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2045
    },
    {
      "completion_length": 202.10000610351562,
      "epoch": 2.1223431829963713,
      "grad_norm": 0.14067551493644714,
      "kl": 0.291259765625,
      "learning_rate": 7.167926880081711e-07,
      "loss": 0.1152,
      "reward": 1.4749999940395355,
      "reward_std": 0.1442900411784649,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2046
    },
    {
      "completion_length": 237.9500026702881,
      "epoch": 2.1233799896319336,
      "grad_norm": 0.26595401763916016,
      "kl": 0.358154296875,
      "learning_rate": 7.152486734495177e-07,
      "loss": 0.1557,
      "reward": 1.550000011920929,
      "reward_std": 0.28485123068094254,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 2047
    },
    {
      "completion_length": 203.95000457763672,
      "epoch": 2.124416796267496,
      "grad_norm": 0.2376692295074463,
      "kl": 0.3544921875,
      "learning_rate": 7.137058028675325e-07,
      "loss": 0.0118,
      "reward": 1.5250000357627869,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 2048
    },
    {
      "completion_length": 302.7000026702881,
      "epoch": 2.1254536029030584,
      "grad_norm": 0.39795634150505066,
      "kl": 0.36181640625,
      "learning_rate": 7.121640785113449e-07,
      "loss": 0.0532,
      "reward": 1.5750000476837158,
      "reward_std": 0.4415663927793503,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2049
    },
    {
      "completion_length": 242.8250026702881,
      "epoch": 2.1264904095386212,
      "grad_norm": 0.4598332345485687,
      "kl": 0.3505859375,
      "learning_rate": 7.106235026284143e-07,
      "loss": 0.1314,
      "reward": 1.7000000178813934,
      "reward_std": 0.38409101217985153,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2050
    },
    {
      "completion_length": 351.5750198364258,
      "epoch": 2.1275272161741836,
      "grad_norm": 0.30183881521224976,
      "kl": 0.361328125,
      "learning_rate": 7.090840774645254e-07,
      "loss": 0.2712,
      "reward": 1.4500000476837158,
      "reward_std": 0.3913119025528431,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2051
    },
    {
      "completion_length": 147.67500114440918,
      "epoch": 2.128564022809746,
      "grad_norm": 0.4420170783996582,
      "kl": 0.3172607421875,
      "learning_rate": 7.075458052637875e-07,
      "loss": 0.0728,
      "reward": 1.6500000059604645,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2052
    },
    {
      "completion_length": 253.17500114440918,
      "epoch": 2.1296008294453084,
      "grad_norm": 0.33052965998649597,
      "kl": 0.2900390625,
      "learning_rate": 7.060086882686252e-07,
      "loss": 0.1018,
      "reward": 1.4500000476837158,
      "reward_std": 0.3804604485630989,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.925000011920929,
      "step": 2053
    },
    {
      "completion_length": 198.04999923706055,
      "epoch": 2.1306376360808708,
      "grad_norm": 0.3080199956893921,
      "kl": 0.26220703125,
      "learning_rate": 7.044727287197831e-07,
      "loss": 0.1128,
      "reward": 1.675000011920929,
      "reward_std": 0.12436702102422714,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2054
    },
    {
      "completion_length": 249.3000087738037,
      "epoch": 2.1316744427164336,
      "grad_norm": 0.4225241541862488,
      "kl": 0.278564453125,
      "learning_rate": 7.029379288563166e-07,
      "loss": 0.1006,
      "reward": 1.5000000149011612,
      "reward_std": 0.3966546095907688,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2055
    },
    {
      "completion_length": 262.0249996185303,
      "epoch": 2.132711249351996,
      "grad_norm": 0.2810532748699188,
      "kl": 0.30517578125,
      "learning_rate": 7.014042909155904e-07,
      "loss": 0.1282,
      "reward": 1.675000011920929,
      "reward_std": 0.31719937175512314,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 2056
    },
    {
      "completion_length": 313.3750057220459,
      "epoch": 2.1337480559875583,
      "grad_norm": 0.3446415364742279,
      "kl": 0.42333984375,
      "learning_rate": 6.998718171332753e-07,
      "loss": 0.1523,
      "reward": 1.5000000298023224,
      "reward_std": 0.36053742840886116,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2057
    },
    {
      "completion_length": 341.1500072479248,
      "epoch": 2.1347848626231207,
      "grad_norm": 0.4387199580669403,
      "kl": 0.333984375,
      "learning_rate": 6.983405097433468e-07,
      "loss": 0.1715,
      "reward": 1.2250000536441803,
      "reward_std": 0.5031153000891209,
      "rewards/accuracy_reward": 0.37500000186264515,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2058
    },
    {
      "completion_length": 259.9250030517578,
      "epoch": 2.135821669258683,
      "grad_norm": 0.19477449357509613,
      "kl": 0.36328125,
      "learning_rate": 6.968103709780783e-07,
      "loss": 0.1761,
      "reward": 1.6000000536441803,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2059
    },
    {
      "completion_length": 229.69999885559082,
      "epoch": 2.136858475894246,
      "grad_norm": 0.36783385276794434,
      "kl": 0.3450927734375,
      "learning_rate": 6.952814030680402e-07,
      "loss": 0.0892,
      "reward": 1.600000023841858,
      "reward_std": 0.292072132229805,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.925000011920929,
      "step": 2060
    },
    {
      "completion_length": 398.27501106262207,
      "epoch": 2.1378952825298083,
      "grad_norm": 0.32776206731796265,
      "kl": 0.43212890625,
      "learning_rate": 6.937536082420951e-07,
      "loss": 0.2079,
      "reward": 1.5,
      "reward_std": 0.5282425247132778,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2061
    },
    {
      "completion_length": 267.40001106262207,
      "epoch": 2.1389320891653707,
      "grad_norm": 0.33373355865478516,
      "kl": 0.396240234375,
      "learning_rate": 6.922269887273986e-07,
      "loss": 0.1273,
      "reward": 1.3500000536441803,
      "reward_std": 0.3856646716594696,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.925000011920929,
      "step": 2062
    },
    {
      "completion_length": 156.65000343322754,
      "epoch": 2.139968895800933,
      "grad_norm": 0.3265523910522461,
      "kl": 0.19677734375,
      "learning_rate": 6.907015467493883e-07,
      "loss": 0.0581,
      "reward": 1.75,
      "reward_std": 0.1604842059314251,
      "rewards/accuracy_reward": 0.7750000096857548,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2063
    },
    {
      "completion_length": 358.0249996185303,
      "epoch": 2.1410057024364955,
      "grad_norm": 0.32270118594169617,
      "kl": 0.29443359375,
      "learning_rate": 6.891772845317896e-07,
      "loss": 0.254,
      "reward": 1.725000023841858,
      "reward_std": 0.6149186678230762,
      "rewards/accuracy_reward": 0.8500000238418579,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2064
    },
    {
      "completion_length": 412.72500801086426,
      "epoch": 2.1420425090720583,
      "grad_norm": 0.19310106337070465,
      "kl": 0.326416015625,
      "learning_rate": 6.876542042966056e-07,
      "loss": 0.2197,
      "reward": 1.550000011920929,
      "reward_std": 0.26129765436053276,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2065
    },
    {
      "completion_length": 344.4000015258789,
      "epoch": 2.1430793157076207,
      "grad_norm": 0.36721178889274597,
      "kl": 0.36376953125,
      "learning_rate": 6.861323082641166e-07,
      "loss": 0.2453,
      "reward": 1.3000000268220901,
      "reward_std": 0.5084580145776272,
      "rewards/accuracy_reward": 0.42500001564621925,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2066
    },
    {
      "completion_length": 205.97500228881836,
      "epoch": 2.144116122343183,
      "grad_norm": 0.1803424209356308,
      "kl": 0.29443359375,
      "learning_rate": 6.846115986528782e-07,
      "loss": 0.1113,
      "reward": 1.3000000268220901,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2067
    },
    {
      "completion_length": 98.2750015258789,
      "epoch": 2.1451529289787454,
      "grad_norm": 0.4333617389202118,
      "kl": 0.219970703125,
      "learning_rate": 6.830920776797146e-07,
      "loss": 0.008,
      "reward": 1.875,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 1.0,
      "step": 2068
    },
    {
      "completion_length": 252.5000057220459,
      "epoch": 2.146189735614308,
      "grad_norm": 0.37118276953697205,
      "kl": 0.307373046875,
      "learning_rate": 6.815737475597182e-07,
      "loss": 0.1497,
      "reward": 1.8000000417232513,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2069
    },
    {
      "completion_length": 364.87500953674316,
      "epoch": 2.14722654224987,
      "grad_norm": 0.2745435833930969,
      "kl": 0.322509765625,
      "learning_rate": 6.800566105062445e-07,
      "loss": 0.1428,
      "reward": 1.725000023841858,
      "reward_std": 0.40450849011540413,
      "rewards/accuracy_reward": 0.8500000238418579,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2070
    },
    {
      "completion_length": 397.97501373291016,
      "epoch": 2.148263348885433,
      "grad_norm": 0.44514235854148865,
      "kl": 0.3564453125,
      "learning_rate": 6.785406687309116e-07,
      "loss": 0.3016,
      "reward": 1.5250000357627869,
      "reward_std": 0.6400459185242653,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2071
    },
    {
      "completion_length": 269.00000762939453,
      "epoch": 2.1493001555209954,
      "grad_norm": 0.3160283863544464,
      "kl": 0.276611328125,
      "learning_rate": 6.770259244435939e-07,
      "loss": 0.1818,
      "reward": 1.3750000447034836,
      "reward_std": 0.45977723225951195,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2072
    },
    {
      "completion_length": 269.7000045776367,
      "epoch": 2.150336962156558,
      "grad_norm": 0.32224565744400024,
      "kl": 0.359375,
      "learning_rate": 6.755123798524199e-07,
      "loss": 0.1479,
      "reward": 1.3250000327825546,
      "reward_std": 0.36053745448589325,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2073
    },
    {
      "completion_length": 141.6750030517578,
      "epoch": 2.15137376879212,
      "grad_norm": 0.13151563704013824,
      "kl": 0.197265625,
      "learning_rate": 6.740000371637694e-07,
      "loss": 0.0554,
      "reward": 1.7249999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2074
    },
    {
      "completion_length": 216.5250015258789,
      "epoch": 2.1524105754276825,
      "grad_norm": 0.09155783802270889,
      "kl": 0.276123046875,
      "learning_rate": 6.724888985822719e-07,
      "loss": 0.0918,
      "reward": 1.5750000178813934,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2075
    },
    {
      "completion_length": 251.07500457763672,
      "epoch": 2.1534473820632454,
      "grad_norm": 0.47157299518585205,
      "kl": 0.274169921875,
      "learning_rate": 6.70978966310798e-07,
      "loss": 0.1757,
      "reward": 1.475000023841858,
      "reward_std": 0.3966546133160591,
      "rewards/accuracy_reward": 0.5500000081956387,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2076
    },
    {
      "completion_length": 197.69999885559082,
      "epoch": 2.1544841886988078,
      "grad_norm": 0.389840692281723,
      "kl": 0.246337890625,
      "learning_rate": 6.694702425504638e-07,
      "loss": 0.1361,
      "reward": 1.3250000327825546,
      "reward_std": 0.36053745076060295,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2077
    },
    {
      "completion_length": 274.1500053405762,
      "epoch": 2.15552099533437,
      "grad_norm": 0.40455520153045654,
      "kl": 0.30615234375,
      "learning_rate": 6.679627295006206e-07,
      "loss": 0.1146,
      "reward": 1.5750000178813934,
      "reward_std": 0.4849044382572174,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2078
    },
    {
      "completion_length": 303.9250030517578,
      "epoch": 2.1565578019699325,
      "grad_norm": 0.28625062108039856,
      "kl": 0.2314453125,
      "learning_rate": 6.664564293588565e-07,
      "loss": 0.1768,
      "reward": 1.3250000029802322,
      "reward_std": 0.30463575199246407,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2079
    },
    {
      "completion_length": 306.2750015258789,
      "epoch": 2.157594608605495,
      "grad_norm": 0.35751137137413025,
      "kl": 0.313720703125,
      "learning_rate": 6.649513443209901e-07,
      "loss": 0.206,
      "reward": 1.6250000149011612,
      "reward_std": 0.3678968548774719,
      "rewards/accuracy_reward": 0.7250000182539225,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2080
    },
    {
      "completion_length": 197.67500686645508,
      "epoch": 2.1586314152410577,
      "grad_norm": 0.3194804787635803,
      "kl": 0.2451171875,
      "learning_rate": 6.634474765810709e-07,
      "loss": 0.1003,
      "reward": 1.5750000029802322,
      "reward_std": 0.20019173622131348,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2081
    },
    {
      "completion_length": 348.7250099182129,
      "epoch": 2.15966822187662,
      "grad_norm": 0.2810622453689575,
      "kl": 0.3701171875,
      "learning_rate": 6.619448283313717e-07,
      "loss": 0.1539,
      "reward": 1.4500000327825546,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2082
    },
    {
      "completion_length": 245.60000038146973,
      "epoch": 2.1607050285121825,
      "grad_norm": 0.3394882082939148,
      "kl": 0.2900390625,
      "learning_rate": 6.604434017623885e-07,
      "loss": 0.0689,
      "reward": 1.675000011920929,
      "reward_std": 0.34448179602622986,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2083
    },
    {
      "completion_length": 142.0250072479248,
      "epoch": 2.161741835147745,
      "grad_norm": 0.3429841995239258,
      "kl": 0.23486328125,
      "learning_rate": 6.589431990628355e-07,
      "loss": 0.0539,
      "reward": 1.5250000059604645,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2084
    },
    {
      "completion_length": 260.2000045776367,
      "epoch": 2.1627786417833073,
      "grad_norm": 0.20053693652153015,
      "kl": 0.28369140625,
      "learning_rate": 6.574442224196453e-07,
      "loss": 0.1194,
      "reward": 1.300000011920929,
      "reward_std": 0.24151314795017242,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.925000011920929,
      "step": 2085
    },
    {
      "completion_length": 292.49999809265137,
      "epoch": 2.16381544841887,
      "grad_norm": 0.19939592480659485,
      "kl": 0.271240234375,
      "learning_rate": 6.559464740179594e-07,
      "loss": 0.1884,
      "reward": 1.3500000089406967,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2086
    },
    {
      "completion_length": 139.00000381469727,
      "epoch": 2.1648522550544325,
      "grad_norm": 0.1418561488389969,
      "kl": 0.235595703125,
      "learning_rate": 6.544499560411323e-07,
      "loss": 0.0593,
      "reward": 1.8250000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2087
    },
    {
      "completion_length": 314.9250144958496,
      "epoch": 2.165889061689995,
      "grad_norm": 0.2700906991958618,
      "kl": 0.28857421875,
      "learning_rate": 6.529546706707231e-07,
      "loss": 0.0947,
      "reward": 1.5499999821186066,
      "reward_std": 0.25609344989061356,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2088
    },
    {
      "completion_length": 98.1500015258789,
      "epoch": 2.1669258683255572,
      "grad_norm": 0.32122209668159485,
      "kl": 0.3193359375,
      "learning_rate": 6.514606200864943e-07,
      "loss": -0.0563,
      "reward": 1.600000023841858,
      "reward_std": 0.23617040365934372,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2089
    },
    {
      "completion_length": 204.65000534057617,
      "epoch": 2.1679626749611196,
      "grad_norm": 0.407219797372818,
      "kl": 0.3095703125,
      "learning_rate": 6.499678064664077e-07,
      "loss": 0.1286,
      "reward": 1.6250000298023224,
      "reward_std": 0.391311913728714,
      "rewards/accuracy_reward": 0.6750000212341547,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2090
    },
    {
      "completion_length": 151.42500114440918,
      "epoch": 2.168999481596682,
      "grad_norm": 1.0500727891921997,
      "kl": 0.338134765625,
      "learning_rate": 6.484762319866241e-07,
      "loss": 0.0649,
      "reward": 1.4500000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2091
    },
    {
      "completion_length": 167.9000072479248,
      "epoch": 2.170036288232245,
      "grad_norm": 0.11143656075000763,
      "kl": 0.275634765625,
      "learning_rate": 6.469858988214956e-07,
      "loss": 0.0624,
      "reward": 1.5999999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2092
    },
    {
      "completion_length": 247.0250072479248,
      "epoch": 2.171073094867807,
      "grad_norm": 0.3200322687625885,
      "kl": 0.34814453125,
      "learning_rate": 6.454968091435661e-07,
      "loss": 0.1342,
      "reward": 1.5250000059604645,
      "reward_std": 0.335410188883543,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2093
    },
    {
      "completion_length": 187.12500381469727,
      "epoch": 2.1721099015033696,
      "grad_norm": 0.17621248960494995,
      "kl": 0.26220703125,
      "learning_rate": 6.440089651235656e-07,
      "loss": 0.0806,
      "reward": 1.675000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2094
    },
    {
      "completion_length": 201.05000495910645,
      "epoch": 2.173146708138932,
      "grad_norm": 0.3452743589878082,
      "kl": 0.2109375,
      "learning_rate": 6.425223689304104e-07,
      "loss": 0.0968,
      "reward": 1.875,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.9250000268220901,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2095
    },
    {
      "completion_length": 163.3250026702881,
      "epoch": 2.1741835147744943,
      "grad_norm": 0.10241902619600296,
      "kl": 0.18212890625,
      "learning_rate": 6.410370227311947e-07,
      "loss": 0.0607,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2096
    },
    {
      "completion_length": 174.2750072479248,
      "epoch": 2.175220321410057,
      "grad_norm": 0.1164451539516449,
      "kl": 0.34033203125,
      "learning_rate": 6.395529286911933e-07,
      "loss": 0.0646,
      "reward": 1.8250000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2097
    },
    {
      "completion_length": 222.6500015258789,
      "epoch": 2.1762571280456195,
      "grad_norm": 0.3327724039554596,
      "kl": 0.267578125,
      "learning_rate": 6.380700889738543e-07,
      "loss": 0.1271,
      "reward": 1.5750000178813934,
      "reward_std": 0.31199516355991364,
      "rewards/accuracy_reward": 0.6500000152736902,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2098
    },
    {
      "completion_length": 203.1750030517578,
      "epoch": 2.177293934681182,
      "grad_norm": 0.21910661458969116,
      "kl": 0.255859375,
      "learning_rate": 6.365885057407974e-07,
      "loss": 0.1105,
      "reward": 1.5000000298023224,
      "reward_std": 0.22360678017139435,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2099
    },
    {
      "completion_length": 195.25000762939453,
      "epoch": 2.1783307413167443,
      "grad_norm": 0.3655296564102173,
      "kl": 0.31005859375,
      "learning_rate": 6.351081811518103e-07,
      "loss": 0.1045,
      "reward": 1.475000023841858,
      "reward_std": 0.27228760346770287,
      "rewards/accuracy_reward": 0.5250000040978193,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2100
    },
    {
      "completion_length": 250.64999961853027,
      "epoch": 2.1793675479523067,
      "grad_norm": 0.36043232679367065,
      "kl": 0.253662109375,
      "learning_rate": 6.336291173648476e-07,
      "loss": 0.1367,
      "reward": 1.5250000059604645,
      "reward_std": 0.3354101963341236,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2101
    },
    {
      "completion_length": 305.9750061035156,
      "epoch": 2.1804043545878695,
      "grad_norm": 0.2912316620349884,
      "kl": 0.33251953125,
      "learning_rate": 6.321513165360239e-07,
      "loss": 0.2028,
      "reward": 1.5000000447034836,
      "reward_std": 0.36789683997631073,
      "rewards/accuracy_reward": 0.6000000182539225,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2102
    },
    {
      "completion_length": 216.20001029968262,
      "epoch": 2.181441161223432,
      "grad_norm": 0.3792998194694519,
      "kl": 0.28515625,
      "learning_rate": 6.30674780819614e-07,
      "loss": 0.0864,
      "reward": 1.5000000298023224,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2103
    },
    {
      "completion_length": 359.725004196167,
      "epoch": 2.1824779678589943,
      "grad_norm": 0.29253607988357544,
      "kl": 0.408203125,
      "learning_rate": 6.291995123680472e-07,
      "loss": 0.2289,
      "reward": 1.350000023841858,
      "reward_std": 0.36053742468357086,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2104
    },
    {
      "completion_length": 262.850004196167,
      "epoch": 2.1835147744945567,
      "grad_norm": 0.44387897849082947,
      "kl": 0.31982421875,
      "learning_rate": 6.277255133319073e-07,
      "loss": 0.1013,
      "reward": 1.425000011920929,
      "reward_std": 0.45977722853422165,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.925000011920929,
      "step": 2105
    },
    {
      "completion_length": 201.5000057220459,
      "epoch": 2.184551581130119,
      "grad_norm": 0.29915985465049744,
      "kl": 0.34912109375,
      "learning_rate": 6.262527858599262e-07,
      "loss": 0.1364,
      "reward": 1.4250000268220901,
      "reward_std": 0.21275537088513374,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2106
    },
    {
      "completion_length": 103.75000190734863,
      "epoch": 2.185588387765682,
      "grad_norm": 0.2710251808166504,
      "kl": 0.20166015625,
      "learning_rate": 6.247813320989821e-07,
      "loss": 0.0148,
      "reward": 1.5250000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 1.0,
      "step": 2107
    },
    {
      "completion_length": 163.1750030517578,
      "epoch": 2.1866251944012443,
      "grad_norm": 0.11666197329759598,
      "kl": 0.25732421875,
      "learning_rate": 6.233111541940984e-07,
      "loss": 0.0615,
      "reward": 1.2000000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.22500000894069672,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2108
    },
    {
      "completion_length": 154.90000343322754,
      "epoch": 2.1876620010368066,
      "grad_norm": 0.09576858580112457,
      "kl": 0.2099609375,
      "learning_rate": 6.218422542884348e-07,
      "loss": 0.0634,
      "reward": 1.6000000089406967,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2109
    },
    {
      "completion_length": 357.4250011444092,
      "epoch": 2.188698807672369,
      "grad_norm": 0.306047648191452,
      "kl": 0.39892578125,
      "learning_rate": 6.203746345232921e-07,
      "loss": 0.1261,
      "reward": 1.3500000089406967,
      "reward_std": 0.36053742468357086,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2110
    },
    {
      "completion_length": 153.7750015258789,
      "epoch": 2.1897356143079314,
      "grad_norm": 0.4538502097129822,
      "kl": 0.25732421875,
      "learning_rate": 6.189082970381023e-07,
      "loss": 0.0484,
      "reward": 1.8250000476837158,
      "reward_std": 0.3479738011956215,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2111
    },
    {
      "completion_length": 417.75,
      "epoch": 2.1907724209434942,
      "grad_norm": 0.32154762744903564,
      "kl": 0.40673828125,
      "learning_rate": 6.174432439704295e-07,
      "loss": 0.2898,
      "reward": 1.225000023841858,
      "reward_std": 0.4958944134414196,
      "rewards/accuracy_reward": 0.3750000037252903,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2112
    },
    {
      "completion_length": 102.25000190734863,
      "epoch": 2.1918092275790566,
      "grad_norm": 0.29115691781044006,
      "kl": 0.259033203125,
      "learning_rate": 6.15979477455964e-07,
      "loss": -0.0117,
      "reward": 1.425000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 1.0,
      "step": 2113
    },
    {
      "completion_length": 262.6999969482422,
      "epoch": 2.192846034214619,
      "grad_norm": 0.3237297236919403,
      "kl": 0.31591796875,
      "learning_rate": 6.145169996285228e-07,
      "loss": 0.1138,
      "reward": 1.5000000149011612,
      "reward_std": 0.3479738235473633,
      "rewards/accuracy_reward": 0.6000000238418579,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2114
    },
    {
      "completion_length": 363.0500011444092,
      "epoch": 2.1938828408501814,
      "grad_norm": 0.32679373025894165,
      "kl": 0.333984375,
      "learning_rate": 6.130558126200424e-07,
      "loss": 0.2867,
      "reward": 1.600000023841858,
      "reward_std": 0.5156789273023605,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2115
    },
    {
      "completion_length": 153.77499961853027,
      "epoch": 2.1949196474857438,
      "grad_norm": 0.22133252024650574,
      "kl": 0.31005859375,
      "learning_rate": 6.115959185605784e-07,
      "loss": 0.0542,
      "reward": 1.7500000298023224,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2116
    },
    {
      "completion_length": 270.74999809265137,
      "epoch": 2.1959564541213066,
      "grad_norm": 0.3255959153175354,
      "kl": 0.31396484375,
      "learning_rate": 6.101373195783008e-07,
      "loss": 0.1266,
      "reward": 1.4500000476837158,
      "reward_std": 0.45977720245718956,
      "rewards/accuracy_reward": 0.5750000197440386,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2117
    },
    {
      "completion_length": 123.45000267028809,
      "epoch": 2.196993260756869,
      "grad_norm": 0.2980897128582001,
      "kl": 0.2237548828125,
      "learning_rate": 6.086800177994939e-07,
      "loss": -0.0133,
      "reward": 1.7750000059604645,
      "reward_std": 0.12436700984835625,
      "rewards/accuracy_reward": 0.7750000096857548,
      "rewards/format_reward": 1.0,
      "step": 2118
    },
    {
      "completion_length": 345.12500762939453,
      "epoch": 2.1980300673924313,
      "grad_norm": 0.33550626039505005,
      "kl": 0.34765625,
      "learning_rate": 6.072240153485473e-07,
      "loss": 0.2662,
      "reward": 1.3250000327825546,
      "reward_std": 0.4849044792354107,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2119
    },
    {
      "completion_length": 142.22500038146973,
      "epoch": 2.1990668740279937,
      "grad_norm": 0.1336161345243454,
      "kl": 0.252685546875,
      "learning_rate": 6.057693143479606e-07,
      "loss": 0.0506,
      "reward": 1.8499999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2120
    },
    {
      "completion_length": 301.32500648498535,
      "epoch": 2.200103680663556,
      "grad_norm": 0.3746643662452698,
      "kl": 0.2950439453125,
      "learning_rate": 6.043159169183332e-07,
      "loss": 0.1818,
      "reward": 1.475000023841858,
      "reward_std": 0.4525563195347786,
      "rewards/accuracy_reward": 0.5749999992549419,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2121
    },
    {
      "completion_length": 256.25000381469727,
      "epoch": 2.201140487299119,
      "grad_norm": 0.24403604865074158,
      "kl": 0.324462890625,
      "learning_rate": 6.028638251783655e-07,
      "loss": 0.14,
      "reward": 1.6000000536441803,
      "reward_std": 0.2487340271472931,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2122
    },
    {
      "completion_length": 157.225004196167,
      "epoch": 2.2021772939346813,
      "grad_norm": 0.2791767716407776,
      "kl": 0.265380859375,
      "learning_rate": 6.014130412448538e-07,
      "loss": 0.0635,
      "reward": 1.3250000029802322,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2123
    },
    {
      "completion_length": 156.80000495910645,
      "epoch": 2.2032141005702437,
      "grad_norm": 0.25042498111724854,
      "kl": 0.235107421875,
      "learning_rate": 5.999635672326892e-07,
      "loss": 0.0675,
      "reward": 1.6500000357627869,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2124
    },
    {
      "completion_length": 255.67499923706055,
      "epoch": 2.204250907205806,
      "grad_norm": 0.36296358704566956,
      "kl": 0.30419921875,
      "learning_rate": 5.985154052548521e-07,
      "loss": 0.1147,
      "reward": 1.475000023841858,
      "reward_std": 0.36789682880043983,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.925000011920929,
      "step": 2125
    },
    {
      "completion_length": 275.3500061035156,
      "epoch": 2.2052877138413685,
      "grad_norm": 0.39117908477783203,
      "kl": 0.280517578125,
      "learning_rate": 5.970685574224106e-07,
      "loss": 0.1166,
      "reward": 1.3499999940395355,
      "reward_std": 0.25609347224235535,
      "rewards/accuracy_reward": 0.42500000819563866,
      "rewards/format_reward": 0.925000011920929,
      "step": 2126
    },
    {
      "completion_length": 261.6499996185303,
      "epoch": 2.2063245204769313,
      "grad_norm": 0.19222311675548553,
      "kl": 0.251953125,
      "learning_rate": 5.956230258445166e-07,
      "loss": 0.1518,
      "reward": 1.1749999970197678,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2127
    },
    {
      "completion_length": 170.47500610351562,
      "epoch": 2.2073613271124937,
      "grad_norm": 0.33654657006263733,
      "kl": 0.326904296875,
      "learning_rate": 5.941788126284056e-07,
      "loss": 0.0449,
      "reward": 1.8499999940395355,
      "reward_std": 0.21638590842485428,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2128
    },
    {
      "completion_length": 162.125,
      "epoch": 2.208398133748056,
      "grad_norm": 0.25384798645973206,
      "kl": 0.227294921875,
      "learning_rate": 5.927359198793866e-07,
      "loss": 0.0574,
      "reward": 1.6500000357627869,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2129
    },
    {
      "completion_length": 262.42501068115234,
      "epoch": 2.2094349403836184,
      "grad_norm": 0.277602881193161,
      "kl": 0.286865234375,
      "learning_rate": 5.912943497008486e-07,
      "loss": 0.1151,
      "reward": 1.6250000298023224,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2130
    },
    {
      "completion_length": 252.6750087738037,
      "epoch": 2.210471747019181,
      "grad_norm": 0.23738892376422882,
      "kl": 0.285888671875,
      "learning_rate": 5.8985410419425e-07,
      "loss": 0.1206,
      "reward": 1.475000023841858,
      "reward_std": 0.3047742545604706,
      "rewards/accuracy_reward": 0.5500000156462193,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2131
    },
    {
      "completion_length": 227.2750015258789,
      "epoch": 2.211508553654743,
      "grad_norm": 0.34613609313964844,
      "kl": 0.251708984375,
      "learning_rate": 5.884151854591182e-07,
      "loss": 0.1256,
      "reward": 1.5250000059604645,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2132
    },
    {
      "completion_length": 302.75,
      "epoch": 2.212545360290306,
      "grad_norm": 0.27813300490379333,
      "kl": 0.330322265625,
      "learning_rate": 5.869775955930468e-07,
      "loss": 0.2056,
      "reward": 1.3750000298023224,
      "reward_std": 0.3479738272726536,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2133
    },
    {
      "completion_length": 211.02500534057617,
      "epoch": 2.2135821669258684,
      "grad_norm": 0.372684508562088,
      "kl": 0.263671875,
      "learning_rate": 5.855413366916931e-07,
      "loss": 0.0749,
      "reward": 1.625,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.6750000063329935,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2134
    },
    {
      "completion_length": 297.47499656677246,
      "epoch": 2.214618973561431,
      "grad_norm": 0.3438758850097656,
      "kl": 0.26806640625,
      "learning_rate": 5.841064108487733e-07,
      "loss": 0.1861,
      "reward": 1.6750000417232513,
      "reward_std": 0.44892580062150955,
      "rewards/accuracy_reward": 0.8000000268220901,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2135
    },
    {
      "completion_length": 311.95000648498535,
      "epoch": 2.215655780196993,
      "grad_norm": 0.1857534945011139,
      "kl": 0.373046875,
      "learning_rate": 5.826728201560601e-07,
      "loss": 0.1915,
      "reward": 1.5000000298023224,
      "reward_std": 0.3533165380358696,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2136
    },
    {
      "completion_length": 211.44999885559082,
      "epoch": 2.2166925868325555,
      "grad_norm": 0.32240214943885803,
      "kl": 0.286865234375,
      "learning_rate": 5.812405667033799e-07,
      "loss": 0.1029,
      "reward": 1.850000023841858,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.9000000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2137
    },
    {
      "completion_length": 310.8000030517578,
      "epoch": 2.2177293934681184,
      "grad_norm": 0.31809866428375244,
      "kl": 0.3017578125,
      "learning_rate": 5.798096525786115e-07,
      "loss": 0.107,
      "reward": 1.5250000059604645,
      "reward_std": 0.27228761836886406,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2138
    },
    {
      "completion_length": 268.75000953674316,
      "epoch": 2.2187662001036808,
      "grad_norm": 0.377060204744339,
      "kl": 0.297119140625,
      "learning_rate": 5.783800798676779e-07,
      "loss": 0.205,
      "reward": 1.625,
      "reward_std": 0.4723408371210098,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2139
    },
    {
      "completion_length": 104.60000038146973,
      "epoch": 2.219803006739243,
      "grad_norm": 0.32245108485221863,
      "kl": 0.1865234375,
      "learning_rate": 5.769518506545502e-07,
      "loss": 0.0292,
      "reward": 1.7000000178813934,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 1.0,
      "step": 2140
    },
    {
      "completion_length": 334.575008392334,
      "epoch": 2.2208398133748055,
      "grad_norm": 0.166157066822052,
      "kl": 0.315185546875,
      "learning_rate": 5.755249670212387e-07,
      "loss": 0.1584,
      "reward": 1.7500000298023224,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2141
    },
    {
      "completion_length": 312.1500129699707,
      "epoch": 2.221876620010368,
      "grad_norm": 0.33859431743621826,
      "kl": 0.23828125,
      "learning_rate": 5.740994310477931e-07,
      "loss": 0.0086,
      "reward": 1.574999988079071,
      "reward_std": 0.248872559517622,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2142
    },
    {
      "completion_length": 157.87500190734863,
      "epoch": 2.2229134266459307,
      "grad_norm": 0.25519558787345886,
      "kl": 0.2421875,
      "learning_rate": 5.726752448122975e-07,
      "loss": 0.0424,
      "reward": 1.824999988079071,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2143
    },
    {
      "completion_length": 154.35000228881836,
      "epoch": 2.223950233281493,
      "grad_norm": 0.29643017053604126,
      "kl": 0.246826171875,
      "learning_rate": 5.712524103908704e-07,
      "loss": 0.0445,
      "reward": 1.7750000059604645,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2144
    },
    {
      "completion_length": 243.32500839233398,
      "epoch": 2.2249870399170555,
      "grad_norm": 0.3554222881793976,
      "kl": 0.3359375,
      "learning_rate": 5.698309298576579e-07,
      "loss": 0.1407,
      "reward": 1.2500000149011612,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.3750000074505806,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2145
    },
    {
      "completion_length": 315.47500228881836,
      "epoch": 2.226023846552618,
      "grad_norm": 0.3724534511566162,
      "kl": 0.41796875,
      "learning_rate": 5.684108052848329e-07,
      "loss": 0.1917,
      "reward": 1.4500000178813934,
      "reward_std": 0.39302410185337067,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2146
    },
    {
      "completion_length": 169.40000343322754,
      "epoch": 2.2270606531881803,
      "grad_norm": 0.4188615083694458,
      "kl": 0.215087890625,
      "learning_rate": 5.669920387425926e-07,
      "loss": 0.1191,
      "reward": 1.725000023841858,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.7750000096857548,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2147
    },
    {
      "completion_length": 360.6000099182129,
      "epoch": 2.228097459823743,
      "grad_norm": 0.3144167363643646,
      "kl": 0.37255859375,
      "learning_rate": 5.655746322991531e-07,
      "loss": 0.2439,
      "reward": 1.4750000089406967,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2148
    },
    {
      "completion_length": 264.50000762939453,
      "epoch": 2.2291342664593055,
      "grad_norm": 0.6155608892440796,
      "kl": 0.241943359375,
      "learning_rate": 5.641585880207483e-07,
      "loss": 0.1447,
      "reward": 1.7000000178813934,
      "reward_std": 0.3354102075099945,
      "rewards/accuracy_reward": 0.7749999985098839,
      "rewards/format_reward": 0.925000011920929,
      "step": 2149
    },
    {
      "completion_length": 272.17500495910645,
      "epoch": 2.230171073094868,
      "grad_norm": 0.8062008023262024,
      "kl": 0.348876953125,
      "learning_rate": 5.627439079716266e-07,
      "loss": 0.1833,
      "reward": 1.0750000327825546,
      "reward_std": 0.34075291082262993,
      "rewards/accuracy_reward": 0.17500000447034836,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2150
    },
    {
      "completion_length": 250.22500038146973,
      "epoch": 2.23120787973043,
      "grad_norm": 0.9186306595802307,
      "kl": 0.291015625,
      "learning_rate": 5.613305942140487e-07,
      "loss": 0.1751,
      "reward": 1.8750000298023224,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.9500000178813934,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2151
    },
    {
      "completion_length": 255.07500648498535,
      "epoch": 2.2322446863659926,
      "grad_norm": 1.2115280628204346,
      "kl": 0.3232421875,
      "learning_rate": 5.599186488082806e-07,
      "loss": 0.1282,
      "reward": 1.7750000357627869,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2152
    },
    {
      "completion_length": 204.32500457763672,
      "epoch": 2.233281493001555,
      "grad_norm": 0.9407764077186584,
      "kl": 0.298095703125,
      "learning_rate": 5.585080738125969e-07,
      "loss": 0.0997,
      "reward": 1.625,
      "reward_std": 0.27228760346770287,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2153
    },
    {
      "completion_length": 160.20000076293945,
      "epoch": 2.234318299637118,
      "grad_norm": 2.260845422744751,
      "kl": 0.85009765625,
      "learning_rate": 5.570988712832726e-07,
      "loss": 0.1621,
      "reward": 1.800000011920929,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.8250000327825546,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2154
    },
    {
      "completion_length": 142.3750057220459,
      "epoch": 2.23535510627268,
      "grad_norm": 0.8450815081596375,
      "kl": 0.265380859375,
      "learning_rate": 5.556910432745824e-07,
      "loss": 0.0231,
      "reward": 1.574999988079071,
      "reward_std": 0.18026871606707573,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2155
    },
    {
      "completion_length": 175.40000534057617,
      "epoch": 2.2363919129082426,
      "grad_norm": 80.914306640625,
      "kl": 5.95654296875,
      "learning_rate": 5.542845918387969e-07,
      "loss": 0.5502,
      "reward": 1.6000000536441803,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2156
    },
    {
      "completion_length": 299.80001068115234,
      "epoch": 2.237428719543805,
      "grad_norm": 29383.482421875,
      "kl": 1046.375,
      "learning_rate": 5.528795190261812e-07,
      "loss": 82.4977,
      "reward": 1.7750000655651093,
      "reward_std": 0.4472135901451111,
      "rewards/accuracy_reward": 0.8750000298023224,
      "rewards/format_reward": 0.9000000357627869,
      "step": 2157
    },
    {
      "completion_length": 151.475004196167,
      "epoch": 2.2384655261793673,
      "grad_norm": 1.0604592561721802,
      "kl": 0.28662109375,
      "learning_rate": 5.514758268849896e-07,
      "loss": 0.0816,
      "reward": 1.7000000178813934,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2158
    },
    {
      "completion_length": 156.125,
      "epoch": 2.23950233281493,
      "grad_norm": 1.9279935359954834,
      "kl": 0.68408203125,
      "learning_rate": 5.500735174614638e-07,
      "loss": 0.0971,
      "reward": 1.675000011920929,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2159
    },
    {
      "completion_length": 191.30000114440918,
      "epoch": 2.2405391394504925,
      "grad_norm": 10.897223472595215,
      "kl": 1.947265625,
      "learning_rate": 5.486725927998296e-07,
      "loss": 0.1895,
      "reward": 1.800000011920929,
      "reward_std": 0.2920721024274826,
      "rewards/accuracy_reward": 0.8499999940395355,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2160
    },
    {
      "completion_length": 253.7500057220459,
      "epoch": 2.241575946086055,
      "grad_norm": 28.024011611938477,
      "kl": 4.56396484375,
      "learning_rate": 5.472730549422964e-07,
      "loss": 0.4733,
      "reward": 1.5500000268220901,
      "reward_std": 0.25609343871474266,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2161
    },
    {
      "completion_length": 113.0250015258789,
      "epoch": 2.2426127527216173,
      "grad_norm": 0.38272354006767273,
      "kl": 0.232177734375,
      "learning_rate": 5.45874905929048e-07,
      "loss": -0.0174,
      "reward": 1.7750000059604645,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 1.0,
      "step": 2162
    },
    {
      "completion_length": 255.92500114440918,
      "epoch": 2.2436495593571797,
      "grad_norm": 1.8705600500106812,
      "kl": 0.9794921875,
      "learning_rate": 5.444781477982471e-07,
      "loss": 0.1794,
      "reward": 1.675000011920929,
      "reward_std": 0.4290027730166912,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2163
    },
    {
      "completion_length": 156.6750030517578,
      "epoch": 2.2446863659927425,
      "grad_norm": 1.0310665369033813,
      "kl": 0.486572265625,
      "learning_rate": 5.430827825860273e-07,
      "loss": 0.0747,
      "reward": 1.675000011920929,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2164
    },
    {
      "completion_length": 257.9500102996826,
      "epoch": 2.245723172628305,
      "grad_norm": 18.066238403320312,
      "kl": 2.431884765625,
      "learning_rate": 5.416888123264915e-07,
      "loss": 0.3295,
      "reward": 1.6500000357627869,
      "reward_std": 0.4038754850625992,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2165
    },
    {
      "completion_length": 288.55000495910645,
      "epoch": 2.2467599792638673,
      "grad_norm": 32.37207794189453,
      "kl": 5.1201171875,
      "learning_rate": 5.402962390517095e-07,
      "loss": 0.5701,
      "reward": 1.4500000178813934,
      "reward_std": 0.4525563009083271,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2166
    },
    {
      "completion_length": 170.42500686645508,
      "epoch": 2.2477967858994297,
      "grad_norm": 2.758312940597534,
      "kl": 0.92724609375,
      "learning_rate": 5.389050647917151e-07,
      "loss": 0.1456,
      "reward": 1.4250000417232513,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2167
    },
    {
      "completion_length": 192.54999923706055,
      "epoch": 2.248833592534992,
      "grad_norm": 3.5073254108428955,
      "kl": 2.1776123046875,
      "learning_rate": 5.37515291574502e-07,
      "loss": 0.2531,
      "reward": 1.6500000357627869,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.725000012665987,
      "rewards/format_reward": 0.925000011920929,
      "step": 2168
    },
    {
      "completion_length": 173.8750057220459,
      "epoch": 2.249870399170555,
      "grad_norm": 0.948030412197113,
      "kl": 1.024658203125,
      "learning_rate": 5.361269214260215e-07,
      "loss": 0.0939,
      "reward": 1.4000000059604645,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 2169
    },
    {
      "completion_length": 195.95000076293945,
      "epoch": 2.2509072058061173,
      "grad_norm": 6.445439338684082,
      "kl": 2.384033203125,
      "learning_rate": 5.347399563701796e-07,
      "loss": 0.2534,
      "reward": 1.6500000059604645,
      "reward_std": 0.10458250343799591,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2170
    },
    {
      "completion_length": 218.9250030517578,
      "epoch": 2.2519440124416796,
      "grad_norm": 5.6610941886901855,
      "kl": 2.720703125,
      "learning_rate": 5.333543984288353e-07,
      "loss": 0.2827,
      "reward": 1.5750000178813934,
      "reward_std": 0.28485123068094254,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 2171
    },
    {
      "completion_length": 209.37500381469727,
      "epoch": 2.252980819077242,
      "grad_norm": 1.2526971101760864,
      "kl": 0.732421875,
      "learning_rate": 5.319702496217936e-07,
      "loss": 0.1196,
      "reward": 1.8250000178813934,
      "reward_std": 0.20019173622131348,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2172
    },
    {
      "completion_length": 294.8250026702881,
      "epoch": 2.2540176257128044,
      "grad_norm": 6.543614387512207,
      "kl": 1.595458984375,
      "learning_rate": 5.305875119668083e-07,
      "loss": 0.3047,
      "reward": 1.7500000298023224,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2173
    },
    {
      "completion_length": 254.20000457763672,
      "epoch": 2.255054432348367,
      "grad_norm": 1.375815749168396,
      "kl": 0.976318359375,
      "learning_rate": 5.292061874795742e-07,
      "loss": 0.1873,
      "reward": 1.675000011920929,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.925000011920929,
      "step": 2174
    },
    {
      "completion_length": 288.5249996185303,
      "epoch": 2.2560912389839296,
      "grad_norm": 5.173381328582764,
      "kl": 1.3837890625,
      "learning_rate": 5.278262781737269e-07,
      "loss": 0.2962,
      "reward": 1.7000000476837158,
      "reward_std": 0.3913118839263916,
      "rewards/accuracy_reward": 0.8000000193715096,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2175
    },
    {
      "completion_length": 251.02500343322754,
      "epoch": 2.257128045619492,
      "grad_norm": 3.1237993240356445,
      "kl": 1.4462890625,
      "learning_rate": 5.264477860608378e-07,
      "loss": 0.2814,
      "reward": 1.6000000536441803,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2176
    },
    {
      "completion_length": 290.725004196167,
      "epoch": 2.2581648522550544,
      "grad_norm": 13.25760555267334,
      "kl": 5.87255859375,
      "learning_rate": 5.250707131504149e-07,
      "loss": 0.6291,
      "reward": 1.425000011920929,
      "reward_std": 0.36053744331002235,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2177
    },
    {
      "completion_length": 356.7000045776367,
      "epoch": 2.2592016588906167,
      "grad_norm": 67.65271759033203,
      "kl": 16.390625,
      "learning_rate": 5.236950614498944e-07,
      "loss": 1.4881,
      "reward": 1.6000000536441803,
      "reward_std": 0.5590169802308083,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2178
    },
    {
      "completion_length": 217.39999961853027,
      "epoch": 2.2602384655261796,
      "grad_norm": 12.007269859313965,
      "kl": 3.3828125,
      "learning_rate": 5.223208329646429e-07,
      "loss": 0.2767,
      "reward": 1.5500000417232513,
      "reward_std": 0.45413000136613846,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2179
    },
    {
      "completion_length": 243.05000686645508,
      "epoch": 2.261275272161742,
      "grad_norm": 3.9122519493103027,
      "kl": 2.56591796875,
      "learning_rate": 5.209480296979506e-07,
      "loss": 0.2958,
      "reward": 1.475000023841858,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2180
    },
    {
      "completion_length": 118.94999885559082,
      "epoch": 2.2623120787973043,
      "grad_norm": 0.3715357184410095,
      "kl": 0.23388671875,
      "learning_rate": 5.195766536510324e-07,
      "loss": 0.0032,
      "reward": 1.9000000357627869,
      "reward_std": 0.18026870489120483,
      "rewards/accuracy_reward": 0.9000000059604645,
      "rewards/format_reward": 1.0,
      "step": 2181
    },
    {
      "completion_length": 347.25000953674316,
      "epoch": 2.2633488854328667,
      "grad_norm": 1.2834972143173218,
      "kl": 1.10498046875,
      "learning_rate": 5.182067068230192e-07,
      "loss": 0.231,
      "reward": 1.3250000327825546,
      "reward_std": 0.4415663927793503,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2182
    },
    {
      "completion_length": 150.1750030517578,
      "epoch": 2.264385692068429,
      "grad_norm": 1.8223271369934082,
      "kl": 0.6650390625,
      "learning_rate": 5.168381912109616e-07,
      "loss": 0.0728,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2183
    },
    {
      "completion_length": 294.7500114440918,
      "epoch": 2.265422498703992,
      "grad_norm": 3.047154426574707,
      "kl": 2.095703125,
      "learning_rate": 5.154711088098236e-07,
      "loss": 0.3235,
      "reward": 1.3000000268220901,
      "reward_std": 0.28485123068094254,
      "rewards/accuracy_reward": 0.3999999985098839,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2184
    },
    {
      "completion_length": 147.5250015258789,
      "epoch": 2.2664593053395543,
      "grad_norm": 0.5942182540893555,
      "kl": 1.118408203125,
      "learning_rate": 5.141054616124771e-07,
      "loss": 0.1228,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2185
    },
    {
      "completion_length": 296.425012588501,
      "epoch": 2.2674961119751167,
      "grad_norm": 0.992489218711853,
      "kl": 3.3251953125,
      "learning_rate": 5.127412516097051e-07,
      "loss": 0.4322,
      "reward": 1.3250000178813934,
      "reward_std": 0.30463574081659317,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2186
    },
    {
      "completion_length": 204.00000762939453,
      "epoch": 2.268532918610679,
      "grad_norm": 4.018986701965332,
      "kl": 1.537109375,
      "learning_rate": 5.113784807901937e-07,
      "loss": 0.131,
      "reward": 1.449999988079071,
      "reward_std": 0.3354102075099945,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2187
    },
    {
      "completion_length": 343.50000381469727,
      "epoch": 2.2695697252462415,
      "grad_norm": 1.2946738004684448,
      "kl": 1.6767578125,
      "learning_rate": 5.100171511405316e-07,
      "loss": 0.3447,
      "reward": 1.5500000417232513,
      "reward_std": 0.5408061295747757,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2188
    },
    {
      "completion_length": 270.74999618530273,
      "epoch": 2.2706065318818043,
      "grad_norm": 0.4122595191001892,
      "kl": 0.386962890625,
      "learning_rate": 5.086572646452058e-07,
      "loss": 0.1338,
      "reward": 1.6500000059604645,
      "reward_std": 0.3804604783654213,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.925000011920929,
      "step": 2189
    },
    {
      "completion_length": 211.3750057220459,
      "epoch": 2.2716433385173667,
      "grad_norm": 23.308609008789062,
      "kl": 4.0654296875,
      "learning_rate": 5.072988232866014e-07,
      "loss": 0.4366,
      "reward": 1.600000023841858,
      "reward_std": 0.41657763719558716,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 2190
    },
    {
      "completion_length": 389.350004196167,
      "epoch": 2.272680145152929,
      "grad_norm": 5.1695990562438965,
      "kl": 1.736083984375,
      "learning_rate": 5.059418290449949e-07,
      "loss": 0.3052,
      "reward": 1.3250000178813934,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2191
    },
    {
      "completion_length": 204.5750026702881,
      "epoch": 2.2737169517884914,
      "grad_norm": 4.654054641723633,
      "kl": 1.416015625,
      "learning_rate": 5.045862838985546e-07,
      "loss": 0.2085,
      "reward": 1.6750000417232513,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.7250000182539225,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2192
    },
    {
      "completion_length": 158.4250030517578,
      "epoch": 2.274753758424054,
      "grad_norm": 2.0945351123809814,
      "kl": 0.652099609375,
      "learning_rate": 5.032321898233348e-07,
      "loss": 0.0811,
      "reward": 1.3750000298023224,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2193
    },
    {
      "completion_length": 148.90000534057617,
      "epoch": 2.275790565059616,
      "grad_norm": 1.9934879541397095,
      "kl": 0.64697265625,
      "learning_rate": 5.018795487932769e-07,
      "loss": 0.0919,
      "reward": 1.6250000149011612,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2194
    },
    {
      "completion_length": 241.42500495910645,
      "epoch": 2.276827371695179,
      "grad_norm": 0.4953010678291321,
      "kl": 1.177001953125,
      "learning_rate": 5.005283627802012e-07,
      "loss": 0.1267,
      "reward": 1.6500000208616257,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2195
    },
    {
      "completion_length": 155.32500457763672,
      "epoch": 2.2778641783307414,
      "grad_norm": 1.6908633708953857,
      "kl": 1.57958984375,
      "learning_rate": 4.991786337538095e-07,
      "loss": 0.1748,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2196
    },
    {
      "completion_length": 298.7250003814697,
      "epoch": 2.2789009849663038,
      "grad_norm": 18.517498016357422,
      "kl": 6.825927734375,
      "learning_rate": 4.978303636816784e-07,
      "loss": 0.7049,
      "reward": 1.600000023841858,
      "reward_std": 0.452556312084198,
      "rewards/accuracy_reward": 0.7000000067055225,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2197
    },
    {
      "completion_length": 235.7000026702881,
      "epoch": 2.279937791601866,
      "grad_norm": 10.7755126953125,
      "kl": 4.71875,
      "learning_rate": 4.964835545292579e-07,
      "loss": 0.4763,
      "reward": 1.5750000327825546,
      "reward_std": 0.2920721024274826,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2198
    },
    {
      "completion_length": 102.62500381469727,
      "epoch": 2.2809745982374285,
      "grad_norm": 0.27033478021621704,
      "kl": 0.2291259765625,
      "learning_rate": 4.951382082598676e-07,
      "loss": -0.0118,
      "reward": 1.8250000178813934,
      "reward_std": 0.06846531480550766,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 1.0,
      "step": 2199
    },
    {
      "completion_length": 301.37500190734863,
      "epoch": 2.2820114048729914,
      "grad_norm": 3.6355156898498535,
      "kl": 3.11572265625,
      "learning_rate": 4.937943268346965e-07,
      "loss": 0.3966,
      "reward": 1.6750000417232513,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2200
    },
    {
      "completion_length": 442.5250072479248,
      "epoch": 2.2830482115085537,
      "grad_norm": 1.8923827409744263,
      "kl": 3.740966796875,
      "learning_rate": 4.924519122127965e-07,
      "loss": 0.5951,
      "reward": 1.600000023841858,
      "reward_std": 0.6635995358228683,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.8250000029802322,
      "step": 2201
    },
    {
      "completion_length": 230.4000015258789,
      "epoch": 2.284085018144116,
      "grad_norm": 1.642378568649292,
      "kl": 1.709716796875,
      "learning_rate": 4.911109663510817e-07,
      "loss": 0.2659,
      "reward": 1.5250000208616257,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2202
    },
    {
      "completion_length": 149.45000457763672,
      "epoch": 2.2851218247796785,
      "grad_norm": 1.4732751846313477,
      "kl": 0.551025390625,
      "learning_rate": 4.89771491204325e-07,
      "loss": 0.0793,
      "reward": 1.5750000178813934,
      "reward_std": 0.20019175112247467,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2203
    },
    {
      "completion_length": 140.60000228881836,
      "epoch": 2.286158631415241,
      "grad_norm": 1.0003714561462402,
      "kl": 0.89111328125,
      "learning_rate": 4.884334887251569e-07,
      "loss": 0.1034,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510375499725,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2204
    },
    {
      "completion_length": 244.39999961853027,
      "epoch": 2.2871954380508033,
      "grad_norm": 0.8676134943962097,
      "kl": 2.270263671875,
      "learning_rate": 4.870969608640574e-07,
      "loss": 0.3424,
      "reward": 1.5000000298023224,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 2205
    },
    {
      "completion_length": 289.25000381469727,
      "epoch": 2.288232244686366,
      "grad_norm": 6.302436828613281,
      "kl": 4.60498046875,
      "learning_rate": 4.857619095693612e-07,
      "loss": 0.52,
      "reward": 1.725000023841858,
      "reward_std": 0.2920721024274826,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2206
    },
    {
      "completion_length": 161.85000610351562,
      "epoch": 2.2892690513219285,
      "grad_norm": 0.5147799253463745,
      "kl": 1.262939453125,
      "learning_rate": 4.844283367872476e-07,
      "loss": 0.1388,
      "reward": 1.425000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2207
    },
    {
      "completion_length": 156.875,
      "epoch": 2.290305857957491,
      "grad_norm": 0.9272441267967224,
      "kl": 0.9306640625,
      "learning_rate": 4.83096244461742e-07,
      "loss": 0.1479,
      "reward": 1.7500000298023224,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.7749999985098839,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2208
    },
    {
      "completion_length": 154.52499961853027,
      "epoch": 2.2913426645930532,
      "grad_norm": 0.766848087310791,
      "kl": 0.964111328125,
      "learning_rate": 4.817656345347105e-07,
      "loss": 0.1023,
      "reward": 1.5000000298023224,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2209
    },
    {
      "completion_length": 255.2750072479248,
      "epoch": 2.2923794712286156,
      "grad_norm": 0.37823888659477234,
      "kl": 2.388427734375,
      "learning_rate": 4.804365089458598e-07,
      "loss": 0.1609,
      "reward": 1.550000011920929,
      "reward_std": 0.31719937175512314,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 2210
    },
    {
      "completion_length": 143.92500495910645,
      "epoch": 2.2934162778641785,
      "grad_norm": 0.4562288224697113,
      "kl": 0.341064453125,
      "learning_rate": 4.791088696327317e-07,
      "loss": 0.0147,
      "reward": 1.800000011920929,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2211
    },
    {
      "completion_length": 252.80000114440918,
      "epoch": 2.294453084499741,
      "grad_norm": 0.5665162205696106,
      "kl": 2.70654296875,
      "learning_rate": 4.777827185307013e-07,
      "loss": 0.3299,
      "reward": 1.5250000357627869,
      "reward_std": 0.31199517101049423,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2212
    },
    {
      "completion_length": 155.8499984741211,
      "epoch": 2.295489891135303,
      "grad_norm": 0.3933873474597931,
      "kl": 0.3460693359375,
      "learning_rate": 4.76458057572974e-07,
      "loss": 0.0313,
      "reward": 1.475000023841858,
      "reward_std": 0.2812207192182541,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2213
    },
    {
      "completion_length": 248.6749973297119,
      "epoch": 2.2965266977708656,
      "grad_norm": 1.432715892791748,
      "kl": 3.380859375,
      "learning_rate": 4.751348886905852e-07,
      "loss": 0.4316,
      "reward": 1.600000038743019,
      "reward_std": 0.335410188883543,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2214
    },
    {
      "completion_length": 291.7000045776367,
      "epoch": 2.297563504406428,
      "grad_norm": 0.5518354177474976,
      "kl": 1.302490234375,
      "learning_rate": 4.7381321381239125e-07,
      "loss": 0.2535,
      "reward": 1.600000038743019,
      "reward_std": 0.27950847521424294,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2215
    },
    {
      "completion_length": 114.32500076293945,
      "epoch": 2.298600311041991,
      "grad_norm": 0.2603670060634613,
      "kl": 0.181884765625,
      "learning_rate": 4.7249303486507444e-07,
      "loss": -0.0001,
      "reward": 1.5999999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 1.0,
      "step": 2216
    },
    {
      "completion_length": 291.5249996185303,
      "epoch": 2.299637117677553,
      "grad_norm": 1.783521294593811,
      "kl": 2.10302734375,
      "learning_rate": 4.7117435377313437e-07,
      "loss": 0.3014,
      "reward": 1.6000000536441803,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2217
    },
    {
      "completion_length": 344.7249984741211,
      "epoch": 2.3006739243131156,
      "grad_norm": 6.913201808929443,
      "kl": 2.95703125,
      "learning_rate": 4.6985717245888736e-07,
      "loss": 0.4761,
      "reward": 1.550000011920929,
      "reward_std": 0.4472135826945305,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2218
    },
    {
      "completion_length": 207.9499969482422,
      "epoch": 2.301710730948678,
      "grad_norm": 0.7001081705093384,
      "kl": 1.15478515625,
      "learning_rate": 4.685414928424631e-07,
      "loss": 0.1924,
      "reward": 1.375,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.42499999701976776,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2219
    },
    {
      "completion_length": 399.3500061035156,
      "epoch": 2.3027475375842403,
      "grad_norm": 4.699303150177002,
      "kl": 1.8623046875,
      "learning_rate": 4.6722731684180356e-07,
      "loss": 0.3385,
      "reward": 1.4750000089406967,
      "reward_std": 0.4604101777076721,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2220
    },
    {
      "completion_length": 207.2750015258789,
      "epoch": 2.303784344219803,
      "grad_norm": 2.3309645652770996,
      "kl": 1.44384765625,
      "learning_rate": 4.6591464637265753e-07,
      "loss": 0.1782,
      "reward": 1.4000000059604645,
      "reward_std": 0.40387552604079247,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2221
    },
    {
      "completion_length": 143.5500030517578,
      "epoch": 2.3048211508553655,
      "grad_norm": 0.2181098759174347,
      "kl": 1.136474609375,
      "learning_rate": 4.646034833485789e-07,
      "loss": 0.1306,
      "reward": 1.5999999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2222
    },
    {
      "completion_length": 87.70000267028809,
      "epoch": 2.305857957490928,
      "grad_norm": 0.42760446667671204,
      "kl": 0.263671875,
      "learning_rate": 4.632938296809256e-07,
      "loss": 0.0274,
      "reward": 1.4000000059604645,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 1.0,
      "step": 2223
    },
    {
      "completion_length": 300.95000648498535,
      "epoch": 2.3068947641264903,
      "grad_norm": 2.922344207763672,
      "kl": 3.10693359375,
      "learning_rate": 4.6198568727885383e-07,
      "loss": 0.2676,
      "reward": 1.725000023841858,
      "reward_std": 0.3413858935236931,
      "rewards/accuracy_reward": 0.8250000104308128,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2224
    },
    {
      "completion_length": 200.40000915527344,
      "epoch": 2.3079315707620527,
      "grad_norm": 4.799491882324219,
      "kl": 3.469482421875,
      "learning_rate": 4.606790580493173e-07,
      "loss": 0.3815,
      "reward": 1.600000023841858,
      "reward_std": 0.32818929105997086,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2225
    },
    {
      "completion_length": 170.72500610351562,
      "epoch": 2.3089683773976155,
      "grad_norm": 1.7546030282974243,
      "kl": 1.6416015625,
      "learning_rate": 4.5937394389706313e-07,
      "loss": 0.1215,
      "reward": 1.600000023841858,
      "reward_std": 0.37310105934739113,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2226
    },
    {
      "completion_length": 101.1500015258789,
      "epoch": 2.310005184033178,
      "grad_norm": 0.0402827262878418,
      "kl": 0.22412109375,
      "learning_rate": 4.58070346724632e-07,
      "loss": 0.0086,
      "reward": 1.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 1.0,
      "step": 2227
    },
    {
      "completion_length": 311.70000076293945,
      "epoch": 2.3110419906687403,
      "grad_norm": 6.426229000091553,
      "kl": 3.618896484375,
      "learning_rate": 4.5676826843234986e-07,
      "loss": 0.3734,
      "reward": 1.75,
      "reward_std": 0.2133883535861969,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2228
    },
    {
      "completion_length": 241.65000534057617,
      "epoch": 2.3120787973043027,
      "grad_norm": 0.8814529180526733,
      "kl": 2.28662109375,
      "learning_rate": 4.5546771091833173e-07,
      "loss": 0.2873,
      "reward": 1.625,
      "reward_std": 0.27228758856654167,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 2229
    },
    {
      "completion_length": 150.10000228881836,
      "epoch": 2.313115603939865,
      "grad_norm": 2.1378538608551025,
      "kl": 0.57080078125,
      "learning_rate": 4.5416867607847386e-07,
      "loss": 0.0879,
      "reward": 1.5999999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2230
    },
    {
      "completion_length": 389.7000102996826,
      "epoch": 2.314152410575428,
      "grad_norm": 2.657593250274658,
      "kl": 1.59912109375,
      "learning_rate": 4.528711658064531e-07,
      "loss": 0.3534,
      "reward": 1.5250000357627869,
      "reward_std": 0.4651199132204056,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2231
    },
    {
      "completion_length": 310.6000061035156,
      "epoch": 2.3151892172109902,
      "grad_norm": 1.1992459297180176,
      "kl": 0.96826171875,
      "learning_rate": 4.515751819937238e-07,
      "loss": 0.1808,
      "reward": 1.3250000029802322,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2232
    },
    {
      "completion_length": 197.17500114440918,
      "epoch": 2.3162260238465526,
      "grad_norm": 0.4076596796512604,
      "kl": 0.44775390625,
      "learning_rate": 4.5028072652951594e-07,
      "loss": 0.0825,
      "reward": 1.4500000178813934,
      "reward_std": 0.18026872351765633,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2233
    },
    {
      "completion_length": 198.42499923706055,
      "epoch": 2.317262830482115,
      "grad_norm": 1.838197946548462,
      "kl": 0.779296875,
      "learning_rate": 4.489878013008309e-07,
      "loss": 0.0977,
      "reward": 1.625,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2234
    },
    {
      "completion_length": 244.30000686645508,
      "epoch": 2.3182996371176774,
      "grad_norm": 4.243485450744629,
      "kl": 0.98095703125,
      "learning_rate": 4.476964081924388e-07,
      "loss": 0.2029,
      "reward": 1.350000038743019,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2235
    },
    {
      "completion_length": 253.32500648498535,
      "epoch": 2.31933644375324,
      "grad_norm": 4.615521430969238,
      "kl": 1.64013671875,
      "learning_rate": 4.464065490868768e-07,
      "loss": 0.2729,
      "reward": 1.4249999821186066,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2236
    },
    {
      "completion_length": 306.84999656677246,
      "epoch": 2.3203732503888026,
      "grad_norm": 2.5953807830810547,
      "kl": 1.31103515625,
      "learning_rate": 4.451182258644473e-07,
      "loss": 0.2924,
      "reward": 1.375,
      "reward_std": 0.3533165268599987,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2237
    },
    {
      "completion_length": 243.95000076293945,
      "epoch": 2.321410057024365,
      "grad_norm": 0.9265846014022827,
      "kl": 2.1995849609375,
      "learning_rate": 4.4383144040321035e-07,
      "loss": 0.2851,
      "reward": 1.5250000059604645,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2238
    },
    {
      "completion_length": 201.92500114440918,
      "epoch": 2.3224468636599274,
      "grad_norm": 1.1252025365829468,
      "kl": 1.3828125,
      "learning_rate": 4.4254619457898774e-07,
      "loss": 0.2124,
      "reward": 1.5250000208616257,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2239
    },
    {
      "completion_length": 132.82500076293945,
      "epoch": 2.3234836702954897,
      "grad_norm": 3.8554553985595703,
      "kl": 1.8841552734375,
      "learning_rate": 4.412624902653552e-07,
      "loss": 0.1759,
      "reward": 1.5250000059604645,
      "reward_std": 0.10458250343799591,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2240
    },
    {
      "completion_length": 106.5250015258789,
      "epoch": 2.3245204769310526,
      "grad_norm": 0.056571315973997116,
      "kl": 0.28369140625,
      "learning_rate": 4.399803293336412e-07,
      "loss": 0.0105,
      "reward": 1.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 1.0,
      "step": 2241
    },
    {
      "completion_length": 163.9749984741211,
      "epoch": 2.325557283566615,
      "grad_norm": 2.1101109981536865,
      "kl": 1.6240234375,
      "learning_rate": 4.3869971365292475e-07,
      "loss": 0.203,
      "reward": 1.449999988079071,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2242
    },
    {
      "completion_length": 195.7249984741211,
      "epoch": 2.3265940902021773,
      "grad_norm": 1.644981861114502,
      "kl": 1.5142822265625,
      "learning_rate": 4.3742064509003263e-07,
      "loss": 0.1858,
      "reward": 1.625,
      "reward_std": 0.1604841947555542,
      "rewards/accuracy_reward": 0.6750000007450581,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2243
    },
    {
      "completion_length": 148.6750030517578,
      "epoch": 2.3276308968377397,
      "grad_norm": 0.8315623998641968,
      "kl": 1.2890625,
      "learning_rate": 4.3614312550953573e-07,
      "loss": 0.1348,
      "reward": 1.675000011920929,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2244
    },
    {
      "completion_length": 237.5250072479248,
      "epoch": 2.328667703473302,
      "grad_norm": 0.659866213798523,
      "kl": 1.24560546875,
      "learning_rate": 4.348671567737468e-07,
      "loss": 0.2205,
      "reward": 1.7250000536441803,
      "reward_std": 0.3354102149605751,
      "rewards/accuracy_reward": 0.8000000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2245
    },
    {
      "completion_length": 266.77500343322754,
      "epoch": 2.329704510108865,
      "grad_norm": 2.9652602672576904,
      "kl": 1.91015625,
      "learning_rate": 4.3359274074271776e-07,
      "loss": 0.298,
      "reward": 1.4250000268220901,
      "reward_std": 0.4723408557474613,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2246
    },
    {
      "completion_length": 175.87500190734863,
      "epoch": 2.3307413167444273,
      "grad_norm": 2.8715803623199463,
      "kl": 1.182861328125,
      "learning_rate": 4.3231987927423876e-07,
      "loss": 0.1985,
      "reward": 1.6750000417232513,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2247
    },
    {
      "completion_length": 194.37499809265137,
      "epoch": 2.3317781233799897,
      "grad_norm": 0.7718262672424316,
      "kl": 1.14697265625,
      "learning_rate": 4.310485742238307e-07,
      "loss": 0.1689,
      "reward": 1.2000000029802322,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.25,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2248
    },
    {
      "completion_length": 93.67500114440918,
      "epoch": 2.332814930015552,
      "grad_norm": 0.24666506052017212,
      "kl": 0.2412109375,
      "learning_rate": 4.297788274447484e-07,
      "loss": 0.028,
      "reward": 1.7249999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 1.0,
      "step": 2249
    },
    {
      "completion_length": 107.15000343322754,
      "epoch": 2.3338517366511145,
      "grad_norm": 0.5002438426017761,
      "kl": 0.46240234375,
      "learning_rate": 4.285106407879737e-07,
      "loss": 0.0162,
      "reward": 1.725000023841858,
      "reward_std": 0.27950847521424294,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2250
    },
    {
      "completion_length": 131.50000190734863,
      "epoch": 2.3348885432866773,
      "grad_norm": 0.851094663143158,
      "kl": 0.8470458984375,
      "learning_rate": 4.272440161022143e-07,
      "loss": 0.1095,
      "reward": 1.725000023841858,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.7500000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2251
    },
    {
      "completion_length": 292.35000228881836,
      "epoch": 2.3359253499222397,
      "grad_norm": 0.4606613218784332,
      "kl": 3.3369140625,
      "learning_rate": 4.25978955233901e-07,
      "loss": 0.2805,
      "reward": 1.5999999940395355,
      "reward_std": 0.24873405322432518,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2252
    },
    {
      "completion_length": 153.07500076293945,
      "epoch": 2.336962156557802,
      "grad_norm": 0.5345311164855957,
      "kl": 0.429931640625,
      "learning_rate": 4.2471546002718547e-07,
      "loss": 0.0664,
      "reward": 1.475000023841858,
      "reward_std": 0.1928323358297348,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2253
    },
    {
      "completion_length": 251.87500381469727,
      "epoch": 2.3379989631933644,
      "grad_norm": 2.472031831741333,
      "kl": 2.205322265625,
      "learning_rate": 4.2345353232393637e-07,
      "loss": 0.2868,
      "reward": 1.675000011920929,
      "reward_std": 0.32818930596113205,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 2254
    },
    {
      "completion_length": 153.0999984741211,
      "epoch": 2.339035769828927,
      "grad_norm": 1.701666235923767,
      "kl": 1.35400390625,
      "learning_rate": 4.221931739637376e-07,
      "loss": 0.1172,
      "reward": 1.600000023841858,
      "reward_std": 0.17304782569408417,
      "rewards/accuracy_reward": 0.6250000037252903,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2255
    },
    {
      "completion_length": 253.05000495910645,
      "epoch": 2.340072576464489,
      "grad_norm": 2.656787157058716,
      "kl": 2.70263671875,
      "learning_rate": 4.209343867838846e-07,
      "loss": 0.3486,
      "reward": 1.4000000357627869,
      "reward_std": 0.571580596268177,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2256
    },
    {
      "completion_length": 335.25000381469727,
      "epoch": 2.341109383100052,
      "grad_norm": 1.389851689338684,
      "kl": 3.82275390625,
      "learning_rate": 4.1967717261938466e-07,
      "loss": 0.3971,
      "reward": 1.5250000059604645,
      "reward_std": 0.528242539614439,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2257
    },
    {
      "completion_length": 155.27499961853027,
      "epoch": 2.3421461897356144,
      "grad_norm": 0.6044714450836182,
      "kl": 1.2886962890625,
      "learning_rate": 4.184215333029485e-07,
      "loss": 0.1093,
      "reward": 1.5500000417232513,
      "reward_std": 0.3479738235473633,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 2258
    },
    {
      "completion_length": 219.55000114440918,
      "epoch": 2.3431829963711768,
      "grad_norm": 2.261915922164917,
      "kl": 1.72265625,
      "learning_rate": 4.1716747066499424e-07,
      "loss": 0.1899,
      "reward": 1.7000000029802322,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2259
    },
    {
      "completion_length": 145.65000534057617,
      "epoch": 2.344219803006739,
      "grad_norm": 1.1528451442718506,
      "kl": 0.74462890625,
      "learning_rate": 4.159149865336401e-07,
      "loss": 0.0774,
      "reward": 1.5000000298023224,
      "reward_std": 0.2920721247792244,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2260
    },
    {
      "completion_length": 170.57500457763672,
      "epoch": 2.3452566096423015,
      "grad_norm": 2.8177783489227295,
      "kl": 1.922119140625,
      "learning_rate": 4.1466408273470316e-07,
      "loss": 0.2329,
      "reward": 1.75,
      "reward_std": 0.1928323470056057,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2261
    },
    {
      "completion_length": 107.97500038146973,
      "epoch": 2.3462934162778644,
      "grad_norm": 0.042099762707948685,
      "kl": 0.244384765625,
      "learning_rate": 4.134147610916971e-07,
      "loss": 0.0093,
      "reward": 1.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 1.0,
      "step": 2262
    },
    {
      "completion_length": 214.44999885559082,
      "epoch": 2.3473302229134267,
      "grad_norm": 1.4778048992156982,
      "kl": 1.6986083984375,
      "learning_rate": 4.1216702342582974e-07,
      "loss": 0.2535,
      "reward": 1.6500000357627869,
      "reward_std": 0.3570454269647598,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2263
    },
    {
      "completion_length": 202.80000686645508,
      "epoch": 2.348367029548989,
      "grad_norm": 1.5482327938079834,
      "kl": 1.154052734375,
      "learning_rate": 4.1092087155599906e-07,
      "loss": 0.1527,
      "reward": 1.625,
      "reward_std": 0.36053743213415146,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2264
    },
    {
      "completion_length": 204.44999885559082,
      "epoch": 2.3494038361845515,
      "grad_norm": 1.1103105545043945,
      "kl": 1.162353515625,
      "learning_rate": 4.0967630729879076e-07,
      "loss": 0.1286,
      "reward": 1.5750000178813934,
      "reward_std": 0.2920721136033535,
      "rewards/accuracy_reward": 0.6250000111758709,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2265
    },
    {
      "completion_length": 260.9750003814697,
      "epoch": 2.350440642820114,
      "grad_norm": 1.8445768356323242,
      "kl": 2.577392578125,
      "learning_rate": 4.084333324684782e-07,
      "loss": 0.3042,
      "reward": 1.5999999940395355,
      "reward_std": 0.18090170249342918,
      "rewards/accuracy_reward": 0.6750000007450581,
      "rewards/format_reward": 0.925000011920929,
      "step": 2266
    },
    {
      "completion_length": 238.15000343322754,
      "epoch": 2.3514774494556763,
      "grad_norm": 1.9114346504211426,
      "kl": 1.80224609375,
      "learning_rate": 4.07191948877016e-07,
      "loss": 0.2691,
      "reward": 1.475000038743019,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.925000011920929,
      "step": 2267
    },
    {
      "completion_length": 266.75,
      "epoch": 2.352514256091239,
      "grad_norm": 5.558021068572998,
      "kl": 4.21142578125,
      "learning_rate": 4.059521583340396e-07,
      "loss": 0.4675,
      "reward": 1.4749999940395355,
      "reward_std": 0.40921826660633087,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2268
    },
    {
      "completion_length": 135.9250030517578,
      "epoch": 2.3535510627268015,
      "grad_norm": 0.6464206576347351,
      "kl": 0.9725341796875,
      "learning_rate": 4.047139626468618e-07,
      "loss": 0.0995,
      "reward": 1.9500000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.9750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2269
    },
    {
      "completion_length": 243.55000114440918,
      "epoch": 2.354587869362364,
      "grad_norm": 1.67167067527771,
      "kl": 2.059814453125,
      "learning_rate": 4.034773636204722e-07,
      "loss": 0.2732,
      "reward": 1.2750000059604645,
      "reward_std": 0.2974148541688919,
      "rewards/accuracy_reward": 0.3499999977648258,
      "rewards/format_reward": 0.925000011920929,
      "step": 2270
    },
    {
      "completion_length": 136.87500190734863,
      "epoch": 2.3556246759979262,
      "grad_norm": 0.7818654775619507,
      "kl": 1.02197265625,
      "learning_rate": 4.022423630575299e-07,
      "loss": 0.1335,
      "reward": 1.5500000268220901,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2271
    },
    {
      "completion_length": 191.72500610351562,
      "epoch": 2.3566614826334886,
      "grad_norm": 1.5709656476974487,
      "kl": 1.54736328125,
      "learning_rate": 4.01008962758367e-07,
      "loss": 0.0451,
      "reward": 1.75,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.8000000007450581,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2272
    },
    {
      "completion_length": 213.50000190734863,
      "epoch": 2.3576982892690515,
      "grad_norm": 3.527369499206543,
      "kl": 1.7890625,
      "learning_rate": 3.997771645209804e-07,
      "loss": 0.1347,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.8500000238418579,
      "rewards/format_reward": 0.925000011920929,
      "step": 2273
    },
    {
      "completion_length": 227.90000343322754,
      "epoch": 2.358735095904614,
      "grad_norm": 8.278703689575195,
      "kl": 2.4296875,
      "learning_rate": 3.9854697014103284e-07,
      "loss": 0.2935,
      "reward": 1.2750000357627869,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2274
    },
    {
      "completion_length": 254.1250057220459,
      "epoch": 2.359771902540176,
      "grad_norm": 1.2080625295639038,
      "kl": 1.75830078125,
      "learning_rate": 3.973183814118482e-07,
      "loss": 0.1996,
      "reward": 1.6500000357627869,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2275
    },
    {
      "completion_length": 271.8000087738037,
      "epoch": 2.3608087091757386,
      "grad_norm": 2.1522293090820312,
      "kl": 2.3916015625,
      "learning_rate": 3.9609140012441087e-07,
      "loss": 0.3868,
      "reward": 1.1750000417232513,
      "reward_std": 0.3479738272726536,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2276
    },
    {
      "completion_length": 345.5500030517578,
      "epoch": 2.361845515811301,
      "grad_norm": 4.218634128570557,
      "kl": 3.22265625,
      "learning_rate": 3.948660280673612e-07,
      "loss": 0.4648,
      "reward": 1.225000038743019,
      "reward_std": 0.5282425321638584,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2277
    },
    {
      "completion_length": 157.4500026702881,
      "epoch": 2.362882322446864,
      "grad_norm": 0.4688565135002136,
      "kl": 1.18994140625,
      "learning_rate": 3.9364226702699334e-07,
      "loss": 0.1205,
      "reward": 1.675000011920929,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2278
    },
    {
      "completion_length": 206.52500343322754,
      "epoch": 2.363919129082426,
      "grad_norm": 2.2750909328460693,
      "kl": 2.321533203125,
      "learning_rate": 3.924201187872534e-07,
      "loss": 0.2404,
      "reward": 1.3250000029802322,
      "reward_std": 0.20019177719950676,
      "rewards/accuracy_reward": 0.37500000186264515,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2279
    },
    {
      "completion_length": 273.4249954223633,
      "epoch": 2.3649559357179886,
      "grad_norm": 3.191754102706909,
      "kl": 3.87841796875,
      "learning_rate": 3.9119958512973743e-07,
      "loss": 0.2735,
      "reward": 1.425000011920929,
      "reward_std": 0.3533165343105793,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2280
    },
    {
      "completion_length": 204.22500038146973,
      "epoch": 2.365992742353551,
      "grad_norm": 1.4717326164245605,
      "kl": 2.7279052734375,
      "learning_rate": 3.8998066783368507e-07,
      "loss": 0.2241,
      "reward": 1.8000000417232513,
      "reward_std": 0.4038754999637604,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 2281
    },
    {
      "completion_length": 199.57499885559082,
      "epoch": 2.3670295489891133,
      "grad_norm": 0.78553706407547,
      "kl": 1.5052490234375,
      "learning_rate": 3.8876336867598304e-07,
      "loss": 0.0066,
      "reward": 1.7750000059604645,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2282
    },
    {
      "completion_length": 223.57501029968262,
      "epoch": 2.368066355624676,
      "grad_norm": 1.0880004167556763,
      "kl": 1.29931640625,
      "learning_rate": 3.875476894311567e-07,
      "loss": 0.213,
      "reward": 1.675000011920929,
      "reward_std": 0.21638591587543488,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2283
    },
    {
      "completion_length": 334.5500011444092,
      "epoch": 2.3691031622602385,
      "grad_norm": 1.3197311162948608,
      "kl": 3.641357421875,
      "learning_rate": 3.8633363187137126e-07,
      "loss": 0.4619,
      "reward": 1.5500000417232513,
      "reward_std": 0.4729738309979439,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2284
    },
    {
      "completion_length": 399.67501068115234,
      "epoch": 2.370139968895801,
      "grad_norm": 1.812806248664856,
      "kl": 5.3671875,
      "learning_rate": 3.851211977664269e-07,
      "loss": 0.7023,
      "reward": 1.2500000149011612,
      "reward_std": 0.5210216604173183,
      "rewards/accuracy_reward": 0.4249999988824129,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2285
    },
    {
      "completion_length": 408.67501068115234,
      "epoch": 2.3711767755313633,
      "grad_norm": 2.6503489017486572,
      "kl": 3.47265625,
      "learning_rate": 3.839103888837588e-07,
      "loss": 0.4759,
      "reward": 1.3250000178813934,
      "reward_std": 0.5282425470650196,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2286
    },
    {
      "completion_length": 162.6500072479248,
      "epoch": 2.3722135821669257,
      "grad_norm": 1.2067577838897705,
      "kl": 0.76025390625,
      "learning_rate": 3.827012069884312e-07,
      "loss": 0.1154,
      "reward": 1.800000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2287
    },
    {
      "completion_length": 211.45000457763672,
      "epoch": 2.3732503888024885,
      "grad_norm": 1.447800874710083,
      "kl": 1.662353515625,
      "learning_rate": 3.814936538431378e-07,
      "loss": 0.2414,
      "reward": 1.75,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 2288
    },
    {
      "completion_length": 258.87500381469727,
      "epoch": 2.374287195438051,
      "grad_norm": 0.9774056673049927,
      "kl": 2.126953125,
      "learning_rate": 3.8028773120819694e-07,
      "loss": 0.3161,
      "reward": 1.2000000178813934,
      "reward_std": 0.40387552231550217,
      "rewards/accuracy_reward": 0.3000000063329935,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2289
    },
    {
      "completion_length": 288.7750072479248,
      "epoch": 2.3753240020736133,
      "grad_norm": 2.715700626373291,
      "kl": 3.4560546875,
      "learning_rate": 3.7908344084155176e-07,
      "loss": 0.3767,
      "reward": 1.7250000536441803,
      "reward_std": 0.4290027767419815,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2290
    },
    {
      "completion_length": 253.850004196167,
      "epoch": 2.3763608087091757,
      "grad_norm": 0.961925745010376,
      "kl": 2.11962890625,
      "learning_rate": 3.778807844987635e-07,
      "loss": 0.2933,
      "reward": 1.525000050663948,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.6000000238418579,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2291
    },
    {
      "completion_length": 223.47500610351562,
      "epoch": 2.377397615344738,
      "grad_norm": 4.471120834350586,
      "kl": 3.923828125,
      "learning_rate": 3.766797639330137e-07,
      "loss": 0.4438,
      "reward": 1.6250000298023224,
      "reward_std": 0.391311913728714,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2292
    },
    {
      "completion_length": 504.2250061035156,
      "epoch": 2.378434421980301,
      "grad_norm": 10.488146781921387,
      "kl": 7.79296875,
      "learning_rate": 3.754803808950983e-07,
      "loss": 0.9638,
      "reward": 1.3500000089406967,
      "reward_std": 0.4849044792354107,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.800000011920929,
      "step": 2293
    },
    {
      "completion_length": 109.45000076293945,
      "epoch": 2.3794712286158632,
      "grad_norm": 0.040921587496995926,
      "kl": 0.2236328125,
      "learning_rate": 3.742826371334259e-07,
      "loss": 0.0082,
      "reward": 1.5,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 1.0,
      "step": 2294
    },
    {
      "completion_length": 290.2250099182129,
      "epoch": 2.3805080352514256,
      "grad_norm": 2.5741422176361084,
      "kl": 1.95263671875,
      "learning_rate": 3.7308653439401547e-07,
      "loss": 0.3087,
      "reward": 1.6750000417232513,
      "reward_std": 0.36053743958473206,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2295
    },
    {
      "completion_length": 200.4500026702881,
      "epoch": 2.381544841886988,
      "grad_norm": 1.7534135580062866,
      "kl": 1.39404296875,
      "learning_rate": 3.718920744204948e-07,
      "loss": 0.2131,
      "reward": 1.75,
      "reward_std": 0.2795085050165653,
      "rewards/accuracy_reward": 0.8000000268220901,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2296
    },
    {
      "completion_length": 365.7250118255615,
      "epoch": 2.3825816485225504,
      "grad_norm": 18.31806755065918,
      "kl": 8.953125,
      "learning_rate": 3.70699258954096e-07,
      "loss": 0.9118,
      "reward": 1.4750000536441803,
      "reward_std": 0.6651731804013252,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.800000011920929,
      "step": 2297
    },
    {
      "completion_length": 323.1250057220459,
      "epoch": 2.383618455158113,
      "grad_norm": 1.1554791927337646,
      "kl": 3.6015625,
      "learning_rate": 3.6950808973365376e-07,
      "loss": 0.4902,
      "reward": 1.5750000178813934,
      "reward_std": 0.497468076646328,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2298
    },
    {
      "completion_length": 294.00000762939453,
      "epoch": 2.3846552617936756,
      "grad_norm": 6.341114044189453,
      "kl": 3.67919921875,
      "learning_rate": 3.6831856849560284e-07,
      "loss": 0.4459,
      "reward": 1.475000038743019,
      "reward_std": 0.2920721024274826,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2299
    },
    {
      "completion_length": 340.6250114440918,
      "epoch": 2.385692068429238,
      "grad_norm": 6.670530319213867,
      "kl": 5.348388671875,
      "learning_rate": 3.6713069697397763e-07,
      "loss": 0.4886,
      "reward": 1.4000000357627869,
      "reward_std": 0.5048274993896484,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2300
    },
    {
      "completion_length": 252.3500099182129,
      "epoch": 2.3867288750648004,
      "grad_norm": 1.6615840196609497,
      "kl": 1.8544921875,
      "learning_rate": 3.6594447690040385e-07,
      "loss": 0.2679,
      "reward": 1.350000038743019,
      "reward_std": 0.3804604895412922,
      "rewards/accuracy_reward": 0.4250000137835741,
      "rewards/format_reward": 0.925000011920929,
      "step": 2301
    },
    {
      "completion_length": 196.80000495910645,
      "epoch": 2.3877656817003627,
      "grad_norm": 1.0518519878387451,
      "kl": 1.4334716796875,
      "learning_rate": 3.6475991000410333e-07,
      "loss": 0.1939,
      "reward": 1.725000023841858,
      "reward_std": 0.3354101702570915,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2302
    },
    {
      "completion_length": 224.15000534057617,
      "epoch": 2.3888024883359256,
      "grad_norm": 1.8225899934768677,
      "kl": 1.6796875,
      "learning_rate": 3.635769980118872e-07,
      "loss": 0.2316,
      "reward": 1.7750000059604645,
      "reward_std": 0.38409097865223885,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2303
    },
    {
      "completion_length": 266.6500053405762,
      "epoch": 2.389839294971488,
      "grad_norm": 2.874328374862671,
      "kl": 3.658203125,
      "learning_rate": 3.6239574264815186e-07,
      "loss": 0.4571,
      "reward": 1.725000023841858,
      "reward_std": 0.4472135789692402,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2304
    },
    {
      "completion_length": 482.6500015258789,
      "epoch": 2.3908761016070503,
      "grad_norm": 4.6452484130859375,
      "kl": 4.349609375,
      "learning_rate": 3.6121614563488195e-07,
      "loss": 0.4556,
      "reward": 1.5,
      "reward_std": 0.6651731692254543,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.800000011920929,
      "step": 2305
    },
    {
      "completion_length": 134.20000457763672,
      "epoch": 2.3919129082426127,
      "grad_norm": 0.15382592380046844,
      "kl": 0.824951171875,
      "learning_rate": 3.600382086916428e-07,
      "loss": 0.1034,
      "reward": 1.9500000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.9750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2306
    },
    {
      "completion_length": 154.82500457763672,
      "epoch": 2.392949714878175,
      "grad_norm": 0.7913884520530701,
      "kl": 1.654296875,
      "learning_rate": 3.5886193353557983e-07,
      "loss": 0.1719,
      "reward": 1.725000023841858,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2307
    },
    {
      "completion_length": 442.5749988555908,
      "epoch": 2.393986521513738,
      "grad_norm": 2.501725912094116,
      "kl": 3.158447265625,
      "learning_rate": 3.5768732188141597e-07,
      "loss": 0.5003,
      "reward": 1.5750000178813934,
      "reward_std": 0.4849044643342495,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.8250000029802322,
      "step": 2308
    },
    {
      "completion_length": 448.7750072479248,
      "epoch": 2.3950233281493003,
      "grad_norm": 1.1683588027954102,
      "kl": 3.13330078125,
      "learning_rate": 3.5651437544145043e-07,
      "loss": 0.3143,
      "reward": 1.350000023841858,
      "reward_std": 0.37310105562210083,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.8250000029802322,
      "step": 2309
    },
    {
      "completion_length": 205.85000228881836,
      "epoch": 2.3960601347848627,
      "grad_norm": 2.88496732711792,
      "kl": 1.298583984375,
      "learning_rate": 3.5534309592555354e-07,
      "loss": 0.216,
      "reward": 1.475000023841858,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 2310
    },
    {
      "completion_length": 297.350004196167,
      "epoch": 2.397096941420425,
      "grad_norm": 1.2751463651657104,
      "kl": 1.91650390625,
      "learning_rate": 3.541734850411659e-07,
      "loss": 0.2189,
      "reward": 1.7000000178813934,
      "reward_std": 0.3046357296407223,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2311
    },
    {
      "completion_length": 302.87500762939453,
      "epoch": 2.3981337480559874,
      "grad_norm": 1.0832476615905762,
      "kl": 2.6064453125,
      "learning_rate": 3.5300554449329573e-07,
      "loss": 0.3916,
      "reward": 1.4750000089406967,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2312
    },
    {
      "completion_length": 285.4500026702881,
      "epoch": 2.3991705546915503,
      "grad_norm": 1.040643572807312,
      "kl": 2.483154296875,
      "learning_rate": 3.518392759845176e-07,
      "loss": 0.234,
      "reward": 1.800000011920929,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.9000000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2313
    },
    {
      "completion_length": 202.875,
      "epoch": 2.4002073613271127,
      "grad_norm": 1.2668492794036865,
      "kl": 1.83642578125,
      "learning_rate": 3.506746812149657e-07,
      "loss": 0.1939,
      "reward": 1.449999988079071,
      "reward_std": 0.27228758856654167,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2314
    },
    {
      "completion_length": 348.1750011444092,
      "epoch": 2.401244167962675,
      "grad_norm": 4.579446315765381,
      "kl": 2.7060546875,
      "learning_rate": 3.4951176188233743e-07,
      "loss": 0.4072,
      "reward": 1.725000023841858,
      "reward_std": 0.47234082967042923,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2315
    },
    {
      "completion_length": 392.90000915527344,
      "epoch": 2.4022809745982374,
      "grad_norm": 4.644406318664551,
      "kl": 4.51953125,
      "learning_rate": 3.48350519681886e-07,
      "loss": 0.5998,
      "reward": 1.2500000447034836,
      "reward_std": 0.5571388378739357,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2316
    },
    {
      "completion_length": 341.9499988555908,
      "epoch": 2.4033177812338,
      "grad_norm": 3.313523530960083,
      "kl": 3.36865234375,
      "learning_rate": 3.471909563064202e-07,
      "loss": 0.4151,
      "reward": 1.100000038743019,
      "reward_std": 0.4472135826945305,
      "rewards/accuracy_reward": 0.25000000186264515,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2317
    },
    {
      "completion_length": 214.25000381469727,
      "epoch": 2.404354587869362,
      "grad_norm": 1.278300166130066,
      "kl": 1.544189453125,
      "learning_rate": 3.4603307344630135e-07,
      "loss": 0.1792,
      "reward": 1.3749999850988388,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2318
    },
    {
      "completion_length": 404.77500915527344,
      "epoch": 2.405391394504925,
      "grad_norm": 1.3889893293380737,
      "kl": 3.466796875,
      "learning_rate": 3.4487687278944184e-07,
      "loss": 0.5459,
      "reward": 1.2750000059604645,
      "reward_std": 0.6202614307403564,
      "rewards/accuracy_reward": 0.44999999925494194,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2319
    },
    {
      "completion_length": 303.9000129699707,
      "epoch": 2.4064282011404874,
      "grad_norm": 1.4112569093704224,
      "kl": 2.0166015625,
      "learning_rate": 3.4372235602130095e-07,
      "loss": 0.2811,
      "reward": 1.4250000417232513,
      "reward_std": 0.3930240832269192,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2320
    },
    {
      "completion_length": 198.1999969482422,
      "epoch": 2.4074650077760498,
      "grad_norm": 2.4360010623931885,
      "kl": 3.114013671875,
      "learning_rate": 3.425695248248833e-07,
      "loss": 0.1731,
      "reward": 1.3750000447034836,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2321
    },
    {
      "completion_length": 290.9000015258789,
      "epoch": 2.408501814411612,
      "grad_norm": 2.1025843620300293,
      "kl": 1.75,
      "learning_rate": 3.41418380880736e-07,
      "loss": 0.3233,
      "reward": 1.600000023841858,
      "reward_std": 0.3354102075099945,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9000000357627869,
      "step": 2322
    },
    {
      "completion_length": 349.12500381469727,
      "epoch": 2.4095386210471745,
      "grad_norm": 1.5481314659118652,
      "kl": 3.046630859375,
      "learning_rate": 3.4026892586694863e-07,
      "loss": 0.4629,
      "reward": 1.4500000327825546,
      "reward_std": 0.3731010630726814,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2323
    },
    {
      "completion_length": 292.2250061035156,
      "epoch": 2.4105754276827374,
      "grad_norm": 1.0657193660736084,
      "kl": 2.52734375,
      "learning_rate": 3.3912116145914544e-07,
      "loss": 0.3211,
      "reward": 1.3000000268220901,
      "reward_std": 0.3479738235473633,
      "rewards/accuracy_reward": 0.4000000096857548,
      "rewards/format_reward": 0.9000000357627869,
      "step": 2324
    },
    {
      "completion_length": 203.67500495910645,
      "epoch": 2.4116122343182997,
      "grad_norm": 0.5969396233558655,
      "kl": 1.52783203125,
      "learning_rate": 3.3797508933048895e-07,
      "loss": 0.1703,
      "reward": 1.5750000327825546,
      "reward_std": 0.22894952818751335,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2325
    },
    {
      "completion_length": 237.97500228881836,
      "epoch": 2.412649040953862,
      "grad_norm": 0.9664300680160522,
      "kl": 2.80078125,
      "learning_rate": 3.3683071115167344e-07,
      "loss": 0.1842,
      "reward": 1.725000023841858,
      "reward_std": 0.24873404204845428,
      "rewards/accuracy_reward": 0.7999999970197678,
      "rewards/format_reward": 0.925000011920929,
      "step": 2326
    },
    {
      "completion_length": 297.52500343322754,
      "epoch": 2.4136858475894245,
      "grad_norm": 1.7112267017364502,
      "kl": 2.087890625,
      "learning_rate": 3.356880285909243e-07,
      "loss": 0.3628,
      "reward": 1.4750000536441803,
      "reward_std": 0.4472136162221432,
      "rewards/accuracy_reward": 0.5750000197440386,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2327
    },
    {
      "completion_length": 144.82499885559082,
      "epoch": 2.414722654224987,
      "grad_norm": 1.2764661312103271,
      "kl": 0.9000244140625,
      "learning_rate": 3.345470433139943e-07,
      "loss": 0.1393,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2328
    },
    {
      "completion_length": 203.00000190734863,
      "epoch": 2.4157594608605493,
      "grad_norm": 1.1932405233383179,
      "kl": 1.4002685546875,
      "learning_rate": 3.3340775698416327e-07,
      "loss": 0.1987,
      "reward": 1.675000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2329
    },
    {
      "completion_length": 244.45001220703125,
      "epoch": 2.416796267496112,
      "grad_norm": 1.6320277452468872,
      "kl": 2.1358642578125,
      "learning_rate": 3.3227017126223376e-07,
      "loss": 0.2749,
      "reward": 1.5000000149011612,
      "reward_std": 0.1928323470056057,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 2330
    },
    {
      "completion_length": 106.20000267028809,
      "epoch": 2.4178330741316745,
      "grad_norm": 0.2150980830192566,
      "kl": 0.2099609375,
      "learning_rate": 3.311342878065289e-07,
      "loss": 0.0163,
      "reward": 1.5999999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 1.0,
      "step": 2331
    },
    {
      "completion_length": 193.37500381469727,
      "epoch": 2.418869880767237,
      "grad_norm": 1.8777616024017334,
      "kl": 0.917236328125,
      "learning_rate": 3.300001082728901e-07,
      "loss": 0.1585,
      "reward": 1.5750000178813934,
      "reward_std": 0.27228759601712227,
      "rewards/accuracy_reward": 0.6250000186264515,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2332
    },
    {
      "completion_length": 186.8750057220459,
      "epoch": 2.4199066874027992,
      "grad_norm": 1.403493046760559,
      "kl": 1.2109375,
      "learning_rate": 3.288676343146771e-07,
      "loss": 0.2036,
      "reward": 1.675000011920929,
      "reward_std": 0.3479738049209118,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2333
    },
    {
      "completion_length": 199.1500072479248,
      "epoch": 2.4209434940383616,
      "grad_norm": 2.4488933086395264,
      "kl": 0.776611328125,
      "learning_rate": 3.2773686758275974e-07,
      "loss": 0.1664,
      "reward": 1.4000000208616257,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2334
    },
    {
      "completion_length": 263.30000495910645,
      "epoch": 2.4219803006739244,
      "grad_norm": 2.7692065238952637,
      "kl": 1.05224609375,
      "learning_rate": 3.2660780972552213e-07,
      "loss": 0.1881,
      "reward": 1.625,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 2335
    },
    {
      "completion_length": 364.70000648498535,
      "epoch": 2.423017107309487,
      "grad_norm": 0.7779722213745117,
      "kl": 2.97900390625,
      "learning_rate": 3.2548046238885605e-07,
      "loss": 0.4893,
      "reward": 1.5750000178813934,
      "reward_std": 0.41643912717700005,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2336
    },
    {
      "completion_length": 291.6499996185303,
      "epoch": 2.424053913945049,
      "grad_norm": 2.762993812561035,
      "kl": 2.61962890625,
      "learning_rate": 3.243548272161596e-07,
      "loss": 0.3786,
      "reward": 1.5250000208616257,
      "reward_std": 0.36053745076060295,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2337
    },
    {
      "completion_length": 148.850004196167,
      "epoch": 2.4250907205806116,
      "grad_norm": 0.37188079953193665,
      "kl": 0.6761474609375,
      "learning_rate": 3.232309058483349e-07,
      "loss": 0.0936,
      "reward": 1.925000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.9500000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2338
    },
    {
      "completion_length": 191.27500343322754,
      "epoch": 2.426127527216174,
      "grad_norm": 10.821712493896484,
      "kl": 3.979736328125,
      "learning_rate": 3.2210869992378706e-07,
      "loss": 0.4034,
      "reward": 1.699999988079071,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2339
    },
    {
      "completion_length": 193.85000610351562,
      "epoch": 2.427164333851737,
      "grad_norm": 1.010215163230896,
      "kl": 0.516357421875,
      "learning_rate": 3.209882110784189e-07,
      "loss": 0.0911,
      "reward": 1.5750000178813934,
      "reward_std": 0.31199517101049423,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2340
    },
    {
      "completion_length": 385.07501220703125,
      "epoch": 2.428201140487299,
      "grad_norm": 6.709095478057861,
      "kl": 5.19921875,
      "learning_rate": 3.1986944094563044e-07,
      "loss": 0.6463,
      "reward": 1.225000023841858,
      "reward_std": 0.38409100472927094,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2341
    },
    {
      "completion_length": 100.69999885559082,
      "epoch": 2.4292379471228616,
      "grad_norm": 0.339815229177475,
      "kl": 0.13818359375,
      "learning_rate": 3.1875239115631735e-07,
      "loss": 0.0053,
      "reward": 1.875,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 1.0,
      "step": 2342
    },
    {
      "completion_length": 289.4000072479248,
      "epoch": 2.430274753758424,
      "grad_norm": 2.1046578884124756,
      "kl": 3.15185546875,
      "learning_rate": 3.176370633388665e-07,
      "loss": 0.4601,
      "reward": 1.5250000357627869,
      "reward_std": 0.5031152591109276,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2343
    },
    {
      "completion_length": 187.17500686645508,
      "epoch": 2.4313115603939863,
      "grad_norm": 1.7078900337219238,
      "kl": 1.69189453125,
      "learning_rate": 3.165234591191544e-07,
      "loss": 0.1986,
      "reward": 1.8250000178813934,
      "reward_std": 0.20019173622131348,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2344
    },
    {
      "completion_length": 346.3000144958496,
      "epoch": 2.432348367029549,
      "grad_norm": 0.6840152740478516,
      "kl": 2.170166015625,
      "learning_rate": 3.154115801205451e-07,
      "loss": 0.1528,
      "reward": 1.3250000327825546,
      "reward_std": 0.26129766181111336,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2345
    },
    {
      "completion_length": 175.0000057220459,
      "epoch": 2.4333851736651115,
      "grad_norm": 3.0889248847961426,
      "kl": 1.138916015625,
      "learning_rate": 3.143014279638888e-07,
      "loss": 0.1689,
      "reward": 1.475000023841858,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2346
    },
    {
      "completion_length": 100.20000076293945,
      "epoch": 2.434421980300674,
      "grad_norm": 0.589874804019928,
      "kl": 1.05419921875,
      "learning_rate": 3.1319300426751636e-07,
      "loss": -0.0187,
      "reward": 1.475000023841858,
      "reward_std": 0.22894951701164246,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2347
    },
    {
      "completion_length": 242.35000038146973,
      "epoch": 2.4354587869362363,
      "grad_norm": 1.2983113527297974,
      "kl": 1.19287109375,
      "learning_rate": 3.120863106472409e-07,
      "loss": 0.248,
      "reward": 1.6250000298023224,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.7000000067055225,
      "rewards/format_reward": 0.925000011920929,
      "step": 2348
    },
    {
      "completion_length": 235.82500839233398,
      "epoch": 2.4364955935717987,
      "grad_norm": 1.5528539419174194,
      "kl": 1.769287109375,
      "learning_rate": 3.109813487163525e-07,
      "loss": 0.2449,
      "reward": 1.550000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.925000011920929,
      "step": 2349
    },
    {
      "completion_length": 291.2749996185303,
      "epoch": 2.4375324002073615,
      "grad_norm": 0.5441251993179321,
      "kl": 0.70654296875,
      "learning_rate": 3.098781200856169e-07,
      "loss": 0.1037,
      "reward": 1.3750000149011612,
      "reward_std": 0.26129766926169395,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2350
    },
    {
      "completion_length": 437.6500015258789,
      "epoch": 2.438569206842924,
      "grad_norm": 1.4629409313201904,
      "kl": 1.1337890625,
      "learning_rate": 3.087766263632729e-07,
      "loss": 0.2119,
      "reward": 1.3250000476837158,
      "reward_std": 0.604067251086235,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.8000000268220901,
      "step": 2351
    },
    {
      "completion_length": 292.95001220703125,
      "epoch": 2.4396060134784863,
      "grad_norm": 3.109811782836914,
      "kl": 0.76708984375,
      "learning_rate": 3.0767686915503137e-07,
      "loss": 0.2371,
      "reward": 1.5750000178813934,
      "reward_std": 0.3354101777076721,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9000000357627869,
      "step": 2352
    },
    {
      "completion_length": 388.70001220703125,
      "epoch": 2.4406428201140487,
      "grad_norm": 3.7494449615478516,
      "kl": 3.681640625,
      "learning_rate": 3.065788500640706e-07,
      "loss": 0.5241,
      "reward": 1.5250000357627869,
      "reward_std": 0.45977718755602837,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2353
    },
    {
      "completion_length": 288.0750102996826,
      "epoch": 2.441679626749611,
      "grad_norm": 5.217642307281494,
      "kl": 3.9384765625,
      "learning_rate": 3.054825706910356e-07,
      "loss": 0.446,
      "reward": 1.3499999940395355,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2354
    },
    {
      "completion_length": 140.20000457763672,
      "epoch": 2.442716433385174,
      "grad_norm": 0.8968407511711121,
      "kl": 1.45458984375,
      "learning_rate": 3.043880326340346e-07,
      "loss": 0.1372,
      "reward": 1.3500000089406967,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2355
    },
    {
      "completion_length": 243.60000801086426,
      "epoch": 2.4437532400207362,
      "grad_norm": 0.7036870121955872,
      "kl": 1.313232421875,
      "learning_rate": 3.0329523748863944e-07,
      "loss": 0.2331,
      "reward": 1.7250000536441803,
      "reward_std": 0.3913118839263916,
      "rewards/accuracy_reward": 0.8000000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2356
    },
    {
      "completion_length": 231.90000343322754,
      "epoch": 2.4447900466562986,
      "grad_norm": 1.1117327213287354,
      "kl": 1.30419921875,
      "learning_rate": 3.022041868478781e-07,
      "loss": 0.2512,
      "reward": 1.5000000149011612,
      "reward_std": 0.3966546133160591,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2357
    },
    {
      "completion_length": 284.87500762939453,
      "epoch": 2.445826853291861,
      "grad_norm": 0.6404615044593811,
      "kl": 1.6473388671875,
      "learning_rate": 3.011148823022382e-07,
      "loss": 0.1735,
      "reward": 1.4000000208616257,
      "reward_std": 0.28485121577978134,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2358
    },
    {
      "completion_length": 100.65000343322754,
      "epoch": 2.4468636599274234,
      "grad_norm": 0.22570554912090302,
      "kl": 0.203857421875,
      "learning_rate": 3.000273254396609e-07,
      "loss": -0.0092,
      "reward": 1.5999999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 1.0,
      "step": 2359
    },
    {
      "completion_length": 236.10000228881836,
      "epoch": 2.447900466562986,
      "grad_norm": 1.8055659532546997,
      "kl": 1.50927734375,
      "learning_rate": 2.989415178455394e-07,
      "loss": 0.2624,
      "reward": 1.5000000298023224,
      "reward_std": 0.3678968548774719,
      "rewards/accuracy_reward": 0.575000012293458,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2360
    },
    {
      "completion_length": 192.07500457763672,
      "epoch": 2.4489372731985486,
      "grad_norm": 0.592898964881897,
      "kl": 0.760498046875,
      "learning_rate": 2.9785746110271714e-07,
      "loss": 0.1156,
      "reward": 1.725000023841858,
      "reward_std": 0.21638590842485428,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2361
    },
    {
      "completion_length": 232.2500057220459,
      "epoch": 2.449974079834111,
      "grad_norm": 1.8508909940719604,
      "kl": 1.441650390625,
      "learning_rate": 2.9677515679148595e-07,
      "loss": 0.2427,
      "reward": 1.5000000298023224,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2362
    },
    {
      "completion_length": 138.0750026702881,
      "epoch": 2.4510108864696734,
      "grad_norm": 1.6418184041976929,
      "kl": 0.80517578125,
      "learning_rate": 2.95694606489582e-07,
      "loss": 0.111,
      "reward": 1.725000023841858,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2363
    },
    {
      "completion_length": 288.75000953674316,
      "epoch": 2.4520476931052357,
      "grad_norm": 1.4929782152175903,
      "kl": 2.129150390625,
      "learning_rate": 2.9461581177218496e-07,
      "loss": 0.334,
      "reward": 1.5250000059604645,
      "reward_std": 0.3413859084248543,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2364
    },
    {
      "completion_length": 199.27500343322754,
      "epoch": 2.4530844997407986,
      "grad_norm": 1.2114561796188354,
      "kl": 1.0078125,
      "learning_rate": 2.935387742119149e-07,
      "loss": 0.1603,
      "reward": 1.725000023841858,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.7749999985098839,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2365
    },
    {
      "completion_length": 242.62500190734863,
      "epoch": 2.454121306376361,
      "grad_norm": 9.375083923339844,
      "kl": 3.9327392578125,
      "learning_rate": 2.9246349537883187e-07,
      "loss": 0.4171,
      "reward": 1.725000023841858,
      "reward_std": 0.36053744331002235,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 2366
    },
    {
      "completion_length": 358.72499656677246,
      "epoch": 2.4551581130119233,
      "grad_norm": 15.948465347290039,
      "kl": 6.984375,
      "learning_rate": 2.913899768404297e-07,
      "loss": 0.656,
      "reward": 1.3250000178813934,
      "reward_std": 0.45413001626729965,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2367
    },
    {
      "completion_length": 203.20000076293945,
      "epoch": 2.4561949196474857,
      "grad_norm": 0.4468858540058136,
      "kl": 0.7708740234375,
      "learning_rate": 2.9031822016163847e-07,
      "loss": 0.1453,
      "reward": 1.7000000178813934,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2368
    },
    {
      "completion_length": 348.5750045776367,
      "epoch": 2.457231726283048,
      "grad_norm": 5.419402122497559,
      "kl": 4.77734375,
      "learning_rate": 2.8924822690481843e-07,
      "loss": 0.5542,
      "reward": 1.400000050663948,
      "reward_std": 0.3407529331743717,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2369
    },
    {
      "completion_length": 278.49999809265137,
      "epoch": 2.458268532918611,
      "grad_norm": 1.3491854667663574,
      "kl": 2.1357421875,
      "learning_rate": 2.881799986297597e-07,
      "loss": 0.3503,
      "reward": 1.6500000357627869,
      "reward_std": 0.50311528891325,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2370
    },
    {
      "completion_length": 201.225004196167,
      "epoch": 2.4593053395541733,
      "grad_norm": 0.6863802671432495,
      "kl": 0.5845947265625,
      "learning_rate": 2.8711353689367905e-07,
      "loss": 0.1523,
      "reward": 1.425000011920929,
      "reward_std": 0.3479738086462021,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2371
    },
    {
      "completion_length": 383.0500068664551,
      "epoch": 2.4603421461897357,
      "grad_norm": 3.002951145172119,
      "kl": 3.07421875,
      "learning_rate": 2.860488432512192e-07,
      "loss": 0.466,
      "reward": 1.3000000268220901,
      "reward_std": 0.49746808409690857,
      "rewards/accuracy_reward": 0.44999999552965164,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2372
    },
    {
      "completion_length": 403.5,
      "epoch": 2.461378952825298,
      "grad_norm": 3.643052577972412,
      "kl": 3.5087890625,
      "learning_rate": 2.849859192544444e-07,
      "loss": 0.5191,
      "reward": 1.350000038743019,
      "reward_std": 0.5210216343402863,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2373
    },
    {
      "completion_length": 242.92500114440918,
      "epoch": 2.4624157594608604,
      "grad_norm": 3.1055939197540283,
      "kl": 1.591064453125,
      "learning_rate": 2.8392476645283927e-07,
      "loss": 0.2492,
      "reward": 1.5250000357627869,
      "reward_std": 0.3354101963341236,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.925000011920929,
      "step": 2374
    },
    {
      "completion_length": 160.94999885559082,
      "epoch": 2.463452566096423,
      "grad_norm": 0.4246121644973755,
      "kl": 0.412841796875,
      "learning_rate": 2.828653863933063e-07,
      "loss": 0.0714,
      "reward": 1.7750000357627869,
      "reward_std": 0.18026870489120483,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2375
    },
    {
      "completion_length": 261.2250099182129,
      "epoch": 2.4644893727319857,
      "grad_norm": 1.694840431213379,
      "kl": 1.4765625,
      "learning_rate": 2.818077806201653e-07,
      "loss": 0.1536,
      "reward": 1.675000011920929,
      "reward_std": 0.30463573709130287,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2376
    },
    {
      "completion_length": 124.25000381469727,
      "epoch": 2.465526179367548,
      "grad_norm": 0.8132579326629639,
      "kl": 0.361328125,
      "learning_rate": 2.8075195067514683e-07,
      "loss": 0.061,
      "reward": 1.800000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2377
    },
    {
      "completion_length": 184.52500534057617,
      "epoch": 2.4665629860031104,
      "grad_norm": 0.5665480494499207,
      "kl": 0.82275390625,
      "learning_rate": 2.796978980973954e-07,
      "loss": 0.1502,
      "reward": 1.675000011920929,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2378
    },
    {
      "completion_length": 254.7500057220459,
      "epoch": 2.467599792638673,
      "grad_norm": 0.7989780902862549,
      "kl": 2.2333984375,
      "learning_rate": 2.7864562442346304e-07,
      "loss": 0.3123,
      "reward": 1.3750000447034836,
      "reward_std": 0.3913118652999401,
      "rewards/accuracy_reward": 0.45000001043081284,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2379
    },
    {
      "completion_length": 192.64999961853027,
      "epoch": 2.468636599274235,
      "grad_norm": 1.6563884019851685,
      "kl": 1.677001953125,
      "learning_rate": 2.7759513118730854e-07,
      "loss": 0.177,
      "reward": 1.7750000357627869,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2380
    },
    {
      "completion_length": 193.0000057220459,
      "epoch": 2.469673405909798,
      "grad_norm": 2.466038942337036,
      "kl": 2.145263671875,
      "learning_rate": 2.7654641992029664e-07,
      "loss": 0.2738,
      "reward": 1.5250000059604645,
      "reward_std": 0.335410188883543,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2381
    },
    {
      "completion_length": 273.12500762939453,
      "epoch": 2.4707102125453604,
      "grad_norm": 0.6439417600631714,
      "kl": 1.188232421875,
      "learning_rate": 2.7549949215119295e-07,
      "loss": 0.2338,
      "reward": 1.6500000357627869,
      "reward_std": 0.41643913835287094,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2382
    },
    {
      "completion_length": 195.9500026702881,
      "epoch": 2.4717470191809228,
      "grad_norm": 0.7071945667266846,
      "kl": 1.03564453125,
      "learning_rate": 2.7445434940616365e-07,
      "loss": 0.1715,
      "reward": 1.675000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2383
    },
    {
      "completion_length": 297.8999996185303,
      "epoch": 2.472783825816485,
      "grad_norm": 0.6132912039756775,
      "kl": 1.6171875,
      "learning_rate": 2.7341099320877235e-07,
      "loss": 0.2878,
      "reward": 1.100000023841858,
      "reward_std": 0.337122417986393,
      "rewards/accuracy_reward": 0.22500000149011612,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2384
    },
    {
      "completion_length": 385.5000057220459,
      "epoch": 2.4738206324520475,
      "grad_norm": 2.6434671878814697,
      "kl": 2.9677734375,
      "learning_rate": 2.723694250799795e-07,
      "loss": 0.3079,
      "reward": 1.300000011920929,
      "reward_std": 0.46511993557214737,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2385
    },
    {
      "completion_length": 255.3000087738037,
      "epoch": 2.47485743908761,
      "grad_norm": 1.1256232261657715,
      "kl": 1.26513671875,
      "learning_rate": 2.71329646538138e-07,
      "loss": 0.2244,
      "reward": 1.4500000476837158,
      "reward_std": 0.3354102149605751,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2386
    },
    {
      "completion_length": 288.75000381469727,
      "epoch": 2.4758942457231727,
      "grad_norm": 2.9735267162323,
      "kl": 1.58203125,
      "learning_rate": 2.7029165909899216e-07,
      "loss": 0.2941,
      "reward": 1.475000023841858,
      "reward_std": 0.4237985350191593,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2387
    },
    {
      "completion_length": 301.4500026702881,
      "epoch": 2.476931052358735,
      "grad_norm": 1.974998116493225,
      "kl": 1.06884765625,
      "learning_rate": 2.6925546427567495e-07,
      "loss": 0.2343,
      "reward": 1.5250000357627869,
      "reward_std": 0.41643913835287094,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2388
    },
    {
      "completion_length": 161.25000190734863,
      "epoch": 2.4779678589942975,
      "grad_norm": 0.959082305431366,
      "kl": 0.863525390625,
      "learning_rate": 2.682210635787075e-07,
      "loss": 0.107,
      "reward": 1.6500000059604645,
      "reward_std": 0.1369306445121765,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2389
    },
    {
      "completion_length": 188.2000026702881,
      "epoch": 2.47900466562986,
      "grad_norm": 1.0125101804733276,
      "kl": 1.053955078125,
      "learning_rate": 2.671884585159934e-07,
      "loss": 0.1855,
      "reward": 1.550000011920929,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2390
    },
    {
      "completion_length": 489.1250171661377,
      "epoch": 2.4800414722654223,
      "grad_norm": 3.6530213356018066,
      "kl": 3.310546875,
      "learning_rate": 2.661576505928206e-07,
      "loss": 0.4268,
      "reward": 1.2000000178813934,
      "reward_std": 0.6579522900283337,
      "rewards/accuracy_reward": 0.4000000059604645,
      "rewards/format_reward": 0.800000011920929,
      "step": 2391
    },
    {
      "completion_length": 246.20000839233398,
      "epoch": 2.481078278900985,
      "grad_norm": 1.3209348917007446,
      "kl": 1.00927734375,
      "learning_rate": 2.6512864131185634e-07,
      "loss": 0.1936,
      "reward": 1.6500000357627869,
      "reward_std": 0.4038754999637604,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2392
    },
    {
      "completion_length": 242.0250015258789,
      "epoch": 2.4821150855365475,
      "grad_norm": 1.5393630266189575,
      "kl": 1.8310546875,
      "learning_rate": 2.641014321731458e-07,
      "loss": 0.2787,
      "reward": 1.425000011920929,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 2393
    },
    {
      "completion_length": 335.4749984741211,
      "epoch": 2.48315189217211,
      "grad_norm": 2.023813486099243,
      "kl": 2.30859375,
      "learning_rate": 2.630760246741101e-07,
      "loss": 0.4196,
      "reward": 1.5250000059604645,
      "reward_std": 0.5031152926385403,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2394
    },
    {
      "completion_length": 344.1750011444092,
      "epoch": 2.4841886988076722,
      "grad_norm": 2.498627185821533,
      "kl": 1.908935546875,
      "learning_rate": 2.620524203095446e-07,
      "loss": 0.2783,
      "reward": 1.7750000357627869,
      "reward_std": 0.40387551486492157,
      "rewards/accuracy_reward": 0.9000000208616257,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2395
    },
    {
      "completion_length": 256.6750011444092,
      "epoch": 2.4852255054432346,
      "grad_norm": 0.825520932674408,
      "kl": 1.21484375,
      "learning_rate": 2.6103062057161534e-07,
      "loss": 0.2454,
      "reward": 1.675000011920929,
      "reward_std": 0.3913118876516819,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2396
    },
    {
      "completion_length": 197.72500038146973,
      "epoch": 2.4862623120787974,
      "grad_norm": 1.734663963317871,
      "kl": 1.119873046875,
      "learning_rate": 2.6001062694985793e-07,
      "loss": 0.1658,
      "reward": 1.2249999940395355,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.27500000037252903,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2397
    },
    {
      "completion_length": 177.8750057220459,
      "epoch": 2.48729911871436,
      "grad_norm": 1.4318667650222778,
      "kl": 1.071044921875,
      "learning_rate": 2.589924409311749e-07,
      "loss": 0.1696,
      "reward": 1.6500000059604645,
      "reward_std": 0.22360678762197495,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2398
    },
    {
      "completion_length": 156.4000015258789,
      "epoch": 2.488335925349922,
      "grad_norm": 0.4744577705860138,
      "kl": 0.6318359375,
      "learning_rate": 2.5797606399983483e-07,
      "loss": 0.0869,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2399
    },
    {
      "completion_length": 249.62500381469727,
      "epoch": 2.4893727319854846,
      "grad_norm": 4.381124019622803,
      "kl": 1.009033203125,
      "learning_rate": 2.5696149763746696e-07,
      "loss": 0.1847,
      "reward": 1.350000023841858,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.925000011920929,
      "step": 2400
    },
    {
      "completion_length": 168.65000343322754,
      "epoch": 2.490409538621047,
      "grad_norm": 0.2993444502353668,
      "kl": 0.6865234375,
      "learning_rate": 2.559487433230633e-07,
      "loss": 0.0833,
      "reward": 1.675000011920929,
      "reward_std": 0.20916500687599182,
      "rewards/accuracy_reward": 0.7249999940395355,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2401
    },
    {
      "completion_length": 293.475004196167,
      "epoch": 2.49144634525661,
      "grad_norm": 0.4146404266357422,
      "kl": 1.395263671875,
      "learning_rate": 2.5493780253297314e-07,
      "loss": 0.2614,
      "reward": 1.675000011920929,
      "reward_std": 0.36053742840886116,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2402
    },
    {
      "completion_length": 294.5500068664551,
      "epoch": 2.492483151892172,
      "grad_norm": 0.5028513073921204,
      "kl": 1.027587890625,
      "learning_rate": 2.539286767409026e-07,
      "loss": 0.1734,
      "reward": 1.3000000268220901,
      "reward_std": 0.36053744703531265,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2403
    },
    {
      "completion_length": 253.22500038146973,
      "epoch": 2.4935199585277346,
      "grad_norm": 1.7324107885360718,
      "kl": 0.6953125,
      "learning_rate": 2.529213674179112e-07,
      "loss": 0.1841,
      "reward": 1.2750000208616257,
      "reward_std": 0.31199514120817184,
      "rewards/accuracy_reward": 0.35000001080334187,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2404
    },
    {
      "completion_length": 153.62500381469727,
      "epoch": 2.494556765163297,
      "grad_norm": 0.7748910188674927,
      "kl": 0.33056640625,
      "learning_rate": 2.519158760324118e-07,
      "loss": 0.0673,
      "reward": 1.7750000357627869,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2405
    },
    {
      "completion_length": 205.15000534057617,
      "epoch": 2.4955935717988593,
      "grad_norm": 0.6636053919792175,
      "kl": 0.780517578125,
      "learning_rate": 2.509122040501661e-07,
      "loss": 0.1209,
      "reward": 1.5750000029802322,
      "reward_std": 0.20019175112247467,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2406
    },
    {
      "completion_length": 197.9000015258789,
      "epoch": 2.496630378434422,
      "grad_norm": 0.39183738827705383,
      "kl": 0.5146484375,
      "learning_rate": 2.4991035293428385e-07,
      "loss": 0.0774,
      "reward": 1.5999999940395355,
      "reward_std": 0.10458250343799591,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2407
    },
    {
      "completion_length": 361.7500057220459,
      "epoch": 2.4976671850699845,
      "grad_norm": 1.5787347555160522,
      "kl": 3.111083984375,
      "learning_rate": 2.4891032414522013e-07,
      "loss": 0.4915,
      "reward": 1.1750000268220901,
      "reward_std": 0.47234082594513893,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2408
    },
    {
      "completion_length": 148.72500228881836,
      "epoch": 2.498703991705547,
      "grad_norm": 1.6090366840362549,
      "kl": 0.8369140625,
      "learning_rate": 2.4791211914077477e-07,
      "loss": 0.1036,
      "reward": 1.5999999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2409
    },
    {
      "completion_length": 159.70000457763672,
      "epoch": 2.4997407983411093,
      "grad_norm": 0.1528596132993698,
      "kl": 0.277587890625,
      "learning_rate": 2.4691573937608675e-07,
      "loss": 0.0575,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2410
    },
    {
      "completion_length": 197.0,
      "epoch": 2.5007776049766717,
      "grad_norm": 0.40553736686706543,
      "kl": 1.471435546875,
      "learning_rate": 2.459211863036365e-07,
      "loss": 0.19,
      "reward": 1.3250000029802322,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2411
    },
    {
      "completion_length": 101.45000076293945,
      "epoch": 2.5018144116122345,
      "grad_norm": 0.40257906913757324,
      "kl": 0.265869140625,
      "learning_rate": 2.449284613732404e-07,
      "loss": 0.0027,
      "reward": 1.6749999821186066,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 1.0,
      "step": 2412
    },
    {
      "completion_length": 207.05000686645508,
      "epoch": 2.502851218247797,
      "grad_norm": 4.426974296569824,
      "kl": 2.2987060546875,
      "learning_rate": 2.439375660320499e-07,
      "loss": 0.2726,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2413
    },
    {
      "completion_length": 350.7500057220459,
      "epoch": 2.5038880248833593,
      "grad_norm": 1.7486178874969482,
      "kl": 1.25927734375,
      "learning_rate": 2.429485017245492e-07,
      "loss": 0.1596,
      "reward": 1.4750000089406967,
      "reward_std": 0.39822830632328987,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2414
    },
    {
      "completion_length": 206.975004196167,
      "epoch": 2.5049248315189216,
      "grad_norm": 2.0769124031066895,
      "kl": 1.19677734375,
      "learning_rate": 2.4196126989255454e-07,
      "loss": 0.1787,
      "reward": 1.4250000268220901,
      "reward_std": 0.23617040738463402,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.925000011920929,
      "step": 2415
    },
    {
      "completion_length": 141.7000026702881,
      "epoch": 2.505961638154484,
      "grad_norm": 0.2743751108646393,
      "kl": 0.590576171875,
      "learning_rate": 2.409758719752093e-07,
      "loss": 0.0843,
      "reward": 1.800000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2416
    },
    {
      "completion_length": 202.75000190734863,
      "epoch": 2.506998444790047,
      "grad_norm": 2.0179221630096436,
      "kl": 0.6356201171875,
      "learning_rate": 2.399923094089844e-07,
      "loss": 0.1385,
      "reward": 1.725000023841858,
      "reward_std": 0.3047742620110512,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 2417
    },
    {
      "completion_length": 250.22500610351562,
      "epoch": 2.5080352514256092,
      "grad_norm": 8.600875854492188,
      "kl": 4.88623046875,
      "learning_rate": 2.3901058362767425e-07,
      "loss": 0.4912,
      "reward": 1.4000000059604645,
      "reward_std": 0.31199513375759125,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2418
    },
    {
      "completion_length": 211.55000686645508,
      "epoch": 2.5090720580611716,
      "grad_norm": 0.4297401010990143,
      "kl": 0.8759765625,
      "learning_rate": 2.3803069606239753e-07,
      "loss": 0.041,
      "reward": 1.7750000059604645,
      "reward_std": 0.1369306445121765,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2419
    },
    {
      "completion_length": 191.15000343322754,
      "epoch": 2.510108864696734,
      "grad_norm": 0.8180016875267029,
      "kl": 0.778564453125,
      "learning_rate": 2.370526481415914e-07,
      "loss": 0.1761,
      "reward": 1.5000000298023224,
      "reward_std": 0.26865706592798233,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2420
    },
    {
      "completion_length": 358.64999771118164,
      "epoch": 2.5111456713322964,
      "grad_norm": 9.843328475952148,
      "kl": 5.0234375,
      "learning_rate": 2.3607644129101168e-07,
      "loss": 0.6082,
      "reward": 1.5000000149011612,
      "reward_std": 0.36053745821118355,
      "rewards/accuracy_reward": 0.6500000096857548,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2421
    },
    {
      "completion_length": 299.17501068115234,
      "epoch": 2.512182477967859,
      "grad_norm": 2.3310999870300293,
      "kl": 0.71435546875,
      "learning_rate": 2.3510207693373193e-07,
      "loss": 0.2773,
      "reward": 1.600000023841858,
      "reward_std": 0.4237985499203205,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000357627869,
      "step": 2422
    },
    {
      "completion_length": 207.00000762939453,
      "epoch": 2.5132192846034216,
      "grad_norm": 0.5458081364631653,
      "kl": 0.855712890625,
      "learning_rate": 2.341295564901369e-07,
      "loss": 0.0813,
      "reward": 1.6749999821186066,
      "reward_std": 0.36053744703531265,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2423
    },
    {
      "completion_length": 220.22500801086426,
      "epoch": 2.514256091238984,
      "grad_norm": 3.1432647705078125,
      "kl": 1.585693359375,
      "learning_rate": 2.3315888137792606e-07,
      "loss": 0.2521,
      "reward": 1.6750000417232513,
      "reward_std": 0.40387552976608276,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 2424
    },
    {
      "completion_length": 161.0,
      "epoch": 2.5152928978745464,
      "grad_norm": 0.4819316565990448,
      "kl": 0.228759765625,
      "learning_rate": 2.3219005301210754e-07,
      "loss": 0.0361,
      "reward": 1.7750000357627869,
      "reward_std": 0.3046357296407223,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2425
    },
    {
      "completion_length": 155.600004196167,
      "epoch": 2.5163297045101087,
      "grad_norm": 0.42437878251075745,
      "kl": 0.428955078125,
      "learning_rate": 2.3122307280499743e-07,
      "loss": 0.0144,
      "reward": 1.5250000208616257,
      "reward_std": 0.1928323432803154,
      "rewards/accuracy_reward": 0.5750000011175871,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2426
    },
    {
      "completion_length": 447.52500915527344,
      "epoch": 2.5173665111456716,
      "grad_norm": 4.043467044830322,
      "kl": 2.103515625,
      "learning_rate": 2.3025794216621748e-07,
      "loss": 0.4791,
      "reward": 1.3750000596046448,
      "reward_std": 0.6076977849006653,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2427
    },
    {
      "completion_length": 144.74999809265137,
      "epoch": 2.518403317781234,
      "grad_norm": 2.267932891845703,
      "kl": 0.782470703125,
      "learning_rate": 2.2929466250269442e-07,
      "loss": 0.089,
      "reward": 1.800000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2428
    },
    {
      "completion_length": 349.4500160217285,
      "epoch": 2.5194401244167963,
      "grad_norm": 1.6534165143966675,
      "kl": 1.912109375,
      "learning_rate": 2.2833323521865535e-07,
      "loss": 0.3491,
      "reward": 1.375,
      "reward_std": 0.3245587758719921,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2429
    },
    {
      "completion_length": 347.6000099182129,
      "epoch": 2.5204769310523587,
      "grad_norm": 2.545100212097168,
      "kl": 2.7041015625,
      "learning_rate": 2.2737366171562745e-07,
      "loss": 0.4399,
      "reward": 1.3250000029802322,
      "reward_std": 0.3966546282172203,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2430
    },
    {
      "completion_length": 247.72500801086426,
      "epoch": 2.521513737687921,
      "grad_norm": 0.5369862914085388,
      "kl": 0.9951171875,
      "learning_rate": 2.2641594339243533e-07,
      "loss": 0.1218,
      "reward": 1.600000023841858,
      "reward_std": 0.4922638423740864,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2431
    },
    {
      "completion_length": 157.50000190734863,
      "epoch": 2.522550544323484,
      "grad_norm": 0.46581900119781494,
      "kl": 0.801513671875,
      "learning_rate": 2.2546008164520065e-07,
      "loss": 0.114,
      "reward": 1.824999988079071,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8500000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2432
    },
    {
      "completion_length": 300.5749988555908,
      "epoch": 2.5235873509590463,
      "grad_norm": 1.6311348676681519,
      "kl": 1.603515625,
      "learning_rate": 2.2450607786733585e-07,
      "loss": 0.3072,
      "reward": 1.8000000417232513,
      "reward_std": 0.4472135603427887,
      "rewards/accuracy_reward": 0.9000000208616257,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2433
    },
    {
      "completion_length": 197.70000839233398,
      "epoch": 2.5246241575946087,
      "grad_norm": 2.101179599761963,
      "kl": 1.89111328125,
      "learning_rate": 2.2355393344954795e-07,
      "loss": 0.2222,
      "reward": 1.675000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2434
    },
    {
      "completion_length": 263.550012588501,
      "epoch": 2.525660964230171,
      "grad_norm": 0.7821170687675476,
      "kl": 1.248291015625,
      "learning_rate": 2.2260364977983134e-07,
      "loss": 0.2396,
      "reward": 1.6750000268220901,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.7750000096857548,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2435
    },
    {
      "completion_length": 295.6750011444092,
      "epoch": 2.5266977708657334,
      "grad_norm": 2.059645652770996,
      "kl": 2.422119140625,
      "learning_rate": 2.2165522824346873e-07,
      "loss": 0.3395,
      "reward": 1.4500000327825546,
      "reward_std": 0.3209684118628502,
      "rewards/accuracy_reward": 0.5500000063329935,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2436
    },
    {
      "completion_length": 290.22500228881836,
      "epoch": 2.5277345775012963,
      "grad_norm": 1.7699933052062988,
      "kl": 2.114013671875,
      "learning_rate": 2.2070867022302787e-07,
      "loss": 0.3448,
      "reward": 1.6500000357627869,
      "reward_std": 0.45977721363306046,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2437
    },
    {
      "completion_length": 212.00000381469727,
      "epoch": 2.528771384136858,
      "grad_norm": 1.1779626607894897,
      "kl": 0.392578125,
      "learning_rate": 2.1976397709836083e-07,
      "loss": 0.111,
      "reward": 1.7750000357627869,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2438
    },
    {
      "completion_length": 298.3000011444092,
      "epoch": 2.529808190772421,
      "grad_norm": 2.9897303581237793,
      "kl": 2.455078125,
      "learning_rate": 2.1882115024660031e-07,
      "loss": 0.3847,
      "reward": 1.3750000298023224,
      "reward_std": 0.3678968474268913,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2439
    },
    {
      "completion_length": 298.87500762939453,
      "epoch": 2.5308449974079834,
      "grad_norm": 2.722738265991211,
      "kl": 2.830078125,
      "learning_rate": 2.1788019104215878e-07,
      "loss": 0.3575,
      "reward": 1.600000023841858,
      "reward_std": 0.447213601320982,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2440
    },
    {
      "completion_length": 200.75000953674316,
      "epoch": 2.531881804043546,
      "grad_norm": 0.410030722618103,
      "kl": 0.4796142578125,
      "learning_rate": 2.169411008567253e-07,
      "loss": 0.1449,
      "reward": 1.350000023841858,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2441
    },
    {
      "completion_length": 220.3250026702881,
      "epoch": 2.5329186106791086,
      "grad_norm": 2.248901605606079,
      "kl": 1.46533203125,
      "learning_rate": 2.160038810592665e-07,
      "loss": 0.1829,
      "reward": 1.0250000059604645,
      "reward_std": 0.24873405322432518,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2442
    },
    {
      "completion_length": 410.2250118255615,
      "epoch": 2.5339554173146706,
      "grad_norm": 2.0582664012908936,
      "kl": 2.76171875,
      "learning_rate": 2.150685330160193e-07,
      "loss": 0.3368,
      "reward": 1.2999999970197678,
      "reward_std": 0.4092182479798794,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2443
    },
    {
      "completion_length": 211.92500495910645,
      "epoch": 2.5349922239502334,
      "grad_norm": 1.1172244548797607,
      "kl": 1.110107421875,
      "learning_rate": 2.1413505809049504e-07,
      "loss": 0.017,
      "reward": 1.4000000059604645,
      "reward_std": 0.29270508512854576,
      "rewards/accuracy_reward": 0.4750000052154064,
      "rewards/format_reward": 0.925000011920929,
      "step": 2444
    },
    {
      "completion_length": 354.07500648498535,
      "epoch": 2.5360290305857958,
      "grad_norm": 5.937859058380127,
      "kl": 2.6923828125,
      "learning_rate": 2.1320345764347287e-07,
      "loss": 0.405,
      "reward": 1.600000023841858,
      "reward_std": 0.4092182293534279,
      "rewards/accuracy_reward": 0.725000012665987,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2445
    },
    {
      "completion_length": 151.4000015258789,
      "epoch": 2.537065837221358,
      "grad_norm": 1.519677996635437,
      "kl": 0.740966796875,
      "learning_rate": 2.122737330329998e-07,
      "loss": 0.0922,
      "reward": 1.800000011920929,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2446
    },
    {
      "completion_length": 257.0250072479248,
      "epoch": 2.5381026438569205,
      "grad_norm": 1.4506206512451172,
      "kl": 1.140869140625,
      "learning_rate": 2.1134588561438783e-07,
      "loss": 0.1718,
      "reward": 1.8250000178813934,
      "reward_std": 0.292072132229805,
      "rewards/accuracy_reward": 0.9000000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 2447
    },
    {
      "completion_length": 193.92500495910645,
      "epoch": 2.539139450492483,
      "grad_norm": 1.0832746028900146,
      "kl": 1.373291015625,
      "learning_rate": 2.1041991674021356e-07,
      "loss": 0.2085,
      "reward": 1.5250000059604645,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2448
    },
    {
      "completion_length": 162.375,
      "epoch": 2.5401762571280457,
      "grad_norm": 1.0275706052780151,
      "kl": 0.849365234375,
      "learning_rate": 2.094958277603145e-07,
      "loss": 0.1154,
      "reward": 1.725000023841858,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2449
    },
    {
      "completion_length": 127.47500228881836,
      "epoch": 2.541213063763608,
      "grad_norm": 0.4823669493198395,
      "kl": 0.912353515625,
      "learning_rate": 2.0857362002178748e-07,
      "loss": 0.0776,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2450
    },
    {
      "completion_length": 163.20000457763672,
      "epoch": 2.5422498703991705,
      "grad_norm": 0.5657950639724731,
      "kl": 0.384765625,
      "learning_rate": 2.0765329486898692e-07,
      "loss": 0.0448,
      "reward": 1.425000011920929,
      "reward_std": 0.15685366839170456,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2451
    },
    {
      "completion_length": 146.375,
      "epoch": 2.543286677034733,
      "grad_norm": 4.356749534606934,
      "kl": 2.1279296875,
      "learning_rate": 2.0673485364352418e-07,
      "loss": 0.1607,
      "reward": 1.5000000298023224,
      "reward_std": 0.2920721285045147,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2452
    },
    {
      "completion_length": 297.27500343322754,
      "epoch": 2.5443234836702953,
      "grad_norm": 2.0256128311157227,
      "kl": 1.6573486328125,
      "learning_rate": 2.0581829768426192e-07,
      "loss": 0.2897,
      "reward": 1.7750000059604645,
      "reward_std": 0.41643911972641945,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2453
    },
    {
      "completion_length": 161.80000114440918,
      "epoch": 2.545360290305858,
      "grad_norm": 6.711730480194092,
      "kl": 2.5654296875,
      "learning_rate": 2.0490362832731703e-07,
      "loss": 0.2177,
      "reward": 1.4000000059604645,
      "reward_std": 0.21638589352369308,
      "rewards/accuracy_reward": 0.4499999973922968,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2454
    },
    {
      "completion_length": 242.82499885559082,
      "epoch": 2.5463970969414205,
      "grad_norm": 1.7809377908706665,
      "kl": 1.859375,
      "learning_rate": 2.0399084690605475e-07,
      "loss": 0.2834,
      "reward": 1.600000023841858,
      "reward_std": 0.3354101963341236,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2455
    },
    {
      "completion_length": 206.40000534057617,
      "epoch": 2.547433903576983,
      "grad_norm": 4.902795791625977,
      "kl": 2.54443359375,
      "learning_rate": 2.0307995475108837e-07,
      "loss": 0.2417,
      "reward": 1.3750000298023224,
      "reward_std": 0.2920721098780632,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2456
    },
    {
      "completion_length": 149.37500381469727,
      "epoch": 2.5484707102125452,
      "grad_norm": 0.20117555558681488,
      "kl": 0.6728515625,
      "learning_rate": 2.0217095319027706e-07,
      "loss": 0.0984,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2457
    },
    {
      "completion_length": 165.55000114440918,
      "epoch": 2.5495075168481076,
      "grad_norm": 1.5324474573135376,
      "kl": 1.467041015625,
      "learning_rate": 2.0126384354872468e-07,
      "loss": 0.1844,
      "reward": 1.475000023841858,
      "reward_std": 0.23617041110992432,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2458
    },
    {
      "completion_length": 251.2500057220459,
      "epoch": 2.5505443234836704,
      "grad_norm": 1.3593672513961792,
      "kl": 1.3779296875,
      "learning_rate": 2.0035862714877628e-07,
      "loss": 0.2421,
      "reward": 1.5000000447034836,
      "reward_std": 0.27950849011540413,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2459
    },
    {
      "completion_length": 248.45000457763672,
      "epoch": 2.551581130119233,
      "grad_norm": 2.4053030014038086,
      "kl": 2.6328125,
      "learning_rate": 1.9945530531001681e-07,
      "loss": 0.3059,
      "reward": 1.6750000417232513,
      "reward_std": 0.4472135864198208,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2460
    },
    {
      "completion_length": 249.30000495910645,
      "epoch": 2.552617936754795,
      "grad_norm": 1.3433115482330322,
      "kl": 2.554931640625,
      "learning_rate": 1.9855387934927094e-07,
      "loss": 0.3007,
      "reward": 1.7000000178813934,
      "reward_std": 0.5156789049506187,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2461
    },
    {
      "completion_length": 343.4500045776367,
      "epoch": 2.5536547433903576,
      "grad_norm": 2.275127410888672,
      "kl": 2.16015625,
      "learning_rate": 1.97654350580598e-07,
      "loss": 0.3847,
      "reward": 1.5000000596046448,
      "reward_std": 0.4723408445715904,
      "rewards/accuracy_reward": 0.6250000223517418,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2462
    },
    {
      "completion_length": 247.95000648498535,
      "epoch": 2.55469155002592,
      "grad_norm": 2.2795634269714355,
      "kl": 1.2908935546875,
      "learning_rate": 1.9675672031529225e-07,
      "loss": 0.2241,
      "reward": 1.625,
      "reward_std": 0.22894952073693275,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 2463
    },
    {
      "completion_length": 288.8000030517578,
      "epoch": 2.555728356661483,
      "grad_norm": 3.2034480571746826,
      "kl": 1.3037109375,
      "learning_rate": 1.958609898618803e-07,
      "loss": 0.2907,
      "reward": 1.7000000476837158,
      "reward_std": 0.3913119100034237,
      "rewards/accuracy_reward": 0.8000000268220901,
      "rewards/format_reward": 0.9000000357627869,
      "step": 2464
    },
    {
      "completion_length": 157.20000648498535,
      "epoch": 2.556765163297045,
      "grad_norm": 1.0970335006713867,
      "kl": 0.60205078125,
      "learning_rate": 1.9496716052612012e-07,
      "loss": 0.0676,
      "reward": 1.75,
      "reward_std": 0.3407529331743717,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2465
    },
    {
      "completion_length": 139.75000190734863,
      "epoch": 2.5578019699326076,
      "grad_norm": 0.9260252714157104,
      "kl": 0.343505859375,
      "learning_rate": 1.940752336109965e-07,
      "loss": 0.0922,
      "reward": 1.625,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2466
    },
    {
      "completion_length": 161.4000015258789,
      "epoch": 2.55883877656817,
      "grad_norm": 1.3939898014068604,
      "kl": 0.65576171875,
      "learning_rate": 1.93185210416723e-07,
      "loss": 0.1041,
      "reward": 1.625,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2467
    },
    {
      "completion_length": 257.87500953674316,
      "epoch": 2.5598755832037323,
      "grad_norm": 2.3138604164123535,
      "kl": 0.784912109375,
      "learning_rate": 1.9229709224073662e-07,
      "loss": 0.2016,
      "reward": 1.6500000357627869,
      "reward_std": 0.403875507414341,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 2468
    },
    {
      "completion_length": 335.100004196167,
      "epoch": 2.560912389839295,
      "grad_norm": 1.0027917623519897,
      "kl": 2.3531494140625,
      "learning_rate": 1.9141088037769806e-07,
      "loss": 0.4347,
      "reward": 1.6000000536441803,
      "reward_std": 0.5054604634642601,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2469
    },
    {
      "completion_length": 390.94999504089355,
      "epoch": 2.5619491964748575,
      "grad_norm": 1.4724165201187134,
      "kl": 3.13818359375,
      "learning_rate": 1.905265761194881e-07,
      "loss": 0.3548,
      "reward": 1.3999999910593033,
      "reward_std": 0.4922638610005379,
      "rewards/accuracy_reward": 0.5500000063329935,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2470
    },
    {
      "completion_length": 150.72500228881836,
      "epoch": 2.56298600311042,
      "grad_norm": 0.501360297203064,
      "kl": 0.68994140625,
      "learning_rate": 1.8964418075520824e-07,
      "loss": 0.1017,
      "reward": 1.4500000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2471
    },
    {
      "completion_length": 412.37500381469727,
      "epoch": 2.5640228097459823,
      "grad_norm": 3.1552064418792725,
      "kl": 2.12353515625,
      "learning_rate": 1.88763695571176e-07,
      "loss": 0.3557,
      "reward": 1.3250000327825546,
      "reward_std": 0.49746808782219887,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2472
    },
    {
      "completion_length": 156.25000190734863,
      "epoch": 2.5650596163815447,
      "grad_norm": 0.8898472785949707,
      "kl": 0.760986328125,
      "learning_rate": 1.878851218509247e-07,
      "loss": 0.1242,
      "reward": 1.550000011920929,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2473
    },
    {
      "completion_length": 203.9500026702881,
      "epoch": 2.5660964230171075,
      "grad_norm": 1.1891981363296509,
      "kl": 1.92431640625,
      "learning_rate": 1.870084608752012e-07,
      "loss": 0.2618,
      "reward": 1.5000000149011612,
      "reward_std": 0.31199515610933304,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2474
    },
    {
      "completion_length": 257.6750030517578,
      "epoch": 2.56713322965267,
      "grad_norm": 0.5489189624786377,
      "kl": 0.83642578125,
      "learning_rate": 1.8613371392196483e-07,
      "loss": 0.1977,
      "reward": 1.6500000059604645,
      "reward_std": 0.5156789161264896,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2475
    },
    {
      "completion_length": 219.37500381469727,
      "epoch": 2.5681700362882323,
      "grad_norm": 0.6120697259902954,
      "kl": 0.81640625,
      "learning_rate": 1.8526088226638276e-07,
      "loss": 0.1443,
      "reward": 1.3750000298023224,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2476
    },
    {
      "completion_length": 190.6250057220459,
      "epoch": 2.5692068429237946,
      "grad_norm": 0.4692276418209076,
      "kl": 0.5260009765625,
      "learning_rate": 1.843899671808324e-07,
      "loss": 0.1302,
      "reward": 1.800000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2477
    },
    {
      "completion_length": 240.24999809265137,
      "epoch": 2.570243649559357,
      "grad_norm": 0.4318724274635315,
      "kl": 0.756103515625,
      "learning_rate": 1.8352096993489582e-07,
      "loss": 0.1839,
      "reward": 1.625,
      "reward_std": 0.1928323470056057,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 2478
    },
    {
      "completion_length": 294.5500011444092,
      "epoch": 2.57128045619492,
      "grad_norm": 1.0414520502090454,
      "kl": 2.3583984375,
      "learning_rate": 1.826538917953597e-07,
      "loss": 0.3722,
      "reward": 1.5500000566244125,
      "reward_std": 0.436362162232399,
      "rewards/accuracy_reward": 0.6500000152736902,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2479
    },
    {
      "completion_length": 297.925012588501,
      "epoch": 2.5723172628304822,
      "grad_norm": 0.8396540880203247,
      "kl": 1.47265625,
      "learning_rate": 1.8178873402621295e-07,
      "loss": 0.2585,
      "reward": 1.5500000417232513,
      "reward_std": 0.4296357333660126,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2480
    },
    {
      "completion_length": 213.0999984741211,
      "epoch": 2.5733540694660446,
      "grad_norm": 0.303814560174942,
      "kl": 1.07958984375,
      "learning_rate": 1.8092549788864605e-07,
      "loss": 0.2173,
      "reward": 1.7000000476837158,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2481
    },
    {
      "completion_length": 186.37500381469727,
      "epoch": 2.574390876101607,
      "grad_norm": 0.6133432984352112,
      "kl": 0.62255859375,
      "learning_rate": 1.800641846410469e-07,
      "loss": 0.0968,
      "reward": 1.6500000059604645,
      "reward_std": 0.40387552231550217,
      "rewards/accuracy_reward": 0.700000025331974,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2482
    },
    {
      "completion_length": 151.67499923706055,
      "epoch": 2.5754276827371694,
      "grad_norm": 0.5538397431373596,
      "kl": 0.316650390625,
      "learning_rate": 1.7920479553900083e-07,
      "loss": 0.0406,
      "reward": 1.5999999940395355,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.6250000223517418,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2483
    },
    {
      "completion_length": 293.9000072479248,
      "epoch": 2.576464489372732,
      "grad_norm": 8.192227363586426,
      "kl": 4.710693359375,
      "learning_rate": 1.7834733183528819e-07,
      "loss": 0.5385,
      "reward": 1.574999988079071,
      "reward_std": 0.30463573709130287,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2484
    },
    {
      "completion_length": 160.57500076293945,
      "epoch": 2.5775012960082946,
      "grad_norm": 0.4454459547996521,
      "kl": 0.6864013671875,
      "learning_rate": 1.774917947798834e-07,
      "loss": 0.0911,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2485
    },
    {
      "completion_length": 352.65000915527344,
      "epoch": 2.578538102643857,
      "grad_norm": 2.296649932861328,
      "kl": 2.75,
      "learning_rate": 1.7663818561995028e-07,
      "loss": 0.333,
      "reward": 1.6250000298023224,
      "reward_std": 0.5590169914066792,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2486
    },
    {
      "completion_length": 244.62500762939453,
      "epoch": 2.5795749092794193,
      "grad_norm": 0.9369842410087585,
      "kl": 0.935546875,
      "learning_rate": 1.7578650559984432e-07,
      "loss": 0.2175,
      "reward": 1.5,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 2487
    },
    {
      "completion_length": 183.32500457763672,
      "epoch": 2.5806117159149817,
      "grad_norm": 0.23203659057617188,
      "kl": 0.950439453125,
      "learning_rate": 1.749367559611078e-07,
      "loss": 0.1333,
      "reward": 1.6250000298023224,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2488
    },
    {
      "completion_length": 95.25000190734863,
      "epoch": 2.5816485225505446,
      "grad_norm": 0.3252972662448883,
      "kl": 0.1888427734375,
      "learning_rate": 1.74088937942469e-07,
      "loss": 0.0135,
      "reward": 1.7000000178813934,
      "reward_std": 0.06846531480550766,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 1.0,
      "step": 2489
    },
    {
      "completion_length": 101.52499961853027,
      "epoch": 2.582685329186107,
      "grad_norm": 0.22385184466838837,
      "kl": 0.307373046875,
      "learning_rate": 1.732430527798401e-07,
      "loss": 0.0108,
      "reward": 1.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 1.0,
      "step": 2490
    },
    {
      "completion_length": 146.85000610351562,
      "epoch": 2.5837221358216693,
      "grad_norm": 0.8692068457603455,
      "kl": 0.401611328125,
      "learning_rate": 1.723991017063169e-07,
      "loss": 0.0919,
      "reward": 1.7000000178813934,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2491
    },
    {
      "completion_length": 140.04999923706055,
      "epoch": 2.5847589424572317,
      "grad_norm": 0.6108434200286865,
      "kl": 0.669677734375,
      "learning_rate": 1.7155708595217461e-07,
      "loss": 0.08,
      "reward": 1.6250000298023224,
      "reward_std": 0.1928323358297348,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2492
    },
    {
      "completion_length": 99.80000114440918,
      "epoch": 2.585795749092794,
      "grad_norm": 0.22735685110092163,
      "kl": 0.251220703125,
      "learning_rate": 1.7071700674486722e-07,
      "loss": 0.0102,
      "reward": 1.7750000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.775000000372529,
      "rewards/format_reward": 1.0,
      "step": 2493
    },
    {
      "completion_length": 146.47500228881836,
      "epoch": 2.586832555728357,
      "grad_norm": 0.5281745195388794,
      "kl": 0.3465576171875,
      "learning_rate": 1.6987886530902596e-07,
      "loss": 0.0641,
      "reward": 1.925000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.9500000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2494
    },
    {
      "completion_length": 146.6250057220459,
      "epoch": 2.5878693623639193,
      "grad_norm": 0.5651646852493286,
      "kl": 0.44287109375,
      "learning_rate": 1.6904266286645814e-07,
      "loss": 0.0849,
      "reward": 1.475000023841858,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2495
    },
    {
      "completion_length": 256.0000057220459,
      "epoch": 2.5889061689994817,
      "grad_norm": 1.038611650466919,
      "kl": 0.97900390625,
      "learning_rate": 1.6820840063614217e-07,
      "loss": 0.2229,
      "reward": 1.8250000476837158,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.9000000208616257,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2496
    },
    {
      "completion_length": 185.05000114440918,
      "epoch": 2.589942975635044,
      "grad_norm": 0.42412254214286804,
      "kl": 0.578125,
      "learning_rate": 1.6737607983423092e-07,
      "loss": 0.1259,
      "reward": 1.7000000178813934,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2497
    },
    {
      "completion_length": 188.32500457763672,
      "epoch": 2.5909797822706064,
      "grad_norm": 2.296624183654785,
      "kl": 0.845458984375,
      "learning_rate": 1.665457016740453e-07,
      "loss": 0.1409,
      "reward": 1.5250000059604645,
      "reward_std": 0.1369306445121765,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2498
    },
    {
      "completion_length": 141.37500381469727,
      "epoch": 2.5920165889061693,
      "grad_norm": 0.716563880443573,
      "kl": 0.598876953125,
      "learning_rate": 1.657172673660744e-07,
      "loss": 0.045,
      "reward": 1.7750000357627869,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2499
    },
    {
      "completion_length": 336.2750072479248,
      "epoch": 2.593053395541731,
      "grad_norm": 2.1289913654327393,
      "kl": 1.87841796875,
      "learning_rate": 1.648907781179748e-07,
      "loss": 0.3576,
      "reward": 1.6500000059604645,
      "reward_std": 0.5915036089718342,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2500
    },
    {
      "completion_length": 199.65000343322754,
      "epoch": 2.594090202177294,
      "grad_norm": 0.9478914737701416,
      "kl": 1.04638671875,
      "learning_rate": 1.6406623513456666e-07,
      "loss": 0.1576,
      "reward": 1.6500000059604645,
      "reward_std": 0.18090170249342918,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2501
    },
    {
      "completion_length": 157.75000190734863,
      "epoch": 2.5951270088128564,
      "grad_norm": 0.3171178102493286,
      "kl": 0.689208984375,
      "learning_rate": 1.6324363961783307e-07,
      "loss": 0.1326,
      "reward": 1.699999988079071,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2502
    },
    {
      "completion_length": 295.02500343322754,
      "epoch": 2.596163815448419,
      "grad_norm": 0.9354013800621033,
      "kl": 1.55810546875,
      "learning_rate": 1.6242299276691803e-07,
      "loss": 0.2669,
      "reward": 1.4500000178813934,
      "reward_std": 0.30463573709130287,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2503
    },
    {
      "completion_length": 379.0500030517578,
      "epoch": 2.5972006220839816,
      "grad_norm": 1.1785205602645874,
      "kl": 1.189453125,
      "learning_rate": 1.6160429577812574e-07,
      "loss": 0.2367,
      "reward": 1.4000000208616257,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2504
    },
    {
      "completion_length": 397.5249996185303,
      "epoch": 2.5982374287195436,
      "grad_norm": 7.383028030395508,
      "kl": 5.77880859375,
      "learning_rate": 1.6078754984491722e-07,
      "loss": 0.5118,
      "reward": 1.375,
      "reward_std": 0.45977720245718956,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.8499999940395355,
      "step": 2505
    },
    {
      "completion_length": 348.85000801086426,
      "epoch": 2.5992742353551064,
      "grad_norm": 10.178338050842285,
      "kl": 5.079345703125,
      "learning_rate": 1.5997275615790956e-07,
      "loss": 0.5044,
      "reward": 1.2000000327825546,
      "reward_std": 0.4472135677933693,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2506
    },
    {
      "completion_length": 241.8250026702881,
      "epoch": 2.6003110419906688,
      "grad_norm": 0.9275565147399902,
      "kl": 3.25,
      "learning_rate": 1.5915991590487332e-07,
      "loss": 0.4125,
      "reward": 1.5500000417232513,
      "reward_std": 0.4472135864198208,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2507
    },
    {
      "completion_length": 353.8500061035156,
      "epoch": 2.601347848626231,
      "grad_norm": 1.4856528043746948,
      "kl": 1.8779296875,
      "learning_rate": 1.58349030270733e-07,
      "loss": 0.3466,
      "reward": 1.4250000417232513,
      "reward_std": 0.472340852022171,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2508
    },
    {
      "completion_length": 198.45000648498535,
      "epoch": 2.6023846552617935,
      "grad_norm": 1.1513665914535522,
      "kl": 0.89697265625,
      "learning_rate": 1.575401004375615e-07,
      "loss": 0.07,
      "reward": 1.725000023841858,
      "reward_std": 0.41643914580345154,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2509
    },
    {
      "completion_length": 146.0500030517578,
      "epoch": 2.603421461897356,
      "grad_norm": 0.6724655032157898,
      "kl": 1.073974609375,
      "learning_rate": 1.567331275845829e-07,
      "loss": 0.1486,
      "reward": 1.425000011920929,
      "reward_std": 0.25609346106648445,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2510
    },
    {
      "completion_length": 189.22500228881836,
      "epoch": 2.6044582685329187,
      "grad_norm": 0.6389365196228027,
      "kl": 0.9525146484375,
      "learning_rate": 1.5592811288816722e-07,
      "loss": 0.16,
      "reward": 1.5999999940395355,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2511
    },
    {
      "completion_length": 182.70000457763672,
      "epoch": 2.605495075168481,
      "grad_norm": 0.820368766784668,
      "kl": 1.23486328125,
      "learning_rate": 1.5512505752182987e-07,
      "loss": 0.1582,
      "reward": 1.3250000327825546,
      "reward_std": 0.27228759601712227,
      "rewards/accuracy_reward": 0.4000000022351742,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2512
    },
    {
      "completion_length": 344.07500648498535,
      "epoch": 2.6065318818040435,
      "grad_norm": 1.3516125679016113,
      "kl": 1.3974609375,
      "learning_rate": 1.5432396265623043e-07,
      "loss": 0.2858,
      "reward": 1.550000011920929,
      "reward_std": 0.5967078544199467,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2513
    },
    {
      "completion_length": 182.92500686645508,
      "epoch": 2.607568688439606,
      "grad_norm": 0.5100857019424438,
      "kl": 0.299560546875,
      "learning_rate": 1.535248294591708e-07,
      "loss": 0.1021,
      "reward": 1.7750000357627869,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2514
    },
    {
      "completion_length": 240.20001029968262,
      "epoch": 2.6086054950751683,
      "grad_norm": 0.9690713882446289,
      "kl": 0.4453125,
      "learning_rate": 1.5272765909559272e-07,
      "loss": 0.1679,
      "reward": 1.6500000059604645,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2515
    },
    {
      "completion_length": 98.5,
      "epoch": 2.609642301710731,
      "grad_norm": 0.041470013558864594,
      "kl": 0.2998046875,
      "learning_rate": 1.5193245272757667e-07,
      "loss": 0.0106,
      "reward": 1.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 1.0,
      "step": 2516
    },
    {
      "completion_length": 190.20000457763672,
      "epoch": 2.6106791083462935,
      "grad_norm": 1.3450777530670166,
      "kl": 0.696533203125,
      "learning_rate": 1.5113921151434018e-07,
      "loss": 0.1078,
      "reward": 1.725000023841858,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2517
    },
    {
      "completion_length": 127.20000076293945,
      "epoch": 2.611715914981856,
      "grad_norm": 0.3597521483898163,
      "kl": 0.3255615234375,
      "learning_rate": 1.503479366122368e-07,
      "loss": 0.0565,
      "reward": 1.6500000357627869,
      "reward_std": 0.18026872724294662,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2518
    },
    {
      "completion_length": 292.47500228881836,
      "epoch": 2.6127527216174182,
      "grad_norm": 7.4777140617370605,
      "kl": 4.23486328125,
      "learning_rate": 1.4955862917475195e-07,
      "loss": 0.4481,
      "reward": 1.7250000536441803,
      "reward_std": 0.4290027469396591,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2519
    },
    {
      "completion_length": 236.0000057220459,
      "epoch": 2.6137895282529806,
      "grad_norm": 10.597983360290527,
      "kl": 4.14990234375,
      "learning_rate": 1.4877129035250475e-07,
      "loss": 0.4512,
      "reward": 1.600000023841858,
      "reward_std": 0.36053743213415146,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.925000011920929,
      "step": 2520
    },
    {
      "completion_length": 103.57500076293945,
      "epoch": 2.6148263348885434,
      "grad_norm": 0.04561980068683624,
      "kl": 0.3236083984375,
      "learning_rate": 1.4798592129324346e-07,
      "loss": 0.0102,
      "reward": 1.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 1.0,
      "step": 2521
    },
    {
      "completion_length": 167.7750015258789,
      "epoch": 2.615863141524106,
      "grad_norm": 0.532559871673584,
      "kl": 0.419189453125,
      "learning_rate": 1.4720252314184558e-07,
      "loss": 0.0683,
      "reward": 1.4000000059604645,
      "reward_std": 0.23617040365934372,
      "rewards/accuracy_reward": 0.42500001564621925,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2522
    },
    {
      "completion_length": 110.77500343322754,
      "epoch": 2.616899948159668,
      "grad_norm": 0.8255411982536316,
      "kl": 1.947021484375,
      "learning_rate": 1.4642109704031442e-07,
      "loss": -0.0001,
      "reward": 1.8000000417232513,
      "reward_std": 0.32818929851055145,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2523
    },
    {
      "completion_length": 142.0750026702881,
      "epoch": 2.6179367547952306,
      "grad_norm": 1.331695795059204,
      "kl": 0.9169921875,
      "learning_rate": 1.456416441277803e-07,
      "loss": 0.0951,
      "reward": 1.550000011920929,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.5750000197440386,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2524
    },
    {
      "completion_length": 138.17500686645508,
      "epoch": 2.618973561430793,
      "grad_norm": 0.2185720056295395,
      "kl": 0.37744140625,
      "learning_rate": 1.4486416554049546e-07,
      "loss": 0.065,
      "reward": 1.3250000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.3500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2525
    },
    {
      "completion_length": 130.39999771118164,
      "epoch": 2.620010368066356,
      "grad_norm": 1.1363494396209717,
      "kl": 0.65234375,
      "learning_rate": 1.4408866241183484e-07,
      "loss": 0.0917,
      "reward": 1.7000000178813934,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2526
    },
    {
      "completion_length": 97.60000038146973,
      "epoch": 2.621047174701918,
      "grad_norm": 0.07096569985151291,
      "kl": 0.150634765625,
      "learning_rate": 1.4331513587229321e-07,
      "loss": 0.0057,
      "reward": 1.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 1.0,
      "step": 2527
    },
    {
      "completion_length": 184.72500228881836,
      "epoch": 2.6220839813374806,
      "grad_norm": 1.59454345703125,
      "kl": 1.71142578125,
      "learning_rate": 1.42543587049485e-07,
      "loss": 0.2033,
      "reward": 1.5500000417232513,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.6000000108033419,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2528
    },
    {
      "completion_length": 235.95001029968262,
      "epoch": 2.623120787973043,
      "grad_norm": 1.0381091833114624,
      "kl": 1.610595703125,
      "learning_rate": 1.417740170681398e-07,
      "loss": 0.2611,
      "reward": 1.5750000476837158,
      "reward_std": 0.391311913728714,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2529
    },
    {
      "completion_length": 188.27500534057617,
      "epoch": 2.6241575946086053,
      "grad_norm": 0.8474007248878479,
      "kl": 1.332763671875,
      "learning_rate": 1.410064270501044e-07,
      "loss": 0.1945,
      "reward": 1.7750000357627869,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2530
    },
    {
      "completion_length": 164.70000839233398,
      "epoch": 2.625194401244168,
      "grad_norm": 0.18072180449962616,
      "kl": 0.326171875,
      "learning_rate": 1.4024081811433809e-07,
      "loss": 0.0673,
      "reward": 1.550000011920929,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2531
    },
    {
      "completion_length": 141.6250057220459,
      "epoch": 2.6262312078797305,
      "grad_norm": 0.4177073538303375,
      "kl": 0.4189453125,
      "learning_rate": 1.3947719137691257e-07,
      "loss": 0.0606,
      "reward": 1.5000000149011612,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2532
    },
    {
      "completion_length": 295.6750087738037,
      "epoch": 2.627268014515293,
      "grad_norm": 1.028651475906372,
      "kl": 0.92431640625,
      "learning_rate": 1.3871554795100972e-07,
      "loss": 0.2564,
      "reward": 1.4250000417232513,
      "reward_std": 0.44533543661236763,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2533
    },
    {
      "completion_length": 183.12500381469727,
      "epoch": 2.6283048211508553,
      "grad_norm": 0.6243255734443665,
      "kl": 0.385986328125,
      "learning_rate": 1.3795588894692123e-07,
      "loss": 0.111,
      "reward": 1.7750000357627869,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2534
    },
    {
      "completion_length": 240.10000801086426,
      "epoch": 2.6293416277864177,
      "grad_norm": 0.9825228452682495,
      "kl": 0.7509765625,
      "learning_rate": 1.371982154720448e-07,
      "loss": 0.1828,
      "reward": 1.625,
      "reward_std": 0.4850429818034172,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2535
    },
    {
      "completion_length": 413.07501220703125,
      "epoch": 2.6303784344219805,
      "grad_norm": 17.761478424072266,
      "kl": 6.6474609375,
      "learning_rate": 1.3644252863088441e-07,
      "loss": 0.734,
      "reward": 1.5250000059604645,
      "reward_std": 0.5048274733126163,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2536
    },
    {
      "completion_length": 233.42500114440918,
      "epoch": 2.631415241057543,
      "grad_norm": 0.3659479022026062,
      "kl": 0.511474609375,
      "learning_rate": 1.356888295250472e-07,
      "loss": 0.1492,
      "reward": 1.725000023841858,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 2537
    },
    {
      "completion_length": 245.32499885559082,
      "epoch": 2.6324520476931053,
      "grad_norm": 0.8607601523399353,
      "kl": 1.2119140625,
      "learning_rate": 1.3493711925324392e-07,
      "loss": 0.2604,
      "reward": 1.3250000476837158,
      "reward_std": 0.447213564068079,
      "rewards/accuracy_reward": 0.450000012293458,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2538
    },
    {
      "completion_length": 234.35000038146973,
      "epoch": 2.6334888543286676,
      "grad_norm": 0.34118545055389404,
      "kl": 0.68408203125,
      "learning_rate": 1.3418739891128505e-07,
      "loss": 0.1452,
      "reward": 1.4000000208616257,
      "reward_std": 0.25609347969293594,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.925000011920929,
      "step": 2539
    },
    {
      "completion_length": 144.25000190734863,
      "epoch": 2.63452566096423,
      "grad_norm": 0.8021541833877563,
      "kl": 0.9150390625,
      "learning_rate": 1.3343966959208053e-07,
      "loss": 0.1147,
      "reward": 1.7000000178813934,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2540
    },
    {
      "completion_length": 150.25000190734863,
      "epoch": 2.635562467599793,
      "grad_norm": 0.40674591064453125,
      "kl": 0.4403076171875,
      "learning_rate": 1.326939323856386e-07,
      "loss": 0.0571,
      "reward": 1.6250000298023224,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2541
    },
    {
      "completion_length": 320.575008392334,
      "epoch": 2.6365992742353552,
      "grad_norm": 1.7479718923568726,
      "kl": 1.26171875,
      "learning_rate": 1.319501883790618e-07,
      "loss": 0.3456,
      "reward": 1.4500000476837158,
      "reward_std": 0.5090910047292709,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2542
    },
    {
      "completion_length": 234.05000495910645,
      "epoch": 2.6376360808709176,
      "grad_norm": 0.6886463165283203,
      "kl": 0.7119140625,
      "learning_rate": 1.3120843865654892e-07,
      "loss": 0.2044,
      "reward": 1.5000000149011612,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 2543
    },
    {
      "completion_length": 198.32500457763672,
      "epoch": 2.63867288750648,
      "grad_norm": 1.4801619052886963,
      "kl": 1.132568359375,
      "learning_rate": 1.3046868429939046e-07,
      "loss": 0.1345,
      "reward": 1.7500000298023224,
      "reward_std": 0.45977721735835075,
      "rewards/accuracy_reward": 0.8250000327825546,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2544
    },
    {
      "completion_length": 404.90001487731934,
      "epoch": 2.6397096941420424,
      "grad_norm": 1.8300055265426636,
      "kl": 2.587890625,
      "learning_rate": 1.2973092638596833e-07,
      "loss": 0.3177,
      "reward": 1.5500000417232513,
      "reward_std": 0.4797002300620079,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2545
    },
    {
      "completion_length": 266.77499771118164,
      "epoch": 2.640746500777605,
      "grad_norm": 0.3893660604953766,
      "kl": 0.8851318359375,
      "learning_rate": 1.289951659917541e-07,
      "loss": 0.2062,
      "reward": 1.7250000536441803,
      "reward_std": 0.3354102149605751,
      "rewards/accuracy_reward": 0.8000000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2546
    },
    {
      "completion_length": 307.2250003814697,
      "epoch": 2.6417833074131676,
      "grad_norm": 0.962757408618927,
      "kl": 1.403076171875,
      "learning_rate": 1.282614041893081e-07,
      "loss": 0.2889,
      "reward": 1.600000023841858,
      "reward_std": 0.3533165454864502,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2547
    },
    {
      "completion_length": 262.9250030517578,
      "epoch": 2.64282011404873,
      "grad_norm": 0.9546141624450684,
      "kl": 1.755615234375,
      "learning_rate": 1.2752964204827622e-07,
      "loss": 0.2759,
      "reward": 1.7000000178813934,
      "reward_std": 0.3913118727505207,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2548
    },
    {
      "completion_length": 249.20000076293945,
      "epoch": 2.6438569206842923,
      "grad_norm": 0.8743767738342285,
      "kl": 1.0859375,
      "learning_rate": 1.2679988063539e-07,
      "loss": 0.254,
      "reward": 1.199999988079071,
      "reward_std": 0.31199514120817184,
      "rewards/accuracy_reward": 0.2750000022351742,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2549
    },
    {
      "completion_length": 149.12500381469727,
      "epoch": 2.6448937273198547,
      "grad_norm": 0.5427263975143433,
      "kl": 0.40478515625,
      "learning_rate": 1.260721210144638e-07,
      "loss": 0.0724,
      "reward": 1.800000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2550
    },
    {
      "completion_length": 219.05000495910645,
      "epoch": 2.6459305339554176,
      "grad_norm": 0.5831300020217896,
      "kl": 1.981201171875,
      "learning_rate": 1.2534636424639517e-07,
      "loss": 0.3007,
      "reward": 1.4500000178813934,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.5250000078231096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2551
    },
    {
      "completion_length": 242.10000228881836,
      "epoch": 2.64696734059098,
      "grad_norm": 0.8162980079650879,
      "kl": 2.068359375,
      "learning_rate": 1.2462261138915963e-07,
      "loss": 0.2914,
      "reward": 1.6500000357627869,
      "reward_std": 0.4472135901451111,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2552
    },
    {
      "completion_length": 245.3249969482422,
      "epoch": 2.6480041472265423,
      "grad_norm": 0.6796807050704956,
      "kl": 0.99072265625,
      "learning_rate": 1.2390086349781426e-07,
      "loss": 0.2283,
      "reward": 1.3499999940395355,
      "reward_std": 0.24873405322432518,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2553
    },
    {
      "completion_length": 156.1500072479248,
      "epoch": 2.6490409538621047,
      "grad_norm": 1.1500164270401,
      "kl": 0.616455078125,
      "learning_rate": 1.2318112162449125e-07,
      "loss": 0.0409,
      "reward": 1.5250000059604645,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.5500000007450581,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2554
    },
    {
      "completion_length": 195.2750015258789,
      "epoch": 2.650077760497667,
      "grad_norm": 0.39828255772590637,
      "kl": 0.58984375,
      "learning_rate": 1.2246338681839948e-07,
      "loss": 0.0957,
      "reward": 1.75,
      "reward_std": 0.1604841947555542,
      "rewards/accuracy_reward": 0.8000000007450581,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2555
    },
    {
      "completion_length": 296.325008392334,
      "epoch": 2.65111456713323,
      "grad_norm": 0.9349043965339661,
      "kl": 1.307373046875,
      "learning_rate": 1.2174766012582166e-07,
      "loss": 0.2695,
      "reward": 1.5250000357627869,
      "reward_std": 0.4723408780992031,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2556
    },
    {
      "completion_length": 95.20000076293945,
      "epoch": 2.6521513737687923,
      "grad_norm": 2.851870059967041,
      "kl": 2.4708251953125,
      "learning_rate": 1.2103394259011374e-07,
      "loss": 0.1096,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2557
    },
    {
      "completion_length": 206.10000801086426,
      "epoch": 2.6531881804043547,
      "grad_norm": 1.3341933488845825,
      "kl": 0.7060546875,
      "learning_rate": 1.203222352517021e-07,
      "loss": 0.1499,
      "reward": 1.850000023841858,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.9000000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2558
    },
    {
      "completion_length": 279.50000381469727,
      "epoch": 2.654224987039917,
      "grad_norm": 2.5046353340148926,
      "kl": 2.52685546875,
      "learning_rate": 1.1961253914808317e-07,
      "loss": 0.2362,
      "reward": 1.5750000476837158,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2559
    },
    {
      "completion_length": 284.70000076293945,
      "epoch": 2.6552617936754794,
      "grad_norm": 0.47059956192970276,
      "kl": 1.3046875,
      "learning_rate": 1.1890485531382123e-07,
      "loss": 0.185,
      "reward": 1.5500000268220901,
      "reward_std": 0.31719935685396194,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2560
    },
    {
      "completion_length": 160.92499923706055,
      "epoch": 2.6562986003110423,
      "grad_norm": 0.7610751390457153,
      "kl": 0.439208984375,
      "learning_rate": 1.181991847805472e-07,
      "loss": 0.1147,
      "reward": 1.675000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2561
    },
    {
      "completion_length": 199.7249984741211,
      "epoch": 2.657335406946604,
      "grad_norm": 1.4734550714492798,
      "kl": 1.25146484375,
      "learning_rate": 1.1749552857695722e-07,
      "loss": 0.1626,
      "reward": 1.3000000268220901,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2562
    },
    {
      "completion_length": 394.5750102996826,
      "epoch": 2.658372213582167,
      "grad_norm": 2.0895891189575195,
      "kl": 2.760009765625,
      "learning_rate": 1.1679388772881111e-07,
      "loss": 0.3481,
      "reward": 1.4500000178813934,
      "reward_std": 0.4092182442545891,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2563
    },
    {
      "completion_length": 153.67500495910645,
      "epoch": 2.6594090202177294,
      "grad_norm": 0.26248764991760254,
      "kl": 0.72021484375,
      "learning_rate": 1.1609426325893069e-07,
      "loss": 0.0964,
      "reward": 1.800000011920929,
      "reward_std": 0.16770510375499725,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2564
    },
    {
      "completion_length": 250.37500953674316,
      "epoch": 2.660445826853292,
      "grad_norm": 0.523044764995575,
      "kl": 1.02783203125,
      "learning_rate": 1.1539665618719814e-07,
      "loss": 0.2389,
      "reward": 1.5250000357627869,
      "reward_std": 0.36053744703531265,
      "rewards/accuracy_reward": 0.600000012665987,
      "rewards/format_reward": 0.925000011920929,
      "step": 2565
    },
    {
      "completion_length": 204.67500495910645,
      "epoch": 2.661482633488854,
      "grad_norm": 0.3818694055080414,
      "kl": 0.9764404296875,
      "learning_rate": 1.1470106753055498e-07,
      "loss": 0.1364,
      "reward": 1.4249999970197678,
      "reward_std": 0.25609343871474266,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 2566
    },
    {
      "completion_length": 264.7250061035156,
      "epoch": 2.6625194401244165,
      "grad_norm": 1.12467622756958,
      "kl": 1.8447265625,
      "learning_rate": 1.1400749830300077e-07,
      "loss": 0.2778,
      "reward": 1.5000000298023224,
      "reward_std": 0.23617041110992432,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2567
    },
    {
      "completion_length": 210.20000076293945,
      "epoch": 2.6635562467599794,
      "grad_norm": 3.986345052719116,
      "kl": 1.958251953125,
      "learning_rate": 1.133159495155906e-07,
      "loss": 0.2048,
      "reward": 1.5750000178813934,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2568
    },
    {
      "completion_length": 151.95000076293945,
      "epoch": 2.6645930533955418,
      "grad_norm": 0.41620081663131714,
      "kl": 0.444580078125,
      "learning_rate": 1.1262642217643437e-07,
      "loss": 0.0878,
      "reward": 1.9000000059604645,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.925000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2569
    },
    {
      "completion_length": 203.17500495910645,
      "epoch": 2.665629860031104,
      "grad_norm": 1.9681460857391357,
      "kl": 2.07275390625,
      "learning_rate": 1.1193891729069555e-07,
      "loss": 0.2135,
      "reward": 1.2750000059604645,
      "reward_std": 0.45977718755602837,
      "rewards/accuracy_reward": 0.37499999813735485,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2570
    },
    {
      "completion_length": 307.100004196167,
      "epoch": 2.6666666666666665,
      "grad_norm": 1.8691552877426147,
      "kl": 1.573486328125,
      "learning_rate": 1.1125343586058894e-07,
      "loss": 0.2999,
      "reward": 1.5500000268220901,
      "reward_std": 0.28485123068094254,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2571
    },
    {
      "completion_length": 207.75000286102295,
      "epoch": 2.667703473302229,
      "grad_norm": 1.729496717453003,
      "kl": 1.1727294921875,
      "learning_rate": 1.1056997888537973e-07,
      "loss": 0.1815,
      "reward": 1.725000023841858,
      "reward_std": 0.25609346106648445,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2572
    },
    {
      "completion_length": 259.8250141143799,
      "epoch": 2.6687402799377917,
      "grad_norm": 2.043745756149292,
      "kl": 1.14990234375,
      "learning_rate": 1.0988854736138243e-07,
      "loss": 0.2332,
      "reward": 1.4750000536441803,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2573
    },
    {
      "completion_length": 247.30000686645508,
      "epoch": 2.669777086573354,
      "grad_norm": 0.6577063798904419,
      "kl": 1.6807861328125,
      "learning_rate": 1.0920914228195827e-07,
      "loss": 0.2901,
      "reward": 1.449999988079071,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2574
    },
    {
      "completion_length": 200.17500114440918,
      "epoch": 2.6708138932089165,
      "grad_norm": 0.8167489171028137,
      "kl": 0.962890625,
      "learning_rate": 1.0853176463751485e-07,
      "loss": 0.1077,
      "reward": 1.7000000178813934,
      "reward_std": 0.273861289024353,
      "rewards/accuracy_reward": 0.7749999910593033,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2575
    },
    {
      "completion_length": 159.27499961853027,
      "epoch": 2.671850699844479,
      "grad_norm": 0.8983427882194519,
      "kl": 0.41455078125,
      "learning_rate": 1.0785641541550379e-07,
      "loss": 0.1063,
      "reward": 1.5750000476837158,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2576
    },
    {
      "completion_length": 114.20000076293945,
      "epoch": 2.6728875064800413,
      "grad_norm": 0.2904658913612366,
      "kl": 0.2080078125,
      "learning_rate": 1.0718309560042055e-07,
      "loss": -0.0075,
      "reward": 1.675000011920929,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 1.0,
      "step": 2577
    },
    {
      "completion_length": 339.27500915527344,
      "epoch": 2.673924313115604,
      "grad_norm": 0.7642090916633606,
      "kl": 1.4765625,
      "learning_rate": 1.0651180617380169e-07,
      "loss": 0.2803,
      "reward": 1.6500000059604645,
      "reward_std": 0.27228760346770287,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2578
    },
    {
      "completion_length": 208.125,
      "epoch": 2.6749611197511665,
      "grad_norm": 1.021611213684082,
      "kl": 1.3607177734375,
      "learning_rate": 1.0584254811422339e-07,
      "loss": 0.2019,
      "reward": 1.6500000059604645,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2579
    },
    {
      "completion_length": 246.6749973297119,
      "epoch": 2.675997926386729,
      "grad_norm": 0.5598390102386475,
      "kl": 1.282958984375,
      "learning_rate": 1.0517532239730226e-07,
      "loss": 0.2426,
      "reward": 1.699999988079071,
      "reward_std": 0.20019175112247467,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2580
    },
    {
      "completion_length": 196.97500228881836,
      "epoch": 2.6770347330222912,
      "grad_norm": 1.9354521036148071,
      "kl": 1.00146484375,
      "learning_rate": 1.0451012999568993e-07,
      "loss": 0.0926,
      "reward": 1.5750000178813934,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.925000011920929,
      "step": 2581
    },
    {
      "completion_length": 198.87499809265137,
      "epoch": 2.6780715396578536,
      "grad_norm": 0.5886673927307129,
      "kl": 0.642333984375,
      "learning_rate": 1.0384697187907638e-07,
      "loss": 0.187,
      "reward": 1.625,
      "reward_std": 0.20019173622131348,
      "rewards/accuracy_reward": 0.6750000063329935,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2582
    },
    {
      "completion_length": 362.700008392334,
      "epoch": 2.6791083462934164,
      "grad_norm": 0.4216178059577942,
      "kl": 1.15234375,
      "learning_rate": 1.0318584901418432e-07,
      "loss": 0.2115,
      "reward": 1.5750000327825546,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2583
    },
    {
      "completion_length": 172.27499961853027,
      "epoch": 2.680145152928979,
      "grad_norm": 0.6898096203804016,
      "kl": 0.62109375,
      "learning_rate": 1.0252676236477016e-07,
      "loss": 0.143,
      "reward": 1.2750000208616257,
      "reward_std": 0.2686570733785629,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2584
    },
    {
      "completion_length": 213.17499923706055,
      "epoch": 2.681181959564541,
      "grad_norm": 1.4069899320602417,
      "kl": 0.825439453125,
      "learning_rate": 1.0186971289162216e-07,
      "loss": 0.1267,
      "reward": 1.600000023841858,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2585
    },
    {
      "completion_length": 152.350004196167,
      "epoch": 2.6822187662001036,
      "grad_norm": 0.3821243941783905,
      "kl": 0.4501953125,
      "learning_rate": 1.0121470155255886e-07,
      "loss": 0.067,
      "reward": 1.7750000357627869,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2586
    },
    {
      "completion_length": 239.4000072479248,
      "epoch": 2.683255572835666,
      "grad_norm": 1.8423457145690918,
      "kl": 1.84375,
      "learning_rate": 1.005617293024274e-07,
      "loss": 0.3049,
      "reward": 1.475000023841858,
      "reward_std": 0.3913118988275528,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2587
    },
    {
      "completion_length": 108.92500114440918,
      "epoch": 2.684292379471229,
      "grad_norm": 0.485578715801239,
      "kl": 0.2506103515625,
      "learning_rate": 9.991079709310258e-08,
      "loss": -0.0146,
      "reward": 1.800000011920929,
      "reward_std": 0.18026871606707573,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 1.0,
      "step": 2588
    },
    {
      "completion_length": 286.50000190734863,
      "epoch": 2.685329186106791,
      "grad_norm": 0.8975526094436646,
      "kl": 1.93505859375,
      "learning_rate": 9.926190587348516e-08,
      "loss": 0.2011,
      "reward": 1.475000023841858,
      "reward_std": 0.4038754850625992,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2589
    },
    {
      "completion_length": 240.85000228881836,
      "epoch": 2.6863659927423535,
      "grad_norm": 0.7761118412017822,
      "kl": 1.408203125,
      "learning_rate": 9.861505658950148e-08,
      "loss": 0.2573,
      "reward": 1.7500000298023224,
      "reward_std": 0.4038754999637604,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 2590
    },
    {
      "completion_length": 143.57500457763672,
      "epoch": 2.687402799377916,
      "grad_norm": 0.15143676102161407,
      "kl": 0.246337890625,
      "learning_rate": 9.797025018409961e-08,
      "loss": 0.0604,
      "reward": 1.7249999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2591
    },
    {
      "completion_length": 264.075008392334,
      "epoch": 2.6884396060134783,
      "grad_norm": 1.6718195676803589,
      "kl": 2.71533203125,
      "learning_rate": 9.732748759725113e-08,
      "loss": 0.3436,
      "reward": 1.2999999970197678,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2592
    },
    {
      "completion_length": 172.1750030517578,
      "epoch": 2.689476412649041,
      "grad_norm": 0.5123502612113953,
      "kl": 0.35595703125,
      "learning_rate": 9.668676976594771e-08,
      "loss": 0.0904,
      "reward": 1.725000023841858,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2593
    },
    {
      "completion_length": 191.12500381469727,
      "epoch": 2.6905132192846035,
      "grad_norm": 0.2367032766342163,
      "kl": 0.46875,
      "learning_rate": 9.604809762419986e-08,
      "loss": 0.1345,
      "reward": 1.7750000357627869,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2594
    },
    {
      "completion_length": 396.35000228881836,
      "epoch": 2.691550025920166,
      "grad_norm": 1.4451581239700317,
      "kl": 1.2039794921875,
      "learning_rate": 9.54114721030362e-08,
      "loss": 0.2284,
      "reward": 1.475000038743019,
      "reward_std": 0.40387552976608276,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2595
    },
    {
      "completion_length": 98.12500190734863,
      "epoch": 2.6925868325557283,
      "grad_norm": 0.3243737518787384,
      "kl": 0.3759765625,
      "learning_rate": 9.47768941305025e-08,
      "loss": 0.0319,
      "reward": 1.625,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 1.0,
      "step": 2596
    },
    {
      "completion_length": 195.27499771118164,
      "epoch": 2.6936236391912907,
      "grad_norm": 0.7048050761222839,
      "kl": 0.3787841796875,
      "learning_rate": 9.414436463165882e-08,
      "loss": 0.135,
      "reward": 1.675000011920929,
      "reward_std": 0.45977721735835075,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2597
    },
    {
      "completion_length": 247.12500381469727,
      "epoch": 2.6946604458268535,
      "grad_norm": 0.8171465396881104,
      "kl": 0.60595703125,
      "learning_rate": 9.351388452857956e-08,
      "loss": 0.1982,
      "reward": 1.5,
      "reward_std": 0.3913118727505207,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 2598
    },
    {
      "completion_length": 137.3750057220459,
      "epoch": 2.695697252462416,
      "grad_norm": 0.5159651637077332,
      "kl": 0.723876953125,
      "learning_rate": 9.28854547403512e-08,
      "loss": 0.1125,
      "reward": 1.7500000298023224,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2599
    },
    {
      "completion_length": 219.20000457763672,
      "epoch": 2.6967340590979783,
      "grad_norm": 2.9190218448638916,
      "kl": 0.69287109375,
      "learning_rate": 9.225907618307228e-08,
      "loss": 0.1502,
      "reward": 1.7750000357627869,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.8250000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2600
    },
    {
      "completion_length": 148.72500228881836,
      "epoch": 2.6977708657335406,
      "grad_norm": 0.644507884979248,
      "kl": 0.41943359375,
      "learning_rate": 9.163474976984938e-08,
      "loss": -0.0118,
      "reward": 1.5000000298023224,
      "reward_std": 0.30463574826717377,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2601
    },
    {
      "completion_length": 143.32500457763672,
      "epoch": 2.698807672369103,
      "grad_norm": 0.1350562423467636,
      "kl": 0.29052734375,
      "learning_rate": 9.101247641079947e-08,
      "loss": 0.0533,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2602
    },
    {
      "completion_length": 339.85000228881836,
      "epoch": 2.699844479004666,
      "grad_norm": 1.1104599237442017,
      "kl": 1.5419921875,
      "learning_rate": 9.039225701304576e-08,
      "loss": 0.2579,
      "reward": 1.4750000089406967,
      "reward_std": 0.3913118988275528,
      "rewards/accuracy_reward": 0.6000000033527613,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2603
    },
    {
      "completion_length": 145.65000343322754,
      "epoch": 2.7008812856402282,
      "grad_norm": 0.30152395367622375,
      "kl": 0.2947998046875,
      "learning_rate": 8.977409248071706e-08,
      "loss": 0.0501,
      "reward": 1.5250000059604645,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2604
    },
    {
      "completion_length": 178.19999885559082,
      "epoch": 2.7019180922757906,
      "grad_norm": 2.1781415939331055,
      "kl": 0.67626953125,
      "learning_rate": 8.915798371494699e-08,
      "loss": 0.0718,
      "reward": 1.7000000178813934,
      "reward_std": 0.22360678762197495,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2605
    },
    {
      "completion_length": 206.37500190734863,
      "epoch": 2.702954898911353,
      "grad_norm": 0.29660096764564514,
      "kl": 0.418701171875,
      "learning_rate": 8.854393161387264e-08,
      "loss": 0.1143,
      "reward": 1.6750000417232513,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.7250000182539225,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2606
    },
    {
      "completion_length": 392.9500045776367,
      "epoch": 2.7039917055469154,
      "grad_norm": 1.440569519996643,
      "kl": 1.970703125,
      "learning_rate": 8.793193707263264e-08,
      "loss": 0.4374,
      "reward": 1.4250000417232513,
      "reward_std": 0.50311528891325,
      "rewards/accuracy_reward": 0.5750000141561031,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2607
    },
    {
      "completion_length": 196.7500057220459,
      "epoch": 2.705028512182478,
      "grad_norm": 0.5869837999343872,
      "kl": 0.4990234375,
      "learning_rate": 8.732200098336618e-08,
      "loss": 0.1236,
      "reward": 1.6000000536441803,
      "reward_std": 0.3479738347232342,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2608
    },
    {
      "completion_length": 143.10000610351562,
      "epoch": 2.7060653188180406,
      "grad_norm": 0.4040537178516388,
      "kl": 1.211181640625,
      "learning_rate": 8.671412423521152e-08,
      "loss": 0.1193,
      "reward": 1.6500000357627869,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2609
    },
    {
      "completion_length": 143.75,
      "epoch": 2.707102125453603,
      "grad_norm": 0.9139902591705322,
      "kl": 0.3228759765625,
      "learning_rate": 8.6108307714306e-08,
      "loss": -0.0152,
      "reward": 1.625,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.650000000372529,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2610
    },
    {
      "completion_length": 110.375,
      "epoch": 2.7081389320891653,
      "grad_norm": 0.3124227821826935,
      "kl": 0.203125,
      "learning_rate": 8.550455230378184e-08,
      "loss": 0.0133,
      "reward": 1.425000011920929,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 1.0,
      "step": 2611
    },
    {
      "completion_length": 363.3000068664551,
      "epoch": 2.7091757387247277,
      "grad_norm": 31.54953956604004,
      "kl": 9.62646484375,
      "learning_rate": 8.490285888376836e-08,
      "loss": 0.9467,
      "reward": 1.4749999940395355,
      "reward_std": 0.40387551113963127,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2612
    },
    {
      "completion_length": 194.60000801086426,
      "epoch": 2.7102125453602905,
      "grad_norm": 13.590709686279297,
      "kl": 4.250732421875,
      "learning_rate": 8.430322833138809e-08,
      "loss": 0.3712,
      "reward": 1.725000023841858,
      "reward_std": 0.2920721396803856,
      "rewards/accuracy_reward": 0.7749999985098839,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2613
    },
    {
      "completion_length": 247.25000381469727,
      "epoch": 2.711249351995853,
      "grad_norm": 26.16112518310547,
      "kl": 8.117919921875,
      "learning_rate": 8.370566152075665e-08,
      "loss": 0.6659,
      "reward": 1.425000011920929,
      "reward_std": 0.21275537461042404,
      "rewards/accuracy_reward": 0.5000000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 2614
    },
    {
      "completion_length": 163.50000381469727,
      "epoch": 2.7122861586314153,
      "grad_norm": 7.690154075622559,
      "kl": 3.583984375,
      "learning_rate": 8.311015932298088e-08,
      "loss": 0.2927,
      "reward": 1.7000000178813934,
      "reward_std": 0.3479738049209118,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2615
    },
    {
      "completion_length": 183.4500026702881,
      "epoch": 2.7133229652669777,
      "grad_norm": 0.47782066464424133,
      "kl": 0.3291015625,
      "learning_rate": 8.251672260615873e-08,
      "loss": 0.1048,
      "reward": 1.7249999940395355,
      "reward_std": 0.24873405322432518,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2616
    },
    {
      "completion_length": 150.42500495910645,
      "epoch": 2.71435977190254,
      "grad_norm": 0.30651357769966125,
      "kl": 0.340087890625,
      "learning_rate": 8.192535223537634e-08,
      "loss": 0.0718,
      "reward": 1.5250000059604645,
      "reward_std": 0.10458250343799591,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2617
    },
    {
      "completion_length": 239.67500495910645,
      "epoch": 2.715396578538103,
      "grad_norm": 1.7098196744918823,
      "kl": 0.57080078125,
      "learning_rate": 8.133604907270781e-08,
      "loss": 0.1673,
      "reward": 1.5000000298023224,
      "reward_std": 0.3913118913769722,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 2618
    },
    {
      "completion_length": 406.4500045776367,
      "epoch": 2.716433385173665,
      "grad_norm": 1.596725344657898,
      "kl": 2.834716796875,
      "learning_rate": 8.074881397721445e-08,
      "loss": 0.2374,
      "reward": 1.3250000327825546,
      "reward_std": 0.3856646940112114,
      "rewards/accuracy_reward": 0.475000012665987,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2619
    },
    {
      "completion_length": 117.37500190734863,
      "epoch": 2.7174701918092277,
      "grad_norm": 0.4847503900527954,
      "kl": 0.1923828125,
      "learning_rate": 8.016364780494185e-08,
      "loss": 0.0029,
      "reward": 1.3499999940395355,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.3750000149011612,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2620
    },
    {
      "completion_length": 168.10000610351562,
      "epoch": 2.71850699844479,
      "grad_norm": 0.451677143573761,
      "kl": 0.46435546875,
      "learning_rate": 7.958055140892013e-08,
      "loss": 0.1252,
      "reward": 1.4500000029802322,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2621
    },
    {
      "completion_length": 204.52499961853027,
      "epoch": 2.7195438050803524,
      "grad_norm": 0.8093661069869995,
      "kl": 0.870849609375,
      "learning_rate": 7.899952563916201e-08,
      "loss": 0.1606,
      "reward": 1.800000011920929,
      "reward_std": 0.25609343126416206,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2622
    },
    {
      "completion_length": 113.6500015258789,
      "epoch": 2.7205806117159153,
      "grad_norm": 0.25541362166404724,
      "kl": 0.26318359375,
      "learning_rate": 7.842057134266218e-08,
      "loss": 0.0062,
      "reward": 1.550000011920929,
      "reward_std": 0.06846532225608826,
      "rewards/accuracy_reward": 0.5499999970197678,
      "rewards/format_reward": 1.0,
      "step": 2623
    },
    {
      "completion_length": 280.65000915527344,
      "epoch": 2.721617418351477,
      "grad_norm": 0.5384974479675293,
      "kl": 0.73828125,
      "learning_rate": 7.784368936339465e-08,
      "loss": 0.2314,
      "reward": 1.5750000476837158,
      "reward_std": 0.3913119100034237,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.9000000357627869,
      "step": 2624
    },
    {
      "completion_length": 243.5750026702881,
      "epoch": 2.72265422498704,
      "grad_norm": 1.1426197290420532,
      "kl": 0.968994140625,
      "learning_rate": 7.726888054231373e-08,
      "loss": 0.1947,
      "reward": 1.7500000298023224,
      "reward_std": 0.4038754999637604,
      "rewards/accuracy_reward": 0.8500000238418579,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2625
    },
    {
      "completion_length": 206.2999973297119,
      "epoch": 2.7236910316226024,
      "grad_norm": 0.9873815774917603,
      "kl": 1.24853515625,
      "learning_rate": 7.66961457173505e-08,
      "loss": 0.0767,
      "reward": 1.7000000178813934,
      "reward_std": 0.47234082967042923,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2626
    },
    {
      "completion_length": 199.57500076293945,
      "epoch": 2.724727838258165,
      "grad_norm": 0.509842574596405,
      "kl": 0.58642578125,
      "learning_rate": 7.612548572341344e-08,
      "loss": 0.1361,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2627
    },
    {
      "completion_length": 297.9250030517578,
      "epoch": 2.725764644893727,
      "grad_norm": 5.29781436920166,
      "kl": 0.7060546875,
      "learning_rate": 7.555690139238574e-08,
      "loss": 0.2287,
      "reward": 1.425000011920929,
      "reward_std": 0.4472135677933693,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2628
    },
    {
      "completion_length": 278.42501068115234,
      "epoch": 2.7268014515292895,
      "grad_norm": 0.28287577629089355,
      "kl": 1.16796875,
      "learning_rate": 7.499039355312543e-08,
      "loss": 0.1801,
      "reward": 1.5500000268220901,
      "reward_std": 0.31719937175512314,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2629
    },
    {
      "completion_length": 146.87500190734863,
      "epoch": 2.7278382581648524,
      "grad_norm": 1.8818683624267578,
      "kl": 0.35400390625,
      "learning_rate": 7.44259630314632e-08,
      "loss": 0.0731,
      "reward": 1.3250000178813934,
      "reward_std": 0.17304782569408417,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2630
    },
    {
      "completion_length": 334.0500068664551,
      "epoch": 2.7288750648004148,
      "grad_norm": 3.4028501510620117,
      "kl": 1.12646484375,
      "learning_rate": 7.386361065020142e-08,
      "loss": 0.2327,
      "reward": 1.350000038743019,
      "reward_std": 0.37323957681655884,
      "rewards/accuracy_reward": 0.47500000707805157,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2631
    },
    {
      "completion_length": 347.25000762939453,
      "epoch": 2.729911871435977,
      "grad_norm": 2.01867413520813,
      "kl": 1.10595703125,
      "learning_rate": 7.330333722911314e-08,
      "loss": 0.3395,
      "reward": 1.4000000208616257,
      "reward_std": 0.6833839789032936,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2632
    },
    {
      "completion_length": 156.42500114440918,
      "epoch": 2.7309486780715395,
      "grad_norm": 0.5674120187759399,
      "kl": 0.2470703125,
      "learning_rate": 7.274514358494122e-08,
      "loss": 0.0702,
      "reward": 1.7750000357627869,
      "reward_std": 0.31199515610933304,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2633
    },
    {
      "completion_length": 191.4000015258789,
      "epoch": 2.731985484707102,
      "grad_norm": 1.371336579322815,
      "kl": 0.868896484375,
      "learning_rate": 7.218903053139558e-08,
      "loss": 0.105,
      "reward": 1.5000000149011612,
      "reward_std": 0.27950847521424294,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.925000011920929,
      "step": 2634
    },
    {
      "completion_length": 347.0750026702881,
      "epoch": 2.7330222913426647,
      "grad_norm": 0.6004615426063538,
      "kl": 1.296875,
      "learning_rate": 7.163499887915426e-08,
      "loss": 0.2385,
      "reward": 1.3749999850988388,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2635
    },
    {
      "completion_length": 187.87500762939453,
      "epoch": 2.734059097978227,
      "grad_norm": 0.9656141996383667,
      "kl": 0.642822265625,
      "learning_rate": 7.108304943586102e-08,
      "loss": 0.1344,
      "reward": 1.675000011920929,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2636
    },
    {
      "completion_length": 168.8250026702881,
      "epoch": 2.7350959046137895,
      "grad_norm": 0.35847458243370056,
      "kl": 0.56982421875,
      "learning_rate": 7.053318300612344e-08,
      "loss": 0.127,
      "reward": 1.5750000178813934,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.6250000018626451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2637
    },
    {
      "completion_length": 203.8750114440918,
      "epoch": 2.736132711249352,
      "grad_norm": 0.4431527256965637,
      "kl": 0.675048828125,
      "learning_rate": 6.99854003915133e-08,
      "loss": 0.1571,
      "reward": 1.6500000357627869,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2638
    },
    {
      "completion_length": 152.62500190734863,
      "epoch": 2.7371695178849142,
      "grad_norm": 0.42283111810684204,
      "kl": 0.684814453125,
      "learning_rate": 6.943970239056452e-08,
      "loss": 0.0821,
      "reward": 1.8250000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2639
    },
    {
      "completion_length": 144.92500114440918,
      "epoch": 2.738206324520477,
      "grad_norm": 1.4384374618530273,
      "kl": 0.50830078125,
      "learning_rate": 6.889608979877243e-08,
      "loss": 0.0044,
      "reward": 1.6500000357627869,
      "reward_std": 0.3804604895412922,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2640
    },
    {
      "completion_length": 187.42500495910645,
      "epoch": 2.7392431311560395,
      "grad_norm": 0.7810163497924805,
      "kl": 0.685302734375,
      "learning_rate": 6.835456340859147e-08,
      "loss": 0.1423,
      "reward": 1.574999988079071,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2641
    },
    {
      "completion_length": 141.82499885559082,
      "epoch": 2.740279937791602,
      "grad_norm": 0.857786238193512,
      "kl": 0.577392578125,
      "learning_rate": 6.781512400943568e-08,
      "loss": 0.0971,
      "reward": 1.5750000178813934,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2642
    },
    {
      "completion_length": 399.00000762939453,
      "epoch": 2.741316744427164,
      "grad_norm": 5.126710414886475,
      "kl": 4.1015625,
      "learning_rate": 6.727777238767718e-08,
      "loss": 0.5017,
      "reward": 1.2750000357627869,
      "reward_std": 0.6526095792651176,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2643
    },
    {
      "completion_length": 197.0250072479248,
      "epoch": 2.7423535510627266,
      "grad_norm": 1.3232945203781128,
      "kl": 0.80029296875,
      "learning_rate": 6.674250932664289e-08,
      "loss": 0.1387,
      "reward": 1.4500000178813934,
      "reward_std": 0.20539595931768417,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2644
    },
    {
      "completion_length": 194.14999771118164,
      "epoch": 2.7433903576982894,
      "grad_norm": 0.9778532981872559,
      "kl": 1.270751953125,
      "learning_rate": 6.620933560661713e-08,
      "loss": 0.2127,
      "reward": 1.475000023841858,
      "reward_std": 0.3407529182732105,
      "rewards/accuracy_reward": 0.5250000134110451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2645
    },
    {
      "completion_length": 235.3250026702881,
      "epoch": 2.744427164333852,
      "grad_norm": 2.600414514541626,
      "kl": 2.482421875,
      "learning_rate": 6.567825200483718e-08,
      "loss": 0.3217,
      "reward": 1.5999999940395355,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.925000011920929,
      "step": 2646
    },
    {
      "completion_length": 137.7000026702881,
      "epoch": 2.745463970969414,
      "grad_norm": 1.0574477910995483,
      "kl": 0.5487060546875,
      "learning_rate": 6.514925929549393e-08,
      "loss": 0.0548,
      "reward": 1.475000023841858,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2647
    },
    {
      "completion_length": 252.40000915527344,
      "epoch": 2.7465007776049766,
      "grad_norm": 1.5630449056625366,
      "kl": 0.852783203125,
      "learning_rate": 6.462235824973e-08,
      "loss": 0.1434,
      "reward": 1.4500000476837158,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2648
    },
    {
      "completion_length": 347.20001220703125,
      "epoch": 2.747537584240539,
      "grad_norm": 1.2366613149642944,
      "kl": 2.0478515625,
      "learning_rate": 6.409754963563935e-08,
      "loss": 0.3818,
      "reward": 1.5750000178813934,
      "reward_std": 0.484904445707798,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2649
    },
    {
      "completion_length": 338.39999771118164,
      "epoch": 2.748574390876102,
      "grad_norm": 1.4604618549346924,
      "kl": 2.15185546875,
      "learning_rate": 6.357483421826515e-08,
      "loss": 0.2864,
      "reward": 1.4500000178813934,
      "reward_std": 0.347973819822073,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2650
    },
    {
      "completion_length": 255.5250072479248,
      "epoch": 2.749611197511664,
      "grad_norm": 0.7960745096206665,
      "kl": 1.47216796875,
      "learning_rate": 6.305421275959938e-08,
      "loss": 0.2757,
      "reward": 1.4750000536441803,
      "reward_std": 0.3354102149605751,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2651
    },
    {
      "completion_length": 148.55000114440918,
      "epoch": 2.7506480041472265,
      "grad_norm": 0.16960236430168152,
      "kl": 0.53857421875,
      "learning_rate": 6.253568601858161e-08,
      "loss": 0.0793,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2652
    },
    {
      "completion_length": 193.80000495910645,
      "epoch": 2.751684810782789,
      "grad_norm": 0.5603395104408264,
      "kl": 0.871826171875,
      "learning_rate": 6.201925475109804e-08,
      "loss": 0.1211,
      "reward": 1.550000011920929,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2653
    },
    {
      "completion_length": 242.40000534057617,
      "epoch": 2.7527216174183513,
      "grad_norm": 1.4470975399017334,
      "kl": 1.82763671875,
      "learning_rate": 6.150491970997929e-08,
      "loss": 0.2805,
      "reward": 1.725000023841858,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 2654
    },
    {
      "completion_length": 298.3000030517578,
      "epoch": 2.753758424053914,
      "grad_norm": 1.442375898361206,
      "kl": 2.0166015625,
      "learning_rate": 6.099268164500143e-08,
      "loss": 0.3526,
      "reward": 1.2500000149011612,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9000000357627869,
      "step": 2655
    },
    {
      "completion_length": 325.1750030517578,
      "epoch": 2.7547952306894765,
      "grad_norm": 1.369773507118225,
      "kl": 3.185791015625,
      "learning_rate": 6.0482541302883e-08,
      "loss": 0.3465,
      "reward": 1.4750000089406967,
      "reward_std": 0.2920721210539341,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2656
    },
    {
      "completion_length": 103.00000190734863,
      "epoch": 2.755832037325039,
      "grad_norm": 0.22802236676216125,
      "kl": 0.2625732421875,
      "learning_rate": 5.997449942728395e-08,
      "loss": 0.0098,
      "reward": 1.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 1.0,
      "step": 2657
    },
    {
      "completion_length": 197.72500038146973,
      "epoch": 2.7568688439606013,
      "grad_norm": 1.557938575744629,
      "kl": 0.98876953125,
      "learning_rate": 5.946855675880669e-08,
      "loss": 0.1964,
      "reward": 1.5250000059604645,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2658
    },
    {
      "completion_length": 149.72500610351562,
      "epoch": 2.7579056505961637,
      "grad_norm": 1.6007956266403198,
      "kl": 2.418212890625,
      "learning_rate": 5.896471403499226e-08,
      "loss": 0.2043,
      "reward": 1.3250000178813934,
      "reward_std": 0.27228760719299316,
      "rewards/accuracy_reward": 0.3999999985098839,
      "rewards/format_reward": 0.925000011920929,
      "step": 2659
    },
    {
      "completion_length": 191.85000038146973,
      "epoch": 2.7589424572317265,
      "grad_norm": 0.9115043878555298,
      "kl": 0.75927734375,
      "learning_rate": 5.84629719903208e-08,
      "loss": 0.1368,
      "reward": 1.6500000059604645,
      "reward_std": 0.4849044606089592,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2660
    },
    {
      "completion_length": 191.67500495910645,
      "epoch": 2.759979263867289,
      "grad_norm": 0.26682189106941223,
      "kl": 0.725830078125,
      "learning_rate": 5.796333135621007e-08,
      "loss": 0.1564,
      "reward": 1.4000000208616257,
      "reward_std": 0.18026871606707573,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2661
    },
    {
      "completion_length": 84.22500228881836,
      "epoch": 2.7610160705028512,
      "grad_norm": 0.5228694677352905,
      "kl": 0.220703125,
      "learning_rate": 5.746579286101511e-08,
      "loss": -0.0144,
      "reward": 1.675000011920929,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.6999999973922968,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2662
    },
    {
      "completion_length": 145.70000076293945,
      "epoch": 2.7620528771384136,
      "grad_norm": 0.4472379982471466,
      "kl": 0.58154296875,
      "learning_rate": 5.6970357230025916e-08,
      "loss": 0.1021,
      "reward": 1.625,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.6499999985098839,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2663
    },
    {
      "completion_length": 188.350004196167,
      "epoch": 2.763089683773976,
      "grad_norm": 1.4383118152618408,
      "kl": 2.2161865234375,
      "learning_rate": 5.64770251854671e-08,
      "loss": 0.2481,
      "reward": 1.6250000298023224,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.6750000193715096,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2664
    },
    {
      "completion_length": 129.95000076293945,
      "epoch": 2.764126490409539,
      "grad_norm": 0.6044561862945557,
      "kl": 0.26171875,
      "learning_rate": 5.59857974464969e-08,
      "loss": 0.0956,
      "reward": 1.600000023841858,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.6250000242143869,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2665
    },
    {
      "completion_length": 287.3750114440918,
      "epoch": 2.765163297045101,
      "grad_norm": 0.731571614742279,
      "kl": 1.1236572265625,
      "learning_rate": 5.5496674729206164e-08,
      "loss": 0.2669,
      "reward": 1.3499999940395355,
      "reward_std": 0.36053743213415146,
      "rewards/accuracy_reward": 0.4499999973922968,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2666
    },
    {
      "completion_length": 182.17500495910645,
      "epoch": 2.7662001036806636,
      "grad_norm": 0.5813654661178589,
      "kl": 1.121337890625,
      "learning_rate": 5.500965774661654e-08,
      "loss": 0.1742,
      "reward": 1.550000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2667
    },
    {
      "completion_length": 174.5,
      "epoch": 2.767236910316226,
      "grad_norm": 0.43639108538627625,
      "kl": 1.35888671875,
      "learning_rate": 5.452474720868061e-08,
      "loss": 0.1145,
      "reward": 1.5750000178813934,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.925000011920929,
      "step": 2668
    },
    {
      "completion_length": 287.6750011444092,
      "epoch": 2.7682737169517884,
      "grad_norm": 0.5966921448707581,
      "kl": 0.599853515625,
      "learning_rate": 5.404194382228012e-08,
      "loss": 0.2336,
      "reward": 1.2750000059604645,
      "reward_std": 0.31199515983462334,
      "rewards/accuracy_reward": 0.37500000186264515,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2669
    },
    {
      "completion_length": 162.87500190734863,
      "epoch": 2.769310523587351,
      "grad_norm": 0.5164470672607422,
      "kl": 0.5185546875,
      "learning_rate": 5.356124829122489e-08,
      "loss": 0.1433,
      "reward": 1.6500000059604645,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.7000000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2670
    },
    {
      "completion_length": 243.45000839233398,
      "epoch": 2.7703473302229136,
      "grad_norm": 0.4556516408920288,
      "kl": 1.1640625,
      "learning_rate": 5.308266131625206e-08,
      "loss": 0.1582,
      "reward": 1.600000023841858,
      "reward_std": 0.335410188883543,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2671
    },
    {
      "completion_length": 293.1750068664551,
      "epoch": 2.771384136858476,
      "grad_norm": 0.5882754921913147,
      "kl": 0.92138671875,
      "learning_rate": 5.260618359502556e-08,
      "loss": 0.187,
      "reward": 1.3250000327825546,
      "reward_std": 0.5282425433397293,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2672
    },
    {
      "completion_length": 185.07500457763672,
      "epoch": 2.7724209434940383,
      "grad_norm": 0.5973179936408997,
      "kl": 1.0390625,
      "learning_rate": 5.213181582213361e-08,
      "loss": 0.1714,
      "reward": 1.5250000208616257,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2673
    },
    {
      "completion_length": 116.60000228881836,
      "epoch": 2.7734577501296007,
      "grad_norm": 0.33007726073265076,
      "kl": 0.365234375,
      "learning_rate": 5.165955868908923e-08,
      "loss": 0.0845,
      "reward": 1.574999988079071,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2674
    },
    {
      "completion_length": 86.4000015258789,
      "epoch": 2.7744945567651635,
      "grad_norm": 0.38368144631385803,
      "kl": 0.2353515625,
      "learning_rate": 5.1189412884328544e-08,
      "loss": 0.0411,
      "reward": 1.425000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 1.0,
      "step": 2675
    },
    {
      "completion_length": 86.45000076293945,
      "epoch": 2.775531363400726,
      "grad_norm": 0.38368967175483704,
      "kl": 0.37255859375,
      "learning_rate": 5.072137909321017e-08,
      "loss": 0.0583,
      "reward": 1.7249999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 1.0,
      "step": 2676
    },
    {
      "completion_length": 139.10000610351562,
      "epoch": 2.7765681700362883,
      "grad_norm": 0.5959401726722717,
      "kl": 0.595458984375,
      "learning_rate": 5.0255457998012844e-08,
      "loss": 0.0737,
      "reward": 1.5999999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2677
    },
    {
      "completion_length": 235.27500343322754,
      "epoch": 2.7776049766718507,
      "grad_norm": 0.7969774007797241,
      "kl": 0.82568359375,
      "learning_rate": 4.9791650277936915e-08,
      "loss": 0.1782,
      "reward": 1.5499999821186066,
      "reward_std": 0.36053744703531265,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2678
    },
    {
      "completion_length": 335.2750015258789,
      "epoch": 2.778641783307413,
      "grad_norm": 1.0074883699417114,
      "kl": 2.220703125,
      "learning_rate": 4.9329956609101225e-08,
      "loss": 0.4489,
      "reward": 1.2750000357627869,
      "reward_std": 0.6274822950363159,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2679
    },
    {
      "completion_length": 88.2750015258789,
      "epoch": 2.779678589942976,
      "grad_norm": 0.07997974753379822,
      "kl": 0.1761474609375,
      "learning_rate": 4.8870377664542555e-08,
      "loss": 0.0063,
      "reward": 1.75,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 1.0,
      "step": 2680
    },
    {
      "completion_length": 175.80000495910645,
      "epoch": 2.780715396578538,
      "grad_norm": 0.32113999128341675,
      "kl": 1.204345703125,
      "learning_rate": 4.84129141142155e-08,
      "loss": 0.1699,
      "reward": 1.5999999940395355,
      "reward_std": 0.21638591587543488,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2681
    },
    {
      "completion_length": 132.05000114440918,
      "epoch": 2.7817522032141007,
      "grad_norm": 0.3984842002391815,
      "kl": 0.275634765625,
      "learning_rate": 4.795756662499079e-08,
      "loss": 0.0779,
      "reward": 1.600000023841858,
      "reward_std": 0.21275537088513374,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2682
    },
    {
      "completion_length": 149.77499961853027,
      "epoch": 2.782789009849663,
      "grad_norm": 0.729710578918457,
      "kl": 1.3125,
      "learning_rate": 4.750433586065428e-08,
      "loss": 0.138,
      "reward": 1.6500000059604645,
      "reward_std": 0.3354101702570915,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2683
    },
    {
      "completion_length": 77.42500305175781,
      "epoch": 2.7838258164852254,
      "grad_norm": 0.4288610517978668,
      "kl": 0.35546875,
      "learning_rate": 4.705322248190647e-08,
      "loss": 0.0225,
      "reward": 1.600000023841858,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 1.0,
      "step": 2684
    },
    {
      "completion_length": 187.72500228881836,
      "epoch": 2.7848626231207882,
      "grad_norm": 0.6534906029701233,
      "kl": 0.772216796875,
      "learning_rate": 4.660422714636031e-08,
      "loss": 0.0955,
      "reward": 1.675000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.7250000163912773,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2685
    },
    {
      "completion_length": 145.05000495910645,
      "epoch": 2.78589942975635,
      "grad_norm": 0.9536347985267639,
      "kl": 0.71923828125,
      "learning_rate": 4.615735050854292e-08,
      "loss": 0.1546,
      "reward": 1.3750000596046448,
      "reward_std": 0.45977722108364105,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2686
    },
    {
      "completion_length": 227.9250030517578,
      "epoch": 2.786936236391913,
      "grad_norm": 1.1452770233154297,
      "kl": 0.818603515625,
      "learning_rate": 4.571259321989085e-08,
      "loss": 0.1905,
      "reward": 1.5000000298023224,
      "reward_std": 0.36067594215273857,
      "rewards/accuracy_reward": 0.5749999973922968,
      "rewards/format_reward": 0.925000011920929,
      "step": 2687
    },
    {
      "completion_length": 189.0500030517578,
      "epoch": 2.7879730430274754,
      "grad_norm": 0.8598825931549072,
      "kl": 0.50341796875,
      "learning_rate": 4.526995592875249e-08,
      "loss": 0.1236,
      "reward": 1.6250000298023224,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2688
    },
    {
      "completion_length": 133.5750026702881,
      "epoch": 2.7890098496630378,
      "grad_norm": 0.36645692586898804,
      "kl": 0.5966796875,
      "learning_rate": 4.482943928038552e-08,
      "loss": 0.0707,
      "reward": 1.5,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2689
    },
    {
      "completion_length": 187.8000087738037,
      "epoch": 2.7900466562986,
      "grad_norm": 0.995021641254425,
      "kl": 1.5439453125,
      "learning_rate": 4.439104391695609e-08,
      "loss": 0.2073,
      "reward": 1.4000000059604645,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2690
    },
    {
      "completion_length": 185.55000591278076,
      "epoch": 2.7910834629341625,
      "grad_norm": 0.3788370192050934,
      "kl": 0.94287109375,
      "learning_rate": 4.395477047753782e-08,
      "loss": 0.1585,
      "reward": 1.3000000268220901,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2691
    },
    {
      "completion_length": 227.77500343322754,
      "epoch": 2.7921202695697254,
      "grad_norm": 1.1315395832061768,
      "kl": 1.89208984375,
      "learning_rate": 4.352061959811149e-08,
      "loss": 0.2352,
      "reward": 1.6500000357627869,
      "reward_std": 0.31719934195280075,
      "rewards/accuracy_reward": 0.725000012665987,
      "rewards/format_reward": 0.925000011920929,
      "step": 2692
    },
    {
      "completion_length": 264.350004196167,
      "epoch": 2.7931570762052877,
      "grad_norm": 0.4569127559661865,
      "kl": 1.173095703125,
      "learning_rate": 4.308859191156367e-08,
      "loss": 0.2762,
      "reward": 1.400000050663948,
      "reward_std": 0.4237985387444496,
      "rewards/accuracy_reward": 0.5000000149011612,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2693
    },
    {
      "completion_length": 170.5250015258789,
      "epoch": 2.79419388284085,
      "grad_norm": 0.9355179071426392,
      "kl": 0.657958984375,
      "learning_rate": 4.265868804768558e-08,
      "loss": 0.1285,
      "reward": 1.6250000298023224,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2694
    },
    {
      "completion_length": 130.92500495910645,
      "epoch": 2.7952306894764125,
      "grad_norm": 0.6848067045211792,
      "kl": 0.645263671875,
      "learning_rate": 4.2230908633172414e-08,
      "loss": 0.1103,
      "reward": 1.7000000178813934,
      "reward_std": 0.1928323395550251,
      "rewards/accuracy_reward": 0.7249999940395355,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2695
    },
    {
      "completion_length": 181.80000495910645,
      "epoch": 2.796267496111975,
      "grad_norm": 0.457476943731308,
      "kl": 0.755859375,
      "learning_rate": 4.1805254291622854e-08,
      "loss": 0.0704,
      "reward": 1.4000000059604645,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.4750000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2696
    },
    {
      "completion_length": 254.30000495910645,
      "epoch": 2.7973043027475377,
      "grad_norm": 1.3537697792053223,
      "kl": 0.93359375,
      "learning_rate": 4.138172564353721e-08,
      "loss": 0.2341,
      "reward": 1.4500000476837158,
      "reward_std": 0.3913119025528431,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9000000357627869,
      "step": 2697
    },
    {
      "completion_length": 134.4000015258789,
      "epoch": 2.7983411093831,
      "grad_norm": 0.8841086030006409,
      "kl": 0.65478515625,
      "learning_rate": 4.096032330631744e-08,
      "loss": 0.084,
      "reward": 1.2750000059604645,
      "reward_std": 0.3245587721467018,
      "rewards/accuracy_reward": 0.3250000011175871,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2698
    },
    {
      "completion_length": 134.90000534057617,
      "epoch": 2.7993779160186625,
      "grad_norm": 0.3711719512939453,
      "kl": 0.4814453125,
      "learning_rate": 4.0541047894265824e-08,
      "loss": 0.0649,
      "reward": 1.6250000298023224,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2699
    },
    {
      "completion_length": 193.6250057220459,
      "epoch": 2.800414722654225,
      "grad_norm": 0.5220799446105957,
      "kl": 0.733154296875,
      "learning_rate": 4.012390001858363e-08,
      "loss": 0.1809,
      "reward": 1.4000000059604645,
      "reward_std": 0.21275537088513374,
      "rewards/accuracy_reward": 0.45000000670552254,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2700
    },
    {
      "completion_length": 237.97500801086426,
      "epoch": 2.8014515292897872,
      "grad_norm": 11.900105476379395,
      "kl": 3.472900390625,
      "learning_rate": 3.970888028737141e-08,
      "loss": 0.3516,
      "reward": 1.7000000178813934,
      "reward_std": 0.5031152740120888,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2701
    },
    {
      "completion_length": 176.6500072479248,
      "epoch": 2.80248833592535,
      "grad_norm": 0.9762704968452454,
      "kl": 1.08251953125,
      "learning_rate": 3.929598930562689e-08,
      "loss": 0.1767,
      "reward": 1.5000000298023224,
      "reward_std": 0.2361704222857952,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2702
    },
    {
      "completion_length": 96.625,
      "epoch": 2.8035251425609125,
      "grad_norm": 0.6630142331123352,
      "kl": 1.26611328125,
      "learning_rate": 3.8885227675244604e-08,
      "loss": 0.1091,
      "reward": 1.4500000178813934,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.47500000335276127,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2703
    },
    {
      "completion_length": 126.67500305175781,
      "epoch": 2.804561949196475,
      "grad_norm": 0.9734599590301514,
      "kl": 0.7880859375,
      "learning_rate": 3.847659599501541e-08,
      "loss": 0.0942,
      "reward": 1.5500000417232513,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.5750000048428774,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2704
    },
    {
      "completion_length": 208.32500839233398,
      "epoch": 2.805598755832037,
      "grad_norm": 0.8408308625221252,
      "kl": 1.108642578125,
      "learning_rate": 3.8070094860624795e-08,
      "loss": 0.1132,
      "reward": 1.4749999940395355,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.925000011920929,
      "step": 2705
    },
    {
      "completion_length": 145.85000610351562,
      "epoch": 2.8066355624675996,
      "grad_norm": 0.3530755043029785,
      "kl": 0.46240234375,
      "learning_rate": 3.7665724864652754e-08,
      "loss": 0.075,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2706
    },
    {
      "completion_length": 89.42500305175781,
      "epoch": 2.8076723691031624,
      "grad_norm": 0.41172271966934204,
      "kl": 0.21240234375,
      "learning_rate": 3.7263486596572415e-08,
      "loss": -0.0196,
      "reward": 1.675000011920929,
      "reward_std": 0.12436701357364655,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 1.0,
      "step": 2707
    },
    {
      "completion_length": 137.67500495910645,
      "epoch": 2.808709175738725,
      "grad_norm": 0.4275289475917816,
      "kl": 0.485595703125,
      "learning_rate": 3.686338064274908e-08,
      "loss": 0.043,
      "reward": 1.7500000298023224,
      "reward_std": 0.2361704185605049,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2708
    },
    {
      "completion_length": 149.50000381469727,
      "epoch": 2.809745982374287,
      "grad_norm": 1.577640175819397,
      "kl": 1.546875,
      "learning_rate": 3.646540758644051e-08,
      "loss": 0.1611,
      "reward": 1.6500000059604645,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2709
    },
    {
      "completion_length": 274.3000087738037,
      "epoch": 2.8107827890098496,
      "grad_norm": 2.477776050567627,
      "kl": 3.36767578125,
      "learning_rate": 3.606956800779415e-08,
      "loss": 0.2846,
      "reward": 1.2750000208616257,
      "reward_std": 0.292072132229805,
      "rewards/accuracy_reward": 0.37500000931322575,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2710
    },
    {
      "completion_length": 285.47500801086426,
      "epoch": 2.811819595645412,
      "grad_norm": 0.6023179292678833,
      "kl": 1.4501953125,
      "learning_rate": 3.5675862483848245e-08,
      "loss": 0.2982,
      "reward": 1.8250000178813934,
      "reward_std": 0.3913118727505207,
      "rewards/accuracy_reward": 0.925000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2711
    },
    {
      "completion_length": 87.9000015258789,
      "epoch": 2.8128564022809748,
      "grad_norm": 0.442744642496109,
      "kl": 0.19482421875,
      "learning_rate": 3.528429158852969e-08,
      "loss": -0.0065,
      "reward": 1.4000000208616257,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2712
    },
    {
      "completion_length": 135.12500381469727,
      "epoch": 2.813893208916537,
      "grad_norm": 0.49087026715278625,
      "kl": 0.91259765625,
      "learning_rate": 3.48948558926534e-08,
      "loss": 0.0843,
      "reward": 1.300000011920929,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.32500000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2713
    },
    {
      "completion_length": 139.6750030517578,
      "epoch": 2.8149300155520995,
      "grad_norm": 0.3492768704891205,
      "kl": 0.388671875,
      "learning_rate": 3.450755596392208e-08,
      "loss": 0.0897,
      "reward": 1.550000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2714
    },
    {
      "completion_length": 187.82500839233398,
      "epoch": 2.815966822187662,
      "grad_norm": 0.617585301399231,
      "kl": 0.54296875,
      "learning_rate": 3.41223923669251e-08,
      "loss": 0.1895,
      "reward": 1.75,
      "reward_std": 0.2795084938406944,
      "rewards/accuracy_reward": 0.8250000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 2715
    },
    {
      "completion_length": 190.22500228881836,
      "epoch": 2.8170036288232243,
      "grad_norm": 1.9915069341659546,
      "kl": 0.815185546875,
      "learning_rate": 3.373936566313718e-08,
      "loss": 0.153,
      "reward": 1.9000000357627869,
      "reward_std": 0.22360678017139435,
      "rewards/accuracy_reward": 0.9500000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2716
    },
    {
      "completion_length": 152.12500381469727,
      "epoch": 2.818040435458787,
      "grad_norm": 0.42638441920280457,
      "kl": 0.3427734375,
      "learning_rate": 3.335847641091816e-08,
      "loss": 0.0714,
      "reward": 1.5250000059604645,
      "reward_std": 0.18026872351765633,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2717
    },
    {
      "completion_length": 227.5000057220459,
      "epoch": 2.8190772420943495,
      "grad_norm": 4.111725807189941,
      "kl": 1.18505859375,
      "learning_rate": 3.297972516551173e-08,
      "loss": 0.2229,
      "reward": 1.7000000178813934,
      "reward_std": 0.3913118876516819,
      "rewards/accuracy_reward": 0.7750000208616257,
      "rewards/format_reward": 0.925000011920929,
      "step": 2718
    },
    {
      "completion_length": 97.75000190734863,
      "epoch": 2.820114048729912,
      "grad_norm": 0.29310405254364014,
      "kl": 0.248046875,
      "learning_rate": 3.260311247904574e-08,
      "loss": 0.0094,
      "reward": 1.5999999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 1.0,
      "step": 2719
    },
    {
      "completion_length": 156.0250072479248,
      "epoch": 2.8211508553654743,
      "grad_norm": 1.5296965837478638,
      "kl": 1.225830078125,
      "learning_rate": 3.222863890052935e-08,
      "loss": 0.147,
      "reward": 1.5500000417232513,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2720
    },
    {
      "completion_length": 134.9500026702881,
      "epoch": 2.8221876620010367,
      "grad_norm": 0.6774704456329346,
      "kl": 0.580078125,
      "learning_rate": 3.185630497585423e-08,
      "loss": 0.0836,
      "reward": 1.625,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2721
    },
    {
      "completion_length": 222.3750057220459,
      "epoch": 2.8232244686365995,
      "grad_norm": 0.5395699739456177,
      "kl": 1.84619140625,
      "learning_rate": 3.148611124779288e-08,
      "loss": 0.3052,
      "reward": 1.5750000178813934,
      "reward_std": 0.3913118876516819,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2722
    },
    {
      "completion_length": 234.0250015258789,
      "epoch": 2.824261275272162,
      "grad_norm": 1.7121381759643555,
      "kl": 2.06640625,
      "learning_rate": 3.111805825599745e-08,
      "loss": 0.3202,
      "reward": 1.5000000298023224,
      "reward_std": 0.447213601320982,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2723
    },
    {
      "completion_length": 134.0250015258789,
      "epoch": 2.8252980819077242,
      "grad_norm": 0.3554559350013733,
      "kl": 0.4794921875,
      "learning_rate": 3.075214653699976e-08,
      "loss": 0.0846,
      "reward": 1.75,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2724
    },
    {
      "completion_length": 272.450008392334,
      "epoch": 2.8263348885432866,
      "grad_norm": 1.0414302349090576,
      "kl": 1.6962890625,
      "learning_rate": 3.0388376624210456e-08,
      "loss": 0.2326,
      "reward": 1.550000011920929,
      "reward_std": 0.4038754962384701,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.875,
      "step": 2725
    },
    {
      "completion_length": 281.52501106262207,
      "epoch": 2.827371695178849,
      "grad_norm": 9.531081199645996,
      "kl": 3.955078125,
      "learning_rate": 3.0026749047917525e-08,
      "loss": 0.4187,
      "reward": 1.7500000298023224,
      "reward_std": 0.42900275811553,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2726
    },
    {
      "completion_length": 221.40000343322754,
      "epoch": 2.828408501814412,
      "grad_norm": 5.895714282989502,
      "kl": 2.6103515625,
      "learning_rate": 2.9667264335285937e-08,
      "loss": 0.3564,
      "reward": 1.600000023841858,
      "reward_std": 0.38409101963043213,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2727
    },
    {
      "completion_length": 378.7000102996826,
      "epoch": 2.829445308449974,
      "grad_norm": 1.7167994976043701,
      "kl": 2.89697265625,
      "learning_rate": 2.930992301035701e-08,
      "loss": 0.4967,
      "reward": 1.475000023841858,
      "reward_std": 0.5282425209879875,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2728
    },
    {
      "completion_length": 234.47500228881836,
      "epoch": 2.8304821150855366,
      "grad_norm": 0.5508548021316528,
      "kl": 0.826904296875,
      "learning_rate": 2.8954725594047716e-08,
      "loss": 0.2113,
      "reward": 1.324999988079071,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2729
    },
    {
      "completion_length": 306.1000118255615,
      "epoch": 2.831518921721099,
      "grad_norm": 0.6908534169197083,
      "kl": 1.283203125,
      "learning_rate": 2.8601672604149365e-08,
      "loss": 0.2826,
      "reward": 1.6500000357627869,
      "reward_std": 0.5288755372166634,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2730
    },
    {
      "completion_length": 140.35000610351562,
      "epoch": 2.8325557283566614,
      "grad_norm": 2.3090004920959473,
      "kl": 0.8212890625,
      "learning_rate": 2.8250764555327436e-08,
      "loss": 0.1201,
      "reward": 1.449999988079071,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2731
    },
    {
      "completion_length": 132.95000457763672,
      "epoch": 2.833592534992224,
      "grad_norm": 0.18538494408130646,
      "kl": 0.462646484375,
      "learning_rate": 2.790200195912057e-08,
      "loss": 0.0739,
      "reward": 1.8250000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2732
    },
    {
      "completion_length": 194.82500839233398,
      "epoch": 2.8346293416277866,
      "grad_norm": 0.6282142996788025,
      "kl": 0.599609375,
      "learning_rate": 2.755538532393992e-08,
      "loss": 0.1336,
      "reward": 1.7750000357627869,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2733
    },
    {
      "completion_length": 146.3250026702881,
      "epoch": 2.835666148263349,
      "grad_norm": 0.7779990434646606,
      "kl": 0.69580078125,
      "learning_rate": 2.721091515506813e-08,
      "loss": 0.1099,
      "reward": 1.5500000417232513,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2734
    },
    {
      "completion_length": 227.64999771118164,
      "epoch": 2.8367029548989113,
      "grad_norm": 0.20258407294750214,
      "kl": 0.668212890625,
      "learning_rate": 2.6868591954659017e-08,
      "loss": 0.0685,
      "reward": 1.7750000357627869,
      "reward_std": 0.18026870489120483,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2735
    },
    {
      "completion_length": 146.82500076293945,
      "epoch": 2.8377397615344737,
      "grad_norm": 0.20224982500076294,
      "kl": 0.669921875,
      "learning_rate": 2.652841622173674e-08,
      "loss": 0.1108,
      "reward": 1.5500000268220901,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2736
    },
    {
      "completion_length": 272.6750030517578,
      "epoch": 2.8387765681700365,
      "grad_norm": 0.49968308210372925,
      "kl": 1.40625,
      "learning_rate": 2.6190388452194623e-08,
      "loss": 0.2931,
      "reward": 1.200000062584877,
      "reward_std": 0.3913119025528431,
      "rewards/accuracy_reward": 0.30000000447034836,
      "rewards/format_reward": 0.9000000357627869,
      "step": 2737
    },
    {
      "completion_length": 249.0250072479248,
      "epoch": 2.839813374805599,
      "grad_norm": 2.41104793548584,
      "kl": 1.678466796875,
      "learning_rate": 2.5854509138795167e-08,
      "loss": 0.3292,
      "reward": 1.4500000178813934,
      "reward_std": 0.3479737974703312,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2738
    },
    {
      "completion_length": 197.42500495910645,
      "epoch": 2.8408501814411613,
      "grad_norm": 0.4147087037563324,
      "kl": 0.55224609375,
      "learning_rate": 2.5520778771168705e-08,
      "loss": 0.147,
      "reward": 1.4749999940395355,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2739
    },
    {
      "completion_length": 282.65000343322754,
      "epoch": 2.8418869880767237,
      "grad_norm": 0.9939954280853271,
      "kl": 1.13671875,
      "learning_rate": 2.5189197835812915e-08,
      "loss": 0.2535,
      "reward": 1.6750000417232513,
      "reward_std": 0.5590169653296471,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.8500000238418579,
      "step": 2740
    },
    {
      "completion_length": 134.30000114440918,
      "epoch": 2.842923794712286,
      "grad_norm": 0.8458899259567261,
      "kl": 0.81494140625,
      "learning_rate": 2.4859766816092144e-08,
      "loss": 0.04,
      "reward": 1.6000000089406967,
      "reward_std": 0.18090170249342918,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.925000011920929,
      "step": 2741
    },
    {
      "completion_length": 202.60000801086426,
      "epoch": 2.843960601347849,
      "grad_norm": 1.121363878250122,
      "kl": 0.6572265625,
      "learning_rate": 2.4532486192236914e-08,
      "loss": 0.1421,
      "reward": 1.7000000476837158,
      "reward_std": 0.3479738309979439,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2742
    },
    {
      "completion_length": 128.04999923706055,
      "epoch": 2.844997407983411,
      "grad_norm": 0.4545036554336548,
      "kl": 0.492919921875,
      "learning_rate": 2.4207356441342587e-08,
      "loss": 0.0907,
      "reward": 1.4000000208616257,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2743
    },
    {
      "completion_length": 141.47500610351562,
      "epoch": 2.8460342146189737,
      "grad_norm": 0.45426639914512634,
      "kl": 0.304443359375,
      "learning_rate": 2.3884378037369703e-08,
      "loss": 0.0783,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2744
    },
    {
      "completion_length": 191.60000801086426,
      "epoch": 2.847071021254536,
      "grad_norm": 0.49299612641334534,
      "kl": 1.090576171875,
      "learning_rate": 2.3563551451141806e-08,
      "loss": 0.171,
      "reward": 1.5750000476837158,
      "reward_std": 0.3354101926088333,
      "rewards/accuracy_reward": 0.6250000223517418,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2745
    },
    {
      "completion_length": 83.45000076293945,
      "epoch": 2.8481078278900984,
      "grad_norm": 0.09031056612730026,
      "kl": 0.201171875,
      "learning_rate": 2.3244877150346288e-08,
      "loss": 0.0076,
      "reward": 1.5,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 1.0,
      "step": 2746
    },
    {
      "completion_length": 199.30000686645508,
      "epoch": 2.849144634525661,
      "grad_norm": 0.2941316068172455,
      "kl": 0.75927734375,
      "learning_rate": 2.2928355599532712e-08,
      "loss": 0.1529,
      "reward": 1.4000000357627869,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2747
    },
    {
      "completion_length": 144.87500190734863,
      "epoch": 2.850181441161223,
      "grad_norm": 1.0592080354690552,
      "kl": 0.5233154296875,
      "learning_rate": 2.2613987260112657e-08,
      "loss": 0.0852,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2748
    },
    {
      "completion_length": 228.600004196167,
      "epoch": 2.851218247796786,
      "grad_norm": 0.3585069477558136,
      "kl": 1.044189453125,
      "learning_rate": 2.230177259035887e-08,
      "loss": 0.216,
      "reward": 1.5250000059604645,
      "reward_std": 0.22360679134726524,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 2749
    },
    {
      "completion_length": 155.72500610351562,
      "epoch": 2.8522550544323484,
      "grad_norm": 0.530051589012146,
      "kl": 0.8980712890625,
      "learning_rate": 2.1991712045404288e-08,
      "loss": 0.1121,
      "reward": 1.7500000298023224,
      "reward_std": 0.27228760346770287,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2750
    },
    {
      "completion_length": 215.60000228881836,
      "epoch": 2.8532918610679108,
      "grad_norm": 2.6323437690734863,
      "kl": 1.85009765625,
      "learning_rate": 2.168380607724152e-08,
      "loss": 0.2744,
      "reward": 1.1500000059604645,
      "reward_std": 0.3354101777076721,
      "rewards/accuracy_reward": 0.22500000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 2751
    },
    {
      "completion_length": 192.975004196167,
      "epoch": 2.854328667703473,
      "grad_norm": 0.1603558361530304,
      "kl": 0.489013671875,
      "learning_rate": 2.1378055134723183e-08,
      "loss": 0.148,
      "reward": 1.3500000089406967,
      "reward_std": 0.1442900486290455,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2752
    },
    {
      "completion_length": 191.67500114440918,
      "epoch": 2.8553654743390355,
      "grad_norm": 2.326918363571167,
      "kl": 1.961181640625,
      "learning_rate": 2.107445966355942e-08,
      "loss": 0.2405,
      "reward": 1.725000023841858,
      "reward_std": 0.2920721024274826,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2753
    },
    {
      "completion_length": 118.25000381469727,
      "epoch": 2.8564022809745984,
      "grad_norm": 0.7669432163238525,
      "kl": 0.515625,
      "learning_rate": 2.0773020106318874e-08,
      "loss": 0.1094,
      "reward": 1.4000000059604645,
      "reward_std": 0.2361704260110855,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2754
    },
    {
      "completion_length": 192.17500495910645,
      "epoch": 2.8574390876101607,
      "grad_norm": 1.0182937383651733,
      "kl": 2.23828125,
      "learning_rate": 2.0473736902426878e-08,
      "loss": 0.2641,
      "reward": 1.8250000476837158,
      "reward_std": 0.2920721173286438,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2755
    },
    {
      "completion_length": 200.7750015258789,
      "epoch": 2.858475894245723,
      "grad_norm": 42.23041915893555,
      "kl": 9.769775390625,
      "learning_rate": 2.0176610488165936e-08,
      "loss": 0.4195,
      "reward": 1.449999988079071,
      "reward_std": 0.30463575571775436,
      "rewards/accuracy_reward": 0.5500000044703484,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2756
    },
    {
      "completion_length": 151.70000457763672,
      "epoch": 2.8595127008812855,
      "grad_norm": 0.5360174179077148,
      "kl": 0.639892578125,
      "learning_rate": 1.9881641296673747e-08,
      "loss": 0.1079,
      "reward": 1.4500000178813934,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5000000018626451,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2757
    },
    {
      "completion_length": 199.82500076293945,
      "epoch": 2.860549507516848,
      "grad_norm": 0.4138404428958893,
      "kl": 0.65869140625,
      "learning_rate": 1.9588829757943848e-08,
      "loss": 0.1265,
      "reward": 1.6749999821186066,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2758
    },
    {
      "completion_length": 111.07500457763672,
      "epoch": 2.8615863141524107,
      "grad_norm": 0.8471639156341553,
      "kl": 0.71484375,
      "learning_rate": 1.9298176298824467e-08,
      "loss": 0.1236,
      "reward": 1.6000000536441803,
      "reward_std": 0.292072132229805,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2759
    },
    {
      "completion_length": 241.1250057220459,
      "epoch": 2.862623120787973,
      "grad_norm": 0.36379024386405945,
      "kl": 0.72509765625,
      "learning_rate": 1.9009681343017503e-08,
      "loss": 0.1672,
      "reward": 1.725000023841858,
      "reward_std": 0.2795085087418556,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 2760
    },
    {
      "completion_length": 84.32500267028809,
      "epoch": 2.8636599274235355,
      "grad_norm": 0.36007457971572876,
      "kl": 0.400390625,
      "learning_rate": 1.8723345311078388e-08,
      "loss": 0.0275,
      "reward": 1.7249999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2761
    },
    {
      "completion_length": 135.9000015258789,
      "epoch": 2.864696734059098,
      "grad_norm": 0.6733860373497009,
      "kl": 0.402099609375,
      "learning_rate": 1.8439168620415725e-08,
      "loss": 0.0882,
      "reward": 1.4750000536441803,
      "reward_std": 0.3966546133160591,
      "rewards/accuracy_reward": 0.5500000193715096,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2762
    },
    {
      "completion_length": 143.9250030517578,
      "epoch": 2.8657335406946602,
      "grad_norm": 0.4623294174671173,
      "kl": 0.418212890625,
      "learning_rate": 1.815715168528981e-08,
      "loss": 0.0611,
      "reward": 1.75,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2763
    },
    {
      "completion_length": 289.100004196167,
      "epoch": 2.866770347330223,
      "grad_norm": 0.32892030477523804,
      "kl": 0.8984375,
      "learning_rate": 1.7877294916812782e-08,
      "loss": 0.2019,
      "reward": 1.25,
      "reward_std": 0.36789684370160103,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2764
    },
    {
      "completion_length": 232.5999984741211,
      "epoch": 2.8678071539657854,
      "grad_norm": 3.9510769844055176,
      "kl": 1.39013671875,
      "learning_rate": 1.7599598722947974e-08,
      "loss": 0.2017,
      "reward": 1.3250000178813934,
      "reward_std": 0.4495587646961212,
      "rewards/accuracy_reward": 0.42500001192092896,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2765
    },
    {
      "completion_length": 199.32500648498535,
      "epoch": 2.868843960601348,
      "grad_norm": 0.4974198341369629,
      "kl": 0.76171875,
      "learning_rate": 1.7324063508508736e-08,
      "loss": 0.0994,
      "reward": 1.375,
      "reward_std": 0.3245587572455406,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2766
    },
    {
      "completion_length": 142.07500553131104,
      "epoch": 2.86988076723691,
      "grad_norm": 2.5557098388671875,
      "kl": 0.552978515625,
      "learning_rate": 1.7050689675158436e-08,
      "loss": 0.0755,
      "reward": 1.675000011920929,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2767
    },
    {
      "completion_length": 208.32500648498535,
      "epoch": 2.8709175738724726,
      "grad_norm": 2.344703435897827,
      "kl": 0.91943359375,
      "learning_rate": 1.67794776214098e-08,
      "loss": 0.1009,
      "reward": 1.6750000417232513,
      "reward_std": 0.40387552976608276,
      "rewards/accuracy_reward": 0.7750000134110451,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2768
    },
    {
      "completion_length": 141.1500015258789,
      "epoch": 2.8719543805080354,
      "grad_norm": 0.23868072032928467,
      "kl": 0.224365234375,
      "learning_rate": 1.6510427742624067e-08,
      "loss": 0.0527,
      "reward": 1.75,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.775000000372529,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2769
    },
    {
      "completion_length": 142.17499923706055,
      "epoch": 2.872991187143598,
      "grad_norm": 2.4860002994537354,
      "kl": 1.144775390625,
      "learning_rate": 1.6243540431010507e-08,
      "loss": 0.1597,
      "reward": 1.2500000149011612,
      "reward_std": 0.38409100100398064,
      "rewards/accuracy_reward": 0.35000000335276127,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2770
    },
    {
      "completion_length": 236.37500381469727,
      "epoch": 2.87402799377916,
      "grad_norm": 0.8142707943916321,
      "kl": 0.69091796875,
      "learning_rate": 1.5978816075625737e-08,
      "loss": 0.1855,
      "reward": 1.6750000417232513,
      "reward_std": 0.31719936430454254,
      "rewards/accuracy_reward": 0.7500000149011612,
      "rewards/format_reward": 0.925000011920929,
      "step": 2771
    },
    {
      "completion_length": 243.29999923706055,
      "epoch": 2.8750648004147226,
      "grad_norm": 0.48338213562965393,
      "kl": 1.3851318359375,
      "learning_rate": 1.5716255062374062e-08,
      "loss": 0.2181,
      "reward": 1.725000023841858,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.800000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2772
    },
    {
      "completion_length": 105.47500419616699,
      "epoch": 2.876101607050285,
      "grad_norm": 0.32397958636283875,
      "kl": 0.234619140625,
      "learning_rate": 1.545585777400499e-08,
      "loss": -0.0128,
      "reward": 1.6500000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.650000000372529,
      "rewards/format_reward": 1.0,
      "step": 2773
    },
    {
      "completion_length": 141.39999961853027,
      "epoch": 2.8771384136858478,
      "grad_norm": 0.7080397009849548,
      "kl": 0.331787109375,
      "learning_rate": 1.51976245901147e-08,
      "loss": 0.0749,
      "reward": 1.8000000417232513,
      "reward_std": 0.24151314049959183,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2774
    },
    {
      "completion_length": 244.10000228881836,
      "epoch": 2.87817522032141,
      "grad_norm": 0.8607521653175354,
      "kl": 0.906494140625,
      "learning_rate": 1.4941555887144576e-08,
      "loss": 0.2003,
      "reward": 1.5000000298023224,
      "reward_std": 0.3913118802011013,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2775
    },
    {
      "completion_length": 148.37500190734863,
      "epoch": 2.8792120269569725,
      "grad_norm": 0.3434334993362427,
      "kl": 0.34033203125,
      "learning_rate": 1.468765203838035e-08,
      "loss": 0.0502,
      "reward": 1.800000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.8250000104308128,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2776
    },
    {
      "completion_length": 152.725004196167,
      "epoch": 2.880248833592535,
      "grad_norm": 0.5830362439155579,
      "kl": 0.48388671875,
      "learning_rate": 1.4435913413952118e-08,
      "loss": 0.0855,
      "reward": 1.5250000357627869,
      "reward_std": 0.31719937920570374,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2777
    },
    {
      "completion_length": 311.0500030517578,
      "epoch": 2.8812856402280973,
      "grad_norm": 1.385844349861145,
      "kl": 0.8330078125,
      "learning_rate": 1.418634038083383e-08,
      "loss": 0.3264,
      "reward": 1.550000011920929,
      "reward_std": 0.4472136050462723,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2778
    },
    {
      "completion_length": 88.72500038146973,
      "epoch": 2.88232244686366,
      "grad_norm": 7.81191349029541,
      "kl": 4.72607421875,
      "learning_rate": 1.3938933302841961e-08,
      "loss": 0.1818,
      "reward": 1.725000023841858,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.7500000186264515,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2779
    },
    {
      "completion_length": 144.77499961853027,
      "epoch": 2.8833592534992225,
      "grad_norm": 0.36271801590919495,
      "kl": 0.22412109375,
      "learning_rate": 1.3693692540636017e-08,
      "loss": 0.0836,
      "reward": 1.125,
      "reward_std": 0.20019177719950676,
      "rewards/accuracy_reward": 0.15000000223517418,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2780
    },
    {
      "completion_length": 111.0250015258789,
      "epoch": 2.884396060134785,
      "grad_norm": 0.36495858430862427,
      "kl": 0.23486328125,
      "learning_rate": 1.3450618451717355e-08,
      "loss": 0.0337,
      "reward": 1.5000000298023224,
      "reward_std": 0.13693063706159592,
      "rewards/accuracy_reward": 0.5000000037252903,
      "rewards/format_reward": 1.0,
      "step": 2781
    },
    {
      "completion_length": 244.600004196167,
      "epoch": 2.8854328667703473,
      "grad_norm": 5.447226047515869,
      "kl": 2.8271484375,
      "learning_rate": 1.3209711390429202e-08,
      "loss": 0.3353,
      "reward": 1.5250000357627869,
      "reward_std": 0.31199515610933304,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2782
    },
    {
      "completion_length": 206.0250072479248,
      "epoch": 2.8864696734059097,
      "grad_norm": 2.8135359287261963,
      "kl": 2.600830078125,
      "learning_rate": 1.2970971707955137e-08,
      "loss": 0.3269,
      "reward": 1.5250000208616257,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2783
    },
    {
      "completion_length": 99.82500267028809,
      "epoch": 2.8875064800414725,
      "grad_norm": 0.4570198655128479,
      "kl": 0.268310546875,
      "learning_rate": 1.273439975231977e-08,
      "loss": -0.0161,
      "reward": 1.8250000178813934,
      "reward_std": 0.18026871234178543,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 1.0,
      "step": 2784
    },
    {
      "completion_length": 255.75000953674316,
      "epoch": 2.888543286677035,
      "grad_norm": 2.8772873878479004,
      "kl": 2.87451171875,
      "learning_rate": 1.249999586838757e-08,
      "loss": 0.3417,
      "reward": 1.8000000417232513,
      "reward_std": 0.3479738160967827,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2785
    },
    {
      "completion_length": 158.10000610351562,
      "epoch": 2.8895800933125972,
      "grad_norm": 0.8739559054374695,
      "kl": 0.499267578125,
      "learning_rate": 1.2267760397862038e-08,
      "loss": 0.0896,
      "reward": 1.550000011920929,
      "reward_std": 0.16770510375499725,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2786
    },
    {
      "completion_length": 284.37500381469727,
      "epoch": 2.8906168999481596,
      "grad_norm": 2.8748176097869873,
      "kl": 2.48779296875,
      "learning_rate": 1.2037693679286532e-08,
      "loss": 0.3573,
      "reward": 1.5250000357627869,
      "reward_std": 0.36053743213415146,
      "rewards/accuracy_reward": 0.6250000111758709,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2787
    },
    {
      "completion_length": 184.07500648498535,
      "epoch": 2.891653706583722,
      "grad_norm": 1.1407324075698853,
      "kl": 1.103759765625,
      "learning_rate": 1.180979604804211e-08,
      "loss": 0.1748,
      "reward": 1.675000011920929,
      "reward_std": 0.25609345734119415,
      "rewards/accuracy_reward": 0.7250000033527613,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2788
    },
    {
      "completion_length": 104.22500038146973,
      "epoch": 2.892690513219285,
      "grad_norm": 0.29753732681274414,
      "kl": 0.54541015625,
      "learning_rate": 1.1584067836348189e-08,
      "loss": 0.0501,
      "reward": 1.7000000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2789
    },
    {
      "completion_length": 348.0499954223633,
      "epoch": 2.893727319854847,
      "grad_norm": 1.2128010988235474,
      "kl": 0.8701171875,
      "learning_rate": 1.1360509373261386e-08,
      "loss": 0.2855,
      "reward": 1.475000023841858,
      "reward_std": 0.6040672585368156,
      "rewards/accuracy_reward": 0.6000000163912773,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2790
    },
    {
      "completion_length": 96.67500305175781,
      "epoch": 2.8947641264904096,
      "grad_norm": 0.31217020750045776,
      "kl": 0.2293701171875,
      "learning_rate": 1.113912098467601e-08,
      "loss": 0.0095,
      "reward": 1.699999988079071,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 1.0,
      "step": 2791
    },
    {
      "completion_length": 262.2000045776367,
      "epoch": 2.895800933125972,
      "grad_norm": 0.5420238971710205,
      "kl": 1.084228515625,
      "learning_rate": 1.0919902993322406e-08,
      "loss": 0.1901,
      "reward": 1.5250000357627869,
      "reward_std": 0.273861289024353,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2792
    },
    {
      "completion_length": 195.25000762939453,
      "epoch": 2.8968377397615344,
      "grad_norm": 0.403917521238327,
      "kl": 0.3302001953125,
      "learning_rate": 1.0702855718766947e-08,
      "loss": 0.1116,
      "reward": 1.5000000149011612,
      "reward_std": 0.18026870489120483,
      "rewards/accuracy_reward": 0.5749999973922968,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2793
    },
    {
      "completion_length": 122.50000190734863,
      "epoch": 2.897874546397097,
      "grad_norm": 0.3548263609409332,
      "kl": 0.253662109375,
      "learning_rate": 1.0487979477411869e-08,
      "loss": 0.085,
      "reward": 1.949999988079071,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.9750000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2794
    },
    {
      "completion_length": 139.22500610351562,
      "epoch": 2.8989113530326596,
      "grad_norm": 1.9345402717590332,
      "kl": 0.69384765625,
      "learning_rate": 1.027527458249461e-08,
      "loss": 0.033,
      "reward": 1.675000011920929,
      "reward_std": 0.16770508512854576,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2795
    },
    {
      "completion_length": 196.20000076293945,
      "epoch": 2.899948159668222,
      "grad_norm": 4.578166484832764,
      "kl": 0.85546875,
      "learning_rate": 1.006474134408697e-08,
      "loss": 0.1599,
      "reward": 1.5250000059604645,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2796
    },
    {
      "completion_length": 153.00000381469727,
      "epoch": 2.9009849663037843,
      "grad_norm": 0.7297366261482239,
      "kl": 0.26953125,
      "learning_rate": 9.856380069095617e-09,
      "loss": 0.0927,
      "reward": 1.4500000029802322,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2797
    },
    {
      "completion_length": 309.95000648498535,
      "epoch": 2.9020217729393467,
      "grad_norm": 0.9720076322555542,
      "kl": 1.0972900390625,
      "learning_rate": 9.650191061260083e-09,
      "loss": 0.286,
      "reward": 1.7500000298023224,
      "reward_std": 0.41643913835287094,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2798
    },
    {
      "completion_length": 205.10000610351562,
      "epoch": 2.9030585795749095,
      "grad_norm": 1.149886131286621,
      "kl": 0.373046875,
      "learning_rate": 9.446174621154435e-09,
      "loss": 0.0847,
      "reward": 1.550000011920929,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.6000000033527613,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2799
    },
    {
      "completion_length": 190.32499885559082,
      "epoch": 2.9040953862104715,
      "grad_norm": 2.1195366382598877,
      "kl": 0.64404296875,
      "learning_rate": 9.24433104618444e-09,
      "loss": 0.1434,
      "reward": 1.6250000298023224,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.6750000268220901,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2800
    },
    {
      "completion_length": 162.1999969482422,
      "epoch": 2.9051321928460343,
      "grad_norm": 0.42822936177253723,
      "kl": 0.375732421875,
      "learning_rate": 9.044660630589396e-09,
      "loss": 0.0637,
      "reward": 1.699999988079071,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.75,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2801
    },
    {
      "completion_length": 149.00000381469727,
      "epoch": 2.9061689994815967,
      "grad_norm": 0.3750414550304413,
      "kl": 0.336181640625,
      "learning_rate": 8.847163665440138e-09,
      "loss": 0.054,
      "reward": 1.625,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2802
    },
    {
      "completion_length": 199.75000190734863,
      "epoch": 2.907205806117159,
      "grad_norm": 0.4641471207141876,
      "kl": 1.182373046875,
      "learning_rate": 8.651840438639202e-09,
      "loss": 0.1998,
      "reward": 1.5000000298023224,
      "reward_std": 0.2795085124671459,
      "rewards/accuracy_reward": 0.550000011920929,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2803
    },
    {
      "completion_length": 245.10000801086426,
      "epoch": 2.908242612752722,
      "grad_norm": 0.3972705006599426,
      "kl": 0.95849609375,
      "learning_rate": 8.458691234920156e-09,
      "loss": 0.2342,
      "reward": 1.5750000476837158,
      "reward_std": 0.3479738086462021,
      "rewards/accuracy_reward": 0.6500000208616257,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2804
    },
    {
      "completion_length": 236.47499656677246,
      "epoch": 2.909279419388284,
      "grad_norm": 0.5866678953170776,
      "kl": 1.632080078125,
      "learning_rate": 8.267716335848107e-09,
      "loss": 0.2723,
      "reward": 1.5250000059604645,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 2805
    },
    {
      "completion_length": 208.20001029968262,
      "epoch": 2.9103162260238467,
      "grad_norm": 1.4535342454910278,
      "kl": 1.75732421875,
      "learning_rate": 8.078916019817362e-09,
      "loss": 0.2335,
      "reward": 1.6000000536441803,
      "reward_std": 0.3354102000594139,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2806
    },
    {
      "completion_length": 198.77499961853027,
      "epoch": 2.911353032659409,
      "grad_norm": 1.431024432182312,
      "kl": 0.57958984375,
      "learning_rate": 7.892290562053427e-09,
      "loss": 0.1265,
      "reward": 1.699999988079071,
      "reward_std": 0.1604842059314251,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2807
    },
    {
      "completion_length": 230.10000610351562,
      "epoch": 2.9123898392949714,
      "grad_norm": 1.2065316438674927,
      "kl": 0.556884765625,
      "learning_rate": 7.707840234610685e-09,
      "loss": 0.1219,
      "reward": 1.550000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2808
    },
    {
      "completion_length": 157.50000381469727,
      "epoch": 2.913426645930534,
      "grad_norm": 0.4459967017173767,
      "kl": 0.5992431640625,
      "learning_rate": 7.525565306372883e-09,
      "loss": 0.1,
      "reward": 1.425000011920929,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.925000011920929,
      "step": 2809
    },
    {
      "completion_length": 254.40000343322754,
      "epoch": 2.914463452566096,
      "grad_norm": 1.3066574335098267,
      "kl": 1.3984375,
      "learning_rate": 7.345466043052473e-09,
      "loss": 0.2282,
      "reward": 1.5250000357627869,
      "reward_std": 0.3966546542942524,
      "rewards/accuracy_reward": 0.600000012665987,
      "rewards/format_reward": 0.925000011920929,
      "step": 2810
    },
    {
      "completion_length": 97.92500114440918,
      "epoch": 2.915500259201659,
      "grad_norm": 0.2889236807823181,
      "kl": 0.19189453125,
      "learning_rate": 7.167542707190444e-09,
      "loss": 0.0092,
      "reward": 1.7750000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 1.0,
      "step": 2811
    },
    {
      "completion_length": 99.50000190734863,
      "epoch": 2.9165370658372214,
      "grad_norm": 0.33364081382751465,
      "kl": 0.185791015625,
      "learning_rate": 6.991795558155822e-09,
      "loss": 0.0524,
      "reward": 1.5250000059604645,
      "reward_std": 0.12436701729893684,
      "rewards/accuracy_reward": 0.5249999985098839,
      "rewards/format_reward": 1.0,
      "step": 2812
    },
    {
      "completion_length": 239.1750030517578,
      "epoch": 2.9175738724727838,
      "grad_norm": 1.0660427808761597,
      "kl": 1.21875,
      "learning_rate": 6.818224852145172e-09,
      "loss": 0.2254,
      "reward": 1.6500000059604645,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.925000011920929,
      "step": 2813
    },
    {
      "completion_length": 157.57500457763672,
      "epoch": 2.918610679108346,
      "grad_norm": 1.377952218055725,
      "kl": 1.01318359375,
      "learning_rate": 6.6468308421824274e-09,
      "loss": 0.1058,
      "reward": 1.5250000357627869,
      "reward_std": 0.248734038323164,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2814
    },
    {
      "completion_length": 241.90000534057617,
      "epoch": 2.9196474857439085,
      "grad_norm": 0.5748494267463684,
      "kl": 0.930419921875,
      "learning_rate": 6.47761377811823e-09,
      "loss": 0.2058,
      "reward": 1.3500000089406967,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.45000000298023224,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2815
    },
    {
      "completion_length": 119.22500419616699,
      "epoch": 2.9206842923794714,
      "grad_norm": 0.4970730245113373,
      "kl": 0.26513671875,
      "learning_rate": 6.310573906630257e-09,
      "loss": 0.0979,
      "reward": 1.4500000178813934,
      "reward_std": 0.2686570733785629,
      "rewards/accuracy_reward": 0.47500001080334187,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2816
    },
    {
      "completion_length": 96.75000190734863,
      "epoch": 2.9217210990150337,
      "grad_norm": 0.35038894414901733,
      "kl": 0.204833984375,
      "learning_rate": 6.145711471221726e-09,
      "loss": 0.0056,
      "reward": 1.4000000059604645,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.40000000037252903,
      "rewards/format_reward": 1.0,
      "step": 2817
    },
    {
      "completion_length": 140.17500114440918,
      "epoch": 2.922757905650596,
      "grad_norm": 0.33249610662460327,
      "kl": 0.255126953125,
      "learning_rate": 5.983026712222228e-09,
      "loss": 0.0675,
      "reward": 1.625,
      "reward_std": 0.11180339753627777,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2818
    },
    {
      "completion_length": 80.72500228881836,
      "epoch": 2.9237947122861585,
      "grad_norm": 0.46397265791893005,
      "kl": 0.224853515625,
      "learning_rate": 5.8225198667862246e-09,
      "loss": 0.0166,
      "reward": 1.8250000178813934,
      "reward_std": 0.06846531480550766,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 1.0,
      "step": 2819
    },
    {
      "completion_length": 160.95000648498535,
      "epoch": 2.924831518921721,
      "grad_norm": 0.5962483882904053,
      "kl": 0.43212890625,
      "learning_rate": 5.664191168894051e-09,
      "loss": 0.1584,
      "reward": 1.5000000298023224,
      "reward_std": 0.27950848266482353,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2820
    },
    {
      "completion_length": 206.32500839233398,
      "epoch": 2.9258683255572837,
      "grad_norm": 1.2632807493209839,
      "kl": 1.361083984375,
      "learning_rate": 5.508040849350415e-09,
      "loss": 0.1779,
      "reward": 1.4500000029802322,
      "reward_std": 0.20019173994660378,
      "rewards/accuracy_reward": 0.5250000096857548,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2821
    },
    {
      "completion_length": 231.77500534057617,
      "epoch": 2.926905132192846,
      "grad_norm": 0.9355142712593079,
      "kl": 0.81689453125,
      "learning_rate": 5.354069135784401e-09,
      "loss": 0.1595,
      "reward": 1.675000011920929,
      "reward_std": 0.2487340345978737,
      "rewards/accuracy_reward": 0.7750000096857548,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2822
    },
    {
      "completion_length": 244.40000915527344,
      "epoch": 2.9279419388284085,
      "grad_norm": 0.7836734652519226,
      "kl": 0.6640625,
      "learning_rate": 5.202276252649296e-09,
      "loss": 0.2317,
      "reward": 1.550000011920929,
      "reward_std": 0.47768355160951614,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2823
    },
    {
      "completion_length": 334.62501525878906,
      "epoch": 2.928978745463971,
      "grad_norm": 0.6796460151672363,
      "kl": 1.3388671875,
      "learning_rate": 5.052662421222432e-09,
      "loss": 0.3511,
      "reward": 1.5250000357627869,
      "reward_std": 0.5031153000891209,
      "rewards/accuracy_reward": 0.6500000134110451,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2824
    },
    {
      "completion_length": 133.44999885559082,
      "epoch": 2.9300155520995332,
      "grad_norm": 0.34717845916748047,
      "kl": 0.214111328125,
      "learning_rate": 4.905227859604178e-09,
      "loss": 0.0649,
      "reward": 1.7750000059604645,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.7999999970197678,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2825
    },
    {
      "completion_length": 234.5500030517578,
      "epoch": 2.931052358735096,
      "grad_norm": 0.960311770439148,
      "kl": 0.935546875,
      "learning_rate": 4.759972782718447e-09,
      "loss": 0.0659,
      "reward": 1.6250000298023224,
      "reward_std": 0.4415663816034794,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.925000011920929,
      "step": 2826
    },
    {
      "completion_length": 183.50000190734863,
      "epoch": 2.9320891653706584,
      "grad_norm": 0.5600541830062866,
      "kl": 1.413330078125,
      "learning_rate": 4.616897402311693e-09,
      "loss": 0.1967,
      "reward": 1.7750000059604645,
      "reward_std": 0.22360680252313614,
      "rewards/accuracy_reward": 0.8250000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2827
    },
    {
      "completion_length": 255.47501182556152,
      "epoch": 2.933125972006221,
      "grad_norm": 0.6802055239677429,
      "kl": 1.326171875,
      "learning_rate": 4.476001926953244e-09,
      "loss": 0.1267,
      "reward": 1.324999988079071,
      "reward_std": 0.4525563344359398,
      "rewards/accuracy_reward": 0.42500000819563866,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2828
    },
    {
      "completion_length": 254.65000343322754,
      "epoch": 2.934162778641783,
      "grad_norm": 0.5169515609741211,
      "kl": 0.796875,
      "learning_rate": 4.337286562034304e-09,
      "loss": 0.2734,
      "reward": 1.3250000476837158,
      "reward_std": 0.4363621920347214,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2829
    },
    {
      "completion_length": 187.80000495910645,
      "epoch": 2.9351995852773456,
      "grad_norm": 0.775519847869873,
      "kl": 1.2828369140625,
      "learning_rate": 4.20075150976812e-09,
      "loss": 0.1916,
      "reward": 1.5250000357627869,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5750000029802322,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2830
    },
    {
      "completion_length": 86.05000114440918,
      "epoch": 2.9362363919129084,
      "grad_norm": 0.3250177800655365,
      "kl": 0.4072265625,
      "learning_rate": 4.066396969189645e-09,
      "loss": 0.0489,
      "reward": 1.5,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.525000000372529,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2831
    },
    {
      "completion_length": 81.80000114440918,
      "epoch": 2.937273198548471,
      "grad_norm": 0.4437037706375122,
      "kl": 0.35986328125,
      "learning_rate": 3.9342231361552124e-09,
      "loss": 0.0202,
      "reward": 1.7249999940395355,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 1.0,
      "step": 2832
    },
    {
      "completion_length": 124.42500114440918,
      "epoch": 2.938310005184033,
      "grad_norm": 0.519273579120636,
      "kl": 1.2021484375,
      "learning_rate": 3.804230203342029e-09,
      "loss": 0.1129,
      "reward": 1.7000000178813934,
      "reward_std": 0.22360680997371674,
      "rewards/accuracy_reward": 0.7250000238418579,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2833
    },
    {
      "completion_length": 122.04999923706055,
      "epoch": 2.9393468118195956,
      "grad_norm": 0.3046756684780121,
      "kl": 0.4521484375,
      "learning_rate": 3.676418360248346e-09,
      "loss": 0.0764,
      "reward": 1.6000000089406967,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2834
    },
    {
      "completion_length": 143.12500190734863,
      "epoch": 2.940383618455158,
      "grad_norm": 0.5832608938217163,
      "kl": 0.783203125,
      "learning_rate": 3.5507877931926225e-09,
      "loss": 0.0932,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2835
    },
    {
      "completion_length": 130.62499809265137,
      "epoch": 2.9414204250907208,
      "grad_norm": 0.969624400138855,
      "kl": 0.72802734375,
      "learning_rate": 3.4273386853136968e-09,
      "loss": 0.0897,
      "reward": 1.475000023841858,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.5000000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2836
    },
    {
      "completion_length": 82.77499961853027,
      "epoch": 2.942457231726283,
      "grad_norm": 0.3338201642036438,
      "kl": 0.337890625,
      "learning_rate": 3.3060712165702834e-09,
      "loss": 0.0334,
      "reward": 1.475000023841858,
      "reward_std": 0.13693063706159592,
      "rewards/accuracy_reward": 0.4750000014901161,
      "rewards/format_reward": 1.0,
      "step": 2837
    },
    {
      "completion_length": 89.05000114440918,
      "epoch": 2.9434940383618455,
      "grad_norm": 0.4555501937866211,
      "kl": 0.365478515625,
      "learning_rate": 3.1869855637411407e-09,
      "loss": 0.0213,
      "reward": 1.4000000357627869,
      "reward_std": 0.2361704297363758,
      "rewards/accuracy_reward": 0.42500000447034836,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2838
    },
    {
      "completion_length": 178.47500610351562,
      "epoch": 2.944530844997408,
      "grad_norm": 0.36635392904281616,
      "kl": 0.44091796875,
      "learning_rate": 3.0700819004239044e-09,
      "loss": 0.095,
      "reward": 1.3250000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.375,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2839
    },
    {
      "completion_length": 129.97500228881836,
      "epoch": 2.9455676516329703,
      "grad_norm": 0.3499339520931244,
      "kl": 0.26953125,
      "learning_rate": 2.955360397035922e-09,
      "loss": 0.0732,
      "reward": 1.75,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2840
    },
    {
      "completion_length": 134.2000026702881,
      "epoch": 2.946604458268533,
      "grad_norm": 0.13274773955345154,
      "kl": 0.16943359375,
      "learning_rate": 2.842821220813252e-09,
      "loss": 0.0566,
      "reward": 1.8250000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2841
    },
    {
      "completion_length": 133.850004196167,
      "epoch": 2.9476412649040955,
      "grad_norm": 0.5422719120979309,
      "kl": 0.361328125,
      "learning_rate": 2.732464535810497e-09,
      "loss": 0.0432,
      "reward": 1.600000023841858,
      "reward_std": 0.16770510748028755,
      "rewards/accuracy_reward": 0.6250000093132257,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2842
    },
    {
      "completion_length": 240.77500343322754,
      "epoch": 2.948678071539658,
      "grad_norm": 1.8303571939468384,
      "kl": 0.79150390625,
      "learning_rate": 2.6242905029008058e-09,
      "loss": 0.2086,
      "reward": 1.5999999940395355,
      "reward_std": 0.3354101777076721,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 2843
    },
    {
      "completion_length": 203.72500801086426,
      "epoch": 2.9497148781752203,
      "grad_norm": 0.8652533292770386,
      "kl": 1.025390625,
      "learning_rate": 2.518299279775871e-09,
      "loss": 0.1494,
      "reward": 1.675000011920929,
      "reward_std": 0.16770509257912636,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2844
    },
    {
      "completion_length": 145.2750015258789,
      "epoch": 2.9507516848107826,
      "grad_norm": 0.29992929100990295,
      "kl": 0.36328125,
      "learning_rate": 2.4144910209450977e-09,
      "loss": 0.0549,
      "reward": 1.5999999940395355,
      "reward_std": 0.16770508885383606,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2845
    },
    {
      "completion_length": 153.225004196167,
      "epoch": 2.9517884914463455,
      "grad_norm": 0.47833842039108276,
      "kl": 1.03076171875,
      "learning_rate": 2.3128658777356038e-09,
      "loss": 0.1489,
      "reward": 1.375,
      "reward_std": 0.20019174367189407,
      "rewards/accuracy_reward": 0.4250000063329935,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2846
    },
    {
      "completion_length": 143.3000030517578,
      "epoch": 2.952825298081908,
      "grad_norm": 0.35573238134384155,
      "kl": 0.379638671875,
      "learning_rate": 2.2134239982920524e-09,
      "loss": 0.0814,
      "reward": 1.8250000178813934,
      "reward_std": 0.20019173622131348,
      "rewards/accuracy_reward": 0.8750000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2847
    },
    {
      "completion_length": 219.35000610351562,
      "epoch": 2.9538621047174702,
      "grad_norm": 4.35397481918335,
      "kl": 2.2724609375,
      "learning_rate": 2.116165527576819e-09,
      "loss": 0.2353,
      "reward": 1.6500000059604645,
      "reward_std": 0.24873404949903488,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.925000011920929,
      "step": 2848
    },
    {
      "completion_length": 138.05000686645508,
      "epoch": 2.9548989113530326,
      "grad_norm": 0.15453967452049255,
      "kl": 0.4384765625,
      "learning_rate": 2.021090607368825e-09,
      "loss": 0.0691,
      "reward": 1.3250000178813934,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.3500000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2849
    },
    {
      "completion_length": 140.0,
      "epoch": 2.955935717988595,
      "grad_norm": 0.2405967265367508,
      "kl": 0.451171875,
      "learning_rate": 1.928199376264206e-09,
      "loss": 0.0679,
      "reward": 1.8250000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2850
    },
    {
      "completion_length": 81.25000190734863,
      "epoch": 2.956972524624158,
      "grad_norm": 0.557878851890564,
      "kl": 0.178955078125,
      "learning_rate": 1.837491969675975e-09,
      "loss": -0.0163,
      "reward": 1.6500000059604645,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 1.0,
      "step": 2851
    },
    {
      "completion_length": 195.42500686645508,
      "epoch": 2.95800933125972,
      "grad_norm": 1.190239429473877,
      "kl": 0.794677734375,
      "learning_rate": 1.7489685198333606e-09,
      "loss": 0.0728,
      "reward": 1.4000000059604645,
      "reward_std": 0.337122417986393,
      "rewards/accuracy_reward": 0.4750000163912773,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2852
    },
    {
      "completion_length": 305.4500045776367,
      "epoch": 2.9590461378952826,
      "grad_norm": 0.31447848677635193,
      "kl": 2.4403076171875,
      "learning_rate": 1.662629155782136e-09,
      "loss": 0.2207,
      "reward": 1.7500000298023224,
      "reward_std": 0.23617041483521461,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2853
    },
    {
      "completion_length": 310.62500381469727,
      "epoch": 2.960082944530845,
      "grad_norm": 1.2219738960266113,
      "kl": 2.1181640625,
      "learning_rate": 1.5784740033839562e-09,
      "loss": 0.4002,
      "reward": 1.5250000059604645,
      "reward_std": 0.38409099727869034,
      "rewards/accuracy_reward": 0.6500000059604645,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2854
    },
    {
      "completion_length": 240.79999923706055,
      "epoch": 2.9611197511664074,
      "grad_norm": 0.5787314176559448,
      "kl": 0.685546875,
      "learning_rate": 1.4965031853166888e-09,
      "loss": 0.1793,
      "reward": 1.6500000059604645,
      "reward_std": 0.18026871979236603,
      "rewards/accuracy_reward": 0.7250000014901161,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2855
    },
    {
      "completion_length": 163.94999885559082,
      "epoch": 2.96215655780197,
      "grad_norm": 0.35660290718078613,
      "kl": 0.41552734375,
      "learning_rate": 1.4167168210739157e-09,
      "loss": 0.0947,
      "reward": 1.800000011920929,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9500000029802322,
      "step": 2856
    },
    {
      "completion_length": 248.4250087738037,
      "epoch": 2.9631933644375326,
      "grad_norm": 1.139894723892212,
      "kl": 2.2462158203125,
      "learning_rate": 1.3391150269645991e-09,
      "loss": 0.3164,
      "reward": 1.5250000357627869,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.6000000108033419,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2857
    },
    {
      "completion_length": 233.9500026702881,
      "epoch": 2.964230171073095,
      "grad_norm": 0.6167012453079224,
      "kl": 0.506103515625,
      "learning_rate": 1.2636979161134155e-09,
      "loss": 0.2058,
      "reward": 1.600000023841858,
      "reward_std": 0.34075290337204933,
      "rewards/accuracy_reward": 0.6749999970197678,
      "rewards/format_reward": 0.925000011920929,
      "step": 2858
    },
    {
      "completion_length": 158.32500648498535,
      "epoch": 2.9652669777086573,
      "grad_norm": 0.6147255897521973,
      "kl": 1.21533203125,
      "learning_rate": 1.1904655984599222e-09,
      "loss": 0.1776,
      "reward": 1.5250000357627869,
      "reward_std": 0.22360679507255554,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2859
    },
    {
      "completion_length": 146.89999961853027,
      "epoch": 2.9663037843442197,
      "grad_norm": 0.6912396550178528,
      "kl": 0.95849609375,
      "learning_rate": 1.1194181807593907e-09,
      "loss": 0.1554,
      "reward": 1.7000000178813934,
      "reward_std": 0.22360679134726524,
      "rewards/accuracy_reward": 0.7250000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2860
    },
    {
      "completion_length": 117.72500419616699,
      "epoch": 2.9673405909797825,
      "grad_norm": 0.09659355133771896,
      "kl": 0.303955078125,
      "learning_rate": 1.05055576658164e-09,
      "loss": 0.0548,
      "reward": 1.4750000089406967,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2861
    },
    {
      "completion_length": 96.67500114440918,
      "epoch": 2.9683773976153445,
      "grad_norm": 1.5440837144851685,
      "kl": 1.53271484375,
      "learning_rate": 9.838784563110382e-10,
      "loss": 0.1419,
      "reward": 1.8499999940395355,
      "reward_std": 0.05590169504284859,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2862
    },
    {
      "completion_length": 292.34999656677246,
      "epoch": 2.9694142042509073,
      "grad_norm": 2.0172176361083984,
      "kl": 2.24755859375,
      "learning_rate": 9.193863471471664e-10,
      "loss": 0.3591,
      "reward": 1.425000011920929,
      "reward_std": 0.3913118839263916,
      "rewards/accuracy_reward": 0.5250000059604645,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2863
    },
    {
      "completion_length": 253.80000495910645,
      "epoch": 2.9704510108864697,
      "grad_norm": 1.3698315620422363,
      "kl": 0.389892578125,
      "learning_rate": 8.570795331038217e-10,
      "loss": 0.1943,
      "reward": 1.850000023841858,
      "reward_std": 0.3354102112352848,
      "rewards/accuracy_reward": 0.9500000178813934,
      "rewards/format_reward": 0.9000000208616257,
      "step": 2864
    },
    {
      "completion_length": 242.725004196167,
      "epoch": 2.971487817522032,
      "grad_norm": 0.8438093662261963,
      "kl": 0.37890625,
      "learning_rate": 7.969581050090158e-10,
      "loss": 0.1864,
      "reward": 1.5999999940395355,
      "reward_std": 0.3354101777076721,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.925000011920929,
      "step": 2865
    },
    {
      "completion_length": 85.62500190734863,
      "epoch": 2.972524624157595,
      "grad_norm": 0.2306171953678131,
      "kl": 0.1533203125,
      "learning_rate": 7.390221505053086e-10,
      "loss": 0.0031,
      "reward": 1.7750000059604645,
      "reward_std": 0.055901702493429184,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 1.0,
      "step": 2866
    },
    {
      "completion_length": 353.65000343322754,
      "epoch": 2.973561430793157,
      "grad_norm": 2.580122470855713,
      "kl": 1.95556640625,
      "learning_rate": 6.832717540493083e-10,
      "loss": 0.2743,
      "reward": 1.5250000059604645,
      "reward_std": 0.640045914798975,
      "rewards/accuracy_reward": 0.6750000044703484,
      "rewards/format_reward": 0.8500000089406967,
      "step": 2867
    },
    {
      "completion_length": 181.02499771118164,
      "epoch": 2.9745982374287196,
      "grad_norm": 2.9224021434783936,
      "kl": 1.35205078125,
      "learning_rate": 6.297069969113389e-10,
      "loss": 0.1824,
      "reward": 1.5750000178813934,
      "reward_std": 0.20019175112247467,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2868
    },
    {
      "completion_length": 241.10000228881836,
      "epoch": 2.975635044064282,
      "grad_norm": 0.7921134829521179,
      "kl": 1.12548828125,
      "learning_rate": 5.783279571759392e-10,
      "loss": 0.2041,
      "reward": 1.6750000417232513,
      "reward_std": 0.31719937175512314,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 2869
    },
    {
      "completion_length": 311.5249996185303,
      "epoch": 2.9766718506998444,
      "grad_norm": 4.77922248840332,
      "kl": 2.12890625,
      "learning_rate": 5.291347097411969e-10,
      "loss": 0.4039,
      "reward": 1.3250000178813934,
      "reward_std": 0.5841442309319973,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 0.8250000178813934,
      "step": 2870
    },
    {
      "completion_length": 340.5749969482422,
      "epoch": 2.977708657335407,
      "grad_norm": 1.384347915649414,
      "kl": 1.21826171875,
      "learning_rate": 4.821273263189152e-10,
      "loss": 0.3328,
      "reward": 1.5,
      "reward_std": 0.5084579922258854,
      "rewards/accuracy_reward": 0.6250000149011612,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2871
    },
    {
      "completion_length": 243.92500495910645,
      "epoch": 2.978745463970969,
      "grad_norm": 0.6709572076797485,
      "kl": 1.2880859375,
      "learning_rate": 4.3730587543461263e-10,
      "loss": 0.2284,
      "reward": 1.5000000447034836,
      "reward_std": 0.2795085050165653,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2872
    },
    {
      "completion_length": 156.45000457763672,
      "epoch": 2.979782270606532,
      "grad_norm": 0.14977742731571198,
      "kl": 0.394775390625,
      "learning_rate": 3.9467042242685714e-10,
      "loss": 0.0536,
      "reward": 1.4000000059604645,
      "reward_std": 0.10458250343799591,
      "rewards/accuracy_reward": 0.42500000074505806,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2873
    },
    {
      "completion_length": 243.6250057220459,
      "epoch": 2.9808190772420944,
      "grad_norm": 0.6860036253929138,
      "kl": 0.5439453125,
      "learning_rate": 3.5422102944776547e-10,
      "loss": 0.1678,
      "reward": 1.6250000298023224,
      "reward_std": 0.3913118951022625,
      "rewards/accuracy_reward": 0.7000000104308128,
      "rewards/format_reward": 0.925000011920929,
      "step": 2874
    },
    {
      "completion_length": 242.75,
      "epoch": 2.9818558838776568,
      "grad_norm": 0.25144824385643005,
      "kl": 0.4071044921875,
      "learning_rate": 3.159577554628368e-10,
      "loss": 0.1496,
      "reward": 1.2500000298023224,
      "reward_std": 0.30463573709130287,
      "rewards/accuracy_reward": 0.3250000048428774,
      "rewards/format_reward": 0.925000011920929,
      "step": 2875
    },
    {
      "completion_length": 195.8249969482422,
      "epoch": 2.982892690513219,
      "grad_norm": 1.252168893814087,
      "kl": 1.340576171875,
      "learning_rate": 2.79880656250453e-10,
      "loss": 0.2097,
      "reward": 1.550000011920929,
      "reward_std": 0.4290027692914009,
      "rewards/accuracy_reward": 0.6250000074505806,
      "rewards/format_reward": 0.925000011920929,
      "step": 2876
    },
    {
      "completion_length": 193.0500030517578,
      "epoch": 2.9839294971487815,
      "grad_norm": 2.5504438877105713,
      "kl": 1.7529296875,
      "learning_rate": 2.459897844023784e-10,
      "loss": 0.2422,
      "reward": 1.6750000417232513,
      "reward_std": 0.4525563046336174,
      "rewards/accuracy_reward": 0.7500000074505806,
      "rewards/format_reward": 0.9250000268220901,
      "step": 2877
    },
    {
      "completion_length": 288.375,
      "epoch": 2.9849663037843444,
      "grad_norm": 2.51222825050354,
      "kl": 0.78076171875,
      "learning_rate": 2.1428518932342656e-10,
      "loss": 0.2922,
      "reward": 1.6500000357627869,
      "reward_std": 0.4472136124968529,
      "rewards/accuracy_reward": 0.7750000134110451,
      "rewards/format_reward": 0.8750000149011612,
      "step": 2878
    },
    {
      "completion_length": 185.90000343322754,
      "epoch": 2.9860031104199067,
      "grad_norm": 0.34298786520957947,
      "kl": 1.3505859375,
      "learning_rate": 1.8476691723096074e-10,
      "loss": 0.0755,
      "reward": 1.5,
      "reward_std": 0.22360679879784584,
      "rewards/accuracy_reward": 0.5750000104308128,
      "rewards/format_reward": 0.9249999970197678,
      "step": 2879
    },
    {
      "completion_length": 101.65000343322754,
      "epoch": 2.987039917055469,
      "grad_norm": 0.3778788447380066,
      "kl": 0.203857421875,
      "learning_rate": 1.574350111553935e-10,
      "loss": 0.0306,
      "reward": 1.7750000059604645,
      "reward_std": 0.12436702474951744,
      "rewards/accuracy_reward": 0.7750000059604645,
      "rewards/format_reward": 1.0,
      "step": 2880
    },
    {
      "completion_length": 290.25000953674316,
      "epoch": 2.9880767236910315,
      "grad_norm": 0.4994600713253021,
      "kl": 0.572265625,
      "learning_rate": 1.322895109401867e-10,
      "loss": 0.1779,
      "reward": 1.75,
      "reward_std": 0.4290027432143688,
      "rewards/accuracy_reward": 0.8500000089406967,
      "rewards/format_reward": 0.9000000059604645,
      "step": 2881
    },
    {
      "completion_length": 96.65000343322754,
      "epoch": 2.989113530326594,
      "grad_norm": 1.489416241645813,
      "kl": 0.731689453125,
      "learning_rate": 1.0933045324135194e-10,
      "loss": 0.0763,
      "reward": 1.6500000059604645,
      "reward_std": 0.16770510002970695,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2882
    },
    {
      "completion_length": 157.79999923706055,
      "epoch": 2.9901503369621567,
      "grad_norm": 0.1294565200805664,
      "kl": 0.175537109375,
      "learning_rate": 8.855787152745043e-11,
      "loss": 0.0595,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6000000014901161,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2883
    },
    {
      "completion_length": 199.10000228881836,
      "epoch": 2.991187143597719,
      "grad_norm": 1.046727180480957,
      "kl": 1.052734375,
      "learning_rate": 6.997179608009274e-11,
      "loss": 0.1919,
      "reward": 1.600000023841858,
      "reward_std": 0.3354101851582527,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 0.925000011920929,
      "step": 2884
    },
    {
      "completion_length": 236.225004196167,
      "epoch": 2.9922239502332815,
      "grad_norm": 2.2165963649749756,
      "kl": 1.4754638671875,
      "learning_rate": 5.3572253992939527e-11,
      "loss": 0.2091,
      "reward": 1.850000023841858,
      "reward_std": 0.2795084975659847,
      "rewards/accuracy_reward": 0.9250000268220901,
      "rewards/format_reward": 0.925000011920929,
      "step": 2885
    },
    {
      "completion_length": 166.1750030517578,
      "epoch": 2.993260756868844,
      "grad_norm": 0.5232186317443848,
      "kl": 0.455078125,
      "learning_rate": 3.935926917286725e-11,
      "loss": 0.0581,
      "reward": 1.8499999940395355,
      "reward_std": 0.3354102037847042,
      "rewards/accuracy_reward": 0.8750000298023224,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2886
    },
    {
      "completion_length": 149.10000610351562,
      "epoch": 2.9942975635044062,
      "grad_norm": 0.16788578033447266,
      "kl": 0.578125,
      "learning_rate": 2.73328623388025e-11,
      "loss": 0.0915,
      "reward": 1.5750000178813934,
      "reward_std": 0.11180340498685837,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2887
    },
    {
      "completion_length": 256.15000915527344,
      "epoch": 2.995334370139969,
      "grad_norm": 0.9628523588180542,
      "kl": 2.640625,
      "learning_rate": 1.749305102222154e-11,
      "loss": 0.3374,
      "reward": 1.3750000298023224,
      "reward_std": 0.43999267742037773,
      "rewards/accuracy_reward": 0.5000000111758709,
      "rewards/format_reward": 0.8750000298023224,
      "step": 2888
    },
    {
      "completion_length": 149.35000038146973,
      "epoch": 2.9963711767755314,
      "grad_norm": 0.49316930770874023,
      "kl": 0.220458984375,
      "learning_rate": 9.839849567483406e-12,
      "loss": 0.0361,
      "reward": 1.574999988079071,
      "reward_std": 0.22360680624842644,
      "rewards/accuracy_reward": 0.6000000089406967,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2889
    },
    {
      "completion_length": 103.15000343322754,
      "epoch": 2.997407983411094,
      "grad_norm": 0.04041944816708565,
      "kl": 0.1763916015625,
      "learning_rate": 4.373269130664159e-12,
      "loss": 0.0067,
      "reward": 1.875,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 1.0,
      "step": 2890
    },
    {
      "completion_length": 143.90000534057617,
      "epoch": 2.998444790046656,
      "grad_norm": 0.9387772679328918,
      "kl": 1.1044921875,
      "learning_rate": 1.0933176810556945e-12,
      "loss": 0.1227,
      "reward": 1.800000011920929,
      "reward_std": 0.11180339008569717,
      "rewards/accuracy_reward": 0.8250000029802322,
      "rewards/format_reward": 0.9750000089406967,
      "step": 2891
    },
    {
      "completion_length": 202.0750026702881,
      "epoch": 2.9994815966822186,
      "grad_norm": 0.4531560242176056,
      "kl": 0.42236328125,
      "learning_rate": 0.0,
      "loss": 0.1038,
      "reward": 1.6500000357627869,
      "reward_std": 0.292072094976902,
      "rewards/accuracy_reward": 0.7000000178813934,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2892
    }
  ],
  "logging_steps": 1,
  "max_steps": 2892,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "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": 2,
  "trial_name": null,
  "trial_params": null
}