{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 100,
  "global_step": 2500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 427.15834655761716,
      "epoch": 0.002,
      "grad_norm": 1.8633235523265679,
      "kl": 0.0001178741455078125,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.0,
      "reward": 0.6833333551883698,
      "reward_std": 0.4267221510410309,
      "rewards/accuracy_reward": 0.1875000052154064,
      "rewards/format_reward": 0.49583334624767306,
      "step": 5
    },
    {
      "completion_length": 408.5041732788086,
      "epoch": 0.004,
      "grad_norm": 1.2450032484475768,
      "kl": 0.00022439956665039064,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.0,
      "reward": 0.6250000119209289,
      "reward_std": 0.452184134721756,
      "rewards/accuracy_reward": 0.15416667014360427,
      "rewards/format_reward": 0.4708333447575569,
      "step": 10
    },
    {
      "completion_length": 386.3708465576172,
      "epoch": 0.006,
      "grad_norm": 1.5807375329115385,
      "kl": 0.0007476806640625,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.0,
      "reward": 0.7041666805744171,
      "reward_std": 0.4529884338378906,
      "rewards/accuracy_reward": 0.20833333991467953,
      "rewards/format_reward": 0.4958333492279053,
      "step": 15
    },
    {
      "completion_length": 321.80001068115234,
      "epoch": 0.008,
      "grad_norm": 2.3656753548221094,
      "kl": 0.01384429931640625,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.0006,
      "reward": 0.8583333551883697,
      "reward_std": 0.42067554742097857,
      "rewards/accuracy_reward": 0.1666666731238365,
      "rewards/format_reward": 0.6916666835546493,
      "step": 20
    },
    {
      "completion_length": 283.2541748046875,
      "epoch": 0.01,
      "grad_norm": 1.752088104129668,
      "kl": 0.03404541015625,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0014,
      "reward": 0.9416666984558105,
      "reward_std": 0.2924864612519741,
      "rewards/accuracy_reward": 0.07916666977107525,
      "rewards/format_reward": 0.8625000238418579,
      "step": 25
    },
    {
      "completion_length": 231.33334045410157,
      "epoch": 0.012,
      "grad_norm": 2.4751987426338737,
      "kl": 0.05478515625,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.0022,
      "reward": 0.9500000298023223,
      "reward_std": 0.23476256504654885,
      "rewards/accuracy_reward": 0.05000000074505806,
      "rewards/format_reward": 0.900000023841858,
      "step": 30
    },
    {
      "completion_length": 227.18750915527343,
      "epoch": 0.014,
      "grad_norm": 1.0730481146387303,
      "kl": 0.041900634765625,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0017,
      "reward": 0.979166692495346,
      "reward_std": 0.2765915520489216,
      "rewards/accuracy_reward": 0.07916666977107525,
      "rewards/format_reward": 0.9000000298023224,
      "step": 35
    },
    {
      "completion_length": 236.82500762939452,
      "epoch": 0.016,
      "grad_norm": 2.287053976433467,
      "kl": 0.028759765625,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0011,
      "reward": 1.0333333671092988,
      "reward_std": 0.30562743321061137,
      "rewards/accuracy_reward": 0.12500000409781933,
      "rewards/format_reward": 0.9083333551883698,
      "step": 40
    },
    {
      "completion_length": 251.8541717529297,
      "epoch": 0.018,
      "grad_norm": 2.002581508005556,
      "kl": 0.02606201171875,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.001,
      "reward": 0.9375000238418579,
      "reward_std": 0.33919562846422197,
      "rewards/accuracy_reward": 0.10000000260770321,
      "rewards/format_reward": 0.8375000178813934,
      "step": 45
    },
    {
      "completion_length": 211.71667556762696,
      "epoch": 0.02,
      "grad_norm": 1.260198604571592,
      "kl": 0.025537109375,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.001,
      "reward": 0.9625000476837158,
      "reward_std": 0.29876374155282975,
      "rewards/accuracy_reward": 0.08333333544433116,
      "rewards/format_reward": 0.8791666924953461,
      "step": 50
    },
    {
      "completion_length": 164.14583740234374,
      "epoch": 0.022,
      "grad_norm": 0.8963074266016013,
      "kl": 0.04442138671875,
      "learning_rate": 4.4e-06,
      "loss": 0.0018,
      "reward": 1.066666704416275,
      "reward_std": 0.2502697631716728,
      "rewards/accuracy_reward": 0.11666666939854622,
      "rewards/format_reward": 0.9500000178813934,
      "step": 55
    },
    {
      "completion_length": 183.88750686645508,
      "epoch": 0.024,
      "grad_norm": 1.2329398710686466,
      "kl": 0.064013671875,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0026,
      "reward": 1.025000023841858,
      "reward_std": 0.23044276759028434,
      "rewards/accuracy_reward": 0.07916666902601718,
      "rewards/format_reward": 0.9458333492279053,
      "step": 60
    },
    {
      "completion_length": 248.5875045776367,
      "epoch": 0.026,
      "grad_norm": 0.8205760814794363,
      "kl": 0.03978271484375,
      "learning_rate": 5.2e-06,
      "loss": 0.0016,
      "reward": 1.0583333551883698,
      "reward_std": 0.20980931594967842,
      "rewards/accuracy_reward": 0.10000000335276127,
      "rewards/format_reward": 0.9583333551883697,
      "step": 65
    },
    {
      "completion_length": 217.50833740234376,
      "epoch": 0.028,
      "grad_norm": 1.1658067578625744,
      "kl": 0.0590087890625,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.0024,
      "reward": 1.1916666984558106,
      "reward_std": 0.2719326362013817,
      "rewards/accuracy_reward": 0.2041666742414236,
      "rewards/format_reward": 0.987500011920929,
      "step": 70
    },
    {
      "completion_length": 207.8291748046875,
      "epoch": 0.03,
      "grad_norm": 1.496731257519424,
      "kl": 0.0521484375,
      "learning_rate": 6e-06,
      "loss": 0.0021,
      "reward": 1.0708333730697632,
      "reward_std": 0.11254597306251526,
      "rewards/accuracy_reward": 0.07916666977107525,
      "rewards/format_reward": 0.9916666686534882,
      "step": 75
    },
    {
      "completion_length": 199.8000045776367,
      "epoch": 0.032,
      "grad_norm": 1.8251055882491147,
      "kl": 0.065869140625,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.0026,
      "reward": 1.0916667103767395,
      "reward_std": 0.182654220610857,
      "rewards/accuracy_reward": 0.10416666865348816,
      "rewards/format_reward": 0.987500011920929,
      "step": 80
    },
    {
      "completion_length": 250.7291702270508,
      "epoch": 0.034,
      "grad_norm": 1.1675575803274054,
      "kl": 0.196044921875,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.0078,
      "reward": 1.2291666984558105,
      "reward_std": 0.2886375203728676,
      "rewards/accuracy_reward": 0.23750000447034836,
      "rewards/format_reward": 0.9916666746139526,
      "step": 85
    },
    {
      "completion_length": 375.57500610351565,
      "epoch": 0.036,
      "grad_norm": 0.6674123847300312,
      "kl": 0.06044921875,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.0024,
      "reward": 1.062500035762787,
      "reward_std": 0.30103383138775824,
      "rewards/accuracy_reward": 0.13750000670552254,
      "rewards/format_reward": 0.9250000238418579,
      "step": 90
    },
    {
      "completion_length": 417.9250091552734,
      "epoch": 0.038,
      "grad_norm": 0.9023928645148662,
      "kl": 0.180029296875,
      "learning_rate": 7.600000000000001e-06,
      "loss": 0.0072,
      "reward": 1.1375000417232513,
      "reward_std": 0.3338733673095703,
      "rewards/accuracy_reward": 0.18750000782310963,
      "rewards/format_reward": 0.9500000238418579,
      "step": 95
    },
    {
      "completion_length": 485.583349609375,
      "epoch": 0.04,
      "grad_norm": 0.9766684287780194,
      "kl": 0.089501953125,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0036,
      "reward": 1.1083333611488342,
      "reward_std": 0.2985624983906746,
      "rewards/accuracy_reward": 0.17500000558793544,
      "rewards/format_reward": 0.9333333492279052,
      "step": 100
    },
    {
      "epoch": 0.04,
      "eval_completion_length": 485.7147545447716,
      "eval_kl": 0.0946514423076923,
      "eval_loss": 0.0033234155271202326,
      "eval_reward": 1.195512863305899,
      "eval_reward_std": 0.3841831271465008,
      "eval_rewards/accuracy_reward": 0.278846161869856,
      "eval_rewards/format_reward": 0.916666681949909,
      "eval_runtime": 228.106,
      "eval_samples_per_second": 0.434,
      "eval_steps_per_second": 0.013,
      "step": 100
    },
    {
      "completion_length": 545.5041870117187,
      "epoch": 0.042,
      "grad_norm": 0.7346390959845385,
      "kl": 0.0919921875,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.0037,
      "reward": 1.0458333671092988,
      "reward_std": 0.2975893825292587,
      "rewards/accuracy_reward": 0.1375000048428774,
      "rewards/format_reward": 0.9083333611488342,
      "step": 105
    },
    {
      "completion_length": 442.87501525878906,
      "epoch": 0.044,
      "grad_norm": 0.9100982084517627,
      "kl": 0.10302734375,
      "learning_rate": 8.8e-06,
      "loss": 0.0041,
      "reward": 1.1416666865348817,
      "reward_std": 0.3882675111293793,
      "rewards/accuracy_reward": 0.2583333428949118,
      "rewards/format_reward": 0.8833333551883698,
      "step": 110
    },
    {
      "completion_length": 436.5208404541016,
      "epoch": 0.046,
      "grad_norm": 1.0013758124491203,
      "kl": 0.079736328125,
      "learning_rate": 9.200000000000002e-06,
      "loss": 0.0032,
      "reward": 0.9875000238418579,
      "reward_std": 0.3188182398676872,
      "rewards/accuracy_reward": 0.09166666977107525,
      "rewards/format_reward": 0.8958333551883697,
      "step": 115
    },
    {
      "completion_length": 382.6458435058594,
      "epoch": 0.048,
      "grad_norm": 0.8780597651919092,
      "kl": 0.0803466796875,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.0032,
      "reward": 1.0708333730697632,
      "reward_std": 0.2953009992837906,
      "rewards/accuracy_reward": 0.16250000074505805,
      "rewards/format_reward": 0.9083333492279053,
      "step": 120
    },
    {
      "completion_length": 390.9791778564453,
      "epoch": 0.05,
      "grad_norm": 0.9115434520496168,
      "kl": 0.095361328125,
      "learning_rate": 1e-05,
      "loss": 0.0038,
      "reward": 1.1208333730697633,
      "reward_std": 0.38652298897504805,
      "rewards/accuracy_reward": 0.22500000819563865,
      "rewards/format_reward": 0.8958333492279053,
      "step": 125
    },
    {
      "completion_length": 384.08750915527344,
      "epoch": 0.052,
      "grad_norm": 1.097771502265398,
      "kl": 0.194140625,
      "learning_rate": 1.04e-05,
      "loss": 0.0078,
      "reward": 1.2000000536441804,
      "reward_std": 0.3203377678990364,
      "rewards/accuracy_reward": 0.24166667386889457,
      "rewards/format_reward": 0.9583333492279053,
      "step": 130
    },
    {
      "completion_length": 339.4041778564453,
      "epoch": 0.054,
      "grad_norm": 5.345967756487471,
      "kl": 0.180859375,
      "learning_rate": 1.0800000000000002e-05,
      "loss": 0.0072,
      "reward": 1.2583333790302276,
      "reward_std": 0.3293462932109833,
      "rewards/accuracy_reward": 0.3416666749864817,
      "rewards/format_reward": 0.9166666865348816,
      "step": 135
    },
    {
      "completion_length": 405.9458465576172,
      "epoch": 0.056,
      "grad_norm": 110.13539552794225,
      "kl": 0.408203125,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.0163,
      "reward": 1.075000023841858,
      "reward_std": 0.40778897404670716,
      "rewards/accuracy_reward": 0.21666667386889457,
      "rewards/format_reward": 0.8583333551883697,
      "step": 140
    },
    {
      "completion_length": 428.1416717529297,
      "epoch": 0.058,
      "grad_norm": 127.4037667780564,
      "kl": 0.7431640625,
      "learning_rate": 1.16e-05,
      "loss": 0.0297,
      "reward": 1.0958333790302277,
      "reward_std": 0.4978458255529404,
      "rewards/accuracy_reward": 0.2541666742414236,
      "rewards/format_reward": 0.8416666924953461,
      "step": 145
    },
    {
      "completion_length": 346.7416748046875,
      "epoch": 0.06,
      "grad_norm": 1.024737259335268,
      "kl": 0.44423828125,
      "learning_rate": 1.2e-05,
      "loss": 0.0178,
      "reward": 1.1333333671092987,
      "reward_std": 0.3713921308517456,
      "rewards/accuracy_reward": 0.2583333425223827,
      "rewards/format_reward": 0.8750000178813935,
      "step": 150
    },
    {
      "completion_length": 349.40001373291017,
      "epoch": 0.062,
      "grad_norm": 1.1085565147403205,
      "kl": 0.1763671875,
      "learning_rate": 1.2400000000000002e-05,
      "loss": 0.0071,
      "reward": 1.191666704416275,
      "reward_std": 0.4108735501766205,
      "rewards/accuracy_reward": 0.27916667312383653,
      "rewards/format_reward": 0.9125000238418579,
      "step": 155
    },
    {
      "completion_length": 410.98334350585935,
      "epoch": 0.064,
      "grad_norm": 0.9679449380134207,
      "kl": 0.24033203125,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.0096,
      "reward": 1.0500000238418579,
      "reward_std": 0.44029667377471926,
      "rewards/accuracy_reward": 0.18333333656191825,
      "rewards/format_reward": 0.8666666865348815,
      "step": 160
    },
    {
      "completion_length": 440.6083465576172,
      "epoch": 0.066,
      "grad_norm": 3.6330210669194543,
      "kl": 0.3822265625,
      "learning_rate": 1.3200000000000002e-05,
      "loss": 0.0153,
      "reward": 1.337500023841858,
      "reward_std": 0.45285614132881163,
      "rewards/accuracy_reward": 0.43333334624767306,
      "rewards/format_reward": 0.9041666924953461,
      "step": 165
    },
    {
      "completion_length": 371.08750915527344,
      "epoch": 0.068,
      "grad_norm": 1.4615175546831627,
      "kl": 16.89931640625,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.6758,
      "reward": 0.9250000417232513,
      "reward_std": 0.5015424624085426,
      "rewards/accuracy_reward": 0.20000000670552254,
      "rewards/format_reward": 0.7250000208616256,
      "step": 170
    },
    {
      "completion_length": 339.0166717529297,
      "epoch": 0.07,
      "grad_norm": 2.9473137749285523,
      "kl": 0.35244140625,
      "learning_rate": 1.4e-05,
      "loss": 0.0141,
      "reward": 1.154166692495346,
      "reward_std": 0.4933777332305908,
      "rewards/accuracy_reward": 0.30416667759418486,
      "rewards/format_reward": 0.8500000178813935,
      "step": 175
    },
    {
      "completion_length": 992.5208557128906,
      "epoch": 0.072,
      "grad_norm": 4039.653812849788,
      "kl": 619.38203125,
      "learning_rate": 1.4400000000000001e-05,
      "loss": 24.8631,
      "reward": 0.5500000134110451,
      "reward_std": 0.49858903884887695,
      "rewards/accuracy_reward": 0.22083333805203437,
      "rewards/format_reward": 0.32916667461395266,
      "step": 180
    },
    {
      "completion_length": 1968.1625366210938,
      "epoch": 0.074,
      "grad_norm": 19.777563892939803,
      "kl": 1548.0078125,
      "learning_rate": 1.48e-05,
      "loss": 61.9522,
      "reward": 0.10416666977107525,
      "reward_std": 0.18909453824162484,
      "rewards/accuracy_reward": 0.08333333544433116,
      "rewards/format_reward": 0.02083333395421505,
      "step": 185
    },
    {
      "completion_length": 1960.37919921875,
      "epoch": 0.076,
      "grad_norm": 1.388860761683851,
      "kl": 1.40361328125,
      "learning_rate": 1.5200000000000002e-05,
      "loss": 0.0562,
      "reward": 0.08333333730697631,
      "reward_std": 0.14930254891514777,
      "rewards/accuracy_reward": 0.05000000074505806,
      "rewards/format_reward": 0.03333333432674408,
      "step": 190
    },
    {
      "completion_length": 1365.8667114257812,
      "epoch": 0.078,
      "grad_norm": 3.225794828015121,
      "kl": 1.9984375,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 0.0799,
      "reward": 0.21666667200624942,
      "reward_std": 0.32874541357159615,
      "rewards/accuracy_reward": 0.02500000037252903,
      "rewards/format_reward": 0.19166667200624943,
      "step": 195
    },
    {
      "completion_length": 360.0791717529297,
      "epoch": 0.08,
      "grad_norm": 1.8347244902773923,
      "kl": 23.698828125,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.9454,
      "reward": 0.7083333611488343,
      "reward_std": 0.5879518806934356,
      "rewards/accuracy_reward": 0.10000000186264515,
      "rewards/format_reward": 0.6083333551883697,
      "step": 200
    },
    {
      "epoch": 0.08,
      "eval_completion_length": 283.7147533710186,
      "eval_kl": 4.560396634615385,
      "eval_loss": 0.19470153748989105,
      "eval_reward": 0.9615384982182429,
      "eval_reward_std": 0.4933858124109415,
      "eval_rewards/accuracy_reward": 0.17147436222204795,
      "eval_rewards/format_reward": 0.7900641147906964,
      "eval_runtime": 174.8426,
      "eval_samples_per_second": 0.566,
      "eval_steps_per_second": 0.017,
      "step": 200
    },
    {
      "completion_length": 356.27500610351564,
      "epoch": 0.082,
      "grad_norm": 1.2924069593835132,
      "kl": 417792.38505859376,
      "learning_rate": 1.64e-05,
      "loss": 16711.4297,
      "reward": 0.9083333611488342,
      "reward_std": 0.38548276126384734,
      "rewards/accuracy_reward": 0.09166666939854622,
      "rewards/format_reward": 0.8166666805744172,
      "step": 205
    },
    {
      "completion_length": 590.3875183105469,
      "epoch": 0.084,
      "grad_norm": 3.8553857289399875,
      "kl": 440014.5115234375,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 17588.25,
      "reward": 0.9541666984558106,
      "reward_std": 0.4752687841653824,
      "rewards/accuracy_reward": 0.19166667051613331,
      "rewards/format_reward": 0.7625000059604645,
      "step": 210
    },
    {
      "completion_length": 562.529183959961,
      "epoch": 0.086,
      "grad_norm": 1118.0087943573474,
      "kl": 25.59375,
      "learning_rate": 1.72e-05,
      "loss": 1.0267,
      "reward": 1.0333333730697631,
      "reward_std": 0.5306622147560119,
      "rewards/accuracy_reward": 0.2458333384245634,
      "rewards/format_reward": 0.7875000238418579,
      "step": 215
    },
    {
      "completion_length": 472.8041748046875,
      "epoch": 0.088,
      "grad_norm": 0.7246088323208753,
      "kl": 0.5654296875,
      "learning_rate": 1.76e-05,
      "loss": 0.0226,
      "reward": 1.0250000357627869,
      "reward_std": 0.3851146653294563,
      "rewards/accuracy_reward": 0.16250000521540642,
      "rewards/format_reward": 0.8625000238418579,
      "step": 220
    },
    {
      "completion_length": 367.68333892822267,
      "epoch": 0.09,
      "grad_norm": 0.8354974174200616,
      "kl": 0.2619140625,
      "learning_rate": 1.8e-05,
      "loss": 0.0105,
      "reward": 1.212500023841858,
      "reward_std": 0.2969694033265114,
      "rewards/accuracy_reward": 0.29583334140479567,
      "rewards/format_reward": 0.9166666746139527,
      "step": 225
    },
    {
      "completion_length": 310.6916702270508,
      "epoch": 0.092,
      "grad_norm": 189.01184940389123,
      "kl": 2.4029296875,
      "learning_rate": 1.8400000000000003e-05,
      "loss": 0.096,
      "reward": 1.1291666984558106,
      "reward_std": 0.3020160228013992,
      "rewards/accuracy_reward": 0.18750000186264515,
      "rewards/format_reward": 0.9416666865348816,
      "step": 230
    },
    {
      "completion_length": 369.8333404541016,
      "epoch": 0.094,
      "grad_norm": 37.154199182097,
      "kl": 1.755859375,
      "learning_rate": 1.88e-05,
      "loss": 0.0702,
      "reward": 0.9958333671092987,
      "reward_std": 0.33207904547452927,
      "rewards/accuracy_reward": 0.15000000186264514,
      "rewards/format_reward": 0.8458333551883698,
      "step": 235
    },
    {
      "completion_length": 461.2375091552734,
      "epoch": 0.096,
      "grad_norm": 2.2527967247045093,
      "kl": 1.1052734375,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.0442,
      "reward": 0.9791667044162751,
      "reward_std": 0.46096948683261874,
      "rewards/accuracy_reward": 0.20000000447034835,
      "rewards/format_reward": 0.7791666865348816,
      "step": 240
    },
    {
      "completion_length": 445.3166809082031,
      "epoch": 0.098,
      "grad_norm": 3.932319563336608,
      "kl": 4.178515625,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.1673,
      "reward": 1.0208333551883697,
      "reward_std": 0.45676523745059966,
      "rewards/accuracy_reward": 0.18333333656191825,
      "rewards/format_reward": 0.8375000178813934,
      "step": 245
    },
    {
      "completion_length": 363.8000091552734,
      "epoch": 0.1,
      "grad_norm": 2.0338776721959007,
      "kl": 1.2009765625,
      "learning_rate": 2e-05,
      "loss": 0.048,
      "reward": 1.0333333671092988,
      "reward_std": 0.411553718149662,
      "rewards/accuracy_reward": 0.15416667126119138,
      "rewards/format_reward": 0.8791666924953461,
      "step": 250
    },
    {
      "completion_length": 290.4333435058594,
      "epoch": 0.102,
      "grad_norm": 1.6678034955896959,
      "kl": 0.4701171875,
      "learning_rate": 1.9999756307053947e-05,
      "loss": 0.0188,
      "reward": 1.0000000417232513,
      "reward_std": 0.30890622958540914,
      "rewards/accuracy_reward": 0.1000000037252903,
      "rewards/format_reward": 0.9000000119209289,
      "step": 255
    },
    {
      "completion_length": 216.17084197998048,
      "epoch": 0.104,
      "grad_norm": 8.82960885357952,
      "kl": 0.9080078125,
      "learning_rate": 1.9999025240093045e-05,
      "loss": 0.0363,
      "reward": 0.8416666924953461,
      "reward_std": 0.47480377554893494,
      "rewards/accuracy_reward": 0.14166666977107525,
      "rewards/format_reward": 0.7000000089406967,
      "step": 260
    },
    {
      "completion_length": 222.05417022705078,
      "epoch": 0.106,
      "grad_norm": 3.461116528554778,
      "kl": 0.9873046875,
      "learning_rate": 1.9997806834748455e-05,
      "loss": 0.0395,
      "reward": 1.0333333730697631,
      "reward_std": 0.36838603541255,
      "rewards/accuracy_reward": 0.19583333618938922,
      "rewards/format_reward": 0.8375000119209289,
      "step": 265
    },
    {
      "completion_length": 217.8041732788086,
      "epoch": 0.108,
      "grad_norm": 1.086426317188572,
      "kl": 0.4533203125,
      "learning_rate": 1.9996101150403543e-05,
      "loss": 0.0181,
      "reward": 1.0750000298023223,
      "reward_std": 0.2211028292775154,
      "rewards/accuracy_reward": 0.12916667051613331,
      "rewards/format_reward": 0.9458333551883698,
      "step": 270
    },
    {
      "completion_length": 210.55833740234374,
      "epoch": 0.11,
      "grad_norm": 6.780769531887777,
      "kl": 1.0130859375,
      "learning_rate": 1.999390827019096e-05,
      "loss": 0.0406,
      "reward": 1.0833333730697632,
      "reward_std": 0.2511788114905357,
      "rewards/accuracy_reward": 0.13750000521540642,
      "rewards/format_reward": 0.9458333611488342,
      "step": 275
    },
    {
      "completion_length": 256.6375076293945,
      "epoch": 0.112,
      "grad_norm": 1.7774790738197452,
      "kl": 0.76171875,
      "learning_rate": 1.9991228300988586e-05,
      "loss": 0.0305,
      "reward": 1.1208333671092987,
      "reward_std": 0.3549239993095398,
      "rewards/accuracy_reward": 0.19583334177732467,
      "rewards/format_reward": 0.9250000238418579,
      "step": 280
    },
    {
      "completion_length": 272.54583740234375,
      "epoch": 0.114,
      "grad_norm": 14.821428692684155,
      "kl": 1.717578125,
      "learning_rate": 1.9988061373414342e-05,
      "loss": 0.0688,
      "reward": 0.9875000178813934,
      "reward_std": 0.3375275403261185,
      "rewards/accuracy_reward": 0.10416666902601719,
      "rewards/format_reward": 0.8833333492279053,
      "step": 285
    },
    {
      "completion_length": 238.64167175292968,
      "epoch": 0.116,
      "grad_norm": 53.728519178410956,
      "kl": 1.67421875,
      "learning_rate": 1.9984407641819812e-05,
      "loss": 0.0671,
      "reward": 1.0708333611488343,
      "reward_std": 0.33341423869132997,
      "rewards/accuracy_reward": 0.1750000026077032,
      "rewards/format_reward": 0.8958333492279053,
      "step": 290
    },
    {
      "completion_length": 242.93334197998047,
      "epoch": 0.118,
      "grad_norm": 2.576799775213784,
      "kl": 1.2841796875,
      "learning_rate": 1.9980267284282718e-05,
      "loss": 0.0514,
      "reward": 1.075000035762787,
      "reward_std": 0.39172317981719973,
      "rewards/accuracy_reward": 0.18750000596046448,
      "rewards/format_reward": 0.8875000178813934,
      "step": 295
    },
    {
      "completion_length": 232.62500762939453,
      "epoch": 0.12,
      "grad_norm": 3.9883395042058307,
      "kl": 9.513671875,
      "learning_rate": 1.9975640502598243e-05,
      "loss": 0.3803,
      "reward": 0.9750000417232514,
      "reward_std": 0.44072358012199403,
      "rewards/accuracy_reward": 0.19166667088866235,
      "rewards/format_reward": 0.7833333432674408,
      "step": 300
    },
    {
      "epoch": 0.12,
      "eval_completion_length": 222.86699030949518,
      "eval_kl": 1.2294170673076923,
      "eval_loss": 0.0528143048286438,
      "eval_reward": 1.0256410607924829,
      "eval_reward_std": 0.3615907682822301,
      "eval_rewards/accuracy_reward": 0.15224359270471793,
      "eval_rewards/format_reward": 0.8733974511806781,
      "eval_runtime": 198.2465,
      "eval_samples_per_second": 0.499,
      "eval_steps_per_second": 0.015,
      "step": 300
    },
    {
      "completion_length": 207.7291702270508,
      "epoch": 0.122,
      "grad_norm": 2.620279026401797,
      "kl": 2.27109375,
      "learning_rate": 1.9970527522269204e-05,
      "loss": 0.0908,
      "reward": 1.0500000238418579,
      "reward_std": 0.29396146908402443,
      "rewards/accuracy_reward": 0.1375000048428774,
      "rewards/format_reward": 0.9125000238418579,
      "step": 305
    },
    {
      "completion_length": 175.1833381652832,
      "epoch": 0.124,
      "grad_norm": 0.7567148884073016,
      "kl": 0.3833984375,
      "learning_rate": 1.9964928592495046e-05,
      "loss": 0.0153,
      "reward": 1.1708333849906922,
      "reward_std": 0.15448497384786605,
      "rewards/accuracy_reward": 0.1708333395421505,
      "rewards/format_reward": 1.0,
      "step": 310
    },
    {
      "completion_length": 187.89583892822264,
      "epoch": 0.126,
      "grad_norm": 1.1365022325057257,
      "kl": 0.334765625,
      "learning_rate": 1.9958843986159705e-05,
      "loss": 0.0134,
      "reward": 1.1583333611488342,
      "reward_std": 0.16067556515336037,
      "rewards/accuracy_reward": 0.15833333805203437,
      "rewards/format_reward": 1.0,
      "step": 315
    },
    {
      "completion_length": 158.3625015258789,
      "epoch": 0.128,
      "grad_norm": 1.0991071204841167,
      "kl": 0.416796875,
      "learning_rate": 1.9952273999818312e-05,
      "loss": 0.0167,
      "reward": 1.1583333611488342,
      "reward_std": 0.17899299263954163,
      "rewards/accuracy_reward": 0.16666667237877847,
      "rewards/format_reward": 0.9916666746139526,
      "step": 320
    },
    {
      "completion_length": 216.79584121704102,
      "epoch": 0.13,
      "grad_norm": 0.8275882765202969,
      "kl": 0.419921875,
      "learning_rate": 1.9945218953682736e-05,
      "loss": 0.0168,
      "reward": 1.0916666984558105,
      "reward_std": 0.15066706389188766,
      "rewards/accuracy_reward": 0.10000000186264515,
      "rewards/format_reward": 0.9916666746139526,
      "step": 325
    },
    {
      "completion_length": 258.4791732788086,
      "epoch": 0.132,
      "grad_norm": 1.1654941388041056,
      "kl": 0.3876953125,
      "learning_rate": 1.9937679191605964e-05,
      "loss": 0.0155,
      "reward": 1.087500023841858,
      "reward_std": 0.15936369448900223,
      "rewards/accuracy_reward": 0.10000000186264515,
      "rewards/format_reward": 0.987500011920929,
      "step": 330
    },
    {
      "completion_length": 275.38751220703125,
      "epoch": 0.134,
      "grad_norm": 21931.258042501762,
      "kl": 145.1587890625,
      "learning_rate": 1.992965508106537e-05,
      "loss": 5.8162,
      "reward": 1.054166704416275,
      "reward_std": 0.2149122439324856,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.9500000238418579,
      "step": 335
    },
    {
      "completion_length": 1149.975033569336,
      "epoch": 0.136,
      "grad_norm": 398638.02963389497,
      "kl": 9480.2162109375,
      "learning_rate": 1.9921147013144782e-05,
      "loss": 379.039,
      "reward": 0.975000011920929,
      "reward_std": 0.21539116650819778,
      "rewards/accuracy_reward": 0.04583333432674408,
      "rewards/format_reward": 0.9291666924953461,
      "step": 340
    },
    {
      "completion_length": 509.2916748046875,
      "epoch": 0.138,
      "grad_norm": 0.33160914978875666,
      "kl": 5.3359375,
      "learning_rate": 1.991215540251542e-05,
      "loss": 0.2131,
      "reward": 0.9791667103767395,
      "reward_std": 0.2079385258257389,
      "rewards/accuracy_reward": 0.03750000037252903,
      "rewards/format_reward": 0.9416666865348816,
      "step": 345
    },
    {
      "completion_length": 265.97917022705076,
      "epoch": 0.14,
      "grad_norm": 1.35804758291205,
      "kl": 0.4572265625,
      "learning_rate": 1.9902680687415704e-05,
      "loss": 0.0183,
      "reward": 0.954166692495346,
      "reward_std": 0.32328550666570666,
      "rewards/accuracy_reward": 0.08750000111758709,
      "rewards/format_reward": 0.8666666924953461,
      "step": 350
    },
    {
      "completion_length": 264.07083587646486,
      "epoch": 0.142,
      "grad_norm": 0.4918964435152788,
      "kl": 0.4173828125,
      "learning_rate": 1.9892723329629885e-05,
      "loss": 0.0167,
      "reward": 0.9791666984558105,
      "reward_std": 0.19311015233397483,
      "rewards/accuracy_reward": 0.04166666716337204,
      "rewards/format_reward": 0.9375000298023224,
      "step": 355
    },
    {
      "completion_length": 252.10834350585938,
      "epoch": 0.144,
      "grad_norm": 0.5874567666668539,
      "kl": 0.4216796875,
      "learning_rate": 1.988228381446553e-05,
      "loss": 0.0169,
      "reward": 1.0708333671092987,
      "reward_std": 0.16931553408503533,
      "rewards/accuracy_reward": 0.08750000186264514,
      "rewards/format_reward": 0.9833333373069764,
      "step": 360
    },
    {
      "completion_length": 252.58750915527344,
      "epoch": 0.146,
      "grad_norm": 0.9063762518758791,
      "kl": 0.40546875,
      "learning_rate": 1.987136265072988e-05,
      "loss": 0.0162,
      "reward": 1.0833333730697632,
      "reward_std": 0.16870678812265397,
      "rewards/accuracy_reward": 0.10833333507180214,
      "rewards/format_reward": 0.9750000178813935,
      "step": 365
    },
    {
      "completion_length": 188.80833892822267,
      "epoch": 0.148,
      "grad_norm": 0.787412634085505,
      "kl": 0.47890625,
      "learning_rate": 1.985996037070505e-05,
      "loss": 0.0192,
      "reward": 1.0583333730697633,
      "reward_std": 0.1203794963657856,
      "rewards/accuracy_reward": 0.07083333544433117,
      "rewards/format_reward": 0.9875000059604645,
      "step": 370
    },
    {
      "completion_length": 261.20001220703125,
      "epoch": 0.15,
      "grad_norm": 311.4303294674164,
      "kl": 9.298828125,
      "learning_rate": 1.9848077530122083e-05,
      "loss": 0.3741,
      "reward": 0.9875000357627869,
      "reward_std": 0.2795761555433273,
      "rewards/accuracy_reward": 0.07916666828095913,
      "rewards/format_reward": 0.9083333551883698,
      "step": 375
    },
    {
      "completion_length": 609.112515258789,
      "epoch": 0.152,
      "grad_norm": 3.6435215764773003,
      "kl": 1.7125,
      "learning_rate": 1.983571470813386e-05,
      "loss": 0.0684,
      "reward": 0.7833333611488342,
      "reward_std": 0.4051990956068039,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.7416666924953461,
      "step": 380
    },
    {
      "completion_length": 396.24168395996094,
      "epoch": 0.154,
      "grad_norm": 2.8787576101456454,
      "kl": 0.9396484375,
      "learning_rate": 1.982287250728689e-05,
      "loss": 0.0376,
      "reward": 0.9500000238418579,
      "reward_std": 0.3664922297000885,
      "rewards/accuracy_reward": 0.10416666902601719,
      "rewards/format_reward": 0.8458333492279053,
      "step": 385
    },
    {
      "completion_length": 257.5208404541016,
      "epoch": 0.156,
      "grad_norm": 0.8769009991094786,
      "kl": 0.4197265625,
      "learning_rate": 1.9809551553491918e-05,
      "loss": 0.0168,
      "reward": 1.2250000357627868,
      "reward_std": 0.22191281020641326,
      "rewards/accuracy_reward": 0.2458333395421505,
      "rewards/format_reward": 0.9791666746139527,
      "step": 390
    },
    {
      "completion_length": 234.5916748046875,
      "epoch": 0.158,
      "grad_norm": 0.8642430926189166,
      "kl": 0.4025390625,
      "learning_rate": 1.979575249599344e-05,
      "loss": 0.0161,
      "reward": 0.9750000298023224,
      "reward_std": 0.3880663111805916,
      "rewards/accuracy_reward": 0.1750000085681677,
      "rewards/format_reward": 0.8000000178813934,
      "step": 395
    },
    {
      "completion_length": 195.85000457763672,
      "epoch": 0.16,
      "grad_norm": 1.1058693984901085,
      "kl": 0.441796875,
      "learning_rate": 1.9781476007338058e-05,
      "loss": 0.0177,
      "reward": 1.079166704416275,
      "reward_std": 0.21014036387205123,
      "rewards/accuracy_reward": 0.11666666753590108,
      "rewards/format_reward": 0.9625000119209289,
      "step": 400
    },
    {
      "epoch": 0.16,
      "eval_completion_length": 170.86378596379205,
      "eval_kl": 0.44906850961538464,
      "eval_loss": 0.01773521862924099,
      "eval_reward": 1.054487210053664,
      "eval_reward_std": 0.11891960610563938,
      "eval_rewards/accuracy_reward": 0.06410256615624978,
      "eval_rewards/format_reward": 0.9903846245545608,
      "eval_runtime": 97.7782,
      "eval_samples_per_second": 1.012,
      "eval_steps_per_second": 0.031,
      "step": 400
    },
    {
      "completion_length": 151.82083740234376,
      "epoch": 0.162,
      "grad_norm": 0.8251618816845695,
      "kl": 0.5263671875,
      "learning_rate": 1.9766722783341682e-05,
      "loss": 0.0211,
      "reward": 1.037500023841858,
      "reward_std": 0.08976087272167206,
      "rewards/accuracy_reward": 0.04166666753590107,
      "rewards/format_reward": 0.9958333373069763,
      "step": 405
    },
    {
      "completion_length": 145.5250045776367,
      "epoch": 0.164,
      "grad_norm": 0.18821333790127082,
      "kl": 0.5125,
      "learning_rate": 1.9751493543055634e-05,
      "loss": 0.0205,
      "reward": 1.0458333611488342,
      "reward_std": 0.13215193450450896,
      "rewards/accuracy_reward": 0.05833333544433117,
      "rewards/format_reward": 0.987500011920929,
      "step": 410
    },
    {
      "completion_length": 209.22500610351562,
      "epoch": 0.166,
      "grad_norm": 0.540401444429271,
      "kl": 0.456640625,
      "learning_rate": 1.9735789028731603e-05,
      "loss": 0.0183,
      "reward": 1.0708333849906921,
      "reward_std": 0.1970970779657364,
      "rewards/accuracy_reward": 0.0958333346992731,
      "rewards/format_reward": 0.9750000178813935,
      "step": 415
    },
    {
      "completion_length": 235.44583892822266,
      "epoch": 0.168,
      "grad_norm": 1.4057030696513535,
      "kl": 0.6052734375,
      "learning_rate": 1.9719610005785466e-05,
      "loss": 0.0242,
      "reward": 1.0291667044162751,
      "reward_std": 0.2490410476922989,
      "rewards/accuracy_reward": 0.11250000596046447,
      "rewards/format_reward": 0.916666692495346,
      "step": 420
    },
    {
      "completion_length": 371.56251220703126,
      "epoch": 0.17,
      "grad_norm": 1.8260065247232258,
      "kl": 1.516015625,
      "learning_rate": 1.9702957262759964e-05,
      "loss": 0.0607,
      "reward": 0.9250000298023224,
      "reward_std": 0.4401633307337761,
      "rewards/accuracy_reward": 0.12083333618938923,
      "rewards/format_reward": 0.8041666924953461,
      "step": 425
    },
    {
      "completion_length": 528.0041763305665,
      "epoch": 0.172,
      "grad_norm": 1.0316236154420109,
      "kl": 0.641015625,
      "learning_rate": 1.9685831611286312e-05,
      "loss": 0.0257,
      "reward": 0.8958333611488343,
      "reward_std": 0.34284951016306875,
      "rewards/accuracy_reward": 0.07083333469927311,
      "rewards/format_reward": 0.8250000178813934,
      "step": 430
    },
    {
      "completion_length": 241.72084197998046,
      "epoch": 0.174,
      "grad_norm": 1.3580722638101046,
      "kl": 0.476171875,
      "learning_rate": 1.9668233886044597e-05,
      "loss": 0.0191,
      "reward": 1.137500035762787,
      "reward_std": 0.26374180018901827,
      "rewards/accuracy_reward": 0.1666666716337204,
      "rewards/format_reward": 0.9708333492279053,
      "step": 435
    },
    {
      "completion_length": 224.58750610351564,
      "epoch": 0.176,
      "grad_norm": 1.020250043379181,
      "kl": 0.5072265625,
      "learning_rate": 1.9650164944723116e-05,
      "loss": 0.0203,
      "reward": 1.0666667103767395,
      "reward_std": 0.12149013355374336,
      "rewards/accuracy_reward": 0.07083333544433117,
      "rewards/format_reward": 0.9958333373069763,
      "step": 440
    },
    {
      "completion_length": 229.78334197998046,
      "epoch": 0.178,
      "grad_norm": 0.5657766777808181,
      "kl": 0.5259765625,
      "learning_rate": 1.9631625667976584e-05,
      "loss": 0.021,
      "reward": 1.0666667222976685,
      "reward_std": 0.201721428334713,
      "rewards/accuracy_reward": 0.09166666865348816,
      "rewards/format_reward": 0.9750000178813935,
      "step": 445
    },
    {
      "completion_length": 318.9833419799805,
      "epoch": 0.18,
      "grad_norm": 0.823275318702485,
      "kl": 0.48515625,
      "learning_rate": 1.961261695938319e-05,
      "loss": 0.0194,
      "reward": 1.1000000476837157,
      "reward_std": 0.22587689757347107,
      "rewards/accuracy_reward": 0.12500000409781933,
      "rewards/format_reward": 0.9750000059604644,
      "step": 450
    },
    {
      "completion_length": 703.1000152587891,
      "epoch": 0.182,
      "grad_norm": 398.7784621859931,
      "kl": 160.9056640625,
      "learning_rate": 1.9593139745400575e-05,
      "loss": 6.4618,
      "reward": 1.1000000298023225,
      "reward_std": 0.256098273396492,
      "rewards/accuracy_reward": 0.1708333373069763,
      "rewards/format_reward": 0.9291666805744171,
      "step": 455
    },
    {
      "completion_length": 1592.820848083496,
      "epoch": 0.184,
      "grad_norm": 0.7205262218883979,
      "kl": 3.237109375,
      "learning_rate": 1.9573194975320672e-05,
      "loss": 0.1296,
      "reward": 0.9125000298023224,
      "reward_std": 0.3876332946121693,
      "rewards/accuracy_reward": 0.07500000186264515,
      "rewards/format_reward": 0.8375000178813934,
      "step": 460
    },
    {
      "completion_length": 242.61250915527344,
      "epoch": 0.186,
      "grad_norm": 0.5497982888539013,
      "kl": 0.45546875,
      "learning_rate": 1.9552783621223437e-05,
      "loss": 0.0182,
      "reward": 1.0958333611488342,
      "reward_std": 0.2071055732667446,
      "rewards/accuracy_reward": 0.10833333618938923,
      "rewards/format_reward": 0.987500011920929,
      "step": 465
    },
    {
      "completion_length": 240.020841217041,
      "epoch": 0.188,
      "grad_norm": 0.7431928948219811,
      "kl": 0.4419921875,
      "learning_rate": 1.9531906677929472e-05,
      "loss": 0.0177,
      "reward": 1.0583333849906922,
      "reward_std": 0.10746954828500747,
      "rewards/accuracy_reward": 0.06666666902601719,
      "rewards/format_reward": 0.9916666746139526,
      "step": 470
    },
    {
      "completion_length": 265.0541763305664,
      "epoch": 0.19,
      "grad_norm": 0.7124969155812837,
      "kl": 0.394921875,
      "learning_rate": 1.9510565162951538e-05,
      "loss": 0.0158,
      "reward": 1.054166716337204,
      "reward_std": 0.16784698963165284,
      "rewards/accuracy_reward": 0.07916666865348816,
      "rewards/format_reward": 0.9750000178813935,
      "step": 475
    },
    {
      "completion_length": 324.0500061035156,
      "epoch": 0.192,
      "grad_norm": 0.6002616441993323,
      "kl": 0.4072265625,
      "learning_rate": 1.9488760116444966e-05,
      "loss": 0.0163,
      "reward": 1.100000023841858,
      "reward_std": 0.23401278182864188,
      "rewards/accuracy_reward": 0.1541666679084301,
      "rewards/format_reward": 0.9458333432674408,
      "step": 480
    },
    {
      "completion_length": 310.5416778564453,
      "epoch": 0.194,
      "grad_norm": 0.623552896500724,
      "kl": 0.386328125,
      "learning_rate": 1.9466492601156964e-05,
      "loss": 0.0154,
      "reward": 1.0958333730697631,
      "reward_std": 0.1777988240122795,
      "rewards/accuracy_reward": 0.10833333656191826,
      "rewards/format_reward": 0.9875000059604645,
      "step": 485
    },
    {
      "completion_length": 269.23750762939454,
      "epoch": 0.196,
      "grad_norm": 0.18932160471227394,
      "kl": 0.45546875,
      "learning_rate": 1.944376370237481e-05,
      "loss": 0.0182,
      "reward": 1.1000000357627868,
      "reward_std": 0.12386635020375251,
      "rewards/accuracy_reward": 0.10416666865348816,
      "rewards/format_reward": 0.9958333373069763,
      "step": 490
    },
    {
      "completion_length": 279.84167327880857,
      "epoch": 0.198,
      "grad_norm": 0.7399782553975255,
      "kl": 0.4009765625,
      "learning_rate": 1.942057452787297e-05,
      "loss": 0.0161,
      "reward": 1.1041666865348816,
      "reward_std": 0.18267755210399628,
      "rewards/accuracy_reward": 0.10833333544433117,
      "rewards/format_reward": 0.9958333373069763,
      "step": 495
    },
    {
      "completion_length": 417.3166763305664,
      "epoch": 0.2,
      "grad_norm": 0.7406681135160039,
      "kl": 0.437890625,
      "learning_rate": 1.9396926207859085e-05,
      "loss": 0.0175,
      "reward": 1.0125000298023223,
      "reward_std": 0.21925573498010636,
      "rewards/accuracy_reward": 0.07500000223517418,
      "rewards/format_reward": 0.9375000238418579,
      "step": 500
    },
    {
      "epoch": 0.2,
      "eval_completion_length": 388.1218050443209,
      "eval_kl": 0.3545673076923077,
      "eval_loss": 0.014274503104388714,
      "eval_reward": 1.0881410516225374,
      "eval_reward_std": 0.2549985142854544,
      "eval_rewards/accuracy_reward": 0.13141025946690485,
      "eval_rewards/format_reward": 0.9567307829856873,
      "eval_runtime": 229.97,
      "eval_samples_per_second": 0.43,
      "eval_steps_per_second": 0.013,
      "step": 500
    },
    {
      "completion_length": 383.04168090820315,
      "epoch": 0.202,
      "grad_norm": 1.0637146992798727,
      "kl": 0.3994140625,
      "learning_rate": 1.937281989491892e-05,
      "loss": 0.016,
      "reward": 1.054166704416275,
      "reward_std": 0.26914927959442136,
      "rewards/accuracy_reward": 0.10416666939854621,
      "rewards/format_reward": 0.9500000238418579,
      "step": 505
    },
    {
      "completion_length": 388.8250076293945,
      "epoch": 0.204,
      "grad_norm": 0.7997193167799139,
      "kl": 0.3890625,
      "learning_rate": 1.9348256763960146e-05,
      "loss": 0.0156,
      "reward": 1.041666704416275,
      "reward_std": 0.21228634268045427,
      "rewards/accuracy_reward": 0.09583333507180214,
      "rewards/format_reward": 0.9458333432674408,
      "step": 510
    },
    {
      "completion_length": 307.4666748046875,
      "epoch": 0.206,
      "grad_norm": 0.8574230086065704,
      "kl": 0.436328125,
      "learning_rate": 1.9323238012155125e-05,
      "loss": 0.0175,
      "reward": 1.083333384990692,
      "reward_std": 0.25866183936595916,
      "rewards/accuracy_reward": 0.12083333656191826,
      "rewards/format_reward": 0.962500023841858,
      "step": 515
    },
    {
      "completion_length": 367.52500915527344,
      "epoch": 0.208,
      "grad_norm": 0.6605626143593909,
      "kl": 0.404296875,
      "learning_rate": 1.9297764858882516e-05,
      "loss": 0.0162,
      "reward": 1.0375000476837157,
      "reward_std": 0.2624797485768795,
      "rewards/accuracy_reward": 0.11666667200624943,
      "rewards/format_reward": 0.9208333492279053,
      "step": 520
    },
    {
      "completion_length": 393.74167785644534,
      "epoch": 0.21,
      "grad_norm": 0.8191056592078524,
      "kl": 0.36240234375,
      "learning_rate": 1.9271838545667876e-05,
      "loss": 0.0145,
      "reward": 1.1041667103767394,
      "reward_std": 0.2263012297451496,
      "rewards/accuracy_reward": 0.1375000048428774,
      "rewards/format_reward": 0.9666666865348816,
      "step": 525
    },
    {
      "completion_length": 371.56251220703126,
      "epoch": 0.212,
      "grad_norm": 0.7002256272476179,
      "kl": 0.380859375,
      "learning_rate": 1.9245460336123136e-05,
      "loss": 0.0152,
      "reward": 1.1833333611488341,
      "reward_std": 0.2313358962535858,
      "rewards/accuracy_reward": 0.20000000335276127,
      "rewards/format_reward": 0.9833333492279053,
      "step": 530
    },
    {
      "completion_length": 342.72084655761716,
      "epoch": 0.214,
      "grad_norm": 0.6239304105044072,
      "kl": 0.3921875,
      "learning_rate": 1.9218631515885007e-05,
      "loss": 0.0157,
      "reward": 1.1041666984558105,
      "reward_std": 0.15337080582976342,
      "rewards/accuracy_reward": 0.1208333358168602,
      "rewards/format_reward": 0.9833333432674408,
      "step": 535
    },
    {
      "completion_length": 361.2583404541016,
      "epoch": 0.216,
      "grad_norm": 0.6914248861950051,
      "kl": 0.33828125,
      "learning_rate": 1.9191353392552346e-05,
      "loss": 0.0135,
      "reward": 1.0375000357627868,
      "reward_std": 0.1918194182217121,
      "rewards/accuracy_reward": 0.07083333395421505,
      "rewards/format_reward": 0.9666666805744171,
      "step": 540
    },
    {
      "completion_length": 282.5916748046875,
      "epoch": 0.218,
      "grad_norm": 0.40052866572019247,
      "kl": 0.369921875,
      "learning_rate": 1.9163627295622397e-05,
      "loss": 0.0148,
      "reward": 1.112500047683716,
      "reward_std": 0.23641232177615165,
      "rewards/accuracy_reward": 0.12916666977107524,
      "rewards/format_reward": 0.9833333492279053,
      "step": 545
    },
    {
      "completion_length": 264.9500106811523,
      "epoch": 0.22,
      "grad_norm": 0.6866747301077355,
      "kl": 0.3935546875,
      "learning_rate": 1.913545457642601e-05,
      "loss": 0.0157,
      "reward": 1.0791667222976684,
      "reward_std": 0.18432048112154006,
      "rewards/accuracy_reward": 0.09166666939854622,
      "rewards/format_reward": 0.987500011920929,
      "step": 550
    },
    {
      "completion_length": 259.1375076293945,
      "epoch": 0.222,
      "grad_norm": 0.44464019735481025,
      "kl": 0.39765625,
      "learning_rate": 1.910683660806177e-05,
      "loss": 0.0159,
      "reward": 1.025000023841858,
      "reward_std": 0.08862337395548821,
      "rewards/accuracy_reward": 0.03750000037252903,
      "rewards/format_reward": 0.987500011920929,
      "step": 555
    },
    {
      "completion_length": 251.38333740234376,
      "epoch": 0.224,
      "grad_norm": 0.535685767278896,
      "kl": 0.3630859375,
      "learning_rate": 1.907777478532909e-05,
      "loss": 0.0145,
      "reward": 1.0041666984558106,
      "reward_std": 0.11558076366782188,
      "rewards/accuracy_reward": 0.03333333432674408,
      "rewards/format_reward": 0.9708333492279053,
      "step": 560
    },
    {
      "completion_length": 176.72083587646483,
      "epoch": 0.226,
      "grad_norm": 1.253081808635552,
      "kl": 0.3888671875,
      "learning_rate": 1.9048270524660197e-05,
      "loss": 0.0156,
      "reward": 1.0708333909511567,
      "reward_std": 0.1375594198703766,
      "rewards/accuracy_reward": 0.08750000409781933,
      "rewards/format_reward": 0.9833333432674408,
      "step": 565
    },
    {
      "completion_length": 219.94584045410156,
      "epoch": 0.228,
      "grad_norm": 0.7666359774040074,
      "kl": 0.35390625,
      "learning_rate": 1.901832526405114e-05,
      "loss": 0.0141,
      "reward": 1.0250000298023223,
      "reward_std": 0.22138405814766884,
      "rewards/accuracy_reward": 0.08333333507180214,
      "rewards/format_reward": 0.9416666865348816,
      "step": 570
    },
    {
      "completion_length": 325.93750915527346,
      "epoch": 0.23,
      "grad_norm": 0.6205649765301138,
      "kl": 0.36328125,
      "learning_rate": 1.8987940462991673e-05,
      "loss": 0.0145,
      "reward": 0.975000011920929,
      "reward_std": 0.19853876382112504,
      "rewards/accuracy_reward": 0.03750000074505806,
      "rewards/format_reward": 0.9375000238418579,
      "step": 575
    },
    {
      "completion_length": 335.4583450317383,
      "epoch": 0.232,
      "grad_norm": 0.3814987875030443,
      "kl": 0.3525390625,
      "learning_rate": 1.895711760239413e-05,
      "loss": 0.0141,
      "reward": 0.9916666924953461,
      "reward_std": 0.15066706538200378,
      "rewards/accuracy_reward": 0.04166666828095913,
      "rewards/format_reward": 0.9500000238418579,
      "step": 580
    },
    {
      "completion_length": 207.9416732788086,
      "epoch": 0.234,
      "grad_norm": 0.5782975897753057,
      "kl": 0.3814453125,
      "learning_rate": 1.892585818452126e-05,
      "loss": 0.0153,
      "reward": 1.0375000357627868,
      "reward_std": 0.12116261571645737,
      "rewards/accuracy_reward": 0.05000000186264515,
      "rewards/format_reward": 0.987500011920929,
      "step": 585
    },
    {
      "completion_length": 150.35000305175782,
      "epoch": 0.236,
      "grad_norm": 0.8439037056084903,
      "kl": 0.422265625,
      "learning_rate": 1.889416373291298e-05,
      "loss": 0.0169,
      "reward": 1.100000023841858,
      "reward_std": 0.1805025689303875,
      "rewards/accuracy_reward": 0.10833333656191826,
      "rewards/format_reward": 0.9916666746139526,
      "step": 590
    },
    {
      "completion_length": 94.01666870117188,
      "epoch": 0.238,
      "grad_norm": 0.7394663774383612,
      "kl": 0.395703125,
      "learning_rate": 1.8862035792312148e-05,
      "loss": 0.0158,
      "reward": 1.0750000119209289,
      "reward_std": 0.16469117775559425,
      "rewards/accuracy_reward": 0.09166666716337205,
      "rewards/format_reward": 0.9833333492279053,
      "step": 595
    },
    {
      "completion_length": 77.09166946411133,
      "epoch": 0.24,
      "grad_norm": 0.9471876589521598,
      "kl": 0.4751953125,
      "learning_rate": 1.8829475928589272e-05,
      "loss": 0.019,
      "reward": 1.0708333611488343,
      "reward_std": 0.1359931766986847,
      "rewards/accuracy_reward": 0.07500000186264515,
      "rewards/format_reward": 0.9958333373069763,
      "step": 600
    },
    {
      "epoch": 0.24,
      "eval_completion_length": 124.43750528188852,
      "eval_kl": 0.4358473557692308,
      "eval_loss": 0.017942963168025017,
      "eval_reward": 1.0528846566493695,
      "eval_reward_std": 0.15011845901608467,
      "eval_rewards/accuracy_reward": 0.07692307935884365,
      "eval_rewards/format_reward": 0.9759615522164565,
      "eval_runtime": 169.086,
      "eval_samples_per_second": 0.586,
      "eval_steps_per_second": 0.018,
      "step": 600
    },
    {
      "completion_length": 100.78750305175781,
      "epoch": 0.242,
      "grad_norm": 1.1299382774251023,
      "kl": 0.448046875,
      "learning_rate": 1.879648572866617e-05,
      "loss": 0.0179,
      "reward": 1.0416666865348816,
      "reward_std": 0.07013157680630684,
      "rewards/accuracy_reward": 0.04166666753590107,
      "rewards/format_reward": 1.0,
      "step": 605
    },
    {
      "completion_length": 226.15834045410156,
      "epoch": 0.244,
      "grad_norm": 0.658103597322785,
      "kl": 0.474609375,
      "learning_rate": 1.8763066800438638e-05,
      "loss": 0.019,
      "reward": 0.9958333551883698,
      "reward_std": 0.1575162321329117,
      "rewards/accuracy_reward": 0.05833333544433117,
      "rewards/format_reward": 0.9375000119209289,
      "step": 610
    },
    {
      "completion_length": 134.96250534057617,
      "epoch": 0.246,
      "grad_norm": 0.19951858933164002,
      "kl": 0.413671875,
      "learning_rate": 1.8729220772698096e-05,
      "loss": 0.0165,
      "reward": 1.0500000298023224,
      "reward_std": 0.11287702694535255,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.9875000059604645,
      "step": 615
    },
    {
      "completion_length": 116.44583549499512,
      "epoch": 0.248,
      "grad_norm": 0.3702693586956776,
      "kl": 0.4185546875,
      "learning_rate": 1.869494929505219e-05,
      "loss": 0.0168,
      "reward": 1.0666666984558106,
      "reward_std": 0.09455960318446159,
      "rewards/accuracy_reward": 0.0750000026077032,
      "rewards/format_reward": 0.9916666746139526,
      "step": 620
    },
    {
      "completion_length": 200.26667251586915,
      "epoch": 0.25,
      "grad_norm": 0.2852776482719371,
      "kl": 0.3765625,
      "learning_rate": 1.866025403784439e-05,
      "loss": 0.0151,
      "reward": 1.0041666984558106,
      "reward_std": 0.11226474940776825,
      "rewards/accuracy_reward": 0.03333333395421505,
      "rewards/format_reward": 0.9708333492279053,
      "step": 625
    },
    {
      "completion_length": 215.9041748046875,
      "epoch": 0.252,
      "grad_norm": 0.5976353205994976,
      "kl": 0.3591796875,
      "learning_rate": 1.8625136692072577e-05,
      "loss": 0.0144,
      "reward": 1.0291666865348816,
      "reward_std": 0.1262888640165329,
      "rewards/accuracy_reward": 0.05416666865348816,
      "rewards/format_reward": 0.9750000178813935,
      "step": 630
    },
    {
      "completion_length": 190.21250610351564,
      "epoch": 0.254,
      "grad_norm": 1.1842051623519905,
      "kl": 0.8265625,
      "learning_rate": 1.8589598969306646e-05,
      "loss": 0.0332,
      "reward": 1.0500000238418579,
      "reward_std": 0.24186472296714784,
      "rewards/accuracy_reward": 0.10000000223517418,
      "rewards/format_reward": 0.9500000238418579,
      "step": 635
    },
    {
      "completion_length": 147.13750305175782,
      "epoch": 0.256,
      "grad_norm": 1.0641861499505445,
      "kl": 0.3560546875,
      "learning_rate": 1.855364260160507e-05,
      "loss": 0.0142,
      "reward": 0.9791666805744171,
      "reward_std": 0.19374003261327744,
      "rewards/accuracy_reward": 0.04166666716337204,
      "rewards/format_reward": 0.9375000059604645,
      "step": 640
    },
    {
      "completion_length": 123.84167022705078,
      "epoch": 0.258,
      "grad_norm": 0.4937714387665741,
      "kl": 0.373828125,
      "learning_rate": 1.851726934143048e-05,
      "loss": 0.0149,
      "reward": 1.108333373069763,
      "reward_std": 0.12118594124913215,
      "rewards/accuracy_reward": 0.11666667014360428,
      "rewards/format_reward": 0.9916666746139526,
      "step": 645
    },
    {
      "completion_length": 131.0833366394043,
      "epoch": 0.26,
      "grad_norm": 1.113245608082879,
      "kl": 0.362109375,
      "learning_rate": 1.848048096156426e-05,
      "loss": 0.0145,
      "reward": 1.0375000357627868,
      "reward_std": 0.13965441212058066,
      "rewards/accuracy_reward": 0.0541666679084301,
      "rewards/format_reward": 0.9833333432674408,
      "step": 650
    },
    {
      "completion_length": 231.68333740234374,
      "epoch": 0.262,
      "grad_norm": 0.8526891285926977,
      "kl": 0.376953125,
      "learning_rate": 1.8443279255020153e-05,
      "loss": 0.0151,
      "reward": 1.0333333551883697,
      "reward_std": 0.1361908882856369,
      "rewards/accuracy_reward": 0.05416666753590107,
      "rewards/format_reward": 0.9791666805744171,
      "step": 655
    },
    {
      "completion_length": 415.9833465576172,
      "epoch": 0.264,
      "grad_norm": 0.45480441770598673,
      "kl": 0.3181640625,
      "learning_rate": 1.8405666034956842e-05,
      "loss": 0.0127,
      "reward": 0.9791666865348816,
      "reward_std": 0.23286528065800666,
      "rewards/accuracy_reward": 0.0541666679084301,
      "rewards/format_reward": 0.9250000178813934,
      "step": 660
    },
    {
      "completion_length": 267.25000762939453,
      "epoch": 0.266,
      "grad_norm": 0.6844922760236942,
      "kl": 0.333203125,
      "learning_rate": 1.836764313458962e-05,
      "loss": 0.0133,
      "reward": 1.0416666865348816,
      "reward_std": 0.17620573043823243,
      "rewards/accuracy_reward": 0.0666666679084301,
      "rewards/format_reward": 0.975000011920929,
      "step": 665
    },
    {
      "completion_length": 205.49167175292968,
      "epoch": 0.268,
      "grad_norm": 0.4714832811705634,
      "kl": 0.3427734375,
      "learning_rate": 1.8329212407100996e-05,
      "loss": 0.0137,
      "reward": 1.0583333611488341,
      "reward_std": 0.09054399207234383,
      "rewards/accuracy_reward": 0.06666666753590107,
      "rewards/format_reward": 0.9916666746139526,
      "step": 670
    },
    {
      "completion_length": 189.8000061035156,
      "epoch": 0.27,
      "grad_norm": 1.2221084115892056,
      "kl": 0.383984375,
      "learning_rate": 1.8290375725550417e-05,
      "loss": 0.0154,
      "reward": 1.0541667103767396,
      "reward_std": 0.14698250517249106,
      "rewards/accuracy_reward": 0.06666666865348816,
      "rewards/format_reward": 0.987500011920929,
      "step": 675
    },
    {
      "completion_length": 159.6666717529297,
      "epoch": 0.272,
      "grad_norm": 0.4965291293623947,
      "kl": 0.3876953125,
      "learning_rate": 1.8251134982782952e-05,
      "loss": 0.0155,
      "reward": 1.0625000238418578,
      "reward_std": 0.11656158864498138,
      "rewards/accuracy_reward": 0.07083333469927311,
      "rewards/format_reward": 0.9916666746139526,
      "step": 680
    },
    {
      "completion_length": 151.78333740234376,
      "epoch": 0.274,
      "grad_norm": 0.7900759935925377,
      "kl": 0.4359375,
      "learning_rate": 1.821149209133704e-05,
      "loss": 0.0174,
      "reward": 1.0708333611488343,
      "reward_std": 0.14890312701463698,
      "rewards/accuracy_reward": 0.09166666865348816,
      "rewards/format_reward": 0.9791666805744171,
      "step": 685
    },
    {
      "completion_length": 228.40834121704103,
      "epoch": 0.276,
      "grad_norm": 1.0352924993296302,
      "kl": 0.448828125,
      "learning_rate": 1.8171448983351284e-05,
      "loss": 0.018,
      "reward": 1.0291666865348816,
      "reward_std": 0.2831733852624893,
      "rewards/accuracy_reward": 0.09583333693444729,
      "rewards/format_reward": 0.9333333492279052,
      "step": 690
    },
    {
      "completion_length": 266.6416732788086,
      "epoch": 0.278,
      "grad_norm": 0.4014065877432687,
      "kl": 0.3525390625,
      "learning_rate": 1.8131007610470278e-05,
      "loss": 0.0141,
      "reward": 1.0291667044162751,
      "reward_std": 0.21711408495903015,
      "rewards/accuracy_reward": 0.0791666679084301,
      "rewards/format_reward": 0.950000011920929,
      "step": 695
    },
    {
      "completion_length": 231.82084197998046,
      "epoch": 0.28,
      "grad_norm": 0.4138240924988841,
      "kl": 0.387109375,
      "learning_rate": 1.8090169943749477e-05,
      "loss": 0.0155,
      "reward": 1.0291667103767395,
      "reward_std": 0.1521087557077408,
      "rewards/accuracy_reward": 0.06666666865348816,
      "rewards/format_reward": 0.9625000119209289,
      "step": 700
    },
    {
      "epoch": 0.28,
      "eval_completion_length": 198.6218003493089,
      "eval_kl": 0.3661358173076923,
      "eval_loss": 0.014910033904016018,
      "eval_reward": 1.049679513160999,
      "eval_reward_std": 0.18441820717774904,
      "eval_rewards/accuracy_reward": 0.08012820622668816,
      "eval_rewards/format_reward": 0.969551301919497,
      "eval_runtime": 181.1559,
      "eval_samples_per_second": 0.546,
      "eval_steps_per_second": 0.017,
      "step": 700
    },
    {
      "completion_length": 204.03334274291993,
      "epoch": 0.282,
      "grad_norm": 0.8323688087389707,
      "kl": 0.39140625,
      "learning_rate": 1.804893797355914e-05,
      "loss": 0.0157,
      "reward": 1.0666666984558106,
      "reward_std": 0.21953659653663635,
      "rewards/accuracy_reward": 0.11250000074505806,
      "rewards/format_reward": 0.9541666805744171,
      "step": 705
    },
    {
      "completion_length": 150.07083892822266,
      "epoch": 0.284,
      "grad_norm": 0.8298380461791608,
      "kl": 0.3693359375,
      "learning_rate": 1.8007313709487334e-05,
      "loss": 0.0148,
      "reward": 1.0500000178813935,
      "reward_std": 0.18621423542499543,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.9458333373069763,
      "step": 710
    },
    {
      "completion_length": 121.3583381652832,
      "epoch": 0.286,
      "grad_norm": 0.7950998845132158,
      "kl": 0.447265625,
      "learning_rate": 1.7965299180241963e-05,
      "loss": 0.0179,
      "reward": 1.0833333611488343,
      "reward_std": 0.15956140235066413,
      "rewards/accuracy_reward": 0.09583333432674408,
      "rewards/format_reward": 0.9875000059604645,
      "step": 715
    },
    {
      "completion_length": 140.52083740234374,
      "epoch": 0.288,
      "grad_norm": 0.7795378639075542,
      "kl": 0.45703125,
      "learning_rate": 1.792289643355191e-05,
      "loss": 0.0183,
      "reward": 1.108333373069763,
      "reward_std": 0.17277553156018258,
      "rewards/accuracy_reward": 0.12083333618938923,
      "rewards/format_reward": 0.987500011920929,
      "step": 720
    },
    {
      "completion_length": 148.92917022705078,
      "epoch": 0.29,
      "grad_norm": 1.3272943198870826,
      "kl": 0.498828125,
      "learning_rate": 1.788010753606722e-05,
      "loss": 0.02,
      "reward": 1.0708333849906921,
      "reward_std": 0.15627224519848823,
      "rewards/accuracy_reward": 0.08333333432674409,
      "rewards/format_reward": 0.987500011920929,
      "step": 725
    },
    {
      "completion_length": 175.61667098999024,
      "epoch": 0.292,
      "grad_norm": 0.6424382767772573,
      "kl": 0.464453125,
      "learning_rate": 1.78369345732584e-05,
      "loss": 0.0186,
      "reward": 1.041666716337204,
      "reward_std": 0.1477656200528145,
      "rewards/accuracy_reward": 0.07083333544433117,
      "rewards/format_reward": 0.9708333432674408,
      "step": 730
    },
    {
      "completion_length": 126.27500381469727,
      "epoch": 0.294,
      "grad_norm": 0.7948572131207867,
      "kl": 0.639453125,
      "learning_rate": 1.7793379649314743e-05,
      "loss": 0.0256,
      "reward": 1.1208333611488341,
      "reward_std": 0.18972794637084006,
      "rewards/accuracy_reward": 0.13333333618938922,
      "rewards/format_reward": 0.9875000059604645,
      "step": 735
    },
    {
      "completion_length": 94.61666946411133,
      "epoch": 0.296,
      "grad_norm": 1.0745927572590321,
      "kl": 0.5287109375,
      "learning_rate": 1.7749444887041797e-05,
      "loss": 0.0212,
      "reward": 1.0250000476837158,
      "reward_std": 0.1675926238298416,
      "rewards/accuracy_reward": 0.06666666753590107,
      "rewards/format_reward": 0.9583333432674408,
      "step": 740
    },
    {
      "completion_length": 104.27500381469727,
      "epoch": 0.298,
      "grad_norm": 0.2227604024323154,
      "kl": 0.569140625,
      "learning_rate": 1.7705132427757895e-05,
      "loss": 0.0228,
      "reward": 1.0916667103767395,
      "reward_std": 0.10345393419265747,
      "rewards/accuracy_reward": 0.10000000186264515,
      "rewards/format_reward": 0.9916666746139526,
      "step": 745
    },
    {
      "completion_length": 81.67500152587891,
      "epoch": 0.3,
      "grad_norm": 1.7233432115097311,
      "kl": 0.6505859375,
      "learning_rate": 1.766044443118978e-05,
      "loss": 0.026,
      "reward": 1.0416666865348816,
      "reward_std": 0.0905439905822277,
      "rewards/accuracy_reward": 0.04583333432674408,
      "rewards/format_reward": 0.9958333373069763,
      "step": 750
    },
    {
      "completion_length": 101.40833587646485,
      "epoch": 0.302,
      "grad_norm": 0.8228291476842216,
      "kl": 0.559765625,
      "learning_rate": 1.761538307536737e-05,
      "loss": 0.0224,
      "reward": 1.0875000357627869,
      "reward_std": 0.11575513705611229,
      "rewards/accuracy_reward": 0.10833333656191826,
      "rewards/format_reward": 0.9791666865348816,
      "step": 755
    },
    {
      "completion_length": 103.02917098999023,
      "epoch": 0.304,
      "grad_norm": 1.2801924295313796,
      "kl": 0.5697265625,
      "learning_rate": 1.7569950556517566e-05,
      "loss": 0.0228,
      "reward": 1.0833333730697632,
      "reward_std": 0.16625742092728615,
      "rewards/accuracy_reward": 0.08750000223517418,
      "rewards/format_reward": 0.9958333373069763,
      "step": 760
    },
    {
      "completion_length": 109.6833366394043,
      "epoch": 0.306,
      "grad_norm": 1.231261037139189,
      "kl": 0.5431640625,
      "learning_rate": 1.7524149088957244e-05,
      "loss": 0.0217,
      "reward": 1.0708333611488343,
      "reward_std": 0.09324773102998733,
      "rewards/accuracy_reward": 0.0750000026077032,
      "rewards/format_reward": 0.9958333373069763,
      "step": 765
    },
    {
      "completion_length": 142.60833740234375,
      "epoch": 0.308,
      "grad_norm": 0.5849963492276671,
      "kl": 0.5087890625,
      "learning_rate": 1.747798090498532e-05,
      "loss": 0.0203,
      "reward": 1.062500035762787,
      "reward_std": 0.0674278385937214,
      "rewards/accuracy_reward": 0.06250000074505806,
      "rewards/format_reward": 1.0,
      "step": 770
    },
    {
      "completion_length": 161.4041717529297,
      "epoch": 0.31,
      "grad_norm": 0.9058008345995844,
      "kl": 0.57265625,
      "learning_rate": 1.7431448254773943e-05,
      "loss": 0.0229,
      "reward": 1.075000035762787,
      "reward_std": 0.14718020260334014,
      "rewards/accuracy_reward": 0.09583333730697632,
      "rewards/format_reward": 0.9791666805744171,
      "step": 775
    },
    {
      "completion_length": 157.4166702270508,
      "epoch": 0.312,
      "grad_norm": 1.1313027943159109,
      "kl": 0.4814453125,
      "learning_rate": 1.7384553406258842e-05,
      "loss": 0.0193,
      "reward": 1.0875000357627869,
      "reward_std": 0.11366014108061791,
      "rewards/accuracy_reward": 0.08750000260770321,
      "rewards/format_reward": 1.0,
      "step": 780
    },
    {
      "completion_length": 182.37500457763673,
      "epoch": 0.314,
      "grad_norm": 0.6238900338412194,
      "kl": 0.4681640625,
      "learning_rate": 1.7337298645028764e-05,
      "loss": 0.0187,
      "reward": 1.0333333611488342,
      "reward_std": 0.07763404920697212,
      "rewards/accuracy_reward": 0.03750000111758709,
      "rewards/format_reward": 0.9958333373069763,
      "step": 785
    },
    {
      "completion_length": 149.3458381652832,
      "epoch": 0.316,
      "grad_norm": 0.16296278407454673,
      "kl": 0.462109375,
      "learning_rate": 1.7289686274214116e-05,
      "loss": 0.0185,
      "reward": 1.0458333492279053,
      "reward_std": 0.07685092613101005,
      "rewards/accuracy_reward": 0.05000000149011612,
      "rewards/format_reward": 0.9958333373069763,
      "step": 790
    },
    {
      "completion_length": 130.19583587646486,
      "epoch": 0.318,
      "grad_norm": 0.7784532053538376,
      "kl": 0.43359375,
      "learning_rate": 1.7241718614374678e-05,
      "loss": 0.0173,
      "reward": 1.083333396911621,
      "reward_std": 0.13136881664395333,
      "rewards/accuracy_reward": 0.08750000298023224,
      "rewards/format_reward": 0.9958333373069763,
      "step": 795
    },
    {
      "completion_length": 119.23750228881836,
      "epoch": 0.32,
      "grad_norm": 0.983387641880522,
      "kl": 0.4408203125,
      "learning_rate": 1.7193398003386514e-05,
      "loss": 0.0176,
      "reward": 1.0833333730697632,
      "reward_std": 0.14619938507676125,
      "rewards/accuracy_reward": 0.08333333507180214,
      "rewards/format_reward": 1.0,
      "step": 800
    },
    {
      "epoch": 0.32,
      "eval_completion_length": 115.19872049184946,
      "eval_kl": 0.44921875,
      "eval_loss": 0.01777729205787182,
      "eval_reward": 1.0897436233667226,
      "eval_reward_std": 0.08877197739023429,
      "eval_rewards/accuracy_reward": 0.09134615628192058,
      "eval_rewards/format_reward": 0.9983974374257601,
      "eval_runtime": 85.6402,
      "eval_samples_per_second": 1.156,
      "eval_steps_per_second": 0.035,
      "step": 800
    },
    {
      "completion_length": 113.28750381469726,
      "epoch": 0.322,
      "grad_norm": 0.781174828098305,
      "kl": 0.453125,
      "learning_rate": 1.7144726796328034e-05,
      "loss": 0.0181,
      "reward": 1.145833396911621,
      "reward_std": 0.16145868599414825,
      "rewards/accuracy_reward": 0.14583334065973758,
      "rewards/format_reward": 1.0,
      "step": 805
    },
    {
      "completion_length": 175.64167022705078,
      "epoch": 0.324,
      "grad_norm": 0.5727964458555536,
      "kl": 0.471875,
      "learning_rate": 1.709570736536521e-05,
      "loss": 0.0189,
      "reward": 1.1083333611488342,
      "reward_std": 0.07763404548168182,
      "rewards/accuracy_reward": 0.10833333544433117,
      "rewards/format_reward": 1.0,
      "step": 810
    },
    {
      "completion_length": 246.63750762939452,
      "epoch": 0.326,
      "grad_norm": 1.0091346142420787,
      "kl": 0.4962890625,
      "learning_rate": 1.7046342099635948e-05,
      "loss": 0.0198,
      "reward": 1.0916666984558105,
      "reward_std": 0.16956989765167235,
      "rewards/accuracy_reward": 0.10000000223517418,
      "rewards/format_reward": 0.9916666746139526,
      "step": 815
    },
    {
      "completion_length": 322.5458450317383,
      "epoch": 0.328,
      "grad_norm": 0.7989919925619189,
      "kl": 0.465625,
      "learning_rate": 1.6996633405133656e-05,
      "loss": 0.0186,
      "reward": 1.0750000476837158,
      "reward_std": 0.1654976300895214,
      "rewards/accuracy_reward": 0.08750000149011612,
      "rewards/format_reward": 0.987500011920929,
      "step": 820
    },
    {
      "completion_length": 341.00834503173826,
      "epoch": 0.33,
      "grad_norm": 0.7043145841881899,
      "kl": 0.5212890625,
      "learning_rate": 1.6946583704589973e-05,
      "loss": 0.0208,
      "reward": 1.0333333611488342,
      "reward_std": 0.15038584247231485,
      "rewards/accuracy_reward": 0.07083333395421505,
      "rewards/format_reward": 0.962500023841858,
      "step": 825
    },
    {
      "completion_length": 241.1916732788086,
      "epoch": 0.332,
      "grad_norm": 0.6208263212333841,
      "kl": 0.469140625,
      "learning_rate": 1.68961954373567e-05,
      "loss": 0.0188,
      "reward": 1.0541666865348815,
      "reward_std": 0.1895946055650711,
      "rewards/accuracy_reward": 0.09583333693444729,
      "rewards/format_reward": 0.9583333611488343,
      "step": 830
    },
    {
      "completion_length": 230.2041763305664,
      "epoch": 0.334,
      "grad_norm": 1.0022357187387183,
      "kl": 0.5720703125,
      "learning_rate": 1.684547105928689e-05,
      "loss": 0.0229,
      "reward": 0.991666704416275,
      "reward_std": 0.19037772417068483,
      "rewards/accuracy_reward": 0.05833333507180214,
      "rewards/format_reward": 0.9333333551883698,
      "step": 835
    },
    {
      "completion_length": 264.27084197998045,
      "epoch": 0.336,
      "grad_norm": 0.9244091504002026,
      "kl": 0.5970703125,
      "learning_rate": 1.6794413042615168e-05,
      "loss": 0.0239,
      "reward": 1.0458333671092988,
      "reward_std": 0.23153360486030578,
      "rewards/accuracy_reward": 0.08750000260770321,
      "rewards/format_reward": 0.9583333551883697,
      "step": 840
    },
    {
      "completion_length": 158.6208381652832,
      "epoch": 0.338,
      "grad_norm": 0.6420688589621387,
      "kl": 0.4923828125,
      "learning_rate": 1.6743023875837233e-05,
      "loss": 0.0197,
      "reward": 1.045833373069763,
      "reward_std": 0.11767575517296791,
      "rewards/accuracy_reward": 0.05416666865348816,
      "rewards/format_reward": 0.9916666746139526,
      "step": 845
    },
    {
      "completion_length": 177.85000610351562,
      "epoch": 0.34,
      "grad_norm": 0.5413030087841071,
      "kl": 0.4517578125,
      "learning_rate": 1.6691306063588583e-05,
      "loss": 0.018,
      "reward": 1.0375000357627868,
      "reward_std": 0.09132710695266724,
      "rewards/accuracy_reward": 0.05000000149011612,
      "rewards/format_reward": 0.9875000059604645,
      "step": 850
    },
    {
      "completion_length": 153.7375045776367,
      "epoch": 0.342,
      "grad_norm": 15.670714549698124,
      "kl": 0.607421875,
      "learning_rate": 1.6639262126522417e-05,
      "loss": 0.0243,
      "reward": 1.1208333730697633,
      "reward_std": 0.20280873402953148,
      "rewards/accuracy_reward": 0.15000000596046448,
      "rewards/format_reward": 0.9708333432674408,
      "step": 855
    },
    {
      "completion_length": 153.3083351135254,
      "epoch": 0.344,
      "grad_norm": 0.5724519452372687,
      "kl": 0.4134765625,
      "learning_rate": 1.6586894601186804e-05,
      "loss": 0.0165,
      "reward": 1.0666667103767395,
      "reward_std": 0.11252264827489852,
      "rewards/accuracy_reward": 0.06666666828095913,
      "rewards/format_reward": 1.0,
      "step": 860
    },
    {
      "completion_length": 185.32083740234376,
      "epoch": 0.346,
      "grad_norm": 0.4311092487621537,
      "kl": 0.4388671875,
      "learning_rate": 1.6534206039901057e-05,
      "loss": 0.0176,
      "reward": 1.0958333730697631,
      "reward_std": 0.11017328798770905,
      "rewards/accuracy_reward": 0.10000000111758708,
      "rewards/format_reward": 0.9958333373069763,
      "step": 865
    },
    {
      "completion_length": 158.38750534057618,
      "epoch": 0.348,
      "grad_norm": 0.5107925238050466,
      "kl": 0.4388671875,
      "learning_rate": 1.6481199010631312e-05,
      "loss": 0.0176,
      "reward": 1.0791667103767395,
      "reward_std": 0.09132710918784141,
      "rewards/accuracy_reward": 0.07916666939854622,
      "rewards/format_reward": 1.0,
      "step": 870
    },
    {
      "completion_length": 151.10000534057616,
      "epoch": 0.35,
      "grad_norm": 0.855945600131941,
      "kl": 0.462109375,
      "learning_rate": 1.6427876096865394e-05,
      "loss": 0.0185,
      "reward": 1.0916667103767395,
      "reward_std": 0.1482943795621395,
      "rewards/accuracy_reward": 0.0916666690260172,
      "rewards/format_reward": 1.0,
      "step": 875
    },
    {
      "completion_length": 133.1916717529297,
      "epoch": 0.352,
      "grad_norm": 0.7382780134899118,
      "kl": 0.4478515625,
      "learning_rate": 1.63742398974869e-05,
      "loss": 0.0179,
      "reward": 1.1208333730697633,
      "reward_std": 0.1995551086962223,
      "rewards/accuracy_reward": 0.12916667051613331,
      "rewards/format_reward": 0.9916666746139526,
      "step": 880
    },
    {
      "completion_length": 144.54583816528321,
      "epoch": 0.354,
      "grad_norm": 0.9737960502017814,
      "kl": 0.48125,
      "learning_rate": 1.632029302664851e-05,
      "loss": 0.0193,
      "reward": 1.1041666984558105,
      "reward_std": 0.17280239537358283,
      "rewards/accuracy_reward": 0.10833333544433117,
      "rewards/format_reward": 0.9958333373069763,
      "step": 885
    },
    {
      "completion_length": 92.67500228881836,
      "epoch": 0.356,
      "grad_norm": 0.961078931712405,
      "kl": 0.51484375,
      "learning_rate": 1.6266038113644605e-05,
      "loss": 0.0206,
      "reward": 1.0833333611488343,
      "reward_std": 0.12578697353601456,
      "rewards/accuracy_reward": 0.08750000186264514,
      "rewards/format_reward": 0.9958333373069763,
      "step": 890
    },
    {
      "completion_length": 89.66666793823242,
      "epoch": 0.358,
      "grad_norm": 0.7233667735051951,
      "kl": 0.580859375,
      "learning_rate": 1.6211477802783105e-05,
      "loss": 0.0232,
      "reward": 1.1083333611488342,
      "reward_std": 0.16165639609098434,
      "rewards/accuracy_reward": 0.1166666705161333,
      "rewards/format_reward": 0.9916666746139526,
      "step": 895
    },
    {
      "completion_length": 93.3875015258789,
      "epoch": 0.36,
      "grad_norm": 0.4545503509815593,
      "kl": 0.7234375,
      "learning_rate": 1.6156614753256583e-05,
      "loss": 0.0289,
      "reward": 1.0666667103767395,
      "reward_std": 0.1506670705974102,
      "rewards/accuracy_reward": 0.07500000111758709,
      "rewards/format_reward": 0.9916666746139526,
      "step": 900
    },
    {
      "epoch": 0.36,
      "eval_completion_length": 109.18750293438251,
      "eval_kl": 0.6490384615384616,
      "eval_loss": 0.025919783860445023,
      "eval_reward": 1.0737179609445424,
      "eval_reward_std": 0.12753102469902772,
      "eval_rewards/accuracy_reward": 0.08493589882094127,
      "eval_rewards/format_reward": 0.9887820619803208,
      "eval_runtime": 93.0566,
      "eval_samples_per_second": 1.064,
      "eval_steps_per_second": 0.032,
      "step": 900
    },
    {
      "completion_length": 119.41250228881836,
      "epoch": 0.362,
      "grad_norm": 0.7762304601434084,
      "kl": 0.633984375,
      "learning_rate": 1.610145163901268e-05,
      "loss": 0.0253,
      "reward": 1.0500000417232513,
      "reward_std": 0.14235814958810805,
      "rewards/accuracy_reward": 0.07083333544433117,
      "rewards/format_reward": 0.9791666805744171,
      "step": 905
    },
    {
      "completion_length": 98.3083366394043,
      "epoch": 0.364,
      "grad_norm": 0.1969125008394314,
      "kl": 0.5587890625,
      "learning_rate": 1.6045991148623752e-05,
      "loss": 0.0223,
      "reward": 1.0791666984558106,
      "reward_std": 0.14157502874732017,
      "rewards/accuracy_reward": 0.09583333618938923,
      "rewards/format_reward": 0.9833333432674408,
      "step": 910
    },
    {
      "completion_length": 92.20833587646484,
      "epoch": 0.366,
      "grad_norm": 0.5710961469229422,
      "kl": 0.6203125,
      "learning_rate": 1.599023598515586e-05,
      "loss": 0.0248,
      "reward": 1.1000000476837157,
      "reward_std": 0.11095640808343887,
      "rewards/accuracy_reward": 0.1041666705161333,
      "rewards/format_reward": 0.9958333373069763,
      "step": 915
    },
    {
      "completion_length": 169.32083740234376,
      "epoch": 0.368,
      "grad_norm": 0.32186205273931673,
      "kl": 0.6314453125,
      "learning_rate": 1.5934188866037017e-05,
      "loss": 0.0253,
      "reward": 1.1166667103767396,
      "reward_std": 0.20900286808609964,
      "rewards/accuracy_reward": 0.13750000223517417,
      "rewards/format_reward": 0.9791666805744171,
      "step": 920
    },
    {
      "completion_length": 162.40833587646483,
      "epoch": 0.37,
      "grad_norm": 1.0716265467963713,
      "kl": 0.5517578125,
      "learning_rate": 1.5877852522924733e-05,
      "loss": 0.0221,
      "reward": 1.0875000357627869,
      "reward_std": 0.15046936124563218,
      "rewards/accuracy_reward": 0.09583333730697632,
      "rewards/format_reward": 0.9916666746139526,
      "step": 925
    },
    {
      "completion_length": 117.15416870117187,
      "epoch": 0.372,
      "grad_norm": 0.1718307429317029,
      "kl": 0.58984375,
      "learning_rate": 1.5821229701572897e-05,
      "loss": 0.0236,
      "reward": 1.0500000238418579,
      "reward_std": 0.1483220286667347,
      "rewards/accuracy_reward": 0.06250000111758709,
      "rewards/format_reward": 0.9875000059604645,
      "step": 930
    },
    {
      "completion_length": 101.3708366394043,
      "epoch": 0.374,
      "grad_norm": 3.113823270584589,
      "kl": 0.600390625,
      "learning_rate": 1.5764323161697933e-05,
      "loss": 0.024,
      "reward": 1.075000035762787,
      "reward_std": 0.15289186984300612,
      "rewards/accuracy_reward": 0.08750000335276127,
      "rewards/format_reward": 0.987500011920929,
      "step": 935
    },
    {
      "completion_length": 168.9416717529297,
      "epoch": 0.376,
      "grad_norm": 0.887035405128863,
      "kl": 0.6765625,
      "learning_rate": 1.570713567684432e-05,
      "loss": 0.0271,
      "reward": 1.037500023841858,
      "reward_std": 0.16739491894841194,
      "rewards/accuracy_reward": 0.06666666753590107,
      "rewards/format_reward": 0.9708333492279053,
      "step": 940
    },
    {
      "completion_length": 110.50416793823243,
      "epoch": 0.378,
      "grad_norm": 1.1890810823941778,
      "kl": 0.5552734375,
      "learning_rate": 1.564967003424938e-05,
      "loss": 0.0222,
      "reward": 1.1666666984558105,
      "reward_std": 0.2173117831349373,
      "rewards/accuracy_reward": 0.1833333384245634,
      "rewards/format_reward": 0.9833333492279053,
      "step": 945
    },
    {
      "completion_length": 166.82083892822266,
      "epoch": 0.38,
      "grad_norm": 0.9138187216102012,
      "kl": 0.530078125,
      "learning_rate": 1.5591929034707468e-05,
      "loss": 0.0212,
      "reward": 1.0625000298023224,
      "reward_std": 0.09673458784818649,
      "rewards/accuracy_reward": 0.07916666939854622,
      "rewards/format_reward": 0.9833333432674408,
      "step": 950
    },
    {
      "completion_length": 146.1916702270508,
      "epoch": 0.382,
      "grad_norm": 1.0622486882329054,
      "kl": 0.5103515625,
      "learning_rate": 1.553391549243344e-05,
      "loss": 0.0204,
      "reward": 1.0541666984558105,
      "reward_std": 0.137278188765049,
      "rewards/accuracy_reward": 0.07083333618938922,
      "rewards/format_reward": 0.9833333432674408,
      "step": 955
    },
    {
      "completion_length": 160.28750381469726,
      "epoch": 0.384,
      "grad_norm": 0.752318503568895,
      "kl": 0.520703125,
      "learning_rate": 1.5475632234925505e-05,
      "loss": 0.0208,
      "reward": 1.0375000357627868,
      "reward_std": 0.14157502725720406,
      "rewards/accuracy_reward": 0.06250000223517418,
      "rewards/format_reward": 0.9750000238418579,
      "step": 960
    },
    {
      "completion_length": 140.64583740234374,
      "epoch": 0.386,
      "grad_norm": 0.6324158396336607,
      "kl": 0.5330078125,
      "learning_rate": 1.54170821028274e-05,
      "loss": 0.0213,
      "reward": 1.0500000238418579,
      "reward_std": 0.10286851897835732,
      "rewards/accuracy_reward": 0.06250000335276126,
      "rewards/format_reward": 0.987500011920929,
      "step": 965
    },
    {
      "completion_length": 154.11667098999024,
      "epoch": 0.388,
      "grad_norm": 0.6918653815816405,
      "kl": 0.50625,
      "learning_rate": 1.5358267949789968e-05,
      "loss": 0.0203,
      "reward": 1.0666667222976685,
      "reward_std": 0.16120432689785957,
      "rewards/accuracy_reward": 0.07916666902601718,
      "rewards/format_reward": 0.987500011920929,
      "step": 970
    },
    {
      "completion_length": 182.7916717529297,
      "epoch": 0.39,
      "grad_norm": 0.8349715169939224,
      "kl": 0.5755859375,
      "learning_rate": 1.529919264233205e-05,
      "loss": 0.023,
      "reward": 1.0500000298023224,
      "reward_std": 0.179521743953228,
      "rewards/accuracy_reward": 0.07916666902601718,
      "rewards/format_reward": 0.9708333432674408,
      "step": 975
    },
    {
      "completion_length": 167.6125045776367,
      "epoch": 0.392,
      "grad_norm": 0.5431435446037667,
      "kl": 0.452734375,
      "learning_rate": 1.5239859059700794e-05,
      "loss": 0.0181,
      "reward": 1.0833333611488343,
      "reward_std": 0.17568050399422647,
      "rewards/accuracy_reward": 0.10000000447034836,
      "rewards/format_reward": 0.9833333492279053,
      "step": 980
    },
    {
      "completion_length": 179.76250534057618,
      "epoch": 0.394,
      "grad_norm": 0.8606801988109793,
      "kl": 0.439453125,
      "learning_rate": 1.5180270093731305e-05,
      "loss": 0.0176,
      "reward": 1.0916666984558105,
      "reward_std": 0.17522490546107292,
      "rewards/accuracy_reward": 0.1083333358168602,
      "rewards/format_reward": 0.9833333432674408,
      "step": 985
    },
    {
      "completion_length": 211.83334045410157,
      "epoch": 0.396,
      "grad_norm": 0.9351930624726825,
      "kl": 0.4248046875,
      "learning_rate": 1.5120428648705716e-05,
      "loss": 0.017,
      "reward": 1.0666666865348815,
      "reward_std": 0.20573704093694686,
      "rewards/accuracy_reward": 0.09583333544433117,
      "rewards/format_reward": 0.9708333492279053,
      "step": 990
    },
    {
      "completion_length": 165.30833587646484,
      "epoch": 0.398,
      "grad_norm": 0.6856844913007303,
      "kl": 0.43984375,
      "learning_rate": 1.5060337641211637e-05,
      "loss": 0.0176,
      "reward": 1.1291666865348815,
      "reward_std": 0.16761595159769058,
      "rewards/accuracy_reward": 0.1416666679084301,
      "rewards/format_reward": 0.987500011920929,
      "step": 995
    },
    {
      "completion_length": 150.04167098999022,
      "epoch": 0.4,
      "grad_norm": 0.9635145258572108,
      "kl": 0.4900390625,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.0196,
      "reward": 1.079166716337204,
      "reward_std": 0.1558768317103386,
      "rewards/accuracy_reward": 0.10000000335276127,
      "rewards/format_reward": 0.9791666805744171,
      "step": 1000
    },
    {
      "epoch": 0.4,
      "eval_completion_length": 126.44872283935547,
      "eval_kl": 0.48978365384615385,
      "eval_loss": 0.0195956751704216,
      "eval_reward": 1.112179526915917,
      "eval_reward_std": 0.15462170598598626,
      "eval_rewards/accuracy_reward": 0.12179487499480064,
      "eval_rewards/format_reward": 0.9903846245545608,
      "eval_runtime": 127.709,
      "eval_samples_per_second": 0.775,
      "eval_steps_per_second": 0.023,
      "step": 1000
    },
    {
      "completion_length": 168.5666702270508,
      "epoch": 0.402,
      "grad_norm": 1.0895511993093854,
      "kl": 0.4529296875,
      "learning_rate": 1.493941866584231e-05,
      "loss": 0.0181,
      "reward": 1.0875000476837158,
      "reward_std": 0.2581599496304989,
      "rewards/accuracy_reward": 0.12083333618938923,
      "rewards/format_reward": 0.9666666865348816,
      "step": 1005
    },
    {
      "completion_length": 161.23333663940429,
      "epoch": 0.404,
      "grad_norm": 0.7389397460996036,
      "kl": 0.5572265625,
      "learning_rate": 1.4878596591387329e-05,
      "loss": 0.0223,
      "reward": 1.0583333551883698,
      "reward_std": 0.23565217033028601,
      "rewards/accuracy_reward": 0.09583333656191825,
      "rewards/format_reward": 0.9625000119209289,
      "step": 1010
    },
    {
      "completion_length": 221.5958366394043,
      "epoch": 0.406,
      "grad_norm": 0.928302942173807,
      "kl": 0.5322265625,
      "learning_rate": 1.4817536741017153e-05,
      "loss": 0.0213,
      "reward": 1.0625000238418578,
      "reward_std": 0.26485596522688865,
      "rewards/accuracy_reward": 0.12500000335276126,
      "rewards/format_reward": 0.9375000178813935,
      "step": 1015
    },
    {
      "completion_length": 152.99167175292968,
      "epoch": 0.408,
      "grad_norm": 0.9384235272703269,
      "kl": 0.541796875,
      "learning_rate": 1.4756242090702756e-05,
      "loss": 0.0216,
      "reward": 1.0958333611488342,
      "reward_std": 0.25780556723475456,
      "rewards/accuracy_reward": 0.12500000186264515,
      "rewards/format_reward": 0.9708333551883698,
      "step": 1020
    },
    {
      "completion_length": 146.56250534057617,
      "epoch": 0.41,
      "grad_norm": 0.30172387560515257,
      "kl": 0.519140625,
      "learning_rate": 1.469471562785891e-05,
      "loss": 0.0208,
      "reward": 1.0750000417232513,
      "reward_std": 0.10206207036972045,
      "rewards/accuracy_reward": 0.08750000260770321,
      "rewards/format_reward": 0.9875000059604645,
      "step": 1025
    },
    {
      "completion_length": 152.51666946411132,
      "epoch": 0.412,
      "grad_norm": 0.9294930359078027,
      "kl": 1.005859375,
      "learning_rate": 1.463296035119862e-05,
      "loss": 0.0402,
      "reward": 1.0500000357627868,
      "reward_std": 0.14619938284158707,
      "rewards/accuracy_reward": 0.0666666679084301,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1030
    },
    {
      "completion_length": 180.95000381469725,
      "epoch": 0.414,
      "grad_norm": 0.4453899299979498,
      "kl": 0.487109375,
      "learning_rate": 1.4570979270586944e-05,
      "loss": 0.0195,
      "reward": 1.075000035762787,
      "reward_std": 0.21077675521373748,
      "rewards/accuracy_reward": 0.10000000186264515,
      "rewards/format_reward": 0.975000011920929,
      "step": 1035
    },
    {
      "completion_length": 138.85000457763672,
      "epoch": 0.416,
      "grad_norm": 0.7666780913080924,
      "kl": 0.5107421875,
      "learning_rate": 1.4508775406894308e-05,
      "loss": 0.0204,
      "reward": 1.0916666984558105,
      "reward_std": 0.12927382662892342,
      "rewards/accuracy_reward": 0.09583333730697632,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1040
    },
    {
      "completion_length": 226.35834045410155,
      "epoch": 0.418,
      "grad_norm": 0.6438784140914322,
      "kl": 0.449609375,
      "learning_rate": 1.4446351791849276e-05,
      "loss": 0.018,
      "reward": 1.0666667103767395,
      "reward_std": 0.1238663487136364,
      "rewards/accuracy_reward": 0.08333333730697631,
      "rewards/format_reward": 0.9833333492279053,
      "step": 1045
    },
    {
      "completion_length": 244.09584121704103,
      "epoch": 0.42,
      "grad_norm": 0.5444784393227498,
      "kl": 0.4791015625,
      "learning_rate": 1.4383711467890776e-05,
      "loss": 0.0192,
      "reward": 1.079166704416275,
      "reward_std": 0.2215237647294998,
      "rewards/accuracy_reward": 0.12916667126119136,
      "rewards/format_reward": 0.950000011920929,
      "step": 1050
    },
    {
      "completion_length": 177.3416732788086,
      "epoch": 0.422,
      "grad_norm": 0.9397555923299521,
      "kl": 0.5013671875,
      "learning_rate": 1.4320857488019826e-05,
      "loss": 0.0201,
      "reward": 1.0208333730697632,
      "reward_std": 0.14698250442743302,
      "rewards/accuracy_reward": 0.05416666716337204,
      "rewards/format_reward": 0.9666666805744171,
      "step": 1055
    },
    {
      "completion_length": 173.94583740234376,
      "epoch": 0.424,
      "grad_norm": 0.7522880163840423,
      "kl": 0.4654296875,
      "learning_rate": 1.4257792915650728e-05,
      "loss": 0.0186,
      "reward": 1.0541667103767396,
      "reward_std": 0.12657008767127992,
      "rewards/accuracy_reward": 0.07500000298023224,
      "rewards/format_reward": 0.9791666746139527,
      "step": 1060
    },
    {
      "completion_length": 231.49584045410157,
      "epoch": 0.426,
      "grad_norm": 0.1859104138975983,
      "kl": 0.47578125,
      "learning_rate": 1.4194520824461773e-05,
      "loss": 0.019,
      "reward": 1.0458333492279053,
      "reward_std": 0.16251266971230507,
      "rewards/accuracy_reward": 0.0750000026077032,
      "rewards/format_reward": 0.9708333551883698,
      "step": 1065
    },
    {
      "completion_length": 247.60417709350585,
      "epoch": 0.428,
      "grad_norm": 0.6117317081335039,
      "kl": 0.447265625,
      "learning_rate": 1.413104429824542e-05,
      "loss": 0.0179,
      "reward": 1.0250000417232514,
      "reward_std": 0.16218162104487419,
      "rewards/accuracy_reward": 0.05833333507180214,
      "rewards/format_reward": 0.9666666805744171,
      "step": 1070
    },
    {
      "completion_length": 171.2833366394043,
      "epoch": 0.43,
      "grad_norm": 0.9226683920061354,
      "kl": 0.4724609375,
      "learning_rate": 1.4067366430758004e-05,
      "loss": 0.0189,
      "reward": 1.0708333730697632,
      "reward_std": 0.20978599190711975,
      "rewards/accuracy_reward": 0.0916666679084301,
      "rewards/format_reward": 0.9791666865348816,
      "step": 1075
    },
    {
      "completion_length": 251.2166763305664,
      "epoch": 0.432,
      "grad_norm": 0.6333366911834787,
      "kl": 0.42578125,
      "learning_rate": 1.4003490325568953e-05,
      "loss": 0.017,
      "reward": 1.0416666865348816,
      "reward_std": 0.17009865194559098,
      "rewards/accuracy_reward": 0.08333333693444729,
      "rewards/format_reward": 0.9583333492279053,
      "step": 1080
    },
    {
      "completion_length": 269.02083892822264,
      "epoch": 0.434,
      "grad_norm": 0.8555804437546749,
      "kl": 0.4380859375,
      "learning_rate": 1.3939419095909513e-05,
      "loss": 0.0175,
      "reward": 1.0250000476837158,
      "reward_std": 0.22021322846412658,
      "rewards/accuracy_reward": 0.08750000186264514,
      "rewards/format_reward": 0.9375000178813935,
      "step": 1085
    },
    {
      "completion_length": 165.95000534057618,
      "epoch": 0.436,
      "grad_norm": 0.9717530137454518,
      "kl": 0.4203125,
      "learning_rate": 1.3875155864521031e-05,
      "loss": 0.0168,
      "reward": 1.1333333730697632,
      "reward_std": 0.18290041536092758,
      "rewards/accuracy_reward": 0.15416667014360427,
      "rewards/format_reward": 0.9791666865348816,
      "step": 1090
    },
    {
      "completion_length": 207.62084197998047,
      "epoch": 0.438,
      "grad_norm": 0.5696677734414208,
      "kl": 0.476953125,
      "learning_rate": 1.3810703763502744e-05,
      "loss": 0.0191,
      "reward": 1.0333333611488342,
      "reward_std": 0.13760571405291558,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.9708333492279053,
      "step": 1095
    },
    {
      "completion_length": 214.5041702270508,
      "epoch": 0.44,
      "grad_norm": 0.2890840425623631,
      "kl": 0.4705078125,
      "learning_rate": 1.3746065934159123e-05,
      "loss": 0.0188,
      "reward": 1.0750000178813934,
      "reward_std": 0.26415714919567107,
      "rewards/accuracy_reward": 0.12083333656191826,
      "rewards/format_reward": 0.9541666805744171,
      "step": 1100
    },
    {
      "epoch": 0.44,
      "eval_completion_length": 180.66506958007812,
      "eval_kl": 0.5162259615384616,
      "eval_loss": 0.020862072706222534,
      "eval_reward": 1.09775644999284,
      "eval_reward_std": 0.16826987896974271,
      "eval_rewards/accuracy_reward": 0.12500000472825307,
      "eval_rewards/format_reward": 0.9727564316529494,
      "eval_runtime": 203.1184,
      "eval_samples_per_second": 0.487,
      "eval_steps_per_second": 0.015,
      "step": 1100
    },
    {
      "completion_length": 163.7708396911621,
      "epoch": 0.442,
      "grad_norm": 0.777201350447929,
      "kl": 0.48671875,
      "learning_rate": 1.3681245526846782e-05,
      "loss": 0.0195,
      "reward": 1.0583333551883698,
      "reward_std": 0.16357700973749162,
      "rewards/accuracy_reward": 0.07916666902601718,
      "rewards/format_reward": 0.9791666746139527,
      "step": 1105
    },
    {
      "completion_length": 219.01250305175782,
      "epoch": 0.444,
      "grad_norm": 0.8993155700071122,
      "kl": 0.503125,
      "learning_rate": 1.3616245700820922e-05,
      "loss": 0.0201,
      "reward": 1.1000000298023225,
      "reward_std": 0.26238192319869996,
      "rewards/accuracy_reward": 0.1541666705161333,
      "rewards/format_reward": 0.9458333551883698,
      "step": 1110
    },
    {
      "completion_length": 134.5833396911621,
      "epoch": 0.446,
      "grad_norm": 1.3416395741871388,
      "kl": 0.648046875,
      "learning_rate": 1.3551069624081372e-05,
      "loss": 0.0259,
      "reward": 1.0333333790302277,
      "reward_std": 0.2724094092845917,
      "rewards/accuracy_reward": 0.11250000335276127,
      "rewards/format_reward": 0.9208333551883697,
      "step": 1115
    },
    {
      "completion_length": 143.10417251586915,
      "epoch": 0.448,
      "grad_norm": 0.6893211262244785,
      "kl": 0.4814453125,
      "learning_rate": 1.3485720473218153e-05,
      "loss": 0.0192,
      "reward": 1.041666680574417,
      "reward_std": 0.28478238806128503,
      "rewards/accuracy_reward": 0.1208333358168602,
      "rewards/format_reward": 0.9208333492279053,
      "step": 1120
    },
    {
      "completion_length": 178.47500534057616,
      "epoch": 0.45,
      "grad_norm": 0.750719996343038,
      "kl": 0.4640625,
      "learning_rate": 1.342020143325669e-05,
      "loss": 0.0186,
      "reward": 1.062500035762787,
      "reward_std": 0.20043605640530587,
      "rewards/accuracy_reward": 0.08333333618938923,
      "rewards/format_reward": 0.9791666865348816,
      "step": 1125
    },
    {
      "completion_length": 194.51250457763672,
      "epoch": 0.452,
      "grad_norm": 0.9250247396623772,
      "kl": 0.5109375,
      "learning_rate": 1.3354515697502552e-05,
      "loss": 0.0204,
      "reward": 1.012500035762787,
      "reward_std": 0.11767575666308402,
      "rewards/accuracy_reward": 0.03333333432674408,
      "rewards/format_reward": 0.9791666805744171,
      "step": 1130
    },
    {
      "completion_length": 134.37500457763673,
      "epoch": 0.454,
      "grad_norm": 0.4184828078418494,
      "kl": 0.5029296875,
      "learning_rate": 1.3288666467385834e-05,
      "loss": 0.0201,
      "reward": 1.1083333849906922,
      "reward_std": 0.15370185375213624,
      "rewards/accuracy_reward": 0.11250000186264515,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1135
    },
    {
      "completion_length": 141.8625045776367,
      "epoch": 0.456,
      "grad_norm": 1.540000600383148,
      "kl": 0.534765625,
      "learning_rate": 1.3222656952305113e-05,
      "loss": 0.0214,
      "reward": 1.104166704416275,
      "reward_std": 0.21041805893182755,
      "rewards/accuracy_reward": 0.12083333730697632,
      "rewards/format_reward": 0.9833333373069764,
      "step": 1140
    },
    {
      "completion_length": 103.26250305175782,
      "epoch": 0.458,
      "grad_norm": 0.9729389527580656,
      "kl": 0.4873046875,
      "learning_rate": 1.3156490369471026e-05,
      "loss": 0.0195,
      "reward": 1.1083333969116211,
      "reward_std": 0.1348556809127331,
      "rewards/accuracy_reward": 0.10833333693444729,
      "rewards/format_reward": 1.0,
      "step": 1145
    },
    {
      "completion_length": 100.60416984558105,
      "epoch": 0.46,
      "grad_norm": 0.6226549877758246,
      "kl": 0.5244140625,
      "learning_rate": 1.3090169943749475e-05,
      "loss": 0.021,
      "reward": 1.0666666865348815,
      "reward_std": 0.11078203096985817,
      "rewards/accuracy_reward": 0.07083333507180214,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1150
    },
    {
      "completion_length": 125.1916717529297,
      "epoch": 0.462,
      "grad_norm": 0.7322743259128145,
      "kl": 0.490234375,
      "learning_rate": 1.3023698907504447e-05,
      "loss": 0.0196,
      "reward": 1.0458333551883698,
      "reward_std": 0.1147707849740982,
      "rewards/accuracy_reward": 0.05833333432674408,
      "rewards/format_reward": 0.9875000059604645,
      "step": 1155
    },
    {
      "completion_length": 171.20000762939452,
      "epoch": 0.464,
      "grad_norm": 0.8661955941081932,
      "kl": 0.433203125,
      "learning_rate": 1.2957080500440469e-05,
      "loss": 0.0173,
      "reward": 1.1125000417232513,
      "reward_std": 0.1469825029373169,
      "rewards/accuracy_reward": 0.12083333656191826,
      "rewards/format_reward": 0.9916666686534882,
      "step": 1160
    },
    {
      "completion_length": 363.72084655761716,
      "epoch": 0.466,
      "grad_norm": 0.5690418918424673,
      "kl": 0.3953125,
      "learning_rate": 1.2890317969444716e-05,
      "loss": 0.0158,
      "reward": 0.9916666865348815,
      "reward_std": 0.17782215774059296,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1165
    },
    {
      "completion_length": 340.24583892822267,
      "epoch": 0.468,
      "grad_norm": 0.5868426291161072,
      "kl": 0.3611328125,
      "learning_rate": 1.2823414568428767e-05,
      "loss": 0.0144,
      "reward": 1.041666704416275,
      "reward_std": 0.1920206516981125,
      "rewards/accuracy_reward": 0.0708333358168602,
      "rewards/format_reward": 0.9708333492279053,
      "step": 1170
    },
    {
      "completion_length": 287.6666717529297,
      "epoch": 0.47,
      "grad_norm": 0.7021811883073036,
      "kl": 0.337109375,
      "learning_rate": 1.2756373558169992e-05,
      "loss": 0.0135,
      "reward": 1.045833373069763,
      "reward_std": 0.13058570101857186,
      "rewards/accuracy_reward": 0.05416666828095913,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1175
    },
    {
      "completion_length": 186.4791717529297,
      "epoch": 0.472,
      "grad_norm": 0.6684350780126657,
      "kl": 0.3828125,
      "learning_rate": 1.2689198206152657e-05,
      "loss": 0.0153,
      "reward": 1.0708333730697632,
      "reward_std": 0.11366014555096626,
      "rewards/accuracy_reward": 0.07500000223517418,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1180
    },
    {
      "completion_length": 177.3375045776367,
      "epoch": 0.474,
      "grad_norm": 0.6147456691558177,
      "kl": 0.3947265625,
      "learning_rate": 1.2621891786408648e-05,
      "loss": 0.0158,
      "reward": 1.0958333611488342,
      "reward_std": 0.14349564239382745,
      "rewards/accuracy_reward": 0.0958333346992731,
      "rewards/format_reward": 1.0,
      "step": 1185
    },
    {
      "completion_length": 179.8791702270508,
      "epoch": 0.476,
      "grad_norm": 0.7097100700508777,
      "kl": 0.423046875,
      "learning_rate": 1.2554457579357906e-05,
      "loss": 0.0169,
      "reward": 1.0791666984558106,
      "reward_std": 0.16390806213021278,
      "rewards/accuracy_reward": 0.08333333544433116,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1190
    },
    {
      "completion_length": 175.40417098999023,
      "epoch": 0.478,
      "grad_norm": 0.3216081776765636,
      "kl": 0.401953125,
      "learning_rate": 1.2486898871648552e-05,
      "loss": 0.0161,
      "reward": 1.0625000476837159,
      "reward_std": 0.1101732850074768,
      "rewards/accuracy_reward": 0.06250000223517418,
      "rewards/format_reward": 1.0,
      "step": 1195
    },
    {
      "completion_length": 195.07500457763672,
      "epoch": 0.48,
      "grad_norm": 0.8759106701000279,
      "kl": 0.4146484375,
      "learning_rate": 1.2419218955996677e-05,
      "loss": 0.0166,
      "reward": 1.075000035762787,
      "reward_std": 0.10886141434311866,
      "rewards/accuracy_reward": 0.07916666977107525,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1200
    },
    {
      "epoch": 0.48,
      "eval_completion_length": 196.80609365609976,
      "eval_kl": 0.4260817307692308,
      "eval_loss": 0.017165007069706917,
      "eval_reward": 1.115384647479424,
      "eval_reward_std": 0.14654561418753403,
      "eval_rewards/accuracy_reward": 0.12339744000480725,
      "eval_rewards/format_reward": 0.9919871871288006,
      "eval_runtime": 120.029,
      "eval_samples_per_second": 0.825,
      "eval_steps_per_second": 0.025,
      "step": 1200
    },
    {
      "completion_length": 206.79583892822265,
      "epoch": 0.482,
      "grad_norm": 0.5225537669303568,
      "kl": 0.396875,
      "learning_rate": 1.23514211310259e-05,
      "loss": 0.0159,
      "reward": 1.075000035762787,
      "reward_std": 0.12177135646343232,
      "rewards/accuracy_reward": 0.0750000026077032,
      "rewards/format_reward": 1.0,
      "step": 1205
    },
    {
      "completion_length": 206.87500457763673,
      "epoch": 0.484,
      "grad_norm": 0.344380593464978,
      "kl": 0.4521484375,
      "learning_rate": 1.2283508701106559e-05,
      "loss": 0.0181,
      "reward": 1.125000035762787,
      "reward_std": 0.17358550876379014,
      "rewards/accuracy_reward": 0.12500000260770322,
      "rewards/format_reward": 1.0,
      "step": 1210
    },
    {
      "completion_length": 290.5125076293945,
      "epoch": 0.486,
      "grad_norm": 0.47708923896384847,
      "kl": 0.4142578125,
      "learning_rate": 1.2215484976194675e-05,
      "loss": 0.0166,
      "reward": 1.0833333611488343,
      "reward_std": 0.14718020632863044,
      "rewards/accuracy_reward": 0.09583333618938923,
      "rewards/format_reward": 0.9875000059604645,
      "step": 1215
    },
    {
      "completion_length": 262.9541748046875,
      "epoch": 0.488,
      "grad_norm": 0.7337765212074902,
      "kl": 0.5099609375,
      "learning_rate": 1.2147353271670634e-05,
      "loss": 0.0204,
      "reward": 1.062500035762787,
      "reward_std": 0.2556539088487625,
      "rewards/accuracy_reward": 0.10000000260770321,
      "rewards/format_reward": 0.962500023841858,
      "step": 1220
    },
    {
      "completion_length": 249.69584045410156,
      "epoch": 0.49,
      "grad_norm": 0.7659087296399365,
      "kl": 0.4314453125,
      "learning_rate": 1.2079116908177592e-05,
      "loss": 0.0173,
      "reward": 1.0708333611488343,
      "reward_std": 0.1630980782210827,
      "rewards/accuracy_reward": 0.08333333618938923,
      "rewards/format_reward": 0.987500011920929,
      "step": 1225
    },
    {
      "completion_length": 171.30417022705078,
      "epoch": 0.492,
      "grad_norm": 0.5881563006486139,
      "kl": 0.4369140625,
      "learning_rate": 1.2010779211459649e-05,
      "loss": 0.0175,
      "reward": 1.1333333611488343,
      "reward_std": 0.19801353365182878,
      "rewards/accuracy_reward": 0.1541666716337204,
      "rewards/format_reward": 0.9791666805744171,
      "step": 1230
    },
    {
      "completion_length": 215.20000457763672,
      "epoch": 0.494,
      "grad_norm": 1.0090808424345548,
      "kl": 0.4369140625,
      "learning_rate": 1.194234351219972e-05,
      "loss": 0.0175,
      "reward": 1.0833333611488343,
      "reward_std": 0.23648979663848876,
      "rewards/accuracy_reward": 0.12500000335276126,
      "rewards/format_reward": 0.9583333492279053,
      "step": 1235
    },
    {
      "completion_length": 186.05417404174804,
      "epoch": 0.496,
      "grad_norm": 0.79230261520844,
      "kl": 0.419140625,
      "learning_rate": 1.187381314585725e-05,
      "loss": 0.0168,
      "reward": 1.137500035762787,
      "reward_std": 0.21822471916675568,
      "rewards/accuracy_reward": 0.1541666690260172,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1240
    },
    {
      "completion_length": 170.16250457763672,
      "epoch": 0.498,
      "grad_norm": 0.6709583305642091,
      "kl": 0.455859375,
      "learning_rate": 1.1805191452505602e-05,
      "loss": 0.0182,
      "reward": 1.100000023841858,
      "reward_std": 0.1810879833996296,
      "rewards/accuracy_reward": 0.1250000014901161,
      "rewards/format_reward": 0.975000011920929,
      "step": 1245
    },
    {
      "completion_length": 128.26667175292968,
      "epoch": 0.5,
      "grad_norm": 0.8833573142079981,
      "kl": 0.469921875,
      "learning_rate": 1.1736481776669307e-05,
      "loss": 0.0188,
      "reward": 1.1166667222976685,
      "reward_std": 0.14968623965978622,
      "rewards/accuracy_reward": 0.1166666705161333,
      "rewards/format_reward": 1.0,
      "step": 1250
    },
    {
      "completion_length": 162.34583892822266,
      "epoch": 0.502,
      "grad_norm": 0.3294796625934881,
      "kl": 0.461328125,
      "learning_rate": 1.1667687467161025e-05,
      "loss": 0.0185,
      "reward": 1.1416667222976684,
      "reward_std": 0.20132601708173753,
      "rewards/accuracy_reward": 0.1500000063329935,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1255
    },
    {
      "completion_length": 207.9791732788086,
      "epoch": 0.504,
      "grad_norm": 0.7088847090687599,
      "kl": 0.490234375,
      "learning_rate": 1.159881187691835e-05,
      "loss": 0.0196,
      "reward": 1.087500023841858,
      "reward_std": 0.21892432272434234,
      "rewards/accuracy_reward": 0.11250000223517417,
      "rewards/format_reward": 0.9750000238418579,
      "step": 1260
    },
    {
      "completion_length": 264.69584426879885,
      "epoch": 0.506,
      "grad_norm": 1.0708969943160458,
      "kl": 0.467578125,
      "learning_rate": 1.1529858362840383e-05,
      "loss": 0.0187,
      "reward": 1.062500035762787,
      "reward_std": 0.22483758330345155,
      "rewards/accuracy_reward": 0.10416666828095913,
      "rewards/format_reward": 0.9583333492279053,
      "step": 1265
    },
    {
      "completion_length": 232.33334197998047,
      "epoch": 0.508,
      "grad_norm": 0.6585084882308617,
      "kl": 0.437109375,
      "learning_rate": 1.1460830285624119e-05,
      "loss": 0.0175,
      "reward": 1.075000023841858,
      "reward_std": 0.22963632121682168,
      "rewards/accuracy_reward": 0.10833333693444729,
      "rewards/format_reward": 0.9666666805744171,
      "step": 1270
    },
    {
      "completion_length": 194.81666870117186,
      "epoch": 0.51,
      "grad_norm": 0.5578883725556453,
      "kl": 0.453125,
      "learning_rate": 1.1391731009600655e-05,
      "loss": 0.0181,
      "reward": 1.0458333611488342,
      "reward_std": 0.13407255634665488,
      "rewards/accuracy_reward": 0.06250000186264515,
      "rewards/format_reward": 0.9833333492279053,
      "step": 1275
    },
    {
      "completion_length": 140.2375030517578,
      "epoch": 0.512,
      "grad_norm": 0.9319950337687952,
      "kl": 0.51640625,
      "learning_rate": 1.1322563902571227e-05,
      "loss": 0.0207,
      "reward": 1.100000023841858,
      "reward_std": 0.11636388301849365,
      "rewards/accuracy_reward": 0.10000000074505806,
      "rewards/format_reward": 1.0,
      "step": 1280
    },
    {
      "completion_length": 117.49583740234375,
      "epoch": 0.514,
      "grad_norm": 0.6424020462953824,
      "kl": 0.5283203125,
      "learning_rate": 1.1253332335643043e-05,
      "loss": 0.0211,
      "reward": 1.0666667222976685,
      "reward_std": 0.12386635020375251,
      "rewards/accuracy_reward": 0.07083333469927311,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1285
    },
    {
      "completion_length": 160.48750686645508,
      "epoch": 0.516,
      "grad_norm": 0.09722259646016615,
      "kl": 0.529296875,
      "learning_rate": 1.1184039683065014e-05,
      "loss": 0.0212,
      "reward": 1.037500023841858,
      "reward_std": 0.05643851235508919,
      "rewards/accuracy_reward": 0.0416666679084301,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1290
    },
    {
      "completion_length": 180.3333366394043,
      "epoch": 0.518,
      "grad_norm": 1.4174701949018742,
      "kl": 0.5001953125,
      "learning_rate": 1.1114689322063255e-05,
      "loss": 0.02,
      "reward": 1.0708333730697632,
      "reward_std": 0.12866508215665817,
      "rewards/accuracy_reward": 0.08333333544433116,
      "rewards/format_reward": 0.987500011920929,
      "step": 1295
    },
    {
      "completion_length": 233.85000915527343,
      "epoch": 0.52,
      "grad_norm": 0.7535966288197256,
      "kl": 0.48671875,
      "learning_rate": 1.1045284632676535e-05,
      "loss": 0.0194,
      "reward": 1.066666692495346,
      "reward_std": 0.23691422641277313,
      "rewards/accuracy_reward": 0.1125000048428774,
      "rewards/format_reward": 0.9541666805744171,
      "step": 1300
    },
    {
      "epoch": 0.52,
      "eval_completion_length": 262.862191420335,
      "eval_kl": 0.47250600961538464,
      "eval_loss": 0.019559284672141075,
      "eval_reward": 1.0528846520643969,
      "eval_reward_std": 0.21832889031905395,
      "eval_rewards/accuracy_reward": 0.10096154218682876,
      "eval_rewards/format_reward": 0.951923090677995,
      "eval_runtime": 222.8563,
      "eval_samples_per_second": 0.444,
      "eval_steps_per_second": 0.013,
      "step": 1300
    },
    {
      "completion_length": 204.50000762939453,
      "epoch": 0.522,
      "grad_norm": 0.582966083499623,
      "kl": 0.4453125,
      "learning_rate": 1.0975828997591496e-05,
      "loss": 0.0178,
      "reward": 1.0583333730697633,
      "reward_std": 0.1721936509013176,
      "rewards/accuracy_reward": 0.07916666753590107,
      "rewards/format_reward": 0.9791666805744171,
      "step": 1305
    },
    {
      "completion_length": 184.75417098999023,
      "epoch": 0.524,
      "grad_norm": 0.7734702622967983,
      "kl": 0.4318359375,
      "learning_rate": 1.0906325801977804e-05,
      "loss": 0.0173,
      "reward": 1.0625000536441802,
      "reward_std": 0.2130418173968792,
      "rewards/accuracy_reward": 0.0916666690260172,
      "rewards/format_reward": 0.9708333432674408,
      "step": 1310
    },
    {
      "completion_length": 170.43750457763673,
      "epoch": 0.526,
      "grad_norm": 0.8051360088361499,
      "kl": 0.4501953125,
      "learning_rate": 1.083677843332316e-05,
      "loss": 0.018,
      "reward": 1.0791667103767395,
      "reward_std": 0.1652999296784401,
      "rewards/accuracy_reward": 0.10000000260770321,
      "rewards/format_reward": 0.9791666805744171,
      "step": 1315
    },
    {
      "completion_length": 174.08750686645507,
      "epoch": 0.528,
      "grad_norm": 1.1111528717015793,
      "kl": 0.45625,
      "learning_rate": 1.0767190281268187e-05,
      "loss": 0.0183,
      "reward": 1.0208333671092986,
      "reward_std": 0.14639708772301674,
      "rewards/accuracy_reward": 0.05000000260770321,
      "rewards/format_reward": 0.9708333492279053,
      "step": 1320
    },
    {
      "completion_length": 200.2291717529297,
      "epoch": 0.53,
      "grad_norm": 0.8268417258688071,
      "kl": 0.4689453125,
      "learning_rate": 1.0697564737441254e-05,
      "loss": 0.0188,
      "reward": 1.1083333611488342,
      "reward_std": 0.1567763350903988,
      "rewards/accuracy_reward": 0.1333333369344473,
      "rewards/format_reward": 0.9750000178813935,
      "step": 1325
    },
    {
      "completion_length": 165.32083740234376,
      "epoch": 0.532,
      "grad_norm": 0.2358374954563172,
      "kl": 0.4279296875,
      "learning_rate": 1.0627905195293135e-05,
      "loss": 0.0171,
      "reward": 1.129166704416275,
      "reward_std": 0.1762892484664917,
      "rewards/accuracy_reward": 0.14583333693444728,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1330
    },
    {
      "completion_length": 213.89167251586915,
      "epoch": 0.534,
      "grad_norm": 0.7895004407616136,
      "kl": 0.3943359375,
      "learning_rate": 1.055821504993164e-05,
      "loss": 0.0158,
      "reward": 1.025000023841858,
      "reward_std": 0.17843054831027985,
      "rewards/accuracy_reward": 0.05833333432674408,
      "rewards/format_reward": 0.9666666865348816,
      "step": 1335
    },
    {
      "completion_length": 142.70000381469725,
      "epoch": 0.536,
      "grad_norm": 0.8360376478700801,
      "kl": 0.432421875,
      "learning_rate": 1.0488497697956134e-05,
      "loss": 0.0173,
      "reward": 1.0916667044162751,
      "reward_std": 0.15858057737350464,
      "rewards/accuracy_reward": 0.10833333656191826,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1340
    },
    {
      "completion_length": 133.95417022705078,
      "epoch": 0.538,
      "grad_norm": 0.7793211768536685,
      "kl": 0.434765625,
      "learning_rate": 1.0418756537291996e-05,
      "loss": 0.0174,
      "reward": 1.095833384990692,
      "reward_std": 0.17280238792300223,
      "rewards/accuracy_reward": 0.11250000521540642,
      "rewards/format_reward": 0.9833333492279053,
      "step": 1345
    },
    {
      "completion_length": 123.75417022705078,
      "epoch": 0.54,
      "grad_norm": 0.9625905804344617,
      "kl": 0.43828125,
      "learning_rate": 1.0348994967025012e-05,
      "loss": 0.0175,
      "reward": 1.125000035762787,
      "reward_std": 0.18591004833579064,
      "rewards/accuracy_reward": 0.13333333507180214,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1350
    },
    {
      "completion_length": 166.07500762939452,
      "epoch": 0.542,
      "grad_norm": 0.19769259077551565,
      "kl": 0.444921875,
      "learning_rate": 1.0279216387235691e-05,
      "loss": 0.0178,
      "reward": 1.029166692495346,
      "reward_std": 0.11366014182567596,
      "rewards/accuracy_reward": 0.05416666828095913,
      "rewards/format_reward": 0.975000011920929,
      "step": 1355
    },
    {
      "completion_length": 174.76250457763672,
      "epoch": 0.544,
      "grad_norm": 0.8083428884974917,
      "kl": 0.395703125,
      "learning_rate": 1.0209424198833571e-05,
      "loss": 0.0158,
      "reward": 1.075000023841858,
      "reward_std": 0.16870326176285744,
      "rewards/accuracy_reward": 0.09583333618938923,
      "rewards/format_reward": 0.9791666865348816,
      "step": 1360
    },
    {
      "completion_length": 231.99167251586914,
      "epoch": 0.546,
      "grad_norm": 0.3581588542986591,
      "kl": 0.419140625,
      "learning_rate": 1.0139621803391454e-05,
      "loss": 0.0168,
      "reward": 1.1416667103767395,
      "reward_std": 0.21144871562719345,
      "rewards/accuracy_reward": 0.1791666727513075,
      "rewards/format_reward": 0.9625000119209289,
      "step": 1365
    },
    {
      "completion_length": 184.47084121704103,
      "epoch": 0.548,
      "grad_norm": 0.7966076879518822,
      "kl": 0.4314453125,
      "learning_rate": 1.0069812602979617e-05,
      "loss": 0.0173,
      "reward": 1.0208333790302277,
      "reward_std": 0.1700151339173317,
      "rewards/accuracy_reward": 0.05000000149011612,
      "rewards/format_reward": 0.9708333492279053,
      "step": 1370
    },
    {
      "completion_length": 129.3000045776367,
      "epoch": 0.55,
      "grad_norm": 0.8350147510340598,
      "kl": 0.444921875,
      "learning_rate": 1e-05,
      "loss": 0.0178,
      "reward": 1.1250000238418578,
      "reward_std": 0.17009866163134574,
      "rewards/accuracy_reward": 0.1291666690260172,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1375
    },
    {
      "completion_length": 128.46250381469727,
      "epoch": 0.552,
      "grad_norm": 0.40110881221719935,
      "kl": 0.391796875,
      "learning_rate": 9.930187397020385e-06,
      "loss": 0.0157,
      "reward": 1.0666666984558106,
      "reward_std": 0.09804645925760269,
      "rewards/accuracy_reward": 0.06666666828095913,
      "rewards/format_reward": 1.0,
      "step": 1380
    },
    {
      "completion_length": 148.71250610351564,
      "epoch": 0.554,
      "grad_norm": 0.8104207439007759,
      "kl": 0.4185546875,
      "learning_rate": 9.860378196608549e-06,
      "loss": 0.0167,
      "reward": 1.1000000357627868,
      "reward_std": 0.16956989839673042,
      "rewards/accuracy_reward": 0.10416666902601719,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1385
    },
    {
      "completion_length": 138.62917251586913,
      "epoch": 0.556,
      "grad_norm": 0.8861813782686773,
      "kl": 0.404296875,
      "learning_rate": 9.790575801166432e-06,
      "loss": 0.0162,
      "reward": 1.0875000476837158,
      "reward_std": 0.13599317520856857,
      "rewards/accuracy_reward": 0.09583333730697632,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1390
    },
    {
      "completion_length": 129.62500534057617,
      "epoch": 0.558,
      "grad_norm": 0.7959816358964777,
      "kl": 0.3986328125,
      "learning_rate": 9.720783612764314e-06,
      "loss": 0.0159,
      "reward": 1.1583333730697631,
      "reward_std": 0.14700583443045617,
      "rewards/accuracy_reward": 0.16250000670552253,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1395
    },
    {
      "completion_length": 140.90833587646483,
      "epoch": 0.56,
      "grad_norm": 1.3018520628065517,
      "kl": 0.44375,
      "learning_rate": 9.651005032974994e-06,
      "loss": 0.0178,
      "reward": 1.0708333611488343,
      "reward_std": 0.13808816894888878,
      "rewards/accuracy_reward": 0.08750000149011612,
      "rewards/format_reward": 0.9833333492279053,
      "step": 1400
    },
    {
      "epoch": 0.56,
      "eval_completion_length": 139.72115854116586,
      "eval_kl": 0.42322716346153844,
      "eval_loss": 0.018276819959282875,
      "eval_reward": 1.118589768042931,
      "eval_reward_std": 0.17192490513508135,
      "eval_rewards/accuracy_reward": 0.1314102586072225,
      "eval_rewards/format_reward": 0.987179499406081,
      "eval_runtime": 160.9908,
      "eval_samples_per_second": 0.615,
      "eval_steps_per_second": 0.019,
      "step": 1400
    },
    {
      "completion_length": 120.7958366394043,
      "epoch": 0.562,
      "grad_norm": 0.7447224368194102,
      "kl": 0.4337890625,
      "learning_rate": 9.581243462708007e-06,
      "loss": 0.0174,
      "reward": 1.0791666984558106,
      "reward_std": 0.10075020045042038,
      "rewards/accuracy_reward": 0.0791666679084301,
      "rewards/format_reward": 1.0,
      "step": 1405
    },
    {
      "completion_length": 123.17500381469726,
      "epoch": 0.564,
      "grad_norm": 0.6943691730102636,
      "kl": 0.4384765625,
      "learning_rate": 9.511502302043867e-06,
      "loss": 0.0175,
      "reward": 1.0791667103767395,
      "reward_std": 0.16896116137504577,
      "rewards/accuracy_reward": 0.09166666865348816,
      "rewards/format_reward": 0.987500011920929,
      "step": 1410
    },
    {
      "completion_length": 144.02500381469727,
      "epoch": 0.566,
      "grad_norm": 0.7135215199384104,
      "kl": 0.435546875,
      "learning_rate": 9.441784950068362e-06,
      "loss": 0.0174,
      "reward": 1.0666666984558106,
      "reward_std": 0.11416204124689103,
      "rewards/accuracy_reward": 0.07500000111758709,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1415
    },
    {
      "completion_length": 157.21250534057617,
      "epoch": 0.568,
      "grad_norm": 1.0565897923645307,
      "kl": 0.4328125,
      "learning_rate": 9.372094804706867e-06,
      "loss": 0.0173,
      "reward": 1.0625000476837159,
      "reward_std": 0.13348714262247086,
      "rewards/accuracy_reward": 0.0750000026077032,
      "rewards/format_reward": 0.987500011920929,
      "step": 1420
    },
    {
      "completion_length": 140.8791717529297,
      "epoch": 0.57,
      "grad_norm": 1.0299599122671386,
      "kl": 0.5,
      "learning_rate": 9.302435262558748e-06,
      "loss": 0.02,
      "reward": 1.1166666924953461,
      "reward_std": 0.2273202881217003,
      "rewards/accuracy_reward": 0.14166666977107525,
      "rewards/format_reward": 0.9750000178813935,
      "step": 1425
    },
    {
      "completion_length": 180.52083892822264,
      "epoch": 0.572,
      "grad_norm": 0.7151881659016156,
      "kl": 0.4361328125,
      "learning_rate": 9.232809718731815e-06,
      "loss": 0.0174,
      "reward": 1.091666692495346,
      "reward_std": 0.15498905703425409,
      "rewards/accuracy_reward": 0.11250000074505806,
      "rewards/format_reward": 0.9791666746139527,
      "step": 1430
    },
    {
      "completion_length": 192.38750686645508,
      "epoch": 0.574,
      "grad_norm": 0.629276808623102,
      "kl": 0.4154296875,
      "learning_rate": 9.163221566676847e-06,
      "loss": 0.0166,
      "reward": 1.0916667103767395,
      "reward_std": 0.16951324194669723,
      "rewards/accuracy_reward": 0.1125000048428774,
      "rewards/format_reward": 0.9791666865348816,
      "step": 1435
    },
    {
      "completion_length": 197.38333587646486,
      "epoch": 0.576,
      "grad_norm": 0.8364125432332793,
      "kl": 0.4513671875,
      "learning_rate": 9.093674198022201e-06,
      "loss": 0.0181,
      "reward": 1.108333373069763,
      "reward_std": 0.17300009950995446,
      "rewards/accuracy_reward": 0.12083333842456341,
      "rewards/format_reward": 0.9875000059604645,
      "step": 1440
    },
    {
      "completion_length": 309.7000137329102,
      "epoch": 0.578,
      "grad_norm": 0.596028075225735,
      "kl": 0.4134765625,
      "learning_rate": 9.024171002408507e-06,
      "loss": 0.0165,
      "reward": 1.125000035762787,
      "reward_std": 0.18778266608715058,
      "rewards/accuracy_reward": 0.1416666701436043,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1445
    },
    {
      "completion_length": 327.8416778564453,
      "epoch": 0.58,
      "grad_norm": 0.5434179981843781,
      "kl": 0.403515625,
      "learning_rate": 8.954715367323468e-06,
      "loss": 0.0161,
      "reward": 1.0708333730697632,
      "reward_std": 0.2297428011894226,
      "rewards/accuracy_reward": 0.10833333507180214,
      "rewards/format_reward": 0.962500023841858,
      "step": 1450
    },
    {
      "completion_length": 294.1958435058594,
      "epoch": 0.582,
      "grad_norm": 0.5416204560922091,
      "kl": 0.355078125,
      "learning_rate": 8.885310677936746e-06,
      "loss": 0.0142,
      "reward": 1.112500047683716,
      "reward_std": 0.18459817096590997,
      "rewards/accuracy_reward": 0.11666666939854622,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1455
    },
    {
      "completion_length": 270.0791732788086,
      "epoch": 0.584,
      "grad_norm": 0.6575350557585025,
      "kl": 0.38203125,
      "learning_rate": 8.815960316934991e-06,
      "loss": 0.0153,
      "reward": 1.075000035762787,
      "reward_std": 0.1396777369081974,
      "rewards/accuracy_reward": 0.1000000037252903,
      "rewards/format_reward": 0.9750000178813935,
      "step": 1460
    },
    {
      "completion_length": 206.42917556762694,
      "epoch": 0.586,
      "grad_norm": 0.496919053076702,
      "kl": 0.3818359375,
      "learning_rate": 8.746667664356957e-06,
      "loss": 0.0153,
      "reward": 1.0708333611488343,
      "reward_std": 0.08033778220415115,
      "rewards/accuracy_reward": 0.07916666939854622,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1465
    },
    {
      "completion_length": 216.08333892822264,
      "epoch": 0.588,
      "grad_norm": 0.44350270218884014,
      "kl": 0.3541015625,
      "learning_rate": 8.677436097428775e-06,
      "loss": 0.0142,
      "reward": 1.0500000238418579,
      "reward_std": 0.13078340515494347,
      "rewards/accuracy_reward": 0.07083333469927311,
      "rewards/format_reward": 0.9791666805744171,
      "step": 1470
    },
    {
      "completion_length": 235.42917175292968,
      "epoch": 0.59,
      "grad_norm": 0.31256519639828234,
      "kl": 0.369140625,
      "learning_rate": 8.60826899039935e-06,
      "loss": 0.0148,
      "reward": 1.0958333730697631,
      "reward_std": 0.1650187000632286,
      "rewards/accuracy_reward": 0.11666666977107525,
      "rewards/format_reward": 0.9791666865348816,
      "step": 1475
    },
    {
      "completion_length": 233.67501068115234,
      "epoch": 0.592,
      "grad_norm": 0.3462078333529471,
      "kl": 0.3658203125,
      "learning_rate": 8.539169714375885e-06,
      "loss": 0.0146,
      "reward": 1.0666667103767395,
      "reward_std": 0.1788861334323883,
      "rewards/accuracy_reward": 0.09166667126119137,
      "rewards/format_reward": 0.9750000238418579,
      "step": 1480
    },
    {
      "completion_length": 278.04584045410155,
      "epoch": 0.594,
      "grad_norm": 0.4375001521877564,
      "kl": 0.3640625,
      "learning_rate": 8.47014163715962e-06,
      "loss": 0.0146,
      "reward": 1.062500035762787,
      "reward_std": 0.2277390219271183,
      "rewards/accuracy_reward": 0.10833333730697632,
      "rewards/format_reward": 0.9541666865348816,
      "step": 1485
    },
    {
      "completion_length": 227.55834045410157,
      "epoch": 0.596,
      "grad_norm": 0.4936381128270299,
      "kl": 0.384375,
      "learning_rate": 8.401188123081653e-06,
      "loss": 0.0154,
      "reward": 1.1166666746139526,
      "reward_std": 0.25464622527360914,
      "rewards/accuracy_reward": 0.15000000260770321,
      "rewards/format_reward": 0.9666666805744171,
      "step": 1490
    },
    {
      "completion_length": 211.10417251586915,
      "epoch": 0.598,
      "grad_norm": 0.8293632244296539,
      "kl": 0.4357421875,
      "learning_rate": 8.332312532838978e-06,
      "loss": 0.0174,
      "reward": 1.0833333611488343,
      "reward_std": 0.15178123712539673,
      "rewards/accuracy_reward": 0.10000000074505806,
      "rewards/format_reward": 0.9833333492279053,
      "step": 1495
    },
    {
      "completion_length": 235.98750534057618,
      "epoch": 0.6,
      "grad_norm": 0.7367990374860012,
      "kl": 0.376953125,
      "learning_rate": 8.263518223330698e-06,
      "loss": 0.0151,
      "reward": 1.0416666984558105,
      "reward_std": 0.2215484268963337,
      "rewards/accuracy_reward": 0.0833333346992731,
      "rewards/format_reward": 0.9583333492279053,
      "step": 1500
    },
    {
      "epoch": 0.6,
      "eval_completion_length": 184.29808279184195,
      "eval_kl": 0.38716947115384615,
      "eval_loss": 0.014998482540249825,
      "eval_reward": 1.107371830023252,
      "eval_reward_std": 0.20388395740435675,
      "eval_rewards/accuracy_reward": 0.13141025932362446,
      "eval_rewards/format_reward": 0.9759615522164565,
      "eval_runtime": 170.0138,
      "eval_samples_per_second": 0.582,
      "eval_steps_per_second": 0.018,
      "step": 1500
    },
    {
      "completion_length": 209.3416748046875,
      "epoch": 0.602,
      "grad_norm": 0.8772003587312531,
      "kl": 0.355859375,
      "learning_rate": 8.194808547494401e-06,
      "loss": 0.0142,
      "reward": 1.1166667103767396,
      "reward_std": 0.24232522323727607,
      "rewards/accuracy_reward": 0.1416666701436043,
      "rewards/format_reward": 0.9750000178813935,
      "step": 1505
    },
    {
      "completion_length": 188.06250610351563,
      "epoch": 0.604,
      "grad_norm": 0.4103081085850965,
      "kl": 0.36171875,
      "learning_rate": 8.126186854142752e-06,
      "loss": 0.0145,
      "reward": 1.1583333730697631,
      "reward_std": 0.18080894947052,
      "rewards/accuracy_reward": 0.16666667051613332,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1510
    },
    {
      "completion_length": 230.85834197998048,
      "epoch": 0.606,
      "grad_norm": 0.18081100841917966,
      "kl": 0.355859375,
      "learning_rate": 8.057656487800283e-06,
      "loss": 0.0142,
      "reward": 1.0333333611488342,
      "reward_std": 0.13328943625092507,
      "rewards/accuracy_reward": 0.05416666828095913,
      "rewards/format_reward": 0.9791666746139527,
      "step": 1515
    },
    {
      "completion_length": 233.42083892822265,
      "epoch": 0.608,
      "grad_norm": 0.6317667118077563,
      "kl": 0.3671875,
      "learning_rate": 7.989220788540356e-06,
      "loss": 0.0147,
      "reward": 1.1125000238418579,
      "reward_std": 0.21407929211854934,
      "rewards/accuracy_reward": 0.1375000037252903,
      "rewards/format_reward": 0.975000011920929,
      "step": 1520
    },
    {
      "completion_length": 337.39584197998045,
      "epoch": 0.61,
      "grad_norm": 0.6816123949684233,
      "kl": 0.33828125,
      "learning_rate": 7.92088309182241e-06,
      "loss": 0.0135,
      "reward": 1.0833333730697632,
      "reward_std": 0.240230230987072,
      "rewards/accuracy_reward": 0.12916667051613331,
      "rewards/format_reward": 0.9541666746139527,
      "step": 1525
    },
    {
      "completion_length": 224.86250762939454,
      "epoch": 0.612,
      "grad_norm": 0.5107336103473655,
      "kl": 0.3603515625,
      "learning_rate": 7.852646728329368e-06,
      "loss": 0.0144,
      "reward": 1.0916667103767395,
      "reward_std": 0.20342101380228997,
      "rewards/accuracy_reward": 0.10833333693444729,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1530
    },
    {
      "completion_length": 253.57084197998046,
      "epoch": 0.614,
      "grad_norm": 0.7130506106876767,
      "kl": 0.4244140625,
      "learning_rate": 7.784515023805328e-06,
      "loss": 0.017,
      "reward": 1.137500023841858,
      "reward_std": 0.19924872741103172,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1535
    },
    {
      "completion_length": 246.7541748046875,
      "epoch": 0.616,
      "grad_norm": 0.8202318016415265,
      "kl": 0.3501953125,
      "learning_rate": 7.716491298893443e-06,
      "loss": 0.014,
      "reward": 1.108333373069763,
      "reward_std": 0.1968993701040745,
      "rewards/accuracy_reward": 0.12500000558793545,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1540
    },
    {
      "completion_length": 223.5791763305664,
      "epoch": 0.618,
      "grad_norm": 0.5800169184427162,
      "kl": 0.3650390625,
      "learning_rate": 7.6485788689741e-06,
      "loss": 0.0146,
      "reward": 1.0625000238418578,
      "reward_std": 0.14876977801322938,
      "rewards/accuracy_reward": 0.07500000298023224,
      "rewards/format_reward": 0.987500011920929,
      "step": 1545
    },
    {
      "completion_length": 228.31250457763673,
      "epoch": 0.62,
      "grad_norm": 0.7382503773664825,
      "kl": 0.4041015625,
      "learning_rate": 7.580781044003324e-06,
      "loss": 0.0162,
      "reward": 1.0916667103767395,
      "reward_std": 0.2120341219007969,
      "rewards/accuracy_reward": 0.1125000037252903,
      "rewards/format_reward": 0.9791666746139527,
      "step": 1550
    },
    {
      "completion_length": 240.68334045410157,
      "epoch": 0.622,
      "grad_norm": 0.5083007731042197,
      "kl": 0.4341796875,
      "learning_rate": 7.513101128351454e-06,
      "loss": 0.0174,
      "reward": 1.0833333730697632,
      "reward_std": 0.22213384434580802,
      "rewards/accuracy_reward": 0.12083333805203438,
      "rewards/format_reward": 0.9625000178813934,
      "step": 1555
    },
    {
      "completion_length": 246.29584045410155,
      "epoch": 0.624,
      "grad_norm": 0.7661993634854364,
      "kl": 0.3583984375,
      "learning_rate": 7.445542420642097e-06,
      "loss": 0.0143,
      "reward": 1.1000000298023225,
      "reward_std": 0.25303551703691485,
      "rewards/accuracy_reward": 0.12916667126119136,
      "rewards/format_reward": 0.9708333492279053,
      "step": 1560
    },
    {
      "completion_length": 194.09584045410156,
      "epoch": 0.626,
      "grad_norm": 0.5324394879554415,
      "kl": 0.3693359375,
      "learning_rate": 7.378108213591355e-06,
      "loss": 0.0148,
      "reward": 1.1000000357627868,
      "reward_std": 0.1730000950396061,
      "rewards/accuracy_reward": 0.10833333656191826,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1565
    },
    {
      "completion_length": 228.5166748046875,
      "epoch": 0.628,
      "grad_norm": 0.6241877369872114,
      "kl": 0.36484375,
      "learning_rate": 7.310801793847344e-06,
      "loss": 0.0146,
      "reward": 1.1125000357627868,
      "reward_std": 0.2222403332591057,
      "rewards/accuracy_reward": 0.13750000558793546,
      "rewards/format_reward": 0.9750000238418579,
      "step": 1570
    },
    {
      "completion_length": 197.89167175292968,
      "epoch": 0.63,
      "grad_norm": 0.329436159848971,
      "kl": 0.3484375,
      "learning_rate": 7.243626441830009e-06,
      "loss": 0.0139,
      "reward": 1.083333384990692,
      "reward_std": 0.12386635392904281,
      "rewards/accuracy_reward": 0.08750000223517418,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1575
    },
    {
      "completion_length": 220.6916748046875,
      "epoch": 0.632,
      "grad_norm": 0.9395190691678801,
      "kl": 0.35390625,
      "learning_rate": 7.176585431571235e-06,
      "loss": 0.0141,
      "reward": 1.1083333671092988,
      "reward_std": 0.18841607496142387,
      "rewards/accuracy_reward": 0.12916667014360428,
      "rewards/format_reward": 0.9791666805744171,
      "step": 1580
    },
    {
      "completion_length": 232.3625045776367,
      "epoch": 0.634,
      "grad_norm": 0.788552192585705,
      "kl": 0.3609375,
      "learning_rate": 7.109682030555283e-06,
      "loss": 0.0144,
      "reward": 1.0833333432674408,
      "reward_std": 0.1897010862827301,
      "rewards/accuracy_reward": 0.1041666679084301,
      "rewards/format_reward": 0.9791666805744171,
      "step": 1585
    },
    {
      "completion_length": 347.0833404541016,
      "epoch": 0.636,
      "grad_norm": 0.38934477847250354,
      "kl": 0.3583984375,
      "learning_rate": 7.042919499559538e-06,
      "loss": 0.0144,
      "reward": 1.004166692495346,
      "reward_std": 0.25163830146193505,
      "rewards/accuracy_reward": 0.07916666902601718,
      "rewards/format_reward": 0.9250000238418579,
      "step": 1590
    },
    {
      "completion_length": 289.96250610351564,
      "epoch": 0.638,
      "grad_norm": 1.312643011859593,
      "kl": 0.3607421875,
      "learning_rate": 6.976301092495556e-06,
      "loss": 0.0144,
      "reward": 1.0250000357627869,
      "reward_std": 0.21382493525743484,
      "rewards/accuracy_reward": 0.08333333544433116,
      "rewards/format_reward": 0.9416666865348816,
      "step": 1595
    },
    {
      "completion_length": 244.08750610351564,
      "epoch": 0.64,
      "grad_norm": 0.8343345060342183,
      "kl": 0.3611328125,
      "learning_rate": 6.909830056250527e-06,
      "loss": 0.0144,
      "reward": 1.0625000476837159,
      "reward_std": 0.20852040648460388,
      "rewards/accuracy_reward": 0.09166666939854622,
      "rewards/format_reward": 0.9708333551883698,
      "step": 1600
    },
    {
      "epoch": 0.64,
      "eval_completion_length": 226.89583998460037,
      "eval_kl": 0.3518629807692308,
      "eval_loss": 0.01314304769039154,
      "eval_reward": 1.125000041264754,
      "eval_reward_std": 0.2221679870898907,
      "eval_rewards/accuracy_reward": 0.15224359270471793,
      "eval_rewards/format_reward": 0.9727564316529494,
      "eval_runtime": 188.0925,
      "eval_samples_per_second": 0.526,
      "eval_steps_per_second": 0.016,
      "step": 1600
    },
    {
      "completion_length": 264.52917327880857,
      "epoch": 0.642,
      "grad_norm": 0.8851432430573316,
      "kl": 0.36953125,
      "learning_rate": 6.843509630528977e-06,
      "loss": 0.0148,
      "reward": 1.025000011920929,
      "reward_std": 0.20359539091587067,
      "rewards/accuracy_reward": 0.07500000074505805,
      "rewards/format_reward": 0.9500000178813934,
      "step": 1605
    },
    {
      "completion_length": 219.15416870117187,
      "epoch": 0.644,
      "grad_norm": 0.8558209077276151,
      "kl": 0.3716796875,
      "learning_rate": 6.777343047694891e-06,
      "loss": 0.0148,
      "reward": 1.108333373069763,
      "reward_std": 0.17567697837948798,
      "rewards/accuracy_reward": 0.1375000011175871,
      "rewards/format_reward": 0.9708333551883698,
      "step": 1610
    },
    {
      "completion_length": 234.85000610351562,
      "epoch": 0.646,
      "grad_norm": 0.432451594605514,
      "kl": 0.3794921875,
      "learning_rate": 6.711333532614168e-06,
      "loss": 0.0152,
      "reward": 1.0583333730697633,
      "reward_std": 0.21975762471556665,
      "rewards/accuracy_reward": 0.10833333730697632,
      "rewards/format_reward": 0.9500000238418579,
      "step": 1615
    },
    {
      "completion_length": 213.1041717529297,
      "epoch": 0.648,
      "grad_norm": 1.3247688082665616,
      "kl": 0.3876953125,
      "learning_rate": 6.645484302497452e-06,
      "loss": 0.0155,
      "reward": 1.0958333730697631,
      "reward_std": 0.26502900272607804,
      "rewards/accuracy_reward": 0.1375000037252903,
      "rewards/format_reward": 0.9583333551883697,
      "step": 1620
    },
    {
      "completion_length": 180.1125045776367,
      "epoch": 0.65,
      "grad_norm": 0.6778703705706532,
      "kl": 0.36484375,
      "learning_rate": 6.579798566743314e-06,
      "loss": 0.0146,
      "reward": 1.1000000298023225,
      "reward_std": 0.1525876834988594,
      "rewards/accuracy_reward": 0.12083333656191826,
      "rewards/format_reward": 0.9791666805744171,
      "step": 1625
    },
    {
      "completion_length": 185.93334350585937,
      "epoch": 0.652,
      "grad_norm": 0.6029963592777737,
      "kl": 0.3892578125,
      "learning_rate": 6.5142795267818505e-06,
      "loss": 0.0156,
      "reward": 1.145833384990692,
      "reward_std": 0.19070877507328987,
      "rewards/accuracy_reward": 0.15833333805203437,
      "rewards/format_reward": 0.987500011920929,
      "step": 1630
    },
    {
      "completion_length": 230.4250030517578,
      "epoch": 0.654,
      "grad_norm": 0.5157745546659512,
      "kl": 0.3515625,
      "learning_rate": 6.448930375918632e-06,
      "loss": 0.0141,
      "reward": 1.0833333551883697,
      "reward_std": 0.18237519189715384,
      "rewards/accuracy_reward": 0.11250000335276127,
      "rewards/format_reward": 0.9708333432674408,
      "step": 1635
    },
    {
      "completion_length": 257.05833740234374,
      "epoch": 0.656,
      "grad_norm": 0.7811139925774508,
      "kl": 0.3455078125,
      "learning_rate": 6.383754299179079e-06,
      "loss": 0.0138,
      "reward": 1.1041667103767394,
      "reward_std": 0.24514180719852446,
      "rewards/accuracy_reward": 0.1458333358168602,
      "rewards/format_reward": 0.9583333492279053,
      "step": 1640
    },
    {
      "completion_length": 238.72083892822266,
      "epoch": 0.658,
      "grad_norm": 0.5624873094851147,
      "kl": 0.3234375,
      "learning_rate": 6.318754473153221e-06,
      "loss": 0.0129,
      "reward": 1.1041667103767394,
      "reward_std": 0.1908385895192623,
      "rewards/accuracy_reward": 0.1166666705161333,
      "rewards/format_reward": 0.987500011920929,
      "step": 1645
    },
    {
      "completion_length": 294.3958404541016,
      "epoch": 0.66,
      "grad_norm": 0.4438652564966346,
      "kl": 0.374609375,
      "learning_rate": 6.25393406584088e-06,
      "loss": 0.015,
      "reward": 1.1125000238418579,
      "reward_std": 0.18262089043855667,
      "rewards/accuracy_reward": 0.1416666701436043,
      "rewards/format_reward": 0.9708333551883698,
      "step": 1650
    },
    {
      "completion_length": 280.0666732788086,
      "epoch": 0.662,
      "grad_norm": 0.8501811034862022,
      "kl": 0.4396484375,
      "learning_rate": 6.18929623649726e-06,
      "loss": 0.0176,
      "reward": 1.129166728258133,
      "reward_std": 0.2677574008703232,
      "rewards/accuracy_reward": 0.17083334028720856,
      "rewards/format_reward": 0.9583333551883697,
      "step": 1655
    },
    {
      "completion_length": 256.05001068115234,
      "epoch": 0.664,
      "grad_norm": 0.786466356675622,
      "kl": 0.3712890625,
      "learning_rate": 6.124844135478971e-06,
      "loss": 0.0148,
      "reward": 1.1083333611488342,
      "reward_std": 0.15910932496190072,
      "rewards/accuracy_reward": 0.12916667237877846,
      "rewards/format_reward": 0.9791666746139527,
      "step": 1660
    },
    {
      "completion_length": 269.5541748046875,
      "epoch": 0.666,
      "grad_norm": 0.6054140410900962,
      "kl": 0.4185546875,
      "learning_rate": 6.06058090409049e-06,
      "loss": 0.0167,
      "reward": 1.1041666984558105,
      "reward_std": 0.17600802853703498,
      "rewards/accuracy_reward": 0.11666667014360428,
      "rewards/format_reward": 0.987500011920929,
      "step": 1665
    },
    {
      "completion_length": 195.0500061035156,
      "epoch": 0.668,
      "grad_norm": 0.09195814274917967,
      "kl": 0.35390625,
      "learning_rate": 5.996509674431053e-06,
      "loss": 0.0141,
      "reward": 1.0541666984558105,
      "reward_std": 0.08435339778661728,
      "rewards/accuracy_reward": 0.07083333544433117,
      "rewards/format_reward": 0.9833333492279053,
      "step": 1670
    },
    {
      "completion_length": 176.61250228881835,
      "epoch": 0.67,
      "grad_norm": 0.19096552721075902,
      "kl": 0.3958984375,
      "learning_rate": 5.932633569242e-06,
      "loss": 0.0158,
      "reward": 1.0625000476837159,
      "reward_std": 0.09324773177504539,
      "rewards/accuracy_reward": 0.06666666865348816,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1675
    },
    {
      "completion_length": 221.21250686645507,
      "epoch": 0.672,
      "grad_norm": 0.5670018996024074,
      "kl": 0.3828125,
      "learning_rate": 5.868955701754584e-06,
      "loss": 0.0153,
      "reward": 1.0833333730697632,
      "reward_std": 0.19533313140273095,
      "rewards/accuracy_reward": 0.10833333730697632,
      "rewards/format_reward": 0.9750000238418579,
      "step": 1680
    },
    {
      "completion_length": 200.27083740234374,
      "epoch": 0.674,
      "grad_norm": 0.7884552413539826,
      "kl": 0.36875,
      "learning_rate": 5.8054791755382286e-06,
      "loss": 0.0148,
      "reward": 1.0708333730697632,
      "reward_std": 0.1910129614174366,
      "rewards/accuracy_reward": 0.08750000186264514,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1685
    },
    {
      "completion_length": 166.4583381652832,
      "epoch": 0.676,
      "grad_norm": 1.0096233303682502,
      "kl": 0.3990234375,
      "learning_rate": 5.742207084349274e-06,
      "loss": 0.016,
      "reward": 1.112500047683716,
      "reward_std": 0.20021617263555527,
      "rewards/accuracy_reward": 0.12500000484287738,
      "rewards/format_reward": 0.9875000059604645,
      "step": 1690
    },
    {
      "completion_length": 129.7916702270508,
      "epoch": 0.678,
      "grad_norm": 0.7510832019121336,
      "kl": 0.4009765625,
      "learning_rate": 5.679142511980176e-06,
      "loss": 0.016,
      "reward": 1.1500000476837158,
      "reward_std": 0.16661179661750794,
      "rewards/accuracy_reward": 0.1541666716337204,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1695
    },
    {
      "completion_length": 149.3833381652832,
      "epoch": 0.68,
      "grad_norm": 0.2240382817532387,
      "kl": 0.3921875,
      "learning_rate": 5.616288532109225e-06,
      "loss": 0.0157,
      "reward": 1.1041666924953462,
      "reward_std": 0.17280239164829253,
      "rewards/accuracy_reward": 0.12083333656191826,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1700
    },
    {
      "epoch": 0.68,
      "eval_completion_length": 175.23558044433594,
      "eval_kl": 0.3936298076923077,
      "eval_loss": 0.015872277319431305,
      "eval_reward": 1.10737184377817,
      "eval_reward_std": 0.19909162332232183,
      "eval_rewards/accuracy_reward": 0.1378205187905293,
      "eval_rewards/format_reward": 0.9695512927495517,
      "eval_runtime": 187.6606,
      "eval_samples_per_second": 0.528,
      "eval_steps_per_second": 0.016,
      "step": 1700
    },
    {
      "completion_length": 202.48333892822265,
      "epoch": 0.682,
      "grad_norm": 1.2902182940960338,
      "kl": 0.4435546875,
      "learning_rate": 5.553648208150728e-06,
      "loss": 0.0177,
      "reward": 1.0916666984558105,
      "reward_std": 0.27394800409674647,
      "rewards/accuracy_reward": 0.13333333767950534,
      "rewards/format_reward": 0.9583333551883697,
      "step": 1705
    },
    {
      "completion_length": 217.67917251586914,
      "epoch": 0.684,
      "grad_norm": 0.8042172808984961,
      "kl": 0.4064453125,
      "learning_rate": 5.491224593105695e-06,
      "loss": 0.0163,
      "reward": 1.041666704416275,
      "reward_std": 0.19876980185508727,
      "rewards/accuracy_reward": 0.07916666939854622,
      "rewards/format_reward": 0.9625000178813934,
      "step": 1710
    },
    {
      "completion_length": 214.5916732788086,
      "epoch": 0.686,
      "grad_norm": 0.9930554223957376,
      "kl": 0.380078125,
      "learning_rate": 5.429020729413062e-06,
      "loss": 0.0152,
      "reward": 1.108333373069763,
      "reward_std": 0.26563555002212524,
      "rewards/accuracy_reward": 0.14166666939854622,
      "rewards/format_reward": 0.9666666865348816,
      "step": 1715
    },
    {
      "completion_length": 202.17083816528321,
      "epoch": 0.688,
      "grad_norm": 0.8686412520549874,
      "kl": 0.3771484375,
      "learning_rate": 5.367039648801386e-06,
      "loss": 0.0151,
      "reward": 1.141666716337204,
      "reward_std": 0.26376511752605436,
      "rewards/accuracy_reward": 0.17500000558793544,
      "rewards/format_reward": 0.9666666805744171,
      "step": 1720
    },
    {
      "completion_length": 263.8541732788086,
      "epoch": 0.69,
      "grad_norm": 1.2120904832599089,
      "kl": 0.3544921875,
      "learning_rate": 5.305284372141095e-06,
      "loss": 0.0142,
      "reward": 1.066666716337204,
      "reward_std": 0.18272737711668013,
      "rewards/accuracy_reward": 0.09583333693444729,
      "rewards/format_reward": 0.9708333492279053,
      "step": 1725
    },
    {
      "completion_length": 318.8791748046875,
      "epoch": 0.692,
      "grad_norm": 0.7490910563872372,
      "kl": 0.3595703125,
      "learning_rate": 5.243757909297247e-06,
      "loss": 0.0144,
      "reward": 1.0375000298023225,
      "reward_std": 0.20392291769385337,
      "rewards/accuracy_reward": 0.07500000186264515,
      "rewards/format_reward": 0.962500023841858,
      "step": 1730
    },
    {
      "completion_length": 291.9708450317383,
      "epoch": 0.694,
      "grad_norm": 0.5173379094623531,
      "kl": 0.35703125,
      "learning_rate": 5.1824632589828465e-06,
      "loss": 0.0143,
      "reward": 1.0666666984558106,
      "reward_std": 0.12149013429880143,
      "rewards/accuracy_reward": 0.07916666865348816,
      "rewards/format_reward": 0.987500011920929,
      "step": 1735
    },
    {
      "completion_length": 229.35834197998048,
      "epoch": 0.696,
      "grad_norm": 1.1205426336140032,
      "kl": 0.334765625,
      "learning_rate": 5.121403408612672e-06,
      "loss": 0.0134,
      "reward": 1.1291666984558106,
      "reward_std": 0.15256435573101043,
      "rewards/accuracy_reward": 0.12916667014360428,
      "rewards/format_reward": 1.0,
      "step": 1740
    },
    {
      "completion_length": 242.07083740234376,
      "epoch": 0.698,
      "grad_norm": 0.08229901659914558,
      "kl": 0.3361328125,
      "learning_rate": 5.060581334157693e-06,
      "loss": 0.0134,
      "reward": 1.1666666984558105,
      "reward_std": 0.13328943997621537,
      "rewards/accuracy_reward": 0.16666667237877847,
      "rewards/format_reward": 1.0,
      "step": 1745
    },
    {
      "completion_length": 243.38333740234376,
      "epoch": 0.7,
      "grad_norm": 0.4810002013198661,
      "kl": 0.33203125,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.0133,
      "reward": 1.1208333492279052,
      "reward_std": 0.09865520298480987,
      "rewards/accuracy_reward": 0.12500000223517418,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1750
    },
    {
      "completion_length": 201.52500915527344,
      "epoch": 0.702,
      "grad_norm": 0.7845067243546726,
      "kl": 0.3431640625,
      "learning_rate": 4.939662358788364e-06,
      "loss": 0.0137,
      "reward": 1.1500000357627869,
      "reward_std": 0.1625961884856224,
      "rewards/accuracy_reward": 0.15000000074505807,
      "rewards/format_reward": 1.0,
      "step": 1755
    },
    {
      "completion_length": 204.12083892822267,
      "epoch": 0.704,
      "grad_norm": 0.6792203039767948,
      "kl": 0.315625,
      "learning_rate": 4.879571351294287e-06,
      "loss": 0.0126,
      "reward": 1.1416667103767395,
      "reward_std": 0.16416243016719817,
      "rewards/accuracy_reward": 0.14166667237877845,
      "rewards/format_reward": 1.0,
      "step": 1760
    },
    {
      "completion_length": 187.62500457763673,
      "epoch": 0.706,
      "grad_norm": 0.7070939109048883,
      "kl": 0.3744140625,
      "learning_rate": 4.8197299062687e-06,
      "loss": 0.015,
      "reward": 1.1416667103767395,
      "reward_std": 0.18841607943177224,
      "rewards/accuracy_reward": 0.1458333373069763,
      "rewards/format_reward": 0.9958333373069763,
      "step": 1765
    },
    {
      "completion_length": 218.6666748046875,
      "epoch": 0.708,
      "grad_norm": 1.1283215566490286,
      "kl": 0.3330078125,
      "learning_rate": 4.76014094029921e-06,
      "loss": 0.0133,
      "reward": 1.0958333492279053,
      "reward_std": 0.14890312626957894,
      "rewards/accuracy_reward": 0.0958333346992731,
      "rewards/format_reward": 1.0,
      "step": 1770
    },
    {
      "completion_length": 196.70000610351562,
      "epoch": 0.71,
      "grad_norm": 0.5380849514067082,
      "kl": 0.3201171875,
      "learning_rate": 4.700807357667953e-06,
      "loss": 0.0128,
      "reward": 1.1166667103767396,
      "reward_std": 0.12386635094881057,
      "rewards/accuracy_reward": 0.11666667126119137,
      "rewards/format_reward": 1.0,
      "step": 1775
    },
    {
      "completion_length": 199.36250762939454,
      "epoch": 0.712,
      "grad_norm": 0.776557437915111,
      "kl": 0.3330078125,
      "learning_rate": 4.641732050210032e-06,
      "loss": 0.0133,
      "reward": 1.145833384990692,
      "reward_std": 0.17088177502155305,
      "rewards/accuracy_reward": 0.15416667200624942,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1780
    },
    {
      "completion_length": 199.9291748046875,
      "epoch": 0.714,
      "grad_norm": 0.7199296200116914,
      "kl": 0.344140625,
      "learning_rate": 4.582917897172603e-06,
      "loss": 0.0138,
      "reward": 1.1791666865348815,
      "reward_std": 0.18141550272703172,
      "rewards/accuracy_reward": 0.18750000223517418,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1785
    },
    {
      "completion_length": 240.93750610351563,
      "epoch": 0.716,
      "grad_norm": 0.82997985294821,
      "kl": 0.336328125,
      "learning_rate": 4.524367765074499e-06,
      "loss": 0.0135,
      "reward": 1.0583333849906922,
      "reward_std": 0.16817803531885148,
      "rewards/accuracy_reward": 0.07916666902601718,
      "rewards/format_reward": 0.9791666746139527,
      "step": 1790
    },
    {
      "completion_length": 257.6500076293945,
      "epoch": 0.718,
      "grad_norm": 0.6940011512430222,
      "kl": 0.3623046875,
      "learning_rate": 4.46608450756656e-06,
      "loss": 0.0145,
      "reward": 1.1416667103767395,
      "reward_std": 0.19037772193551064,
      "rewards/accuracy_reward": 0.1791666716337204,
      "rewards/format_reward": 0.9625000298023224,
      "step": 1795
    },
    {
      "completion_length": 346.0041778564453,
      "epoch": 0.72,
      "grad_norm": 0.4358586566289972,
      "kl": 0.315625,
      "learning_rate": 4.408070965292534e-06,
      "loss": 0.0126,
      "reward": 1.0125000298023223,
      "reward_std": 0.2788765422999859,
      "rewards/accuracy_reward": 0.08750000260770321,
      "rewards/format_reward": 0.9250000178813934,
      "step": 1800
    },
    {
      "epoch": 0.72,
      "eval_completion_length": 265.323724599985,
      "eval_kl": 0.3454026442307692,
      "eval_loss": 0.013118223287165165,
      "eval_reward": 1.118589772627904,
      "eval_reward_std": 0.25780055958491105,
      "eval_rewards/accuracy_reward": 0.16346154361963272,
      "eval_rewards/format_reward": 0.9551282341663654,
      "eval_runtime": 239.8483,
      "eval_samples_per_second": 0.413,
      "eval_steps_per_second": 0.013,
      "step": 1800
    },
    {
      "completion_length": 263.9666717529297,
      "epoch": 0.722,
      "grad_norm": 0.884361517453845,
      "kl": 0.3580078125,
      "learning_rate": 4.350329965750622e-06,
      "loss": 0.0143,
      "reward": 1.1083333790302277,
      "reward_std": 0.23719543740153312,
      "rewards/accuracy_reward": 0.14583333767950535,
      "rewards/format_reward": 0.9625000178813934,
      "step": 1805
    },
    {
      "completion_length": 252.55001068115234,
      "epoch": 0.724,
      "grad_norm": 0.911989814522389,
      "kl": 0.3294921875,
      "learning_rate": 4.292864323155684e-06,
      "loss": 0.0132,
      "reward": 1.1416666865348817,
      "reward_std": 0.2740778163075447,
      "rewards/accuracy_reward": 0.17916667312383652,
      "rewards/format_reward": 0.9625000178813934,
      "step": 1810
    },
    {
      "completion_length": 265.45834197998045,
      "epoch": 0.726,
      "grad_norm": 0.3701309923874372,
      "kl": 0.306640625,
      "learning_rate": 4.235676838302069e-06,
      "loss": 0.0123,
      "reward": 1.1125000357627868,
      "reward_std": 0.257410154491663,
      "rewards/accuracy_reward": 0.1583333384245634,
      "rewards/format_reward": 0.9541666865348816,
      "step": 1815
    },
    {
      "completion_length": 227.68334045410157,
      "epoch": 0.728,
      "grad_norm": 0.9371502334838836,
      "kl": 0.323046875,
      "learning_rate": 4.178770298427107e-06,
      "loss": 0.0129,
      "reward": 1.137500035762787,
      "reward_std": 0.27804360538721085,
      "rewards/accuracy_reward": 0.15833333805203437,
      "rewards/format_reward": 0.9791666865348816,
      "step": 1820
    },
    {
      "completion_length": 245.9291732788086,
      "epoch": 0.73,
      "grad_norm": 0.5701038034532563,
      "kl": 0.29423828125,
      "learning_rate": 4.12214747707527e-06,
      "loss": 0.0118,
      "reward": 1.1208333611488341,
      "reward_std": 0.22555280923843385,
      "rewards/accuracy_reward": 0.15416667386889457,
      "rewards/format_reward": 0.9666666805744171,
      "step": 1825
    },
    {
      "completion_length": 233.24167404174804,
      "epoch": 0.732,
      "grad_norm": 0.2591556634095291,
      "kl": 0.3109375,
      "learning_rate": 4.065811133962987e-06,
      "loss": 0.0124,
      "reward": 1.0916666865348816,
      "reward_std": 0.1926060661673546,
      "rewards/accuracy_reward": 0.1250000014901161,
      "rewards/format_reward": 0.9666666805744171,
      "step": 1830
    },
    {
      "completion_length": 238.20417327880858,
      "epoch": 0.734,
      "grad_norm": 0.9770366277050908,
      "kl": 0.34375,
      "learning_rate": 4.009764014844143e-06,
      "loss": 0.0137,
      "reward": 1.179166704416275,
      "reward_std": 0.2617766156792641,
      "rewards/accuracy_reward": 0.21666667088866234,
      "rewards/format_reward": 0.9625000119209289,
      "step": 1835
    },
    {
      "completion_length": 189.57084045410156,
      "epoch": 0.736,
      "grad_norm": 0.5416908718652198,
      "kl": 0.319921875,
      "learning_rate": 3.954008851376252e-06,
      "loss": 0.0128,
      "reward": 1.1250000238418578,
      "reward_std": 0.14619938358664514,
      "rewards/accuracy_reward": 0.13750000223517417,
      "rewards/format_reward": 0.987500011920929,
      "step": 1840
    },
    {
      "completion_length": 219.25417251586913,
      "epoch": 0.738,
      "grad_norm": 0.6006299277423031,
      "kl": 0.3302734375,
      "learning_rate": 3.898548360987325e-06,
      "loss": 0.0132,
      "reward": 1.1000000476837157,
      "reward_std": 0.19881998300552367,
      "rewards/accuracy_reward": 0.12500000409781933,
      "rewards/format_reward": 0.9750000178813935,
      "step": 1845
    },
    {
      "completion_length": 203.9916748046875,
      "epoch": 0.74,
      "grad_norm": 1.0816648176296837,
      "kl": 0.33779296875,
      "learning_rate": 3.8433852467434175e-06,
      "loss": 0.0135,
      "reward": 1.1833333849906922,
      "reward_std": 0.25205365046858785,
      "rewards/accuracy_reward": 0.21250000596046448,
      "rewards/format_reward": 0.9708333432674408,
      "step": 1850
    },
    {
      "completion_length": 254.84584045410156,
      "epoch": 0.742,
      "grad_norm": 0.5651446585039597,
      "kl": 0.33984375,
      "learning_rate": 3.7885221972168974e-06,
      "loss": 0.0136,
      "reward": 1.0333333611488342,
      "reward_std": 0.202835600823164,
      "rewards/accuracy_reward": 0.07500000298023224,
      "rewards/format_reward": 0.9583333551883697,
      "step": 1855
    },
    {
      "completion_length": 243.50000381469727,
      "epoch": 0.744,
      "grad_norm": 0.8741874945615253,
      "kl": 0.3328125,
      "learning_rate": 3.7339618863553983e-06,
      "loss": 0.0133,
      "reward": 1.1041667103767394,
      "reward_std": 0.23992215022444724,
      "rewards/accuracy_reward": 0.1416666705161333,
      "rewards/format_reward": 0.9625000119209289,
      "step": 1860
    },
    {
      "completion_length": 183.93750534057617,
      "epoch": 0.746,
      "grad_norm": 0.7998884228143229,
      "kl": 0.35546875,
      "learning_rate": 3.679706973351491e-06,
      "loss": 0.0142,
      "reward": 1.1166666865348815,
      "reward_std": 0.16580181941390038,
      "rewards/accuracy_reward": 0.1291666693985462,
      "rewards/format_reward": 0.987500011920929,
      "step": 1865
    },
    {
      "completion_length": 197.3291748046875,
      "epoch": 0.748,
      "grad_norm": 0.7195504466555774,
      "kl": 0.3251953125,
      "learning_rate": 3.625760102513103e-06,
      "loss": 0.013,
      "reward": 1.0666667103767395,
      "reward_std": 0.18050257191061975,
      "rewards/accuracy_reward": 0.09166666865348816,
      "rewards/format_reward": 0.9750000178813935,
      "step": 1870
    },
    {
      "completion_length": 186.4416717529297,
      "epoch": 0.75,
      "grad_norm": 0.8437020501088257,
      "kl": 0.3259765625,
      "learning_rate": 3.5721239031346067e-06,
      "loss": 0.013,
      "reward": 1.079166716337204,
      "reward_std": 0.18379171639680864,
      "rewards/accuracy_reward": 0.09583333693444729,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1875
    },
    {
      "completion_length": 180.53333587646483,
      "epoch": 0.752,
      "grad_norm": 0.6175522474967708,
      "kl": 0.3048828125,
      "learning_rate": 3.5188009893686916e-06,
      "loss": 0.0122,
      "reward": 1.1333333849906921,
      "reward_std": 0.18649545460939407,
      "rewards/accuracy_reward": 0.14166667237877845,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1880
    },
    {
      "completion_length": 189.47084045410156,
      "epoch": 0.754,
      "grad_norm": 0.7541283975676629,
      "kl": 0.33828125,
      "learning_rate": 3.4657939600989453e-06,
      "loss": 0.0135,
      "reward": 1.1291666984558106,
      "reward_std": 0.22653717398643494,
      "rewards/accuracy_reward": 0.14166666977107525,
      "rewards/format_reward": 0.9875000059604645,
      "step": 1885
    },
    {
      "completion_length": 236.83334197998047,
      "epoch": 0.756,
      "grad_norm": 0.7192849833359639,
      "kl": 0.3146484375,
      "learning_rate": 3.4131053988131947e-06,
      "loss": 0.0126,
      "reward": 1.0875000298023223,
      "reward_std": 0.24265274703502654,
      "rewards/accuracy_reward": 0.12083333842456341,
      "rewards/format_reward": 0.9666666865348816,
      "step": 1890
    },
    {
      "completion_length": 231.8416778564453,
      "epoch": 0.758,
      "grad_norm": 0.43887813755474286,
      "kl": 0.8873046875,
      "learning_rate": 3.360737873477584e-06,
      "loss": 0.0355,
      "reward": 1.0625000298023224,
      "reward_std": 0.1980368673801422,
      "rewards/accuracy_reward": 0.0958333358168602,
      "rewards/format_reward": 0.9666666865348816,
      "step": 1895
    },
    {
      "completion_length": 273.9916748046875,
      "epoch": 0.76,
      "grad_norm": 0.4947009731663165,
      "kl": 0.334765625,
      "learning_rate": 3.308693936411421e-06,
      "loss": 0.0134,
      "reward": 1.0333333611488342,
      "reward_std": 0.2203465722501278,
      "rewards/accuracy_reward": 0.08333333507180214,
      "rewards/format_reward": 0.950000011920929,
      "step": 1900
    },
    {
      "epoch": 0.76,
      "eval_completion_length": 207.45994215745193,
      "eval_kl": 0.3671875,
      "eval_loss": 0.013619143515825272,
      "eval_reward": 1.1314102869767408,
      "eval_reward_std": 0.21630300896672103,
      "eval_rewards/accuracy_reward": 0.1538461590042481,
      "eval_rewards/format_reward": 0.9775641147906964,
      "eval_runtime": 181.6984,
      "eval_samples_per_second": 0.545,
      "eval_steps_per_second": 0.017,
      "step": 1900
    },
    {
      "completion_length": 215.0666732788086,
      "epoch": 0.762,
      "grad_norm": 0.4462722168761537,
      "kl": 0.32421875,
      "learning_rate": 3.2569761241627694e-06,
      "loss": 0.013,
      "reward": 1.1041667103767394,
      "reward_std": 0.18810799717903137,
      "rewards/accuracy_reward": 0.12916667014360428,
      "rewards/format_reward": 0.9750000238418579,
      "step": 1905
    },
    {
      "completion_length": 214.52083892822264,
      "epoch": 0.764,
      "grad_norm": 0.6751355982112365,
      "kl": 0.34326171875,
      "learning_rate": 3.2055869573848374e-06,
      "loss": 0.0137,
      "reward": 1.1291666865348815,
      "reward_std": 0.2662478253245354,
      "rewards/accuracy_reward": 0.15416667126119138,
      "rewards/format_reward": 0.9750000238418579,
      "step": 1910
    },
    {
      "completion_length": 221.32500381469725,
      "epoch": 0.766,
      "grad_norm": 0.892078258457626,
      "kl": 0.31328125,
      "learning_rate": 3.1545289407131128e-06,
      "loss": 0.0125,
      "reward": 1.200000035762787,
      "reward_std": 0.2645213954150677,
      "rewards/accuracy_reward": 0.2333333406597376,
      "rewards/format_reward": 0.9666666924953461,
      "step": 1915
    },
    {
      "completion_length": 210.50000457763673,
      "epoch": 0.768,
      "grad_norm": 0.623883337401102,
      "kl": 0.3212890625,
      "learning_rate": 3.103804562643302e-06,
      "loss": 0.0129,
      "reward": 1.1000000476837157,
      "reward_std": 0.18917233720421792,
      "rewards/accuracy_reward": 0.12500000298023223,
      "rewards/format_reward": 0.9750000238418579,
      "step": 1920
    },
    {
      "completion_length": 298.3458419799805,
      "epoch": 0.77,
      "grad_norm": 1.0767844089766165,
      "kl": 0.318359375,
      "learning_rate": 3.0534162954100264e-06,
      "loss": 0.0127,
      "reward": 1.016666704416275,
      "reward_std": 0.21427700370550157,
      "rewards/accuracy_reward": 0.07083333432674407,
      "rewards/format_reward": 0.9458333551883698,
      "step": 1925
    },
    {
      "completion_length": 219.76667098999025,
      "epoch": 0.772,
      "grad_norm": 0.8725082553969676,
      "kl": 0.3330078125,
      "learning_rate": 3.003366594866345e-06,
      "loss": 0.0133,
      "reward": 1.062500035762787,
      "reward_std": 0.18512693047523499,
      "rewards/accuracy_reward": 0.0916666690260172,
      "rewards/format_reward": 0.9708333551883698,
      "step": 1930
    },
    {
      "completion_length": 240.48750762939454,
      "epoch": 0.774,
      "grad_norm": 0.6293755791762184,
      "kl": 0.3767578125,
      "learning_rate": 2.953657900364053e-06,
      "loss": 0.0151,
      "reward": 1.0541667103767396,
      "reward_std": 0.23292546570301056,
      "rewards/accuracy_reward": 0.10416666939854621,
      "rewards/format_reward": 0.9500000238418579,
      "step": 1935
    },
    {
      "completion_length": 174.57917175292968,
      "epoch": 0.776,
      "grad_norm": 0.6399864262576418,
      "kl": 0.36796875,
      "learning_rate": 2.9042926346347932e-06,
      "loss": 0.0147,
      "reward": 1.104166704416275,
      "reward_std": 0.13599317967891694,
      "rewards/accuracy_reward": 0.12083333805203438,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1940
    },
    {
      "completion_length": 227.00000381469727,
      "epoch": 0.778,
      "grad_norm": 0.8129271898918838,
      "kl": 0.32607421875,
      "learning_rate": 2.855273203671969e-06,
      "loss": 0.013,
      "reward": 1.0708333492279052,
      "reward_std": 0.2076343297958374,
      "rewards/accuracy_reward": 0.11250000186264515,
      "rewards/format_reward": 0.9583333551883697,
      "step": 1945
    },
    {
      "completion_length": 264.7666717529297,
      "epoch": 0.78,
      "grad_norm": 0.8065918747460696,
      "kl": 0.3568359375,
      "learning_rate": 2.8066019966134907e-06,
      "loss": 0.0143,
      "reward": 1.1333333730697632,
      "reward_std": 0.3189865186810493,
      "rewards/accuracy_reward": 0.1875000037252903,
      "rewards/format_reward": 0.9458333671092987,
      "step": 1950
    },
    {
      "completion_length": 233.88334197998046,
      "epoch": 0.782,
      "grad_norm": 1.0111951381404745,
      "kl": 0.3408203125,
      "learning_rate": 2.7582813856253276e-06,
      "loss": 0.0136,
      "reward": 1.0375000298023225,
      "reward_std": 0.20281226933002472,
      "rewards/accuracy_reward": 0.07083333507180214,
      "rewards/format_reward": 0.9666666924953461,
      "step": 1955
    },
    {
      "completion_length": 168.03333892822266,
      "epoch": 0.784,
      "grad_norm": 0.8785903776163005,
      "kl": 0.534375,
      "learning_rate": 2.7103137257858867e-06,
      "loss": 0.0213,
      "reward": 1.0958333730697631,
      "reward_std": 0.2230503059923649,
      "rewards/accuracy_reward": 0.10833333730697632,
      "rewards/format_reward": 0.987500011920929,
      "step": 1960
    },
    {
      "completion_length": 201.39167327880858,
      "epoch": 0.786,
      "grad_norm": 0.47972834947637666,
      "kl": 0.37578125,
      "learning_rate": 2.6627013549712355e-06,
      "loss": 0.015,
      "reward": 1.0958333671092988,
      "reward_std": 0.19642044231295586,
      "rewards/accuracy_reward": 0.1291666716337204,
      "rewards/format_reward": 0.9666666805744171,
      "step": 1965
    },
    {
      "completion_length": 172.42083740234375,
      "epoch": 0.788,
      "grad_norm": 0.8266813840530551,
      "kl": 0.3822265625,
      "learning_rate": 2.615446593741161e-06,
      "loss": 0.0153,
      "reward": 1.1416667103767395,
      "reward_std": 0.2167830340564251,
      "rewards/accuracy_reward": 0.16250000484287738,
      "rewards/format_reward": 0.9791666865348816,
      "step": 1970
    },
    {
      "completion_length": 171.02917098999023,
      "epoch": 0.79,
      "grad_norm": 1.3199547899633295,
      "kl": 0.3583984375,
      "learning_rate": 2.5685517452260566e-06,
      "loss": 0.0143,
      "reward": 1.108333373069763,
      "reward_std": 0.2030666373670101,
      "rewards/accuracy_reward": 0.12500000596046448,
      "rewards/format_reward": 0.9833333492279053,
      "step": 1975
    },
    {
      "completion_length": 166.7291702270508,
      "epoch": 0.792,
      "grad_norm": 0.9825519168859947,
      "kl": 0.3490234375,
      "learning_rate": 2.522019095014683e-06,
      "loss": 0.014,
      "reward": 1.1291667103767395,
      "reward_std": 0.19670166373252868,
      "rewards/accuracy_reward": 0.14583333693444728,
      "rewards/format_reward": 0.9833333432674408,
      "step": 1980
    },
    {
      "completion_length": 197.48750686645508,
      "epoch": 0.794,
      "grad_norm": 1.0796352348885683,
      "kl": 0.333203125,
      "learning_rate": 2.4758509110427576e-06,
      "loss": 0.0133,
      "reward": 1.1166666865348815,
      "reward_std": 0.22812673598527908,
      "rewards/accuracy_reward": 0.1375000037252903,
      "rewards/format_reward": 0.9791666865348816,
      "step": 1985
    },
    {
      "completion_length": 179.22500305175782,
      "epoch": 0.796,
      "grad_norm": 0.966571682281471,
      "kl": 0.36171875,
      "learning_rate": 2.4300494434824373e-06,
      "loss": 0.0145,
      "reward": 1.1458333730697632,
      "reward_std": 0.16198743879795074,
      "rewards/accuracy_reward": 0.15833333879709244,
      "rewards/format_reward": 0.987500011920929,
      "step": 1990
    },
    {
      "completion_length": 191.09584197998046,
      "epoch": 0.798,
      "grad_norm": 0.9230407106995351,
      "kl": 0.3421875,
      "learning_rate": 2.3846169246326345e-06,
      "loss": 0.0137,
      "reward": 1.1708333611488342,
      "reward_std": 0.24694958627223967,
      "rewards/accuracy_reward": 0.17916667088866234,
      "rewards/format_reward": 0.9916666746139526,
      "step": 1995
    },
    {
      "completion_length": 207.94167251586913,
      "epoch": 0.8,
      "grad_norm": 0.8543610763790189,
      "kl": 0.3279296875,
      "learning_rate": 2.339555568810221e-06,
      "loss": 0.0131,
      "reward": 1.1958333730697632,
      "reward_std": 0.2837587997317314,
      "rewards/accuracy_reward": 0.22083333916962147,
      "rewards/format_reward": 0.9750000238418579,
      "step": 2000
    },
    {
      "epoch": 0.8,
      "eval_completion_length": 193.92628831129807,
      "eval_kl": 0.3342848557692308,
      "eval_loss": 0.014630368910729885,
      "eval_reward": 1.181089772627904,
      "eval_reward_std": 0.23239356164748853,
      "eval_rewards/accuracy_reward": 0.19871795220443836,
      "eval_rewards/format_reward": 0.9823718070983887,
      "eval_runtime": 192.2976,
      "eval_samples_per_second": 0.515,
      "eval_steps_per_second": 0.016,
      "step": 2000
    },
    {
      "completion_length": 195.9958396911621,
      "epoch": 0.802,
      "grad_norm": 0.4271754543182463,
      "kl": 0.34921875,
      "learning_rate": 2.2948675722421086e-06,
      "loss": 0.014,
      "reward": 1.200000035762787,
      "reward_std": 0.20581314116716384,
      "rewards/accuracy_reward": 0.22083333991467952,
      "rewards/format_reward": 0.9791666865348816,
      "step": 2005
    },
    {
      "completion_length": 174.7375045776367,
      "epoch": 0.804,
      "grad_norm": 0.9804456879969184,
      "kl": 0.3189453125,
      "learning_rate": 2.2505551129582047e-06,
      "loss": 0.0128,
      "reward": 1.2041666984558106,
      "reward_std": 0.20763080418109894,
      "rewards/accuracy_reward": 0.21666666939854623,
      "rewards/format_reward": 0.987500011920929,
      "step": 2010
    },
    {
      "completion_length": 140.03333740234376,
      "epoch": 0.806,
      "grad_norm": 0.5962293976596295,
      "kl": 0.369140625,
      "learning_rate": 2.206620350685257e-06,
      "loss": 0.0148,
      "reward": 1.0958333611488342,
      "reward_std": 0.10075019598007202,
      "rewards/accuracy_reward": 0.10000000335276127,
      "rewards/format_reward": 0.9958333373069763,
      "step": 2015
    },
    {
      "completion_length": 204.6291748046875,
      "epoch": 0.808,
      "grad_norm": 0.4573375021288064,
      "kl": 0.3125,
      "learning_rate": 2.163065426741603e-06,
      "loss": 0.0125,
      "reward": 1.0583333849906922,
      "reward_std": 0.13136882334947586,
      "rewards/accuracy_reward": 0.07916666977107525,
      "rewards/format_reward": 0.9791666865348816,
      "step": 2020
    },
    {
      "completion_length": 184.0541732788086,
      "epoch": 0.81,
      "grad_norm": 0.5759896485971067,
      "kl": 0.3591796875,
      "learning_rate": 2.119892463932781e-06,
      "loss": 0.0144,
      "reward": 1.0833333730697632,
      "reward_std": 0.1874762825667858,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.9791666865348816,
      "step": 2025
    },
    {
      "completion_length": 205.35833969116212,
      "epoch": 0.812,
      "grad_norm": 0.8719265640204763,
      "kl": 0.336328125,
      "learning_rate": 2.0771035664480944e-06,
      "loss": 0.0135,
      "reward": 1.079166716337204,
      "reward_std": 0.19070877507328987,
      "rewards/accuracy_reward": 0.10833333805203438,
      "rewards/format_reward": 0.9708333432674408,
      "step": 2030
    },
    {
      "completion_length": 181.62083892822267,
      "epoch": 0.814,
      "grad_norm": 0.7446256204044083,
      "kl": 0.321484375,
      "learning_rate": 2.0347008197580376e-06,
      "loss": 0.0129,
      "reward": 1.1708333849906922,
      "reward_std": 0.17628924921154976,
      "rewards/accuracy_reward": 0.17916667237877845,
      "rewards/format_reward": 0.9916666746139526,
      "step": 2035
    },
    {
      "completion_length": 193.7916748046875,
      "epoch": 0.816,
      "grad_norm": 0.9746509464814747,
      "kl": 0.3251953125,
      "learning_rate": 1.9926862905126663e-06,
      "loss": 0.013,
      "reward": 1.0833333790302277,
      "reward_std": 0.1700986571609974,
      "rewards/accuracy_reward": 0.1041666705161333,
      "rewards/format_reward": 0.9791666805744171,
      "step": 2040
    },
    {
      "completion_length": 173.7375015258789,
      "epoch": 0.818,
      "grad_norm": 0.8445304947872159,
      "kl": 0.367578125,
      "learning_rate": 1.95106202644086e-06,
      "loss": 0.0147,
      "reward": 1.1666667222976685,
      "reward_std": 0.16608304530382156,
      "rewards/accuracy_reward": 0.17500000707805158,
      "rewards/format_reward": 0.9916666746139526,
      "step": 2045
    },
    {
      "completion_length": 243.4041732788086,
      "epoch": 0.82,
      "grad_norm": 0.860558343035885,
      "kl": 0.32216796875,
      "learning_rate": 1.9098300562505266e-06,
      "loss": 0.0129,
      "reward": 1.0791666984558106,
      "reward_std": 0.20652016922831534,
      "rewards/accuracy_reward": 0.1208333358168602,
      "rewards/format_reward": 0.9583333551883697,
      "step": 2050
    },
    {
      "completion_length": 214.7291748046875,
      "epoch": 0.822,
      "grad_norm": 0.3936034046066415,
      "kl": 0.325,
      "learning_rate": 1.8689923895297247e-06,
      "loss": 0.013,
      "reward": 1.1208333611488341,
      "reward_std": 0.21031340062618256,
      "rewards/accuracy_reward": 0.13750000223517417,
      "rewards/format_reward": 0.9833333492279053,
      "step": 2055
    },
    {
      "completion_length": 216.96667251586913,
      "epoch": 0.824,
      "grad_norm": 0.7044260701460072,
      "kl": 0.32158203125,
      "learning_rate": 1.8285510166487154e-06,
      "loss": 0.0129,
      "reward": 1.0708333730697632,
      "reward_std": 0.15930703431367874,
      "rewards/accuracy_reward": 0.09166667088866234,
      "rewards/format_reward": 0.9791666865348816,
      "step": 2060
    },
    {
      "completion_length": 263.62917709350586,
      "epoch": 0.826,
      "grad_norm": 0.5498658754659237,
      "kl": 0.3486328125,
      "learning_rate": 1.7885079086629598e-06,
      "loss": 0.0139,
      "reward": 1.012500035762787,
      "reward_std": 0.18825939893722535,
      "rewards/accuracy_reward": 0.06666666828095913,
      "rewards/format_reward": 0.9458333492279053,
      "step": 2065
    },
    {
      "completion_length": 214.28333740234376,
      "epoch": 0.828,
      "grad_norm": 0.6431284953050042,
      "kl": 0.3193359375,
      "learning_rate": 1.7488650172170496e-06,
      "loss": 0.0128,
      "reward": 1.100000023841858,
      "reward_std": 0.2267348773777485,
      "rewards/accuracy_reward": 0.12500000074505807,
      "rewards/format_reward": 0.9750000178813935,
      "step": 2070
    },
    {
      "completion_length": 269.7416748046875,
      "epoch": 0.83,
      "grad_norm": 0.7105813170175688,
      "kl": 0.3037109375,
      "learning_rate": 1.709624274449584e-06,
      "loss": 0.0121,
      "reward": 1.0833333790302277,
      "reward_std": 0.24560517221689224,
      "rewards/accuracy_reward": 0.13333333656191826,
      "rewards/format_reward": 0.9500000178813934,
      "step": 2075
    },
    {
      "completion_length": 230.74583892822267,
      "epoch": 0.832,
      "grad_norm": 0.4826351659850352,
      "kl": 0.33203125,
      "learning_rate": 1.6707875928990059e-06,
      "loss": 0.0133,
      "reward": 1.0958333730697631,
      "reward_std": 0.19321480840444566,
      "rewards/accuracy_reward": 0.12083333805203438,
      "rewards/format_reward": 0.975000011920929,
      "step": 2080
    },
    {
      "completion_length": 250.7125045776367,
      "epoch": 0.834,
      "grad_norm": 0.8155955947722978,
      "kl": 0.3384765625,
      "learning_rate": 1.6323568654103838e-06,
      "loss": 0.0135,
      "reward": 1.1000000417232514,
      "reward_std": 0.2640728458762169,
      "rewards/accuracy_reward": 0.14583333842456342,
      "rewards/format_reward": 0.9541666865348816,
      "step": 2085
    },
    {
      "completion_length": 234.16250610351562,
      "epoch": 0.836,
      "grad_norm": 0.6349006116990963,
      "kl": 0.334765625,
      "learning_rate": 1.5943339650431578e-06,
      "loss": 0.0134,
      "reward": 1.1000000596046449,
      "reward_std": 0.20900287851691246,
      "rewards/accuracy_reward": 0.12916666977107524,
      "rewards/format_reward": 0.9708333551883698,
      "step": 2090
    },
    {
      "completion_length": 280.3166763305664,
      "epoch": 0.838,
      "grad_norm": 0.5436608157021742,
      "kl": 0.31298828125,
      "learning_rate": 1.5567207449798517e-06,
      "loss": 0.0125,
      "reward": 1.104166704416275,
      "reward_std": 0.22294211983680726,
      "rewards/accuracy_reward": 0.15000000558793544,
      "rewards/format_reward": 0.9541666865348816,
      "step": 2095
    },
    {
      "completion_length": 251.92500915527344,
      "epoch": 0.84,
      "grad_norm": 0.9997986753777716,
      "kl": 0.3080078125,
      "learning_rate": 1.5195190384357405e-06,
      "loss": 0.0123,
      "reward": 1.0583333611488341,
      "reward_std": 0.14829438030719758,
      "rewards/accuracy_reward": 0.08750000186264514,
      "rewards/format_reward": 0.9708333492279053,
      "step": 2100
    },
    {
      "epoch": 0.84,
      "eval_completion_length": 230.33814885066107,
      "eval_kl": 0.32181490384615385,
      "eval_loss": 0.012829490937292576,
      "eval_reward": 1.126602590084076,
      "eval_reward_std": 0.23767680273606226,
      "eval_rewards/accuracy_reward": 0.15544871985912323,
      "eval_rewards/format_reward": 0.9711538599087641,
      "eval_runtime": 195.3635,
      "eval_samples_per_second": 0.507,
      "eval_steps_per_second": 0.015,
      "step": 2100
    },
    {
      "completion_length": 210.4125068664551,
      "epoch": 0.842,
      "grad_norm": 0.3475757290411574,
      "kl": 0.3216796875,
      "learning_rate": 1.4827306585695234e-06,
      "loss": 0.0129,
      "reward": 1.0791666865348817,
      "reward_std": 0.16784698367118836,
      "rewards/accuracy_reward": 0.10000000223517418,
      "rewards/format_reward": 0.9791666865348816,
      "step": 2105
    },
    {
      "completion_length": 200.43333892822267,
      "epoch": 0.844,
      "grad_norm": 0.6385339371882108,
      "kl": 0.3228515625,
      "learning_rate": 1.446357398394934e-06,
      "loss": 0.0129,
      "reward": 1.1625000476837157,
      "reward_std": 0.1953062780201435,
      "rewards/accuracy_reward": 0.1750000063329935,
      "rewards/format_reward": 0.987500011920929,
      "step": 2110
    },
    {
      "completion_length": 233.4250045776367,
      "epoch": 0.846,
      "grad_norm": 3.7096325260067315,
      "kl": 0.3455078125,
      "learning_rate": 1.4104010306933558e-06,
      "loss": 0.0138,
      "reward": 1.1166667103767396,
      "reward_std": 0.17760113030672073,
      "rewards/accuracy_reward": 0.1416666701436043,
      "rewards/format_reward": 0.9750000178813935,
      "step": 2115
    },
    {
      "completion_length": 224.95417175292968,
      "epoch": 0.848,
      "grad_norm": 0.5825427147511982,
      "kl": 0.3935546875,
      "learning_rate": 1.3748633079274254e-06,
      "loss": 0.0157,
      "reward": 1.2041667103767395,
      "reward_std": 0.2667228579521179,
      "rewards/accuracy_reward": 0.22500000484287738,
      "rewards/format_reward": 0.9791666805744171,
      "step": 2120
    },
    {
      "completion_length": 254.05833892822267,
      "epoch": 0.85,
      "grad_norm": 1.0034192906781798,
      "kl": 0.3423828125,
      "learning_rate": 1.339745962155613e-06,
      "loss": 0.0137,
      "reward": 1.1083333611488342,
      "reward_std": 0.21382493153214455,
      "rewards/accuracy_reward": 0.1416666679084301,
      "rewards/format_reward": 0.9666666805744171,
      "step": 2125
    },
    {
      "completion_length": 230.9458366394043,
      "epoch": 0.852,
      "grad_norm": 0.8643081521849603,
      "kl": 0.3384765625,
      "learning_rate": 1.30505070494781e-06,
      "loss": 0.0135,
      "reward": 1.141666704416275,
      "reward_std": 0.2681451141834259,
      "rewards/accuracy_reward": 0.16666667126119136,
      "rewards/format_reward": 0.975000011920929,
      "step": 2130
    },
    {
      "completion_length": 192.07917175292968,
      "epoch": 0.854,
      "grad_norm": 0.7868704806845438,
      "kl": 0.34873046875,
      "learning_rate": 1.2707792273019049e-06,
      "loss": 0.0139,
      "reward": 1.1791666984558105,
      "reward_std": 0.18512692525982857,
      "rewards/accuracy_reward": 0.18750000670552253,
      "rewards/format_reward": 0.9916666746139526,
      "step": 2135
    },
    {
      "completion_length": 251.43750610351563,
      "epoch": 0.856,
      "grad_norm": 0.7371049494838817,
      "kl": 0.349609375,
      "learning_rate": 1.2369331995613664e-06,
      "loss": 0.014,
      "reward": 1.079166728258133,
      "reward_std": 0.2230503149330616,
      "rewards/accuracy_reward": 0.10416666939854621,
      "rewards/format_reward": 0.975000011920929,
      "step": 2140
    },
    {
      "completion_length": 199.56250457763673,
      "epoch": 0.858,
      "grad_norm": 0.8177527730479223,
      "kl": 0.3154296875,
      "learning_rate": 1.2035142713338366e-06,
      "loss": 0.0126,
      "reward": 1.1625000357627868,
      "reward_std": 0.20501058548688889,
      "rewards/accuracy_reward": 0.17083333544433116,
      "rewards/format_reward": 0.9916666746139526,
      "step": 2145
    },
    {
      "completion_length": 260.4166732788086,
      "epoch": 0.86,
      "grad_norm": 0.7068503679121204,
      "kl": 0.3138671875,
      "learning_rate": 1.1705240714107301e-06,
      "loss": 0.0126,
      "reward": 1.1208333730697633,
      "reward_std": 0.25823956951498983,
      "rewards/accuracy_reward": 0.14583333544433116,
      "rewards/format_reward": 0.975000011920929,
      "step": 2150
    },
    {
      "completion_length": 259.20417938232424,
      "epoch": 0.862,
      "grad_norm": 0.42884092133676044,
      "kl": 0.3365234375,
      "learning_rate": 1.1379642076878528e-06,
      "loss": 0.0135,
      "reward": 1.0750000417232513,
      "reward_std": 0.20662664845585824,
      "rewards/accuracy_reward": 0.10416667088866234,
      "rewards/format_reward": 0.9708333551883698,
      "step": 2155
    },
    {
      "completion_length": 262.93333892822267,
      "epoch": 0.864,
      "grad_norm": 0.8152710369638511,
      "kl": 0.3275390625,
      "learning_rate": 1.1058362670870248e-06,
      "loss": 0.0131,
      "reward": 1.0583333730697633,
      "reward_std": 0.17358550801873207,
      "rewards/accuracy_reward": 0.08750000074505807,
      "rewards/format_reward": 0.9708333492279053,
      "step": 2160
    },
    {
      "completion_length": 247.0250045776367,
      "epoch": 0.866,
      "grad_norm": 0.688001798775682,
      "kl": 0.3095703125,
      "learning_rate": 1.0741418154787443e-06,
      "loss": 0.0124,
      "reward": 1.1208333849906922,
      "reward_std": 0.2076909951865673,
      "rewards/accuracy_reward": 0.13750000186264516,
      "rewards/format_reward": 0.9833333432674408,
      "step": 2165
    },
    {
      "completion_length": 288.39583892822264,
      "epoch": 0.868,
      "grad_norm": 0.5822864693647378,
      "kl": 0.30859375,
      "learning_rate": 1.042882397605871e-06,
      "loss": 0.0124,
      "reward": 1.0625000417232513,
      "reward_std": 0.18209213241934777,
      "rewards/accuracy_reward": 0.09583333618938923,
      "rewards/format_reward": 0.9666666865348816,
      "step": 2170
    },
    {
      "completion_length": 245.29584045410155,
      "epoch": 0.87,
      "grad_norm": 0.5072380090682689,
      "kl": 0.311328125,
      "learning_rate": 1.012059537008332e-06,
      "loss": 0.0125,
      "reward": 1.0875000476837158,
      "reward_std": 0.19070877134799957,
      "rewards/accuracy_reward": 0.11250000409781932,
      "rewards/format_reward": 0.9750000238418579,
      "step": 2175
    },
    {
      "completion_length": 276.4583374023438,
      "epoch": 0.872,
      "grad_norm": 0.4273306679972882,
      "kl": 0.32421875,
      "learning_rate": 9.816747359488632e-07,
      "loss": 0.013,
      "reward": 1.1875000476837159,
      "reward_std": 0.23159026056528093,
      "rewards/accuracy_reward": 0.22916667200624943,
      "rewards/format_reward": 0.9583333432674408,
      "step": 2180
    },
    {
      "completion_length": 244.8000045776367,
      "epoch": 0.874,
      "grad_norm": 0.6004900969175896,
      "kl": 0.315625,
      "learning_rate": 9.517294753398066e-07,
      "loss": 0.0126,
      "reward": 1.1333333730697632,
      "reward_std": 0.23348402082920075,
      "rewards/accuracy_reward": 0.15833333879709244,
      "rewards/format_reward": 0.9750000238418579,
      "step": 2185
    },
    {
      "completion_length": 277.64584197998045,
      "epoch": 0.876,
      "grad_norm": 0.49930191441234933,
      "kl": 0.31064453125,
      "learning_rate": 9.222252146709143e-07,
      "loss": 0.0124,
      "reward": 1.1375000298023223,
      "reward_std": 0.19854526817798615,
      "rewards/accuracy_reward": 0.17500000707805158,
      "rewards/format_reward": 0.9625000178813934,
      "step": 2190
    },
    {
      "completion_length": 259.41250762939455,
      "epoch": 0.878,
      "grad_norm": 1.2303820073949265,
      "kl": 0.3111328125,
      "learning_rate": 8.931633919382299e-07,
      "loss": 0.0125,
      "reward": 1.0958333730697631,
      "reward_std": 0.26156365871429443,
      "rewards/accuracy_reward": 0.13750000409781932,
      "rewards/format_reward": 0.9583333492279053,
      "step": 2195
    },
    {
      "completion_length": 248.02083892822264,
      "epoch": 0.88,
      "grad_norm": 0.8176257324608204,
      "kl": 0.318359375,
      "learning_rate": 8.645454235739903e-07,
      "loss": 0.0127,
      "reward": 1.1000000357627868,
      "reward_std": 0.23674337863922118,
      "rewards/accuracy_reward": 0.13333333730697633,
      "rewards/format_reward": 0.9666666805744171,
      "step": 2200
    },
    {
      "epoch": 0.88,
      "eval_completion_length": 236.52725102351263,
      "eval_kl": 0.3191105769230769,
      "eval_loss": 0.013348034583032131,
      "eval_reward": 1.158653892003573,
      "eval_reward_std": 0.21606751072865266,
      "eval_rewards/accuracy_reward": 0.17628205452974027,
      "eval_rewards/format_reward": 0.9823718116833613,
      "eval_runtime": 205.161,
      "eval_samples_per_second": 0.483,
      "eval_steps_per_second": 0.015,
      "step": 2200
    },
    {
      "completion_length": 243.39167175292968,
      "epoch": 0.882,
      "grad_norm": 0.724908011352607,
      "kl": 0.2978515625,
      "learning_rate": 8.363727043776037e-07,
      "loss": 0.0119,
      "reward": 1.112500011920929,
      "reward_std": 0.17489738911390304,
      "rewards/accuracy_reward": 0.12916666753590106,
      "rewards/format_reward": 0.9833333492279053,
      "step": 2205
    },
    {
      "completion_length": 250.98750762939454,
      "epoch": 0.884,
      "grad_norm": 0.6527429694726796,
      "kl": 0.3375,
      "learning_rate": 8.086466074476562e-07,
      "loss": 0.0135,
      "reward": 1.0916667044162751,
      "reward_std": 0.2350437968969345,
      "rewards/accuracy_reward": 0.12916667051613331,
      "rewards/format_reward": 0.9625000119209289,
      "step": 2210
    },
    {
      "completion_length": 203.03750457763672,
      "epoch": 0.886,
      "grad_norm": 0.8656053781596581,
      "kl": 0.3228515625,
      "learning_rate": 7.81368484114996e-07,
      "loss": 0.0129,
      "reward": 1.1625000596046449,
      "reward_std": 0.1544849768280983,
      "rewards/accuracy_reward": 0.17916667200624942,
      "rewards/format_reward": 0.9833333492279053,
      "step": 2215
    },
    {
      "completion_length": 253.7791732788086,
      "epoch": 0.888,
      "grad_norm": 0.7886606594118118,
      "kl": 0.3357421875,
      "learning_rate": 7.545396638768698e-07,
      "loss": 0.0134,
      "reward": 1.0958333730697631,
      "reward_std": 0.24795775413513182,
      "rewards/accuracy_reward": 0.13750000447034835,
      "rewards/format_reward": 0.9583333492279053,
      "step": 2220
    },
    {
      "completion_length": 224.60834121704102,
      "epoch": 0.89,
      "grad_norm": 0.7745032909827045,
      "kl": 0.328125,
      "learning_rate": 7.281614543321269e-07,
      "loss": 0.0131,
      "reward": 1.158333384990692,
      "reward_std": 0.2810515329241753,
      "rewards/accuracy_reward": 0.17916667237877845,
      "rewards/format_reward": 0.9791666805744171,
      "step": 2225
    },
    {
      "completion_length": 179.68333740234374,
      "epoch": 0.892,
      "grad_norm": 0.3205028122076133,
      "kl": 0.31044921875,
      "learning_rate": 7.022351411174866e-07,
      "loss": 0.0124,
      "reward": 1.2250000357627868,
      "reward_std": 0.18841607943177224,
      "rewards/accuracy_reward": 0.2250000037252903,
      "rewards/format_reward": 1.0,
      "step": 2230
    },
    {
      "completion_length": 188.59167022705077,
      "epoch": 0.894,
      "grad_norm": 0.8894666456826428,
      "kl": 0.308203125,
      "learning_rate": 6.767619878448783e-07,
      "loss": 0.0123,
      "reward": 1.2125000357627869,
      "reward_std": 0.18571233451366426,
      "rewards/accuracy_reward": 0.2166666727513075,
      "rewards/format_reward": 0.9958333373069763,
      "step": 2235
    },
    {
      "completion_length": 228.5041717529297,
      "epoch": 0.896,
      "grad_norm": 0.5946306291186744,
      "kl": 0.30029296875,
      "learning_rate": 6.517432360398556e-07,
      "loss": 0.012,
      "reward": 1.1791666984558105,
      "reward_std": 0.2665520176291466,
      "rewards/accuracy_reward": 0.20416666939854622,
      "rewards/format_reward": 0.9750000178813935,
      "step": 2240
    },
    {
      "completion_length": 186.78333740234376,
      "epoch": 0.898,
      "grad_norm": 0.8312102347233332,
      "kl": 0.3220703125,
      "learning_rate": 6.271801050810856e-07,
      "loss": 0.0129,
      "reward": 1.187500035762787,
      "reward_std": 0.18710420057177543,
      "rewards/accuracy_reward": 0.18750000596046448,
      "rewards/format_reward": 1.0,
      "step": 2245
    },
    {
      "completion_length": 216.27083892822264,
      "epoch": 0.9,
      "grad_norm": 1.0188499034367569,
      "kl": 0.31484375,
      "learning_rate": 6.030737921409169e-07,
      "loss": 0.0126,
      "reward": 1.1250000238418578,
      "reward_std": 0.20551247671246528,
      "rewards/accuracy_reward": 0.14583333805203438,
      "rewards/format_reward": 0.9791666746139527,
      "step": 2250
    },
    {
      "completion_length": 229.68334045410157,
      "epoch": 0.902,
      "grad_norm": 0.3466462767720492,
      "kl": 0.297265625,
      "learning_rate": 5.794254721270331e-07,
      "loss": 0.0119,
      "reward": 1.1000000357627868,
      "reward_std": 0.20342101901769638,
      "rewards/accuracy_reward": 0.11250000149011612,
      "rewards/format_reward": 0.987500011920929,
      "step": 2255
    },
    {
      "completion_length": 225.60000610351562,
      "epoch": 0.904,
      "grad_norm": 0.5577180539267683,
      "kl": 0.3216796875,
      "learning_rate": 5.562362976251901e-07,
      "loss": 0.0129,
      "reward": 1.1208333849906922,
      "reward_std": 0.1947810523211956,
      "rewards/accuracy_reward": 0.1416666701436043,
      "rewards/format_reward": 0.9791666865348816,
      "step": 2260
    },
    {
      "completion_length": 253.82501068115235,
      "epoch": 0.906,
      "grad_norm": 0.8118903219041304,
      "kl": 0.31328125,
      "learning_rate": 5.335073988430373e-07,
      "loss": 0.0125,
      "reward": 1.1166667222976685,
      "reward_std": 0.20122136250138284,
      "rewards/accuracy_reward": 0.13750000298023224,
      "rewards/format_reward": 0.9791666805744171,
      "step": 2265
    },
    {
      "completion_length": 167.62500457763673,
      "epoch": 0.908,
      "grad_norm": 0.5609148058927211,
      "kl": 0.30732421875,
      "learning_rate": 5.112398835550348e-07,
      "loss": 0.0123,
      "reward": 1.1500000476837158,
      "reward_std": 0.18108798265457154,
      "rewards/accuracy_reward": 0.15416667200624942,
      "rewards/format_reward": 0.9958333373069763,
      "step": 2270
    },
    {
      "completion_length": 262.2083404541016,
      "epoch": 0.91,
      "grad_norm": 0.40905385430671654,
      "kl": 0.3474609375,
      "learning_rate": 4.894348370484648e-07,
      "loss": 0.0139,
      "reward": 1.1041666984558105,
      "reward_std": 0.2402033679187298,
      "rewards/accuracy_reward": 0.15000000409781933,
      "rewards/format_reward": 0.9541666805744171,
      "step": 2275
    },
    {
      "completion_length": 223.8916763305664,
      "epoch": 0.912,
      "grad_norm": 0.5281590255695859,
      "kl": 0.3130859375,
      "learning_rate": 4.6809332207053083e-07,
      "loss": 0.0125,
      "reward": 1.2041667222976684,
      "reward_std": 0.23390628695487975,
      "rewards/accuracy_reward": 0.2250000037252903,
      "rewards/format_reward": 0.9791666865348816,
      "step": 2280
    },
    {
      "completion_length": 250.34584350585936,
      "epoch": 0.914,
      "grad_norm": 1.0063118812707683,
      "kl": 0.3205078125,
      "learning_rate": 4.4721637877656377e-07,
      "loss": 0.0128,
      "reward": 1.0708333611488343,
      "reward_std": 0.1668095014989376,
      "rewards/accuracy_reward": 0.10000000149011612,
      "rewards/format_reward": 0.9708333551883698,
      "step": 2285
    },
    {
      "completion_length": 219.55417251586914,
      "epoch": 0.916,
      "grad_norm": 0.5679221458389377,
      "kl": 0.322265625,
      "learning_rate": 4.268050246793276e-07,
      "loss": 0.0129,
      "reward": 1.287500023841858,
      "reward_std": 0.2404845893383026,
      "rewards/accuracy_reward": 0.3000000104308128,
      "rewards/format_reward": 0.987500011920929,
      "step": 2290
    },
    {
      "completion_length": 249.7291748046875,
      "epoch": 0.918,
      "grad_norm": 0.37307292971981587,
      "kl": 0.296875,
      "learning_rate": 4.068602545994249e-07,
      "loss": 0.0119,
      "reward": 1.0250000298023223,
      "reward_std": 0.13485567793250083,
      "rewards/accuracy_reward": 0.05000000149011612,
      "rewards/format_reward": 0.9750000059604644,
      "step": 2295
    },
    {
      "completion_length": 275.1625091552734,
      "epoch": 0.92,
      "grad_norm": 1.4329667808726139,
      "kl": 0.314453125,
      "learning_rate": 3.8738304061681107e-07,
      "loss": 0.0126,
      "reward": 1.125000035762787,
      "reward_std": 0.22324448227882385,
      "rewards/accuracy_reward": 0.16250000521540642,
      "rewards/format_reward": 0.9625000178813934,
      "step": 2300
    },
    {
      "epoch": 0.92,
      "eval_completion_length": 234.14583998460037,
      "eval_kl": 0.32827524038461536,
      "eval_loss": 0.014351727440953255,
      "eval_reward": 1.1330128541359534,
      "eval_reward_std": 0.2333221613214566,
      "eval_rewards/accuracy_reward": 0.16346154433603471,
      "eval_rewards/format_reward": 0.9695513065044696,
      "eval_runtime": 215.8259,
      "eval_samples_per_second": 0.459,
      "eval_steps_per_second": 0.014,
      "step": 2300
    },
    {
      "completion_length": 304.00000915527346,
      "epoch": 0.922,
      "grad_norm": 0.8276825063442558,
      "kl": 0.2984375,
      "learning_rate": 3.68374332023419e-07,
      "loss": 0.0119,
      "reward": 1.045833373069763,
      "reward_std": 0.2611467272043228,
      "rewards/accuracy_reward": 0.10416667014360428,
      "rewards/format_reward": 0.9416666984558105,
      "step": 2305
    },
    {
      "completion_length": 239.92084045410155,
      "epoch": 0.924,
      "grad_norm": 0.5776581401348798,
      "kl": 0.3056640625,
      "learning_rate": 3.498350552768859e-07,
      "loss": 0.0122,
      "reward": 1.1250000476837159,
      "reward_std": 0.21731178238987922,
      "rewards/accuracy_reward": 0.14166666977107525,
      "rewards/format_reward": 0.9833333432674408,
      "step": 2310
    },
    {
      "completion_length": 243.52084350585938,
      "epoch": 0.926,
      "grad_norm": 0.5717808613340961,
      "kl": 0.28369140625,
      "learning_rate": 3.3176611395540625e-07,
      "loss": 0.0114,
      "reward": 1.0875000357627869,
      "reward_std": 0.18861377909779548,
      "rewards/accuracy_reward": 0.1083333358168602,
      "rewards/format_reward": 0.9791666805744171,
      "step": 2315
    },
    {
      "completion_length": 230.64167327880858,
      "epoch": 0.928,
      "grad_norm": 0.8434649020063801,
      "kl": 0.33193359375,
      "learning_rate": 3.1416838871368925e-07,
      "loss": 0.0133,
      "reward": 1.2041666984558106,
      "reward_std": 0.2520026877522469,
      "rewards/accuracy_reward": 0.22500000447034835,
      "rewards/format_reward": 0.9791666805744171,
      "step": 2320
    },
    {
      "completion_length": 271.32084197998046,
      "epoch": 0.93,
      "grad_norm": 0.6927657151612084,
      "kl": 0.3126953125,
      "learning_rate": 2.970427372400353e-07,
      "loss": 0.0125,
      "reward": 1.0708333730697632,
      "reward_std": 0.21362721994519235,
      "rewards/accuracy_reward": 0.11250000335276127,
      "rewards/format_reward": 0.9583333492279053,
      "step": 2325
    },
    {
      "completion_length": 287.02084197998045,
      "epoch": 0.932,
      "grad_norm": 0.5476312804811295,
      "kl": 0.31416015625,
      "learning_rate": 2.8038999421453827e-07,
      "loss": 0.0126,
      "reward": 1.1083333551883698,
      "reward_std": 0.21203412041068076,
      "rewards/accuracy_reward": 0.15416667200624942,
      "rewards/format_reward": 0.9541666746139527,
      "step": 2330
    },
    {
      "completion_length": 286.46251220703124,
      "epoch": 0.934,
      "grad_norm": 0.6773822629722608,
      "kl": 0.3408203125,
      "learning_rate": 2.6421097126839714e-07,
      "loss": 0.0137,
      "reward": 1.0875000417232514,
      "reward_std": 0.22804322466254234,
      "rewards/accuracy_reward": 0.12916667126119136,
      "rewards/format_reward": 0.9583333551883697,
      "step": 2335
    },
    {
      "completion_length": 233.5666748046875,
      "epoch": 0.936,
      "grad_norm": 0.9631119728257711,
      "kl": 0.3609375,
      "learning_rate": 2.4850645694436736e-07,
      "loss": 0.0144,
      "reward": 1.1500000357627869,
      "reward_std": 0.23964481428265572,
      "rewards/accuracy_reward": 0.17083333805203438,
      "rewards/format_reward": 0.9791666805744171,
      "step": 2340
    },
    {
      "completion_length": 256.07084197998046,
      "epoch": 0.938,
      "grad_norm": 0.5283231336286257,
      "kl": 0.30322265625,
      "learning_rate": 2.332772166583208e-07,
      "loss": 0.0121,
      "reward": 1.112500047683716,
      "reward_std": 0.1783842422068119,
      "rewards/accuracy_reward": 0.14166667088866233,
      "rewards/format_reward": 0.9708333492279053,
      "step": 2345
    },
    {
      "completion_length": 256.93334045410154,
      "epoch": 0.94,
      "grad_norm": 0.7126466204158886,
      "kl": 0.3,
      "learning_rate": 2.1852399266194312e-07,
      "loss": 0.012,
      "reward": 1.0958333611488342,
      "reward_std": 0.23403963819146156,
      "rewards/accuracy_reward": 0.1208333358168602,
      "rewards/format_reward": 0.9750000178813935,
      "step": 2350
    },
    {
      "completion_length": 238.1291732788086,
      "epoch": 0.942,
      "grad_norm": 0.6637433721941498,
      "kl": 0.305859375,
      "learning_rate": 2.0424750400655947e-07,
      "loss": 0.0122,
      "reward": 1.1291667103767395,
      "reward_std": 0.21683285459876062,
      "rewards/accuracy_reward": 0.1541666716337204,
      "rewards/format_reward": 0.9750000238418579,
      "step": 2355
    },
    {
      "completion_length": 245.8625061035156,
      "epoch": 0.944,
      "grad_norm": 0.8413001069576047,
      "kl": 0.3373046875,
      "learning_rate": 1.9044844650808468e-07,
      "loss": 0.0135,
      "reward": 1.0541666924953461,
      "reward_std": 0.18833255916833877,
      "rewards/accuracy_reward": 0.08750000223517418,
      "rewards/format_reward": 0.9666666865348816,
      "step": 2360
    },
    {
      "completion_length": 262.5250045776367,
      "epoch": 0.946,
      "grad_norm": 0.6316109309990907,
      "kl": 0.3193359375,
      "learning_rate": 1.7712749271311392e-07,
      "loss": 0.0128,
      "reward": 1.0833333671092986,
      "reward_std": 0.20121916756033897,
      "rewards/accuracy_reward": 0.11250000186264515,
      "rewards/format_reward": 0.9708333492279053,
      "step": 2365
    },
    {
      "completion_length": 243.39167175292968,
      "epoch": 0.948,
      "grad_norm": 0.6991473147311043,
      "kl": 0.3158203125,
      "learning_rate": 1.6428529186614195e-07,
      "loss": 0.0126,
      "reward": 1.2166667103767395,
      "reward_std": 0.2617532812058926,
      "rewards/accuracy_reward": 0.23750000447034836,
      "rewards/format_reward": 0.9791666746139527,
      "step": 2370
    },
    {
      "completion_length": 248.2041748046875,
      "epoch": 0.95,
      "grad_norm": 0.9976654981267077,
      "kl": 0.29189453125,
      "learning_rate": 1.519224698779198e-07,
      "loss": 0.0117,
      "reward": 1.129166716337204,
      "reward_std": 0.21251304894685746,
      "rewards/accuracy_reward": 0.1541666705161333,
      "rewards/format_reward": 0.975000011920929,
      "step": 2375
    },
    {
      "completion_length": 236.87084197998047,
      "epoch": 0.952,
      "grad_norm": 1.025244017091307,
      "kl": 0.312890625,
      "learning_rate": 1.400396292949513e-07,
      "loss": 0.0125,
      "reward": 1.0833333730697632,
      "reward_std": 0.17568050250411033,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.9791666865348816,
      "step": 2380
    },
    {
      "completion_length": 239.7541717529297,
      "epoch": 0.954,
      "grad_norm": 0.5280779053111202,
      "kl": 0.29990234375,
      "learning_rate": 1.2863734927012094e-07,
      "loss": 0.012,
      "reward": 1.1083333790302277,
      "reward_std": 0.2078887030482292,
      "rewards/accuracy_reward": 0.13333333879709244,
      "rewards/format_reward": 0.975000011920929,
      "step": 2385
    },
    {
      "completion_length": 257.35001068115236,
      "epoch": 0.956,
      "grad_norm": 0.7506657762258959,
      "kl": 0.33359375,
      "learning_rate": 1.1771618553447217e-07,
      "loss": 0.0133,
      "reward": 1.1083333611488342,
      "reward_std": 0.27584577947854994,
      "rewards/accuracy_reward": 0.15000000260770321,
      "rewards/format_reward": 0.9583333492279053,
      "step": 2390
    },
    {
      "completion_length": 253.93750457763673,
      "epoch": 0.958,
      "grad_norm": 0.42898933144776946,
      "kl": 0.3265625,
      "learning_rate": 1.0727667037011668e-07,
      "loss": 0.0131,
      "reward": 1.062500035762787,
      "reward_std": 0.1832063004374504,
      "rewards/accuracy_reward": 0.08750000223517418,
      "rewards/format_reward": 0.9750000238418579,
      "step": 2395
    },
    {
      "completion_length": 204.62084045410157,
      "epoch": 0.96,
      "grad_norm": 0.7711065338130425,
      "kl": 0.33046875,
      "learning_rate": 9.731931258429638e-08,
      "loss": 0.0132,
      "reward": 1.1500000417232514,
      "reward_std": 0.15125247687101365,
      "rewards/accuracy_reward": 0.1708333395421505,
      "rewards/format_reward": 0.9791666805744171,
      "step": 2400
    },
    {
      "epoch": 0.96,
      "eval_completion_length": 232.98879065880408,
      "eval_kl": 0.32271634615384615,
      "eval_loss": 0.013007663190364838,
      "eval_reward": 1.1506410562075102,
      "eval_reward_std": 0.24151590065314218,
      "eval_rewards/accuracy_reward": 0.17788461996958807,
      "eval_rewards/format_reward": 0.9727564270679767,
      "eval_runtime": 214.5988,
      "eval_samples_per_second": 0.461,
      "eval_steps_per_second": 0.014,
      "step": 2400
    },
    {
      "completion_length": 285.9666748046875,
      "epoch": 0.962,
      "grad_norm": 0.21086359841471036,
      "kl": 0.30234375,
      "learning_rate": 8.784459748458318e-08,
      "loss": 0.0121,
      "reward": 1.0583333611488341,
      "reward_std": 0.1348556734621525,
      "rewards/accuracy_reward": 0.09583333395421505,
      "rewards/format_reward": 0.9625000298023224,
      "step": 2405
    },
    {
      "completion_length": 255.64167327880858,
      "epoch": 0.964,
      "grad_norm": 1.1574014329263573,
      "kl": 0.3009765625,
      "learning_rate": 7.885298685522235e-08,
      "loss": 0.012,
      "reward": 1.0875000357627869,
      "reward_std": 0.23103000968694687,
      "rewards/accuracy_reward": 0.12083333767950535,
      "rewards/format_reward": 0.9666666924953461,
      "step": 2410
    },
    {
      "completion_length": 253.06250915527343,
      "epoch": 0.966,
      "grad_norm": 0.5696899940020976,
      "kl": 0.3279296875,
      "learning_rate": 7.034491893463059e-08,
      "loss": 0.0131,
      "reward": 1.0791667103767395,
      "reward_std": 0.21763930991292,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.9750000178813935,
      "step": 2415
    },
    {
      "completion_length": 225.98750762939454,
      "epoch": 0.968,
      "grad_norm": 0.6929231238436696,
      "kl": 0.318359375,
      "learning_rate": 6.232080839403631e-08,
      "loss": 0.0127,
      "reward": 1.1458333611488343,
      "reward_std": 0.19018001332879067,
      "rewards/accuracy_reward": 0.1625000026077032,
      "rewards/format_reward": 0.9833333492279053,
      "step": 2420
    },
    {
      "completion_length": 219.8541748046875,
      "epoch": 0.97,
      "grad_norm": 0.6209077338105314,
      "kl": 0.3103515625,
      "learning_rate": 5.4781046317267103e-08,
      "loss": 0.0124,
      "reward": 1.1625000357627868,
      "reward_std": 0.2506072849035263,
      "rewards/accuracy_reward": 0.17916667014360427,
      "rewards/format_reward": 0.9833333492279053,
      "step": 2425
    },
    {
      "completion_length": 297.3958374023438,
      "epoch": 0.972,
      "grad_norm": 0.4301111358616339,
      "kl": 0.31669921875,
      "learning_rate": 4.772600018168816e-08,
      "loss": 0.0127,
      "reward": 1.016666692495346,
      "reward_std": 0.20313626676797866,
      "rewards/accuracy_reward": 0.07083333507180214,
      "rewards/format_reward": 0.9458333492279053,
      "step": 2430
    },
    {
      "completion_length": 209.98750762939454,
      "epoch": 0.974,
      "grad_norm": 0.39213515299137125,
      "kl": 0.306640625,
      "learning_rate": 4.115601384029666e-08,
      "loss": 0.0123,
      "reward": 1.1333333730697632,
      "reward_std": 0.18591004163026809,
      "rewards/accuracy_reward": 0.14583334028720857,
      "rewards/format_reward": 0.987500011920929,
      "step": 2435
    },
    {
      "completion_length": 215.95000457763672,
      "epoch": 0.976,
      "grad_norm": 1.130278072056706,
      "kl": 0.3189453125,
      "learning_rate": 3.50714075049563e-08,
      "loss": 0.0128,
      "reward": 1.2416667222976685,
      "reward_std": 0.324382396042347,
      "rewards/accuracy_reward": 0.2666666753590107,
      "rewards/format_reward": 0.975000011920929,
      "step": 2440
    },
    {
      "completion_length": 282.88750915527345,
      "epoch": 0.978,
      "grad_norm": 0.2692108788435084,
      "kl": 0.296875,
      "learning_rate": 2.947247773079753e-08,
      "loss": 0.0119,
      "reward": 1.0833333671092986,
      "reward_std": 0.153701850771904,
      "rewards/accuracy_reward": 0.11666667014360428,
      "rewards/format_reward": 0.9666666805744171,
      "step": 2445
    },
    {
      "completion_length": 242.4875045776367,
      "epoch": 0.98,
      "grad_norm": 0.6295793167108874,
      "kl": 0.29150390625,
      "learning_rate": 2.4359497401758026e-08,
      "loss": 0.0117,
      "reward": 1.2041667103767395,
      "reward_std": 0.2249172128736973,
      "rewards/accuracy_reward": 0.23333334028720856,
      "rewards/format_reward": 0.9708333492279053,
      "step": 2450
    },
    {
      "completion_length": 300.2125076293945,
      "epoch": 0.982,
      "grad_norm": 0.765547881156069,
      "kl": 0.3287109375,
      "learning_rate": 1.973271571728441e-08,
      "loss": 0.0131,
      "reward": 1.0958333730697631,
      "reward_std": 0.30428186282515524,
      "rewards/accuracy_reward": 0.1541666716337204,
      "rewards/format_reward": 0.9416666924953461,
      "step": 2455
    },
    {
      "completion_length": 226.41250762939453,
      "epoch": 0.984,
      "grad_norm": 0.3857577216711371,
      "kl": 0.317578125,
      "learning_rate": 1.5592358180189782e-08,
      "loss": 0.0127,
      "reward": 1.0625000238418578,
      "reward_std": 0.12517823055386543,
      "rewards/accuracy_reward": 0.07500000149011612,
      "rewards/format_reward": 0.987500011920929,
      "step": 2460
    },
    {
      "completion_length": 227.72500610351562,
      "epoch": 0.986,
      "grad_norm": 0.6046244061961439,
      "kl": 0.33046875,
      "learning_rate": 1.1938626585660252e-08,
      "loss": 0.0132,
      "reward": 1.0916667103767395,
      "reward_std": 0.19912418201565743,
      "rewards/accuracy_reward": 0.11250000260770321,
      "rewards/format_reward": 0.9791666805744171,
      "step": 2465
    },
    {
      "completion_length": 221.62083892822267,
      "epoch": 0.988,
      "grad_norm": 0.6204160140005214,
      "kl": 0.29775390625,
      "learning_rate": 8.771699011416169e-09,
      "loss": 0.0119,
      "reward": 1.1333333730697632,
      "reward_std": 0.22539967224001883,
      "rewards/accuracy_reward": 0.15416667237877846,
      "rewards/format_reward": 0.9791666805744171,
      "step": 2470
    },
    {
      "completion_length": 210.60834045410155,
      "epoch": 0.99,
      "grad_norm": 1.1403201560013576,
      "kl": 0.3005859375,
      "learning_rate": 6.091729809042379e-09,
      "loss": 0.012,
      "reward": 1.1958333730697632,
      "reward_std": 0.2867168977856636,
      "rewards/accuracy_reward": 0.20833334028720857,
      "rewards/format_reward": 0.987500011920929,
      "step": 2475
    },
    {
      "completion_length": 230.5500045776367,
      "epoch": 0.992,
      "grad_norm": 0.9268749218220572,
      "kl": 0.32685546875,
      "learning_rate": 3.898849596456477e-09,
      "loss": 0.0131,
      "reward": 1.1000000476837157,
      "reward_std": 0.18649545162916184,
      "rewards/accuracy_reward": 0.1208333358168602,
      "rewards/format_reward": 0.9791666686534881,
      "step": 2480
    },
    {
      "completion_length": 235.82917175292968,
      "epoch": 0.994,
      "grad_norm": 0.49651550175485254,
      "kl": 0.28681640625,
      "learning_rate": 2.193165251545004e-09,
      "loss": 0.0115,
      "reward": 1.075000035762787,
      "reward_std": 0.17952174171805382,
      "rewards/accuracy_reward": 0.0958333346992731,
      "rewards/format_reward": 0.9791666865348816,
      "step": 2485
    },
    {
      "completion_length": 256.3333404541016,
      "epoch": 0.996,
      "grad_norm": 0.1905486064300661,
      "kl": 0.3017578125,
      "learning_rate": 9.74759906957612e-10,
      "loss": 0.0121,
      "reward": 1.0375000357627868,
      "reward_std": 0.16711369305849075,
      "rewards/accuracy_reward": 0.06666666939854622,
      "rewards/format_reward": 0.9708333432674408,
      "step": 2490
    },
    {
      "completion_length": 235.0958396911621,
      "epoch": 0.998,
      "grad_norm": 0.5058367377833421,
      "kl": 0.325,
      "learning_rate": 2.436929460525317e-10,
      "loss": 0.013,
      "reward": 1.1083333492279053,
      "reward_std": 0.16178620904684066,
      "rewards/accuracy_reward": 0.1375000011175871,
      "rewards/format_reward": 0.9708333492279053,
      "step": 2495
    },
    {
      "completion_length": 244.67083892822265,
      "epoch": 1.0,
      "grad_norm": 0.3908725363308062,
      "kl": 0.359765625,
      "learning_rate": 0.0,
      "loss": 0.0144,
      "reward": 1.0833333611488343,
      "reward_std": 0.17759759426116944,
      "rewards/accuracy_reward": 0.10416666977107525,
      "rewards/format_reward": 0.9791666865348816,
      "step": 2500
    },
    {
      "epoch": 1.0,
      "eval_completion_length": 204.63622225247897,
      "eval_kl": 0.3269230769230769,
      "eval_loss": 0.012251908890902996,
      "eval_reward": 1.1634615751413198,
      "eval_reward_std": 0.17546415844788918,
      "eval_rewards/accuracy_reward": 0.1714743645145343,
      "eval_rewards/format_reward": 0.9919871871288006,
      "eval_runtime": 145.9043,
      "eval_samples_per_second": 0.679,
      "eval_steps_per_second": 0.021,
      "step": 2500
    },
    {
      "epoch": 1.0,
      "step": 2500,
      "total_flos": 0.0,
      "train_loss": 69.57985635299504,
      "train_runtime": 92793.5148,
      "train_samples_per_second": 0.216,
      "train_steps_per_second": 0.027
    }
  ],
  "logging_steps": 5,
  "max_steps": 2500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}