{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 6.392694063926941,
  "eval_steps": 500,
  "global_step": 1400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "completion_length": 101.3359375,
      "epoch": 0.0045662100456621,
      "grad_norm": 6.199877738952637,
      "kl": 0.0,
      "learning_rate": 9.995433789954337e-07,
      "loss": -0.0,
      "reward": 0.6208333522081375,
      "reward_std": 0.723351925611496,
      "rewards/accuracy_reward": 0.2614583298563957,
      "rewards/format_reward": 0.359375,
      "step": 1
    },
    {
      "completion_length": 105.796875,
      "epoch": 0.0091324200913242,
      "grad_norm": 13.76264476776123,
      "kl": 0.0005550384521484375,
      "learning_rate": 9.990867579908674e-07,
      "loss": 0.0,
      "reward": 1.1440449953079224,
      "reward_std": 0.6297117471694946,
      "rewards/accuracy_reward": 0.5346700549125671,
      "rewards/format_reward": 0.609375,
      "step": 2
    },
    {
      "completion_length": 106.609375,
      "epoch": 0.0136986301369863,
      "grad_norm": 6.990738868713379,
      "kl": 0.000965118408203125,
      "learning_rate": 9.986301369863014e-07,
      "loss": 0.0,
      "reward": 0.9957855343818665,
      "reward_std": 0.726005494594574,
      "rewards/accuracy_reward": 0.45672303438186646,
      "rewards/format_reward": 0.5390625,
      "step": 3
    },
    {
      "completion_length": 101.4140625,
      "epoch": 0.0182648401826484,
      "grad_norm": 4.9568257331848145,
      "kl": 0.001346588134765625,
      "learning_rate": 9.98173515981735e-07,
      "loss": 0.0001,
      "reward": 0.991195559501648,
      "reward_std": 0.7322099208831787,
      "rewards/accuracy_reward": 0.43650802969932556,
      "rewards/format_reward": 0.5546875,
      "step": 4
    },
    {
      "completion_length": 106.4453125,
      "epoch": 0.0228310502283105,
      "grad_norm": 4.493799209594727,
      "kl": 0.0035247802734375,
      "learning_rate": 9.977168949771688e-07,
      "loss": 0.0001,
      "reward": 0.9764133095741272,
      "reward_std": 0.6022496819496155,
      "rewards/accuracy_reward": 0.421725794672966,
      "rewards/format_reward": 0.5546875,
      "step": 5
    },
    {
      "completion_length": 125.046875,
      "epoch": 0.0273972602739726,
      "grad_norm": 3.48476505279541,
      "kl": 0.00315093994140625,
      "learning_rate": 9.972602739726028e-07,
      "loss": 0.0001,
      "reward": 1.162934124469757,
      "reward_std": 0.5378114879131317,
      "rewards/accuracy_reward": 0.49887165427207947,
      "rewards/format_reward": 0.6640625,
      "step": 6
    },
    {
      "completion_length": 107.1484375,
      "epoch": 0.0319634703196347,
      "grad_norm": 3.6667165756225586,
      "kl": 0.0055999755859375,
      "learning_rate": 9.968036529680365e-07,
      "loss": 0.0002,
      "reward": 1.1229968070983887,
      "reward_std": 0.6023176610469818,
      "rewards/accuracy_reward": 0.4511217921972275,
      "rewards/format_reward": 0.671875,
      "step": 7
    },
    {
      "completion_length": 76.890625,
      "epoch": 0.0365296803652968,
      "grad_norm": 4.247033596038818,
      "kl": 0.017059326171875,
      "learning_rate": 9.963470319634703e-07,
      "loss": 0.0007,
      "reward": 0.9820911884307861,
      "reward_std": 0.6595480144023895,
      "rewards/accuracy_reward": 0.31021616607904434,
      "rewards/format_reward": 0.671875,
      "step": 8
    },
    {
      "completion_length": 99.359375,
      "epoch": 0.0410958904109589,
      "grad_norm": 3.2795424461364746,
      "kl": 0.018402099609375,
      "learning_rate": 9.95890410958904e-07,
      "loss": 0.0007,
      "reward": 1.3125749230384827,
      "reward_std": 0.5887171626091003,
      "rewards/accuracy_reward": 0.5235124826431274,
      "rewards/format_reward": 0.7890625,
      "step": 9
    },
    {
      "completion_length": 68.734375,
      "epoch": 0.045662100456621,
      "grad_norm": 4.997896671295166,
      "kl": 0.04248046875,
      "learning_rate": 9.954337899543377e-07,
      "loss": 0.0017,
      "reward": 1.2461639046669006,
      "reward_std": 0.527179092168808,
      "rewards/accuracy_reward": 0.386788934469223,
      "rewards/format_reward": 0.859375,
      "step": 10
    },
    {
      "completion_length": 98.9453125,
      "epoch": 0.0502283105022831,
      "grad_norm": 7.262642860412598,
      "kl": 0.023773193359375,
      "learning_rate": 9.949771689497717e-07,
      "loss": 0.001,
      "reward": 1.3204909563064575,
      "reward_std": 0.4520634114742279,
      "rewards/accuracy_reward": 0.46892838180065155,
      "rewards/format_reward": 0.8515625,
      "step": 11
    },
    {
      "completion_length": 102.6953125,
      "epoch": 0.0547945205479452,
      "grad_norm": 2.358719825744629,
      "kl": 0.032470703125,
      "learning_rate": 9.945205479452054e-07,
      "loss": 0.0013,
      "reward": 1.3686427474021912,
      "reward_std": 0.45033006370067596,
      "rewards/accuracy_reward": 0.5405177175998688,
      "rewards/format_reward": 0.828125,
      "step": 12
    },
    {
      "completion_length": 83.1171875,
      "epoch": 0.0593607305936073,
      "grad_norm": 3.457000255584717,
      "kl": 0.044189453125,
      "learning_rate": 9.940639269406391e-07,
      "loss": 0.0018,
      "reward": 1.3782268166542053,
      "reward_std": 0.4138905107975006,
      "rewards/accuracy_reward": 0.45635175704956055,
      "rewards/format_reward": 0.921875,
      "step": 13
    },
    {
      "completion_length": 48.734375,
      "epoch": 0.0639269406392694,
      "grad_norm": 3.6976211071014404,
      "kl": 0.056884765625,
      "learning_rate": 9.93607305936073e-07,
      "loss": 0.0023,
      "reward": 1.340489685535431,
      "reward_std": 0.37420646846294403,
      "rewards/accuracy_reward": 0.4029896557331085,
      "rewards/format_reward": 0.9375,
      "step": 14
    },
    {
      "completion_length": 79.2421875,
      "epoch": 0.0684931506849315,
      "grad_norm": 2.8622705936431885,
      "kl": 0.043212890625,
      "learning_rate": 9.931506849315068e-07,
      "loss": 0.0017,
      "reward": 1.2608134746551514,
      "reward_std": 0.45702624320983887,
      "rewards/accuracy_reward": 0.393625944852829,
      "rewards/format_reward": 0.8671875,
      "step": 15
    },
    {
      "completion_length": 101.453125,
      "epoch": 0.0730593607305936,
      "grad_norm": 3.5660035610198975,
      "kl": 0.034423828125,
      "learning_rate": 9.926940639269406e-07,
      "loss": 0.0014,
      "reward": 1.4559400081634521,
      "reward_std": 0.3906751722097397,
      "rewards/accuracy_reward": 0.5418775081634521,
      "rewards/format_reward": 0.9140625,
      "step": 16
    },
    {
      "completion_length": 58.125,
      "epoch": 0.0776255707762557,
      "grad_norm": 3.2858753204345703,
      "kl": 0.052978515625,
      "learning_rate": 9.922374429223745e-07,
      "loss": 0.0021,
      "reward": 1.3879406452178955,
      "reward_std": 0.32537831366062164,
      "rewards/accuracy_reward": 0.4113781452178955,
      "rewards/format_reward": 0.9765625,
      "step": 17
    },
    {
      "completion_length": 77.3671875,
      "epoch": 0.0821917808219178,
      "grad_norm": 2.0785841941833496,
      "kl": 0.0511474609375,
      "learning_rate": 9.917808219178082e-07,
      "loss": 0.002,
      "reward": 1.5014740824699402,
      "reward_std": 0.27246882766485214,
      "rewards/accuracy_reward": 0.540536567568779,
      "rewards/format_reward": 0.9609375,
      "step": 18
    },
    {
      "completion_length": 70.7890625,
      "epoch": 0.0867579908675799,
      "grad_norm": 3.1388633251190186,
      "kl": 0.0772705078125,
      "learning_rate": 9.91324200913242e-07,
      "loss": 0.0031,
      "reward": 1.4673261046409607,
      "reward_std": 0.3659689426422119,
      "rewards/accuracy_reward": 0.5142011493444443,
      "rewards/format_reward": 0.953125,
      "step": 19
    },
    {
      "completion_length": 73.71875,
      "epoch": 0.091324200913242,
      "grad_norm": 3.4941606521606445,
      "kl": 0.0772705078125,
      "learning_rate": 9.908675799086757e-07,
      "loss": 0.0031,
      "reward": 1.2669085264205933,
      "reward_std": 0.41668441891670227,
      "rewards/accuracy_reward": 0.3919084817171097,
      "rewards/format_reward": 0.875,
      "step": 20
    },
    {
      "completion_length": 86.0390625,
      "epoch": 0.0958904109589041,
      "grad_norm": 4.310213565826416,
      "kl": 0.0654296875,
      "learning_rate": 9.904109589041094e-07,
      "loss": 0.0026,
      "reward": 1.4882500767707825,
      "reward_std": 0.30715326964855194,
      "rewards/accuracy_reward": 0.5351250767707825,
      "rewards/format_reward": 0.953125,
      "step": 21
    },
    {
      "completion_length": 71.8359375,
      "epoch": 0.1004566210045662,
      "grad_norm": 3.6013519763946533,
      "kl": 0.09912109375,
      "learning_rate": 9.899543378995434e-07,
      "loss": 0.004,
      "reward": 1.602288007736206,
      "reward_std": 0.2894471287727356,
      "rewards/accuracy_reward": 0.6257254481315613,
      "rewards/format_reward": 0.9765625,
      "step": 22
    },
    {
      "completion_length": 80.5078125,
      "epoch": 0.1050228310502283,
      "grad_norm": 2.9229257106781006,
      "kl": 0.07373046875,
      "learning_rate": 9.894977168949771e-07,
      "loss": 0.0029,
      "reward": 1.5954613089561462,
      "reward_std": 0.29795171320438385,
      "rewards/accuracy_reward": 0.6267113089561462,
      "rewards/format_reward": 0.96875,
      "step": 23
    },
    {
      "completion_length": 78.6953125,
      "epoch": 0.1095890410958904,
      "grad_norm": 3.3861594200134277,
      "kl": 0.0640869140625,
      "learning_rate": 9.89041095890411e-07,
      "loss": 0.0026,
      "reward": 1.54551100730896,
      "reward_std": 0.2194862775504589,
      "rewards/accuracy_reward": 0.5611358880996704,
      "rewards/format_reward": 0.984375,
      "step": 24
    },
    {
      "completion_length": 73.4453125,
      "epoch": 0.1141552511415525,
      "grad_norm": 3.412217140197754,
      "kl": 0.093994140625,
      "learning_rate": 9.885844748858448e-07,
      "loss": 0.0038,
      "reward": 1.3831676244735718,
      "reward_std": 0.3298991322517395,
      "rewards/accuracy_reward": 0.4378551170229912,
      "rewards/format_reward": 0.9453125,
      "step": 25
    },
    {
      "completion_length": 67.359375,
      "epoch": 0.1187214611872146,
      "grad_norm": 2.750807285308838,
      "kl": 0.0947265625,
      "learning_rate": 9.881278538812785e-07,
      "loss": 0.0038,
      "reward": 1.5090773701667786,
      "reward_std": 0.30529043078422546,
      "rewards/accuracy_reward": 0.5559523701667786,
      "rewards/format_reward": 0.953125,
      "step": 26
    },
    {
      "completion_length": 65.1953125,
      "epoch": 0.1232876712328767,
      "grad_norm": 5.965710163116455,
      "kl": 0.10595703125,
      "learning_rate": 9.876712328767123e-07,
      "loss": 0.0042,
      "reward": 1.5281611680984497,
      "reward_std": 0.3281934857368469,
      "rewards/accuracy_reward": 0.5437861680984497,
      "rewards/format_reward": 0.984375,
      "step": 27
    },
    {
      "completion_length": 76.8671875,
      "epoch": 0.1278538812785388,
      "grad_norm": 3.2638497352600098,
      "kl": 0.07958984375,
      "learning_rate": 9.87214611872146e-07,
      "loss": 0.0032,
      "reward": 1.5927692651748657,
      "reward_std": 0.24901984632015228,
      "rewards/accuracy_reward": 0.6318317353725433,
      "rewards/format_reward": 0.9609375,
      "step": 28
    },
    {
      "completion_length": 66.0234375,
      "epoch": 0.1324200913242009,
      "grad_norm": 3.1384220123291016,
      "kl": 0.106689453125,
      "learning_rate": 9.867579908675797e-07,
      "loss": 0.0043,
      "reward": 1.5447565913200378,
      "reward_std": 0.284378357231617,
      "rewards/accuracy_reward": 0.5760065913200378,
      "rewards/format_reward": 0.96875,
      "step": 29
    },
    {
      "completion_length": 56.90625,
      "epoch": 0.136986301369863,
      "grad_norm": 5.531259059906006,
      "kl": 0.11572265625,
      "learning_rate": 9.863013698630137e-07,
      "loss": 0.0046,
      "reward": 1.6361945867538452,
      "reward_std": 0.24622034281492233,
      "rewards/accuracy_reward": 0.6361946165561676,
      "rewards/format_reward": 1.0,
      "step": 30
    },
    {
      "completion_length": 55.5703125,
      "epoch": 0.1415525114155251,
      "grad_norm": 3.289832353591919,
      "kl": 0.10986328125,
      "learning_rate": 9.858447488584474e-07,
      "loss": 0.0044,
      "reward": 1.4083333611488342,
      "reward_std": 0.32854069769382477,
      "rewards/accuracy_reward": 0.45520833134651184,
      "rewards/format_reward": 0.953125,
      "step": 31
    },
    {
      "completion_length": 76.828125,
      "epoch": 0.1461187214611872,
      "grad_norm": 10.165680885314941,
      "kl": 0.091064453125,
      "learning_rate": 9.853881278538814e-07,
      "loss": 0.0036,
      "reward": 1.5438354015350342,
      "reward_std": 0.23943377658724785,
      "rewards/accuracy_reward": 0.5672729313373566,
      "rewards/format_reward": 0.9765625,
      "step": 32
    },
    {
      "completion_length": 58.3515625,
      "epoch": 0.1506849315068493,
      "grad_norm": 5.475100517272949,
      "kl": 0.108642578125,
      "learning_rate": 9.84931506849315e-07,
      "loss": 0.0043,
      "reward": 1.6295759081840515,
      "reward_std": 0.25405153632164,
      "rewards/accuracy_reward": 0.6373884081840515,
      "rewards/format_reward": 0.9921875,
      "step": 33
    },
    {
      "completion_length": 67.96875,
      "epoch": 0.1552511415525114,
      "grad_norm": 2.0600223541259766,
      "kl": 0.085693359375,
      "learning_rate": 9.844748858447488e-07,
      "loss": 0.0034,
      "reward": 1.5968750715255737,
      "reward_std": 0.2177756428718567,
      "rewards/accuracy_reward": 0.5968749821186066,
      "rewards/format_reward": 1.0,
      "step": 34
    },
    {
      "completion_length": 55.0703125,
      "epoch": 0.1598173515981735,
      "grad_norm": 4.03507661819458,
      "kl": 0.134033203125,
      "learning_rate": 9.840182648401826e-07,
      "loss": 0.0054,
      "reward": 1.5210416316986084,
      "reward_std": 0.262370266020298,
      "rewards/accuracy_reward": 0.5288541465997696,
      "rewards/format_reward": 0.9921875,
      "step": 35
    },
    {
      "completion_length": 69.890625,
      "epoch": 0.1643835616438356,
      "grad_norm": 2.3833439350128174,
      "kl": 0.091796875,
      "learning_rate": 9.835616438356163e-07,
      "loss": 0.0037,
      "reward": 1.6906250715255737,
      "reward_std": 0.20069602131843567,
      "rewards/accuracy_reward": 0.6906249821186066,
      "rewards/format_reward": 1.0,
      "step": 36
    },
    {
      "completion_length": 67.2265625,
      "epoch": 0.1689497716894977,
      "grad_norm": 3.298387050628662,
      "kl": 0.126220703125,
      "learning_rate": 9.831050228310502e-07,
      "loss": 0.0051,
      "reward": 1.5833591222763062,
      "reward_std": 0.23695440590381622,
      "rewards/accuracy_reward": 0.5911716222763062,
      "rewards/format_reward": 0.9921875,
      "step": 37
    },
    {
      "completion_length": 75.1875,
      "epoch": 0.1735159817351598,
      "grad_norm": 2.7184557914733887,
      "kl": 0.126953125,
      "learning_rate": 9.82648401826484e-07,
      "loss": 0.0051,
      "reward": 1.6548610925674438,
      "reward_std": 0.18883602693676949,
      "rewards/accuracy_reward": 0.6704860329627991,
      "rewards/format_reward": 0.984375,
      "step": 38
    },
    {
      "completion_length": 64.015625,
      "epoch": 0.1780821917808219,
      "grad_norm": 3.2870519161224365,
      "kl": 0.117431640625,
      "learning_rate": 9.821917808219177e-07,
      "loss": 0.0047,
      "reward": 1.5351698398590088,
      "reward_std": 0.267734594643116,
      "rewards/accuracy_reward": 0.5664198100566864,
      "rewards/format_reward": 0.96875,
      "step": 39
    },
    {
      "completion_length": 52.1171875,
      "epoch": 0.182648401826484,
      "grad_norm": 5.351809024810791,
      "kl": 0.15234375,
      "learning_rate": 9.817351598173517e-07,
      "loss": 0.0061,
      "reward": 1.443321943283081,
      "reward_std": 0.32138869166374207,
      "rewards/accuracy_reward": 0.44332198798656464,
      "rewards/format_reward": 1.0,
      "step": 40
    },
    {
      "completion_length": 70.34375,
      "epoch": 0.1872146118721461,
      "grad_norm": 2.1224160194396973,
      "kl": 0.126220703125,
      "learning_rate": 9.812785388127854e-07,
      "loss": 0.0051,
      "reward": 1.761244773864746,
      "reward_std": 0.13502541184425354,
      "rewards/accuracy_reward": 0.7846822142601013,
      "rewards/format_reward": 0.9765625,
      "step": 41
    },
    {
      "completion_length": 64.1796875,
      "epoch": 0.1917808219178082,
      "grad_norm": 3.5647408962249756,
      "kl": 0.15234375,
      "learning_rate": 9.808219178082191e-07,
      "loss": 0.0061,
      "reward": 1.4638384580612183,
      "reward_std": 0.2666083872318268,
      "rewards/accuracy_reward": 0.47946345806121826,
      "rewards/format_reward": 0.984375,
      "step": 42
    },
    {
      "completion_length": 51.8515625,
      "epoch": 0.1963470319634703,
      "grad_norm": 9.191572189331055,
      "kl": 0.15185546875,
      "learning_rate": 9.803652968036529e-07,
      "loss": 0.0061,
      "reward": 1.5036438703536987,
      "reward_std": 0.22346660494804382,
      "rewards/accuracy_reward": 0.5114563703536987,
      "rewards/format_reward": 0.9921875,
      "step": 43
    },
    {
      "completion_length": 67.1328125,
      "epoch": 0.2009132420091324,
      "grad_norm": 4.356919765472412,
      "kl": 0.18994140625,
      "learning_rate": 9.799086757990868e-07,
      "loss": 0.0076,
      "reward": 1.448957920074463,
      "reward_std": 0.2659284919500351,
      "rewards/accuracy_reward": 0.4645828604698181,
      "rewards/format_reward": 0.984375,
      "step": 44
    },
    {
      "completion_length": 82.8203125,
      "epoch": 0.2054794520547945,
      "grad_norm": 4.919662952423096,
      "kl": 0.074951171875,
      "learning_rate": 9.794520547945205e-07,
      "loss": 0.003,
      "reward": 1.6562500596046448,
      "reward_std": 0.1944543570280075,
      "rewards/accuracy_reward": 0.671875,
      "rewards/format_reward": 0.984375,
      "step": 45
    },
    {
      "completion_length": 79.3125,
      "epoch": 0.2100456621004566,
      "grad_norm": 3.381512403488159,
      "kl": 0.0947265625,
      "learning_rate": 9.789954337899543e-07,
      "loss": 0.0038,
      "reward": 1.5502474904060364,
      "reward_std": 0.18939972668886185,
      "rewards/accuracy_reward": 0.5502474904060364,
      "rewards/format_reward": 1.0,
      "step": 46
    },
    {
      "completion_length": 63.5390625,
      "epoch": 0.2146118721461187,
      "grad_norm": 1.7941492795944214,
      "kl": 0.101806640625,
      "learning_rate": 9.78538812785388e-07,
      "loss": 0.0041,
      "reward": 1.634374976158142,
      "reward_std": 0.13939146511256695,
      "rewards/accuracy_reward": 0.6343749165534973,
      "rewards/format_reward": 1.0,
      "step": 47
    },
    {
      "completion_length": 54.125,
      "epoch": 0.2191780821917808,
      "grad_norm": 3.9591901302337646,
      "kl": 0.2177734375,
      "learning_rate": 9.78082191780822e-07,
      "loss": 0.0087,
      "reward": 1.4007303714752197,
      "reward_std": 0.33134153485298157,
      "rewards/accuracy_reward": 0.4163554012775421,
      "rewards/format_reward": 0.984375,
      "step": 48
    },
    {
      "completion_length": 31.8984375,
      "epoch": 0.2237442922374429,
      "grad_norm": 3.5492911338806152,
      "kl": 0.15966796875,
      "learning_rate": 9.776255707762557e-07,
      "loss": 0.0064,
      "reward": 1.44140625,
      "reward_std": 0.32149194180965424,
      "rewards/accuracy_reward": 0.44140625,
      "rewards/format_reward": 1.0,
      "step": 49
    },
    {
      "completion_length": 59.5625,
      "epoch": 0.228310502283105,
      "grad_norm": 6.129611968994141,
      "kl": 0.15771484375,
      "learning_rate": 9.771689497716894e-07,
      "loss": 0.0063,
      "reward": 1.6152344346046448,
      "reward_std": 0.17767168581485748,
      "rewards/accuracy_reward": 0.623046875,
      "rewards/format_reward": 0.9921875,
      "step": 50
    },
    {
      "completion_length": 51.4453125,
      "epoch": 0.2328767123287671,
      "grad_norm": 1.8558975458145142,
      "kl": 0.1015625,
      "learning_rate": 9.767123287671234e-07,
      "loss": 0.0041,
      "reward": 1.6468749642372131,
      "reward_std": 0.16151440143585205,
      "rewards/accuracy_reward": 0.6468749940395355,
      "rewards/format_reward": 1.0,
      "step": 51
    },
    {
      "completion_length": 64.2734375,
      "epoch": 0.2374429223744292,
      "grad_norm": 1.8877390623092651,
      "kl": 0.122802734375,
      "learning_rate": 9.762557077625571e-07,
      "loss": 0.0049,
      "reward": 1.557812511920929,
      "reward_std": 0.1784707009792328,
      "rewards/accuracy_reward": 0.557812511920929,
      "rewards/format_reward": 1.0,
      "step": 52
    },
    {
      "completion_length": 67.5625,
      "epoch": 0.2420091324200913,
      "grad_norm": 3.6787755489349365,
      "kl": 0.111328125,
      "learning_rate": 9.757990867579908e-07,
      "loss": 0.0045,
      "reward": 1.6884114742279053,
      "reward_std": 0.17782026529312134,
      "rewards/accuracy_reward": 0.6962239742279053,
      "rewards/format_reward": 0.9921875,
      "step": 53
    },
    {
      "completion_length": 63.3671875,
      "epoch": 0.2465753424657534,
      "grad_norm": 2.3778576850891113,
      "kl": 0.0980224609375,
      "learning_rate": 9.753424657534246e-07,
      "loss": 0.0039,
      "reward": 1.6024739742279053,
      "reward_std": 0.19950664788484573,
      "rewards/accuracy_reward": 0.6102864444255829,
      "rewards/format_reward": 0.9921875,
      "step": 54
    },
    {
      "completion_length": 78.6640625,
      "epoch": 0.2511415525114155,
      "grad_norm": 3.8972158432006836,
      "kl": 0.113037109375,
      "learning_rate": 9.748858447488583e-07,
      "loss": 0.0045,
      "reward": 1.591796875,
      "reward_std": 0.17978167533874512,
      "rewards/accuracy_reward": 0.5917968302965164,
      "rewards/format_reward": 1.0,
      "step": 55
    },
    {
      "completion_length": 84.65625,
      "epoch": 0.2557077625570776,
      "grad_norm": 1.890934705734253,
      "kl": 0.086181640625,
      "learning_rate": 9.744292237442923e-07,
      "loss": 0.0035,
      "reward": 1.5904948115348816,
      "reward_std": 0.1708931028842926,
      "rewards/accuracy_reward": 0.5983072519302368,
      "rewards/format_reward": 0.9921875,
      "step": 56
    },
    {
      "completion_length": 68.1015625,
      "epoch": 0.2602739726027397,
      "grad_norm": 2.091691017150879,
      "kl": 0.110107421875,
      "learning_rate": 9.73972602739726e-07,
      "loss": 0.0044,
      "reward": 1.590624988079071,
      "reward_std": 0.21159524470567703,
      "rewards/accuracy_reward": 0.6062500178813934,
      "rewards/format_reward": 0.984375,
      "step": 57
    },
    {
      "completion_length": 58.8671875,
      "epoch": 0.2648401826484018,
      "grad_norm": 2.6050846576690674,
      "kl": 0.10546875,
      "learning_rate": 9.735159817351597e-07,
      "loss": 0.0042,
      "reward": 1.4429687857627869,
      "reward_std": 0.3084883987903595,
      "rewards/accuracy_reward": 0.4585937559604645,
      "rewards/format_reward": 0.984375,
      "step": 58
    },
    {
      "completion_length": 57.8125,
      "epoch": 0.2694063926940639,
      "grad_norm": 6.616756439208984,
      "kl": 0.104736328125,
      "learning_rate": 9.730593607305937e-07,
      "loss": 0.0042,
      "reward": 1.4373698234558105,
      "reward_std": 0.3443310409784317,
      "rewards/accuracy_reward": 0.46861980855464935,
      "rewards/format_reward": 0.96875,
      "step": 59
    },
    {
      "completion_length": 72.84375,
      "epoch": 0.273972602739726,
      "grad_norm": 2.8635833263397217,
      "kl": 0.115478515625,
      "learning_rate": 9.726027397260274e-07,
      "loss": 0.0046,
      "reward": 1.5471354722976685,
      "reward_std": 0.25735440850257874,
      "rewards/accuracy_reward": 0.5549479126930237,
      "rewards/format_reward": 0.9921875,
      "step": 60
    },
    {
      "completion_length": 89.2578125,
      "epoch": 0.2785388127853881,
      "grad_norm": 2.5327181816101074,
      "kl": 0.0791015625,
      "learning_rate": 9.721461187214611e-07,
      "loss": 0.0032,
      "reward": 1.732812523841858,
      "reward_std": 0.17859892547130585,
      "rewards/accuracy_reward": 0.7484374642372131,
      "rewards/format_reward": 0.984375,
      "step": 61
    },
    {
      "completion_length": 55.9453125,
      "epoch": 0.2831050228310502,
      "grad_norm": 2.024601697921753,
      "kl": 0.15771484375,
      "learning_rate": 9.716894977168949e-07,
      "loss": 0.0063,
      "reward": 1.6088541746139526,
      "reward_std": 0.28504033386707306,
      "rewards/accuracy_reward": 0.6401041746139526,
      "rewards/format_reward": 0.96875,
      "step": 62
    },
    {
      "completion_length": 69.328125,
      "epoch": 0.2876712328767123,
      "grad_norm": 4.5766730308532715,
      "kl": 0.111083984375,
      "learning_rate": 9.712328767123286e-07,
      "loss": 0.0044,
      "reward": 1.7034826278686523,
      "reward_std": 0.12682656943798065,
      "rewards/accuracy_reward": 0.719107449054718,
      "rewards/format_reward": 0.984375,
      "step": 63
    },
    {
      "completion_length": 78.0234375,
      "epoch": 0.2922374429223744,
      "grad_norm": 3.555784225463867,
      "kl": 0.109375,
      "learning_rate": 9.707762557077626e-07,
      "loss": 0.0044,
      "reward": 1.5473958253860474,
      "reward_std": 0.23826827853918076,
      "rewards/accuracy_reward": 0.5630208402872086,
      "rewards/format_reward": 0.984375,
      "step": 64
    },
    {
      "completion_length": 46.015625,
      "epoch": 0.2968036529680365,
      "grad_norm": 5.032433032989502,
      "kl": 0.17724609375,
      "learning_rate": 9.703196347031963e-07,
      "loss": 0.0071,
      "reward": 1.401562511920929,
      "reward_std": 0.2856694385409355,
      "rewards/accuracy_reward": 0.40156251192092896,
      "rewards/format_reward": 1.0,
      "step": 65
    },
    {
      "completion_length": 80.3203125,
      "epoch": 0.3013698630136986,
      "grad_norm": 3.9773030281066895,
      "kl": 0.083740234375,
      "learning_rate": 9.6986301369863e-07,
      "loss": 0.0034,
      "reward": 1.6140625476837158,
      "reward_std": 0.1680549457669258,
      "rewards/accuracy_reward": 0.6218749582767487,
      "rewards/format_reward": 0.9921875,
      "step": 66
    },
    {
      "completion_length": 72.734375,
      "epoch": 0.3059360730593607,
      "grad_norm": 4.309236526489258,
      "kl": 0.091796875,
      "learning_rate": 9.69406392694064e-07,
      "loss": 0.0037,
      "reward": 1.4543966054916382,
      "reward_std": 0.27315448969602585,
      "rewards/accuracy_reward": 0.4778340458869934,
      "rewards/format_reward": 0.9765625,
      "step": 67
    },
    {
      "completion_length": 48.328125,
      "epoch": 0.3105022831050228,
      "grad_norm": 2.2709150314331055,
      "kl": 0.17724609375,
      "learning_rate": 9.689497716894977e-07,
      "loss": 0.0071,
      "reward": 1.6068063974380493,
      "reward_std": 0.247873917222023,
      "rewards/accuracy_reward": 0.6380563378334045,
      "rewards/format_reward": 0.96875,
      "step": 68
    },
    {
      "completion_length": 64.4921875,
      "epoch": 0.3150684931506849,
      "grad_norm": 2.840082883834839,
      "kl": 0.12841796875,
      "learning_rate": 9.684931506849314e-07,
      "loss": 0.0051,
      "reward": 1.65234375,
      "reward_std": 0.1649593710899353,
      "rewards/accuracy_reward": 0.6601562201976776,
      "rewards/format_reward": 0.9921875,
      "step": 69
    },
    {
      "completion_length": 65.5,
      "epoch": 0.319634703196347,
      "grad_norm": 2.1552846431732178,
      "kl": 0.14453125,
      "learning_rate": 9.680365296803652e-07,
      "loss": 0.0058,
      "reward": 1.6424851417541504,
      "reward_std": 0.14906217902898788,
      "rewards/accuracy_reward": 0.642485111951828,
      "rewards/format_reward": 1.0,
      "step": 70
    },
    {
      "completion_length": 66.7265625,
      "epoch": 0.3242009132420091,
      "grad_norm": 2.2692887783050537,
      "kl": 0.1513671875,
      "learning_rate": 9.675799086757991e-07,
      "loss": 0.0061,
      "reward": 1.689843773841858,
      "reward_std": 0.17708637565374374,
      "rewards/accuracy_reward": 0.6976562440395355,
      "rewards/format_reward": 0.9921875,
      "step": 71
    },
    {
      "completion_length": 76.78125,
      "epoch": 0.3287671232876712,
      "grad_norm": 2.679795265197754,
      "kl": 0.121337890625,
      "learning_rate": 9.671232876712329e-07,
      "loss": 0.0049,
      "reward": 1.6899740099906921,
      "reward_std": 0.1546846330165863,
      "rewards/accuracy_reward": 0.6977863907814026,
      "rewards/format_reward": 0.9921875,
      "step": 72
    },
    {
      "completion_length": 44.765625,
      "epoch": 0.3333333333333333,
      "grad_norm": 5.212102890014648,
      "kl": 0.125,
      "learning_rate": 9.666666666666666e-07,
      "loss": 0.005,
      "reward": 1.5778015851974487,
      "reward_std": 0.2252170294523239,
      "rewards/accuracy_reward": 0.5778016149997711,
      "rewards/format_reward": 1.0,
      "step": 73
    },
    {
      "completion_length": 79.1953125,
      "epoch": 0.3378995433789954,
      "grad_norm": 2.121600389480591,
      "kl": 0.102783203125,
      "learning_rate": 9.662100456621003e-07,
      "loss": 0.0041,
      "reward": 1.4783854484558105,
      "reward_std": 0.2087068259716034,
      "rewards/accuracy_reward": 0.5018228888511658,
      "rewards/format_reward": 0.9765625,
      "step": 74
    },
    {
      "completion_length": 67.703125,
      "epoch": 0.3424657534246575,
      "grad_norm": 3.5149497985839844,
      "kl": 0.11279296875,
      "learning_rate": 9.657534246575343e-07,
      "loss": 0.0045,
      "reward": 1.655573308467865,
      "reward_std": 0.293765589594841,
      "rewards/accuracy_reward": 0.6868232786655426,
      "rewards/format_reward": 0.96875,
      "step": 75
    },
    {
      "completion_length": 64.078125,
      "epoch": 0.3470319634703196,
      "grad_norm": 3.8966798782348633,
      "kl": 0.150634765625,
      "learning_rate": 9.65296803652968e-07,
      "loss": 0.006,
      "reward": 1.686715006828308,
      "reward_std": 0.2304065003991127,
      "rewards/accuracy_reward": 0.7023399174213409,
      "rewards/format_reward": 0.984375,
      "step": 76
    },
    {
      "completion_length": 62.2890625,
      "epoch": 0.3515981735159817,
      "grad_norm": 3.204103469848633,
      "kl": 0.12451171875,
      "learning_rate": 9.648401826484017e-07,
      "loss": 0.005,
      "reward": 1.4418154954910278,
      "reward_std": 0.305108904838562,
      "rewards/accuracy_reward": 0.5199404954910278,
      "rewards/format_reward": 0.921875,
      "step": 77
    },
    {
      "completion_length": 56.8125,
      "epoch": 0.3561643835616438,
      "grad_norm": 35.58188247680664,
      "kl": 0.1640625,
      "learning_rate": 9.643835616438357e-07,
      "loss": 0.0066,
      "reward": 1.5536458492279053,
      "reward_std": 0.34119510650634766,
      "rewards/accuracy_reward": 0.6161458492279053,
      "rewards/format_reward": 0.9375,
      "step": 78
    },
    {
      "completion_length": 79.78125,
      "epoch": 0.3607305936073059,
      "grad_norm": 2.3489811420440674,
      "kl": 0.107421875,
      "learning_rate": 9.639269406392694e-07,
      "loss": 0.0043,
      "reward": 1.5434371829032898,
      "reward_std": 0.2838926613330841,
      "rewards/accuracy_reward": 0.6059371829032898,
      "rewards/format_reward": 0.9375,
      "step": 79
    },
    {
      "completion_length": 76.34375,
      "epoch": 0.365296803652968,
      "grad_norm": 3.45082426071167,
      "kl": 0.1591796875,
      "learning_rate": 9.634703196347032e-07,
      "loss": 0.0064,
      "reward": 1.510318636894226,
      "reward_std": 0.3496459722518921,
      "rewards/accuracy_reward": 0.5806310772895813,
      "rewards/format_reward": 0.9296875,
      "step": 80
    },
    {
      "completion_length": 83.40625,
      "epoch": 0.3698630136986301,
      "grad_norm": 2.6958389282226562,
      "kl": 0.13671875,
      "learning_rate": 9.630136986301369e-07,
      "loss": 0.0055,
      "reward": 1.4917969107627869,
      "reward_std": 0.29826872050762177,
      "rewards/accuracy_reward": 0.5308593809604645,
      "rewards/format_reward": 0.9609375,
      "step": 81
    },
    {
      "completion_length": 61.8984375,
      "epoch": 0.3744292237442922,
      "grad_norm": 9.85806941986084,
      "kl": 0.15966796875,
      "learning_rate": 9.625570776255706e-07,
      "loss": 0.0064,
      "reward": 1.2881510257720947,
      "reward_std": 0.37539002299308777,
      "rewards/accuracy_reward": 0.3506510257720947,
      "rewards/format_reward": 0.9375,
      "step": 82
    },
    {
      "completion_length": 87.1640625,
      "epoch": 0.3789954337899543,
      "grad_norm": 1.7815834283828735,
      "kl": 0.11572265625,
      "learning_rate": 9.621004566210046e-07,
      "loss": 0.0046,
      "reward": 1.624854028224945,
      "reward_std": 0.19279810786247253,
      "rewards/accuracy_reward": 0.6482914686203003,
      "rewards/format_reward": 0.9765625,
      "step": 83
    },
    {
      "completion_length": 72.078125,
      "epoch": 0.3835616438356164,
      "grad_norm": 3.1070339679718018,
      "kl": 0.14013671875,
      "learning_rate": 9.616438356164383e-07,
      "loss": 0.0056,
      "reward": 1.5429518222808838,
      "reward_std": 0.3384602963924408,
      "rewards/accuracy_reward": 0.5820143222808838,
      "rewards/format_reward": 0.9609375,
      "step": 84
    },
    {
      "completion_length": 81.640625,
      "epoch": 0.3881278538812785,
      "grad_norm": 2.3276116847991943,
      "kl": 0.14990234375,
      "learning_rate": 9.61187214611872e-07,
      "loss": 0.006,
      "reward": 1.5467524528503418,
      "reward_std": 0.23365992307662964,
      "rewards/accuracy_reward": 0.570189893245697,
      "rewards/format_reward": 0.9765625,
      "step": 85
    },
    {
      "completion_length": 72.9296875,
      "epoch": 0.3926940639269406,
      "grad_norm": 13.4576416015625,
      "kl": 0.111572265625,
      "learning_rate": 9.60730593607306e-07,
      "loss": 0.0045,
      "reward": 1.580468773841858,
      "reward_std": 0.3186973035335541,
      "rewards/accuracy_reward": 0.6195312440395355,
      "rewards/format_reward": 0.9609375,
      "step": 86
    },
    {
      "completion_length": 76.4453125,
      "epoch": 0.3972602739726027,
      "grad_norm": 9.36266040802002,
      "kl": 0.135498046875,
      "learning_rate": 9.602739726027397e-07,
      "loss": 0.0054,
      "reward": 1.459251582622528,
      "reward_std": 0.3123939037322998,
      "rewards/accuracy_reward": 0.5061265677213669,
      "rewards/format_reward": 0.953125,
      "step": 87
    },
    {
      "completion_length": 71.3828125,
      "epoch": 0.4018264840182648,
      "grad_norm": 3.115593671798706,
      "kl": 0.110107421875,
      "learning_rate": 9.598173515981735e-07,
      "loss": 0.0044,
      "reward": 1.513058066368103,
      "reward_std": 0.2660725861787796,
      "rewards/accuracy_reward": 0.5521205365657806,
      "rewards/format_reward": 0.9609375,
      "step": 88
    },
    {
      "completion_length": 64.2109375,
      "epoch": 0.4063926940639269,
      "grad_norm": 3.5730538368225098,
      "kl": 0.169921875,
      "learning_rate": 9.593607305936072e-07,
      "loss": 0.0068,
      "reward": 1.4846374988555908,
      "reward_std": 0.25958995521068573,
      "rewards/accuracy_reward": 0.5080749839544296,
      "rewards/format_reward": 0.9765625,
      "step": 89
    },
    {
      "completion_length": 95.4921875,
      "epoch": 0.410958904109589,
      "grad_norm": 2.466034173965454,
      "kl": 0.11083984375,
      "learning_rate": 9.58904109589041e-07,
      "loss": 0.0044,
      "reward": 1.6667287349700928,
      "reward_std": 0.21123456954956055,
      "rewards/accuracy_reward": 0.7057911455631256,
      "rewards/format_reward": 0.9609375,
      "step": 90
    },
    {
      "completion_length": 68.859375,
      "epoch": 0.4155251141552511,
      "grad_norm": 2.8238866329193115,
      "kl": 0.147216796875,
      "learning_rate": 9.584474885844749e-07,
      "loss": 0.0059,
      "reward": 1.4025428891181946,
      "reward_std": 0.2600523456931114,
      "rewards/accuracy_reward": 0.4416054040193558,
      "rewards/format_reward": 0.9609375,
      "step": 91
    },
    {
      "completion_length": 86.1484375,
      "epoch": 0.4200913242009132,
      "grad_norm": 3.0570385456085205,
      "kl": 0.112548828125,
      "learning_rate": 9.579908675799086e-07,
      "loss": 0.0045,
      "reward": 1.394381046295166,
      "reward_std": 0.3288656920194626,
      "rewards/accuracy_reward": 0.44125598669052124,
      "rewards/format_reward": 0.953125,
      "step": 92
    },
    {
      "completion_length": 61.4609375,
      "epoch": 0.4246575342465753,
      "grad_norm": 3.7054836750030518,
      "kl": 0.154541015625,
      "learning_rate": 9.575342465753423e-07,
      "loss": 0.0062,
      "reward": 1.3927083611488342,
      "reward_std": 0.3015543594956398,
      "rewards/accuracy_reward": 0.40052083134651184,
      "rewards/format_reward": 0.9921875,
      "step": 93
    },
    {
      "completion_length": 75.2578125,
      "epoch": 0.4292237442922374,
      "grad_norm": 2.8483920097351074,
      "kl": 0.123046875,
      "learning_rate": 9.570776255707763e-07,
      "loss": 0.0049,
      "reward": 1.5788614153862,
      "reward_std": 0.26594626903533936,
      "rewards/accuracy_reward": 0.6022988557815552,
      "rewards/format_reward": 0.9765625,
      "step": 94
    },
    {
      "completion_length": 51.71875,
      "epoch": 0.4337899543378995,
      "grad_norm": 3.642658233642578,
      "kl": 0.2294921875,
      "learning_rate": 9.5662100456621e-07,
      "loss": 0.0092,
      "reward": 1.3690476417541504,
      "reward_std": 0.32933689653873444,
      "rewards/accuracy_reward": 0.400297611951828,
      "rewards/format_reward": 0.96875,
      "step": 95
    },
    {
      "completion_length": 81.4765625,
      "epoch": 0.4383561643835616,
      "grad_norm": 2.5764758586883545,
      "kl": 0.14013671875,
      "learning_rate": 9.561643835616437e-07,
      "loss": 0.0056,
      "reward": 1.6177189946174622,
      "reward_std": 0.20244715362787247,
      "rewards/accuracy_reward": 0.6333439946174622,
      "rewards/format_reward": 0.984375,
      "step": 96
    },
    {
      "completion_length": 64.3984375,
      "epoch": 0.4429223744292237,
      "grad_norm": 3.5117011070251465,
      "kl": 0.18408203125,
      "learning_rate": 9.557077625570777e-07,
      "loss": 0.0074,
      "reward": 1.3848958611488342,
      "reward_std": 0.33436155319213867,
      "rewards/accuracy_reward": 0.39270834624767303,
      "rewards/format_reward": 0.9921875,
      "step": 97
    },
    {
      "completion_length": 75.0703125,
      "epoch": 0.4474885844748858,
      "grad_norm": 5.965494632720947,
      "kl": 0.16650390625,
      "learning_rate": 9.552511415525114e-07,
      "loss": 0.0067,
      "reward": 1.490625023841858,
      "reward_std": 0.3021724224090576,
      "rewards/accuracy_reward": 0.5140624940395355,
      "rewards/format_reward": 0.9765625,
      "step": 98
    },
    {
      "completion_length": 81.328125,
      "epoch": 0.4520547945205479,
      "grad_norm": 5.425507545471191,
      "kl": 0.12939453125,
      "learning_rate": 9.547945205479452e-07,
      "loss": 0.0052,
      "reward": 1.489062488079071,
      "reward_std": 0.30329059064388275,
      "rewards/accuracy_reward": 0.5046875178813934,
      "rewards/format_reward": 0.984375,
      "step": 99
    },
    {
      "completion_length": 85.65625,
      "epoch": 0.45662100456621,
      "grad_norm": 3.3313958644866943,
      "kl": 0.124755859375,
      "learning_rate": 9.54337899543379e-07,
      "loss": 0.005,
      "reward": 1.4551078081130981,
      "reward_std": 0.2936365008354187,
      "rewards/accuracy_reward": 0.5019826889038086,
      "rewards/format_reward": 0.953125,
      "step": 100
    },
    {
      "completion_length": 101.9921875,
      "epoch": 0.4611872146118721,
      "grad_norm": 7.004181861877441,
      "kl": 0.100830078125,
      "learning_rate": 9.538812785388126e-07,
      "loss": 0.004,
      "reward": 1.8317708373069763,
      "reward_std": 0.11421890184283257,
      "rewards/accuracy_reward": 0.8395832479000092,
      "rewards/format_reward": 0.9921875,
      "step": 101
    },
    {
      "completion_length": 65.2890625,
      "epoch": 0.4657534246575342,
      "grad_norm": 5.00878381729126,
      "kl": 0.13720703125,
      "learning_rate": 9.534246575342465e-07,
      "loss": 0.0055,
      "reward": 1.4119791984558105,
      "reward_std": 0.3915850967168808,
      "rewards/accuracy_reward": 0.45885418355464935,
      "rewards/format_reward": 0.953125,
      "step": 102
    },
    {
      "completion_length": 66.1875,
      "epoch": 0.4703196347031963,
      "grad_norm": 33.22105026245117,
      "kl": 0.1689453125,
      "learning_rate": 9.529680365296803e-07,
      "loss": 0.0068,
      "reward": 1.5824455618858337,
      "reward_std": 0.362550288438797,
      "rewards/accuracy_reward": 0.5980705618858337,
      "rewards/format_reward": 0.984375,
      "step": 103
    },
    {
      "completion_length": 94.6015625,
      "epoch": 0.4748858447488584,
      "grad_norm": 2.8027193546295166,
      "kl": 0.100341796875,
      "learning_rate": 9.525114155251142e-07,
      "loss": 0.004,
      "reward": 1.573001742362976,
      "reward_std": 0.2286304533481598,
      "rewards/accuracy_reward": 0.5886267125606537,
      "rewards/format_reward": 0.984375,
      "step": 104
    },
    {
      "completion_length": 88.09375,
      "epoch": 0.4794520547945205,
      "grad_norm": 3.694000482559204,
      "kl": 0.1162109375,
      "learning_rate": 9.520547945205479e-07,
      "loss": 0.0046,
      "reward": 1.616619348526001,
      "reward_std": 0.24134992063045502,
      "rewards/accuracy_reward": 0.6244317889213562,
      "rewards/format_reward": 0.9921875,
      "step": 105
    },
    {
      "completion_length": 97.0546875,
      "epoch": 0.4840182648401826,
      "grad_norm": 3.0992743968963623,
      "kl": 0.1025390625,
      "learning_rate": 9.515981735159817e-07,
      "loss": 0.0041,
      "reward": 1.6255208253860474,
      "reward_std": 0.2588481456041336,
      "rewards/accuracy_reward": 0.6489583253860474,
      "rewards/format_reward": 0.9765625,
      "step": 106
    },
    {
      "completion_length": 76.7890625,
      "epoch": 0.4885844748858447,
      "grad_norm": 2.923036813735962,
      "kl": 0.12255859375,
      "learning_rate": 9.511415525114155e-07,
      "loss": 0.0049,
      "reward": 1.6026042103767395,
      "reward_std": 0.23826471716165543,
      "rewards/accuracy_reward": 0.6260416805744171,
      "rewards/format_reward": 0.9765625,
      "step": 107
    },
    {
      "completion_length": 78.6796875,
      "epoch": 0.4931506849315068,
      "grad_norm": 2.8310043811798096,
      "kl": 0.1357421875,
      "learning_rate": 9.506849315068493e-07,
      "loss": 0.0054,
      "reward": 1.4074218273162842,
      "reward_std": 0.3311483561992645,
      "rewards/accuracy_reward": 0.44648435711860657,
      "rewards/format_reward": 0.9609375,
      "step": 108
    },
    {
      "completion_length": 76.40625,
      "epoch": 0.4977168949771689,
      "grad_norm": 4.38238525390625,
      "kl": 0.119873046875,
      "learning_rate": 9.50228310502283e-07,
      "loss": 0.0048,
      "reward": 1.5411389470100403,
      "reward_std": 0.22866196930408478,
      "rewards/accuracy_reward": 0.5802014023065567,
      "rewards/format_reward": 0.9609375,
      "step": 109
    },
    {
      "completion_length": 71.234375,
      "epoch": 0.502283105022831,
      "grad_norm": 5.190479755401611,
      "kl": 0.18994140625,
      "learning_rate": 9.497716894977168e-07,
      "loss": 0.0076,
      "reward": 1.5246779322624207,
      "reward_std": 0.31538626551628113,
      "rewards/accuracy_reward": 0.5637404024600983,
      "rewards/format_reward": 0.9609375,
      "step": 110
    },
    {
      "completion_length": 63.21875,
      "epoch": 0.5068493150684932,
      "grad_norm": 7.178065776824951,
      "kl": 0.111083984375,
      "learning_rate": 9.493150684931507e-07,
      "loss": 0.0044,
      "reward": 1.421093761920929,
      "reward_std": 0.3210388720035553,
      "rewards/accuracy_reward": 0.42890626192092896,
      "rewards/format_reward": 0.9921875,
      "step": 111
    },
    {
      "completion_length": 86.609375,
      "epoch": 0.5114155251141552,
      "grad_norm": 3.856398105621338,
      "kl": 0.115478515625,
      "learning_rate": 9.488584474885845e-07,
      "loss": 0.0046,
      "reward": 1.504786729812622,
      "reward_std": 0.22957277297973633,
      "rewards/accuracy_reward": 0.5125992149114609,
      "rewards/format_reward": 0.9921875,
      "step": 112
    },
    {
      "completion_length": 77.28125,
      "epoch": 0.5159817351598174,
      "grad_norm": 8.988385200500488,
      "kl": 0.15869140625,
      "learning_rate": 9.484018264840182e-07,
      "loss": 0.0064,
      "reward": 1.4809608459472656,
      "reward_std": 0.3224972039461136,
      "rewards/accuracy_reward": 0.5043983459472656,
      "rewards/format_reward": 0.9765625,
      "step": 113
    },
    {
      "completion_length": 84.90625,
      "epoch": 0.5205479452054794,
      "grad_norm": 4.3304219245910645,
      "kl": 0.1025390625,
      "learning_rate": 9.47945205479452e-07,
      "loss": 0.0041,
      "reward": 1.5533854961395264,
      "reward_std": 0.2838037610054016,
      "rewards/accuracy_reward": 0.5846354365348816,
      "rewards/format_reward": 0.96875,
      "step": 114
    },
    {
      "completion_length": 94.1328125,
      "epoch": 0.5251141552511416,
      "grad_norm": 2.1755430698394775,
      "kl": 0.11572265625,
      "learning_rate": 9.474885844748858e-07,
      "loss": 0.0046,
      "reward": 1.6908555626869202,
      "reward_std": 0.18422479182481766,
      "rewards/accuracy_reward": 0.7064805030822754,
      "rewards/format_reward": 0.984375,
      "step": 115
    },
    {
      "completion_length": 87.375,
      "epoch": 0.5296803652968036,
      "grad_norm": 2.527374029159546,
      "kl": 0.09033203125,
      "learning_rate": 9.470319634703196e-07,
      "loss": 0.0036,
      "reward": 1.4536417722702026,
      "reward_std": 0.23572513461112976,
      "rewards/accuracy_reward": 0.46926678717136383,
      "rewards/format_reward": 0.984375,
      "step": 116
    },
    {
      "completion_length": 90.4140625,
      "epoch": 0.5342465753424658,
      "grad_norm": 2.9267947673797607,
      "kl": 0.098388671875,
      "learning_rate": 9.465753424657534e-07,
      "loss": 0.0039,
      "reward": 1.52434903383255,
      "reward_std": 0.2611450105905533,
      "rewards/accuracy_reward": 0.5477864444255829,
      "rewards/format_reward": 0.9765625,
      "step": 117
    },
    {
      "completion_length": 108.2890625,
      "epoch": 0.5388127853881278,
      "grad_norm": 4.520658493041992,
      "kl": 0.1044921875,
      "learning_rate": 9.461187214611872e-07,
      "loss": 0.0042,
      "reward": 1.5627976655960083,
      "reward_std": 0.2677070200443268,
      "rewards/accuracy_reward": 0.6252975761890411,
      "rewards/format_reward": 0.9375,
      "step": 118
    },
    {
      "completion_length": 74.1953125,
      "epoch": 0.54337899543379,
      "grad_norm": 4.915615558624268,
      "kl": 0.1650390625,
      "learning_rate": 9.45662100456621e-07,
      "loss": 0.0066,
      "reward": 1.5641247630119324,
      "reward_std": 0.22196320444345474,
      "rewards/accuracy_reward": 0.56412473320961,
      "rewards/format_reward": 1.0,
      "step": 119
    },
    {
      "completion_length": 76.375,
      "epoch": 0.547945205479452,
      "grad_norm": 4.11935567855835,
      "kl": 0.094482421875,
      "learning_rate": 9.452054794520548e-07,
      "loss": 0.0038,
      "reward": 1.5418658256530762,
      "reward_std": 0.2516227439045906,
      "rewards/accuracy_reward": 0.5496782958507538,
      "rewards/format_reward": 0.9921875,
      "step": 120
    },
    {
      "completion_length": 68.0859375,
      "epoch": 0.5525114155251142,
      "grad_norm": 3.256564140319824,
      "kl": 0.14208984375,
      "learning_rate": 9.447488584474885e-07,
      "loss": 0.0057,
      "reward": 1.520312488079071,
      "reward_std": 0.2812953442335129,
      "rewards/accuracy_reward": 0.5281250178813934,
      "rewards/format_reward": 0.9921875,
      "step": 121
    },
    {
      "completion_length": 92.7109375,
      "epoch": 0.5570776255707762,
      "grad_norm": 2.9781503677368164,
      "kl": 0.10546875,
      "learning_rate": 9.442922374429223e-07,
      "loss": 0.0042,
      "reward": 1.6091517806053162,
      "reward_std": 0.23979534208774567,
      "rewards/accuracy_reward": 0.6404017508029938,
      "rewards/format_reward": 0.96875,
      "step": 122
    },
    {
      "completion_length": 55.8984375,
      "epoch": 0.5616438356164384,
      "grad_norm": 4.986601829528809,
      "kl": 0.146484375,
      "learning_rate": 9.438356164383561e-07,
      "loss": 0.0059,
      "reward": 1.562890648841858,
      "reward_std": 0.2664823532104492,
      "rewards/accuracy_reward": 0.5707031190395355,
      "rewards/format_reward": 0.9921875,
      "step": 123
    },
    {
      "completion_length": 66.03125,
      "epoch": 0.5662100456621004,
      "grad_norm": 5.077969551086426,
      "kl": 0.140625,
      "learning_rate": 9.4337899543379e-07,
      "loss": 0.0056,
      "reward": 1.5604352951049805,
      "reward_std": 0.2387000024318695,
      "rewards/accuracy_reward": 0.5604352653026581,
      "rewards/format_reward": 1.0,
      "step": 124
    },
    {
      "completion_length": 75.796875,
      "epoch": 0.5707762557077626,
      "grad_norm": 2.614694356918335,
      "kl": 0.124755859375,
      "learning_rate": 9.429223744292237e-07,
      "loss": 0.005,
      "reward": 1.7544271349906921,
      "reward_std": 0.13757340610027313,
      "rewards/accuracy_reward": 0.7700520753860474,
      "rewards/format_reward": 0.984375,
      "step": 125
    },
    {
      "completion_length": 63.4765625,
      "epoch": 0.5753424657534246,
      "grad_norm": 5.098128795623779,
      "kl": 0.13330078125,
      "learning_rate": 9.424657534246575e-07,
      "loss": 0.0053,
      "reward": 1.545653522014618,
      "reward_std": 0.2823333144187927,
      "rewards/accuracy_reward": 0.5612785220146179,
      "rewards/format_reward": 0.984375,
      "step": 126
    },
    {
      "completion_length": 81.3046875,
      "epoch": 0.5799086757990868,
      "grad_norm": 2.6090292930603027,
      "kl": 0.072021484375,
      "learning_rate": 9.420091324200913e-07,
      "loss": 0.0029,
      "reward": 1.5642718076705933,
      "reward_std": 0.16427360475063324,
      "rewards/accuracy_reward": 0.5642717182636261,
      "rewards/format_reward": 1.0,
      "step": 127
    },
    {
      "completion_length": 63.546875,
      "epoch": 0.5844748858447488,
      "grad_norm": 4.00266170501709,
      "kl": 0.093505859375,
      "learning_rate": 9.41552511415525e-07,
      "loss": 0.0037,
      "reward": 1.5677236318588257,
      "reward_std": 0.30948029458522797,
      "rewards/accuracy_reward": 0.5755361616611481,
      "rewards/format_reward": 0.9921875,
      "step": 128
    },
    {
      "completion_length": 75.8359375,
      "epoch": 0.589041095890411,
      "grad_norm": 3.477132558822632,
      "kl": 0.099609375,
      "learning_rate": 9.410958904109588e-07,
      "loss": 0.004,
      "reward": 1.5374347567558289,
      "reward_std": 0.24908459186553955,
      "rewards/accuracy_reward": 0.5608722269535065,
      "rewards/format_reward": 0.9765625,
      "step": 129
    },
    {
      "completion_length": 76.46875,
      "epoch": 0.593607305936073,
      "grad_norm": 2.5465404987335205,
      "kl": 0.1416015625,
      "learning_rate": 9.406392694063926e-07,
      "loss": 0.0057,
      "reward": 1.6419846415519714,
      "reward_std": 0.1617630198597908,
      "rewards/accuracy_reward": 0.6497970819473267,
      "rewards/format_reward": 0.9921875,
      "step": 130
    },
    {
      "completion_length": 69.015625,
      "epoch": 0.5981735159817352,
      "grad_norm": 4.162069320678711,
      "kl": 0.14794921875,
      "learning_rate": 9.401826484018265e-07,
      "loss": 0.0059,
      "reward": 1.5735609531402588,
      "reward_std": 0.24296356737613678,
      "rewards/accuracy_reward": 0.5735608339309692,
      "rewards/format_reward": 1.0,
      "step": 131
    },
    {
      "completion_length": 63.578125,
      "epoch": 0.6027397260273972,
      "grad_norm": 2.8231759071350098,
      "kl": 0.158935546875,
      "learning_rate": 9.397260273972603e-07,
      "loss": 0.0064,
      "reward": 1.6688734889030457,
      "reward_std": 0.20343666523694992,
      "rewards/accuracy_reward": 0.6766859591007233,
      "rewards/format_reward": 0.9921875,
      "step": 132
    },
    {
      "completion_length": 58.921875,
      "epoch": 0.6073059360730594,
      "grad_norm": 3.4793238639831543,
      "kl": 0.13818359375,
      "learning_rate": 9.39269406392694e-07,
      "loss": 0.0055,
      "reward": 1.4002604484558105,
      "reward_std": 0.2559265270829201,
      "rewards/accuracy_reward": 0.40026040375232697,
      "rewards/format_reward": 1.0,
      "step": 133
    },
    {
      "completion_length": 56.109375,
      "epoch": 0.6118721461187214,
      "grad_norm": 3.3541765213012695,
      "kl": 0.14404296875,
      "learning_rate": 9.388127853881278e-07,
      "loss": 0.0058,
      "reward": 1.470312476158142,
      "reward_std": 0.3302301913499832,
      "rewards/accuracy_reward": 0.4937499612569809,
      "rewards/format_reward": 0.9765625,
      "step": 134
    },
    {
      "completion_length": 70.4609375,
      "epoch": 0.6164383561643836,
      "grad_norm": 2.545574903488159,
      "kl": 0.104248046875,
      "learning_rate": 9.383561643835616e-07,
      "loss": 0.0042,
      "reward": 1.6333333849906921,
      "reward_std": 0.16338077187538147,
      "rewards/accuracy_reward": 0.633333295583725,
      "rewards/format_reward": 1.0,
      "step": 135
    },
    {
      "completion_length": 47.046875,
      "epoch": 0.6210045662100456,
      "grad_norm": 3.1955769062042236,
      "kl": 0.102783203125,
      "learning_rate": 9.378995433789953e-07,
      "loss": 0.0041,
      "reward": 1.459455132484436,
      "reward_std": 0.3040963113307953,
      "rewards/accuracy_reward": 0.47508013248443604,
      "rewards/format_reward": 0.984375,
      "step": 136
    },
    {
      "completion_length": 67.234375,
      "epoch": 0.6255707762557078,
      "grad_norm": 4.687453269958496,
      "kl": 0.13427734375,
      "learning_rate": 9.374429223744292e-07,
      "loss": 0.0054,
      "reward": 1.600705862045288,
      "reward_std": 0.20340244472026825,
      "rewards/accuracy_reward": 0.6085183620452881,
      "rewards/format_reward": 0.9921875,
      "step": 137
    },
    {
      "completion_length": 68.546875,
      "epoch": 0.6301369863013698,
      "grad_norm": 2.4881796836853027,
      "kl": 0.13037109375,
      "learning_rate": 9.36986301369863e-07,
      "loss": 0.0052,
      "reward": 1.5750191807746887,
      "reward_std": 0.21577110141515732,
      "rewards/accuracy_reward": 0.5906441807746887,
      "rewards/format_reward": 0.984375,
      "step": 138
    },
    {
      "completion_length": 58.6171875,
      "epoch": 0.634703196347032,
      "grad_norm": 4.482968807220459,
      "kl": 0.1357421875,
      "learning_rate": 9.365296803652968e-07,
      "loss": 0.0054,
      "reward": 1.6022320985794067,
      "reward_std": 0.25604283809661865,
      "rewards/accuracy_reward": 0.6022321283817291,
      "rewards/format_reward": 1.0,
      "step": 139
    },
    {
      "completion_length": 59.625,
      "epoch": 0.639269406392694,
      "grad_norm": 2.6462514400482178,
      "kl": 0.144287109375,
      "learning_rate": 9.360730593607306e-07,
      "loss": 0.0058,
      "reward": 1.6302083134651184,
      "reward_std": 0.2878073900938034,
      "rewards/accuracy_reward": 0.6458333432674408,
      "rewards/format_reward": 0.984375,
      "step": 140
    },
    {
      "completion_length": 67.953125,
      "epoch": 0.6438356164383562,
      "grad_norm": 14.897021293640137,
      "kl": 0.10888671875,
      "learning_rate": 9.356164383561643e-07,
      "loss": 0.0044,
      "reward": 1.5768229961395264,
      "reward_std": 0.26041025668382645,
      "rewards/accuracy_reward": 0.5846354067325592,
      "rewards/format_reward": 0.9921875,
      "step": 141
    },
    {
      "completion_length": 62.2421875,
      "epoch": 0.6484018264840182,
      "grad_norm": 2.723630428314209,
      "kl": 0.140625,
      "learning_rate": 9.351598173515981e-07,
      "loss": 0.0056,
      "reward": 1.5526549220085144,
      "reward_std": 0.2013382911682129,
      "rewards/accuracy_reward": 0.552654892206192,
      "rewards/format_reward": 1.0,
      "step": 142
    },
    {
      "completion_length": 62.234375,
      "epoch": 0.6529680365296804,
      "grad_norm": 4.193840026855469,
      "kl": 0.12744140625,
      "learning_rate": 9.347031963470319e-07,
      "loss": 0.0051,
      "reward": 1.5169403553009033,
      "reward_std": 0.3250080794095993,
      "rewards/accuracy_reward": 0.5481902956962585,
      "rewards/format_reward": 0.96875,
      "step": 143
    },
    {
      "completion_length": 54.5078125,
      "epoch": 0.6575342465753424,
      "grad_norm": 3.8824496269226074,
      "kl": 0.1943359375,
      "learning_rate": 9.342465753424658e-07,
      "loss": 0.0078,
      "reward": 1.5293877124786377,
      "reward_std": 0.20680496841669083,
      "rewards/accuracy_reward": 0.5293876230716705,
      "rewards/format_reward": 1.0,
      "step": 144
    },
    {
      "completion_length": 70.8359375,
      "epoch": 0.6621004566210046,
      "grad_norm": 2.7752647399902344,
      "kl": 0.109375,
      "learning_rate": 9.337899543378995e-07,
      "loss": 0.0044,
      "reward": 1.6002604365348816,
      "reward_std": 0.16751104593276978,
      "rewards/accuracy_reward": 0.6002604067325592,
      "rewards/format_reward": 1.0,
      "step": 145
    },
    {
      "completion_length": 43.4921875,
      "epoch": 0.6666666666666666,
      "grad_norm": 4.007904529571533,
      "kl": 0.1455078125,
      "learning_rate": 9.333333333333333e-07,
      "loss": 0.0058,
      "reward": 1.4145089387893677,
      "reward_std": 0.37413105368614197,
      "rewards/accuracy_reward": 0.4301339387893677,
      "rewards/format_reward": 0.984375,
      "step": 146
    },
    {
      "completion_length": 55.015625,
      "epoch": 0.6712328767123288,
      "grad_norm": 2.844416618347168,
      "kl": 0.14501953125,
      "learning_rate": 9.328767123287671e-07,
      "loss": 0.0058,
      "reward": 1.434374988079071,
      "reward_std": 0.2953774631023407,
      "rewards/accuracy_reward": 0.44218750298023224,
      "rewards/format_reward": 0.9921875,
      "step": 147
    },
    {
      "completion_length": 63.953125,
      "epoch": 0.6757990867579908,
      "grad_norm": 3.3027069568634033,
      "kl": 0.11376953125,
      "learning_rate": 9.324200913242009e-07,
      "loss": 0.0045,
      "reward": 1.5338541269302368,
      "reward_std": 0.2085491269826889,
      "rewards/accuracy_reward": 0.5494791865348816,
      "rewards/format_reward": 0.984375,
      "step": 148
    },
    {
      "completion_length": 60.046875,
      "epoch": 0.680365296803653,
      "grad_norm": 2.4687771797180176,
      "kl": 0.150146484375,
      "learning_rate": 9.319634703196346e-07,
      "loss": 0.006,
      "reward": 1.4104894995689392,
      "reward_std": 0.2745247557759285,
      "rewards/accuracy_reward": 0.4104894697666168,
      "rewards/format_reward": 1.0,
      "step": 149
    },
    {
      "completion_length": 70.234375,
      "epoch": 0.684931506849315,
      "grad_norm": 3.1265056133270264,
      "kl": 0.112548828125,
      "learning_rate": 9.315068493150684e-07,
      "loss": 0.0045,
      "reward": 1.5757813453674316,
      "reward_std": 0.2977932393550873,
      "rewards/accuracy_reward": 0.5914062261581421,
      "rewards/format_reward": 0.984375,
      "step": 150
    },
    {
      "completion_length": 55.9609375,
      "epoch": 0.6894977168949772,
      "grad_norm": 2.7954444885253906,
      "kl": 0.138916015625,
      "learning_rate": 9.310502283105023e-07,
      "loss": 0.0056,
      "reward": 1.532336711883545,
      "reward_std": 0.2634681910276413,
      "rewards/accuracy_reward": 0.5557741522789001,
      "rewards/format_reward": 0.9765625,
      "step": 151
    },
    {
      "completion_length": 91.7734375,
      "epoch": 0.6940639269406392,
      "grad_norm": 3.276801586151123,
      "kl": 0.09521484375,
      "learning_rate": 9.30593607305936e-07,
      "loss": 0.0038,
      "reward": 1.8068453073501587,
      "reward_std": 0.13430681824684143,
      "rewards/accuracy_reward": 0.8146576881408691,
      "rewards/format_reward": 0.9921875,
      "step": 152
    },
    {
      "completion_length": 71.9453125,
      "epoch": 0.6986301369863014,
      "grad_norm": 5.107734680175781,
      "kl": 0.126708984375,
      "learning_rate": 9.301369863013698e-07,
      "loss": 0.0051,
      "reward": 1.5772569179534912,
      "reward_std": 0.3007543087005615,
      "rewards/accuracy_reward": 0.60069440305233,
      "rewards/format_reward": 0.9765625,
      "step": 153
    },
    {
      "completion_length": 65.1015625,
      "epoch": 0.7031963470319634,
      "grad_norm": 11.66229248046875,
      "kl": 0.142578125,
      "learning_rate": 9.296803652968036e-07,
      "loss": 0.0057,
      "reward": 1.5702590942382812,
      "reward_std": 0.2691568061709404,
      "rewards/accuracy_reward": 0.5780715942382812,
      "rewards/format_reward": 0.9921875,
      "step": 154
    },
    {
      "completion_length": 70.734375,
      "epoch": 0.7077625570776256,
      "grad_norm": 2.577918291091919,
      "kl": 0.101318359375,
      "learning_rate": 9.292237442922374e-07,
      "loss": 0.0041,
      "reward": 1.706770896911621,
      "reward_std": 0.17131806910037994,
      "rewards/accuracy_reward": 0.7145833075046539,
      "rewards/format_reward": 0.9921875,
      "step": 155
    },
    {
      "completion_length": 60.515625,
      "epoch": 0.7123287671232876,
      "grad_norm": 3.4716637134552,
      "kl": 0.119384765625,
      "learning_rate": 9.287671232876712e-07,
      "loss": 0.0048,
      "reward": 1.534895896911621,
      "reward_std": 0.2703789845108986,
      "rewards/accuracy_reward": 0.5817708373069763,
      "rewards/format_reward": 0.953125,
      "step": 156
    },
    {
      "completion_length": 56.15625,
      "epoch": 0.7168949771689498,
      "grad_norm": 3.562089204788208,
      "kl": 0.119140625,
      "learning_rate": 9.28310502283105e-07,
      "loss": 0.0048,
      "reward": 1.586718738079071,
      "reward_std": 0.23555129766464233,
      "rewards/accuracy_reward": 0.5867187678813934,
      "rewards/format_reward": 1.0,
      "step": 157
    },
    {
      "completion_length": 78.2890625,
      "epoch": 0.7214611872146118,
      "grad_norm": 3.0020015239715576,
      "kl": 0.124755859375,
      "learning_rate": 9.278538812785388e-07,
      "loss": 0.005,
      "reward": 1.6171875596046448,
      "reward_std": 0.29713982343673706,
      "rewards/accuracy_reward": 0.65625,
      "rewards/format_reward": 0.9609375,
      "step": 158
    },
    {
      "completion_length": 79.8203125,
      "epoch": 0.726027397260274,
      "grad_norm": 3.9287021160125732,
      "kl": 0.089599609375,
      "learning_rate": 9.273972602739726e-07,
      "loss": 0.0036,
      "reward": 1.5328125953674316,
      "reward_std": 0.3052903413772583,
      "rewards/accuracy_reward": 0.5953125059604645,
      "rewards/format_reward": 0.9375,
      "step": 159
    },
    {
      "completion_length": 81.546875,
      "epoch": 0.730593607305936,
      "grad_norm": 4.853604793548584,
      "kl": 0.142333984375,
      "learning_rate": 9.269406392694063e-07,
      "loss": 0.0057,
      "reward": 1.6752576231956482,
      "reward_std": 0.17081036418676376,
      "rewards/accuracy_reward": 0.6830700039863586,
      "rewards/format_reward": 0.9921875,
      "step": 160
    },
    {
      "completion_length": 62.828125,
      "epoch": 0.7351598173515982,
      "grad_norm": 6.2052903175354,
      "kl": 0.14013671875,
      "learning_rate": 9.264840182648401e-07,
      "loss": 0.0056,
      "reward": 1.4204427003860474,
      "reward_std": 0.3042631521821022,
      "rewards/accuracy_reward": 0.42825521528720856,
      "rewards/format_reward": 0.9921875,
      "step": 161
    },
    {
      "completion_length": 113.703125,
      "epoch": 0.7397260273972602,
      "grad_norm": 0.730038583278656,
      "kl": 0.066162109375,
      "learning_rate": 9.260273972602739e-07,
      "loss": 0.0026,
      "reward": 1.8767098784446716,
      "reward_std": 0.03975973278284073,
      "rewards/accuracy_reward": 0.8845223784446716,
      "rewards/format_reward": 0.9921875,
      "step": 162
    },
    {
      "completion_length": 77.8515625,
      "epoch": 0.7442922374429224,
      "grad_norm": 2.9642155170440674,
      "kl": 0.10009765625,
      "learning_rate": 9.255707762557077e-07,
      "loss": 0.004,
      "reward": 1.702742099761963,
      "reward_std": 0.18021715432405472,
      "rewards/accuracy_reward": 0.7105545401573181,
      "rewards/format_reward": 0.9921875,
      "step": 163
    },
    {
      "completion_length": 73.984375,
      "epoch": 0.7488584474885844,
      "grad_norm": 3.072006940841675,
      "kl": 0.13330078125,
      "learning_rate": 9.251141552511416e-07,
      "loss": 0.0053,
      "reward": 1.768545389175415,
      "reward_std": 0.1539178043603897,
      "rewards/accuracy_reward": 0.768545389175415,
      "rewards/format_reward": 1.0,
      "step": 164
    },
    {
      "completion_length": 65.953125,
      "epoch": 0.7534246575342466,
      "grad_norm": 5.146835803985596,
      "kl": 0.12841796875,
      "learning_rate": 9.246575342465753e-07,
      "loss": 0.0051,
      "reward": 1.503950297832489,
      "reward_std": 0.3056825324892998,
      "rewards/accuracy_reward": 0.5195753127336502,
      "rewards/format_reward": 0.984375,
      "step": 165
    },
    {
      "completion_length": 91.4921875,
      "epoch": 0.7579908675799086,
      "grad_norm": 2.6764745712280273,
      "kl": 0.1031494140625,
      "learning_rate": 9.242009132420091e-07,
      "loss": 0.0041,
      "reward": 1.5589421391487122,
      "reward_std": 0.23093140125274658,
      "rewards/accuracy_reward": 0.5745670646429062,
      "rewards/format_reward": 0.984375,
      "step": 166
    },
    {
      "completion_length": 81.7734375,
      "epoch": 0.7625570776255708,
      "grad_norm": 2.828564405441284,
      "kl": 0.12109375,
      "learning_rate": 9.237442922374429e-07,
      "loss": 0.0048,
      "reward": 1.5663008093833923,
      "reward_std": 0.20618148148059845,
      "rewards/accuracy_reward": 0.5975507497787476,
      "rewards/format_reward": 0.96875,
      "step": 167
    },
    {
      "completion_length": 76.296875,
      "epoch": 0.7671232876712328,
      "grad_norm": 2.9903767108917236,
      "kl": 0.1298828125,
      "learning_rate": 9.232876712328766e-07,
      "loss": 0.0052,
      "reward": 1.645312488079071,
      "reward_std": 0.2312253788113594,
      "rewards/accuracy_reward": 0.6531249731779099,
      "rewards/format_reward": 0.9921875,
      "step": 168
    },
    {
      "completion_length": 77.71875,
      "epoch": 0.771689497716895,
      "grad_norm": 2.3729984760284424,
      "kl": 0.103759765625,
      "learning_rate": 9.228310502283104e-07,
      "loss": 0.0042,
      "reward": 1.5843749642372131,
      "reward_std": 0.26058951020240784,
      "rewards/accuracy_reward": 0.6000000238418579,
      "rewards/format_reward": 0.984375,
      "step": 169
    },
    {
      "completion_length": 109.0234375,
      "epoch": 0.776255707762557,
      "grad_norm": 2.769318103790283,
      "kl": 0.076416015625,
      "learning_rate": 9.223744292237442e-07,
      "loss": 0.0031,
      "reward": 1.7934027910232544,
      "reward_std": 0.09004722535610199,
      "rewards/accuracy_reward": 0.801215261220932,
      "rewards/format_reward": 0.9921875,
      "step": 170
    },
    {
      "completion_length": 75.8203125,
      "epoch": 0.7808219178082192,
      "grad_norm": 2.6014890670776367,
      "kl": 0.131103515625,
      "learning_rate": 9.219178082191781e-07,
      "loss": 0.0052,
      "reward": 1.5569196343421936,
      "reward_std": 0.26341303437948227,
      "rewards/accuracy_reward": 0.580357164144516,
      "rewards/format_reward": 0.9765625,
      "step": 171
    },
    {
      "completion_length": 97.9375,
      "epoch": 0.7853881278538812,
      "grad_norm": 2.9145772457122803,
      "kl": 0.091064453125,
      "learning_rate": 9.214611872146119e-07,
      "loss": 0.0036,
      "reward": 1.6800222992897034,
      "reward_std": 0.15495410561561584,
      "rewards/accuracy_reward": 0.6956472992897034,
      "rewards/format_reward": 0.984375,
      "step": 172
    },
    {
      "completion_length": 82.0234375,
      "epoch": 0.7899543378995434,
      "grad_norm": 5.809459686279297,
      "kl": 0.27392578125,
      "learning_rate": 9.210045662100456e-07,
      "loss": 0.0109,
      "reward": 1.489843726158142,
      "reward_std": 0.22895930707454681,
      "rewards/accuracy_reward": 0.4898437559604645,
      "rewards/format_reward": 1.0,
      "step": 173
    },
    {
      "completion_length": 76.0859375,
      "epoch": 0.7945205479452054,
      "grad_norm": 4.242368698120117,
      "kl": 0.111572265625,
      "learning_rate": 9.205479452054794e-07,
      "loss": 0.0045,
      "reward": 1.4959192276000977,
      "reward_std": 0.2452109009027481,
      "rewards/accuracy_reward": 0.5037316679954529,
      "rewards/format_reward": 0.9921875,
      "step": 174
    },
    {
      "completion_length": 72.4921875,
      "epoch": 0.7990867579908676,
      "grad_norm": 4.817114353179932,
      "kl": 0.12060546875,
      "learning_rate": 9.200913242009132e-07,
      "loss": 0.0048,
      "reward": 1.607180118560791,
      "reward_std": 0.2259521484375,
      "rewards/accuracy_reward": 0.6149925291538239,
      "rewards/format_reward": 0.9921875,
      "step": 175
    },
    {
      "completion_length": 73.3828125,
      "epoch": 0.8036529680365296,
      "grad_norm": 5.523159027099609,
      "kl": 0.11669921875,
      "learning_rate": 9.196347031963469e-07,
      "loss": 0.0047,
      "reward": 1.4424473643302917,
      "reward_std": 0.35005757212638855,
      "rewards/accuracy_reward": 0.46588489413261414,
      "rewards/format_reward": 0.9765625,
      "step": 176
    },
    {
      "completion_length": 83.7109375,
      "epoch": 0.8082191780821918,
      "grad_norm": 6.852357387542725,
      "kl": 0.101318359375,
      "learning_rate": 9.191780821917808e-07,
      "loss": 0.0041,
      "reward": 1.7020823955535889,
      "reward_std": 0.1837347373366356,
      "rewards/accuracy_reward": 0.7020823657512665,
      "rewards/format_reward": 1.0,
      "step": 177
    },
    {
      "completion_length": 95.171875,
      "epoch": 0.8127853881278538,
      "grad_norm": 3.1507697105407715,
      "kl": 0.0908203125,
      "learning_rate": 9.187214611872146e-07,
      "loss": 0.0036,
      "reward": 1.6673295497894287,
      "reward_std": 0.14267417788505554,
      "rewards/accuracy_reward": 0.6673295497894287,
      "rewards/format_reward": 1.0,
      "step": 178
    },
    {
      "completion_length": 110.1640625,
      "epoch": 0.817351598173516,
      "grad_norm": 2.0009899139404297,
      "kl": 0.0653076171875,
      "learning_rate": 9.182648401826484e-07,
      "loss": 0.0026,
      "reward": 1.7578993439674377,
      "reward_std": 0.12590338289737701,
      "rewards/accuracy_reward": 0.765711784362793,
      "rewards/format_reward": 0.9921875,
      "step": 179
    },
    {
      "completion_length": 83.59375,
      "epoch": 0.821917808219178,
      "grad_norm": 2.3591296672821045,
      "kl": 0.09765625,
      "learning_rate": 9.178082191780822e-07,
      "loss": 0.0039,
      "reward": 1.72604501247406,
      "reward_std": 0.1832278147339821,
      "rewards/accuracy_reward": 0.7338575124740601,
      "rewards/format_reward": 0.9921875,
      "step": 180
    },
    {
      "completion_length": 90.8125,
      "epoch": 0.8264840182648402,
      "grad_norm": 4.365949630737305,
      "kl": 0.1068115234375,
      "learning_rate": 9.173515981735159e-07,
      "loss": 0.0043,
      "reward": 1.7241073250770569,
      "reward_std": 0.1286808280274272,
      "rewards/accuracy_reward": 0.7241072654724121,
      "rewards/format_reward": 1.0,
      "step": 181
    },
    {
      "completion_length": 70.1015625,
      "epoch": 0.8310502283105022,
      "grad_norm": 4.564199924468994,
      "kl": 0.1298828125,
      "learning_rate": 9.168949771689497e-07,
      "loss": 0.0052,
      "reward": 1.4811267852783203,
      "reward_std": 0.23117025196552277,
      "rewards/accuracy_reward": 0.4967518150806427,
      "rewards/format_reward": 0.984375,
      "step": 182
    },
    {
      "completion_length": 70.0859375,
      "epoch": 0.8356164383561644,
      "grad_norm": 8.132929801940918,
      "kl": 0.15234375,
      "learning_rate": 9.164383561643835e-07,
      "loss": 0.0061,
      "reward": 1.513009488582611,
      "reward_std": 0.31318600475788116,
      "rewards/accuracy_reward": 0.5208219289779663,
      "rewards/format_reward": 0.9921875,
      "step": 183
    },
    {
      "completion_length": 87.390625,
      "epoch": 0.8401826484018264,
      "grad_norm": 2.4152746200561523,
      "kl": 0.086669921875,
      "learning_rate": 9.159817351598174e-07,
      "loss": 0.0035,
      "reward": 1.6757907271385193,
      "reward_std": 0.16663195937871933,
      "rewards/accuracy_reward": 0.6836032569408417,
      "rewards/format_reward": 0.9921875,
      "step": 184
    },
    {
      "completion_length": 71.40625,
      "epoch": 0.8447488584474886,
      "grad_norm": 3.8499457836151123,
      "kl": 0.14111328125,
      "learning_rate": 9.155251141552511e-07,
      "loss": 0.0057,
      "reward": 1.418749988079071,
      "reward_std": 0.27360689640045166,
      "rewards/accuracy_reward": 0.42656251788139343,
      "rewards/format_reward": 0.9921875,
      "step": 185
    },
    {
      "completion_length": 100.109375,
      "epoch": 0.8493150684931506,
      "grad_norm": 1.6684777736663818,
      "kl": 0.103271484375,
      "learning_rate": 9.150684931506849e-07,
      "loss": 0.0041,
      "reward": 1.820052146911621,
      "reward_std": 0.09526955150067806,
      "rewards/accuracy_reward": 0.8278645575046539,
      "rewards/format_reward": 0.9921875,
      "step": 186
    },
    {
      "completion_length": 110.109375,
      "epoch": 0.8538812785388128,
      "grad_norm": 2.6076531410217285,
      "kl": 0.0618896484375,
      "learning_rate": 9.146118721461187e-07,
      "loss": 0.0025,
      "reward": 1.6446732878684998,
      "reward_std": 0.1990083083510399,
      "rewards/accuracy_reward": 0.6602982878684998,
      "rewards/format_reward": 0.984375,
      "step": 187
    },
    {
      "completion_length": 73.921875,
      "epoch": 0.8584474885844748,
      "grad_norm": 3.0167715549468994,
      "kl": 0.129150390625,
      "learning_rate": 9.141552511415525e-07,
      "loss": 0.0052,
      "reward": 1.4889777302742004,
      "reward_std": 0.310699999332428,
      "rewards/accuracy_reward": 0.520227700471878,
      "rewards/format_reward": 0.96875,
      "step": 188
    },
    {
      "completion_length": 77.5625,
      "epoch": 0.863013698630137,
      "grad_norm": 5.573866844177246,
      "kl": 0.142578125,
      "learning_rate": 9.136986301369862e-07,
      "loss": 0.0057,
      "reward": 1.5417782664299011,
      "reward_std": 0.21660251170396805,
      "rewards/accuracy_reward": 0.5417782664299011,
      "rewards/format_reward": 1.0,
      "step": 189
    },
    {
      "completion_length": 98.015625,
      "epoch": 0.867579908675799,
      "grad_norm": 3.7342417240142822,
      "kl": 0.12890625,
      "learning_rate": 9.1324200913242e-07,
      "loss": 0.0051,
      "reward": 1.5212674140930176,
      "reward_std": 0.3198610842227936,
      "rewards/accuracy_reward": 0.5681423544883728,
      "rewards/format_reward": 0.953125,
      "step": 190
    },
    {
      "completion_length": 87.9609375,
      "epoch": 0.8721461187214612,
      "grad_norm": 5.50727653503418,
      "kl": 0.09033203125,
      "learning_rate": 9.127853881278539e-07,
      "loss": 0.0036,
      "reward": 1.4584822058677673,
      "reward_std": 0.2924446016550064,
      "rewards/accuracy_reward": 0.47410714626312256,
      "rewards/format_reward": 0.984375,
      "step": 191
    },
    {
      "completion_length": 86.59375,
      "epoch": 0.8767123287671232,
      "grad_norm": 4.182096004486084,
      "kl": 0.110595703125,
      "learning_rate": 9.123287671232876e-07,
      "loss": 0.0044,
      "reward": 1.4973958134651184,
      "reward_std": 0.27411043643951416,
      "rewards/accuracy_reward": 0.5130208432674408,
      "rewards/format_reward": 0.984375,
      "step": 192
    },
    {
      "completion_length": 83.3359375,
      "epoch": 0.8812785388127854,
      "grad_norm": 3.033630609512329,
      "kl": 0.1015625,
      "learning_rate": 9.118721461187214e-07,
      "loss": 0.0041,
      "reward": 1.6757813096046448,
      "reward_std": 0.18561583012342453,
      "rewards/accuracy_reward": 0.6835937201976776,
      "rewards/format_reward": 0.9921875,
      "step": 193
    },
    {
      "completion_length": 86.953125,
      "epoch": 0.8858447488584474,
      "grad_norm": 2.5366592407226562,
      "kl": 0.0810546875,
      "learning_rate": 9.114155251141552e-07,
      "loss": 0.0032,
      "reward": 1.605208396911621,
      "reward_std": 0.2767683416604996,
      "rewards/accuracy_reward": 0.6286458075046539,
      "rewards/format_reward": 0.9765625,
      "step": 194
    },
    {
      "completion_length": 87.75,
      "epoch": 0.8904109589041096,
      "grad_norm": 2.646970748901367,
      "kl": 0.121826171875,
      "learning_rate": 9.10958904109589e-07,
      "loss": 0.0049,
      "reward": 1.6739211678504944,
      "reward_std": 0.22777877748012543,
      "rewards/accuracy_reward": 0.6817336082458496,
      "rewards/format_reward": 0.9921875,
      "step": 195
    },
    {
      "completion_length": 85.375,
      "epoch": 0.8949771689497716,
      "grad_norm": 4.142679214477539,
      "kl": 0.116455078125,
      "learning_rate": 9.105022831050228e-07,
      "loss": 0.0047,
      "reward": 1.5367559790611267,
      "reward_std": 0.22284159809350967,
      "rewards/accuracy_reward": 0.5367559492588043,
      "rewards/format_reward": 1.0,
      "step": 196
    },
    {
      "completion_length": 73.3359375,
      "epoch": 0.8995433789954338,
      "grad_norm": 98.02293395996094,
      "kl": 0.123779296875,
      "learning_rate": 9.100456621004566e-07,
      "loss": 0.005,
      "reward": 1.613690435886383,
      "reward_std": 0.3185143321752548,
      "rewards/accuracy_reward": 0.6293154358863831,
      "rewards/format_reward": 0.984375,
      "step": 197
    },
    {
      "completion_length": 68.6796875,
      "epoch": 0.9041095890410958,
      "grad_norm": 3.6209559440612793,
      "kl": 0.13037109375,
      "learning_rate": 9.095890410958904e-07,
      "loss": 0.0052,
      "reward": 1.440234363079071,
      "reward_std": 0.31120626628398895,
      "rewards/accuracy_reward": 0.47148437798023224,
      "rewards/format_reward": 0.96875,
      "step": 198
    },
    {
      "completion_length": 79.1953125,
      "epoch": 0.908675799086758,
      "grad_norm": 5.003497123718262,
      "kl": 0.1123046875,
      "learning_rate": 9.091324200913242e-07,
      "loss": 0.0045,
      "reward": 1.648740530014038,
      "reward_std": 0.1829584613442421,
      "rewards/accuracy_reward": 0.6487405002117157,
      "rewards/format_reward": 1.0,
      "step": 199
    },
    {
      "completion_length": 102.6171875,
      "epoch": 0.91324200913242,
      "grad_norm": 2.219266891479492,
      "kl": 0.0859375,
      "learning_rate": 9.08675799086758e-07,
      "loss": 0.0034,
      "reward": 1.7168915271759033,
      "reward_std": 0.12932297587394714,
      "rewards/accuracy_reward": 0.7247040569782257,
      "rewards/format_reward": 0.9921875,
      "step": 200
    },
    {
      "completion_length": 72.140625,
      "epoch": 0.9178082191780822,
      "grad_norm": 3.0909345149993896,
      "kl": 0.1162109375,
      "learning_rate": 9.082191780821917e-07,
      "loss": 0.0046,
      "reward": 1.4611505270004272,
      "reward_std": 0.3023676201701164,
      "rewards/accuracy_reward": 0.4924005717039108,
      "rewards/format_reward": 0.96875,
      "step": 201
    },
    {
      "completion_length": 94.140625,
      "epoch": 0.9223744292237442,
      "grad_norm": 4.803802967071533,
      "kl": 0.103759765625,
      "learning_rate": 9.077625570776255e-07,
      "loss": 0.0042,
      "reward": 1.6710938215255737,
      "reward_std": 0.18428679555654526,
      "rewards/accuracy_reward": 0.6789062321186066,
      "rewards/format_reward": 0.9921875,
      "step": 202
    },
    {
      "completion_length": 86.5234375,
      "epoch": 0.9269406392694064,
      "grad_norm": 4.511474609375,
      "kl": 0.078125,
      "learning_rate": 9.073059360730593e-07,
      "loss": 0.0031,
      "reward": 1.6398438215255737,
      "reward_std": 0.17492686957120895,
      "rewards/accuracy_reward": 0.6632812321186066,
      "rewards/format_reward": 0.9765625,
      "step": 203
    },
    {
      "completion_length": 77.4296875,
      "epoch": 0.9315068493150684,
      "grad_norm": 3.1800436973571777,
      "kl": 0.1025390625,
      "learning_rate": 9.068493150684932e-07,
      "loss": 0.0041,
      "reward": 1.5895833373069763,
      "reward_std": 0.2831973433494568,
      "rewards/accuracy_reward": 0.6130208373069763,
      "rewards/format_reward": 0.9765625,
      "step": 204
    },
    {
      "completion_length": 78.21875,
      "epoch": 0.9360730593607306,
      "grad_norm": 2.723695993423462,
      "kl": 0.110107421875,
      "learning_rate": 9.063926940639269e-07,
      "loss": 0.0044,
      "reward": 1.508962869644165,
      "reward_std": 0.293969988822937,
      "rewards/accuracy_reward": 0.5324003100395203,
      "rewards/format_reward": 0.9765625,
      "step": 205
    },
    {
      "completion_length": 71.1953125,
      "epoch": 0.9406392694063926,
      "grad_norm": 4.026218414306641,
      "kl": 0.125244140625,
      "learning_rate": 9.059360730593607e-07,
      "loss": 0.005,
      "reward": 1.475000023841858,
      "reward_std": 0.31207825243473053,
      "rewards/accuracy_reward": 0.4828125238418579,
      "rewards/format_reward": 0.9921875,
      "step": 206
    },
    {
      "completion_length": 91.5546875,
      "epoch": 0.9452054794520548,
      "grad_norm": 3.48559832572937,
      "kl": 0.078125,
      "learning_rate": 9.054794520547945e-07,
      "loss": 0.0031,
      "reward": 1.5565290451049805,
      "reward_std": 0.17772940546274185,
      "rewards/accuracy_reward": 0.5721540153026581,
      "rewards/format_reward": 0.984375,
      "step": 207
    },
    {
      "completion_length": 80.015625,
      "epoch": 0.9497716894977168,
      "grad_norm": 2.530951499938965,
      "kl": 0.103759765625,
      "learning_rate": 9.050228310502282e-07,
      "loss": 0.0042,
      "reward": 1.7421875,
      "reward_std": 0.17887144908308983,
      "rewards/accuracy_reward": 0.7499999403953552,
      "rewards/format_reward": 0.9921875,
      "step": 208
    },
    {
      "completion_length": 85.671875,
      "epoch": 0.954337899543379,
      "grad_norm": 3.1076207160949707,
      "kl": 0.084228515625,
      "learning_rate": 9.04566210045662e-07,
      "loss": 0.0034,
      "reward": 1.6094618439674377,
      "reward_std": 0.16175774857401848,
      "rewards/accuracy_reward": 0.617274284362793,
      "rewards/format_reward": 0.9921875,
      "step": 209
    },
    {
      "completion_length": 84.546875,
      "epoch": 0.958904109589041,
      "grad_norm": 2.6074228286743164,
      "kl": 0.076171875,
      "learning_rate": 9.041095890410958e-07,
      "loss": 0.003,
      "reward": 1.5931640267372131,
      "reward_std": 0.18021905422210693,
      "rewards/accuracy_reward": 0.5931640565395355,
      "rewards/format_reward": 1.0,
      "step": 210
    },
    {
      "completion_length": 74.015625,
      "epoch": 0.9634703196347032,
      "grad_norm": 3.9831788539886475,
      "kl": 0.087890625,
      "learning_rate": 9.036529680365297e-07,
      "loss": 0.0035,
      "reward": 1.5644965171813965,
      "reward_std": 0.20533857494592667,
      "rewards/accuracy_reward": 0.5644965171813965,
      "rewards/format_reward": 1.0,
      "step": 211
    },
    {
      "completion_length": 60.9296875,
      "epoch": 0.9680365296803652,
      "grad_norm": 2.8256754875183105,
      "kl": 0.10986328125,
      "learning_rate": 9.031963470319635e-07,
      "loss": 0.0044,
      "reward": 1.5828006267547607,
      "reward_std": 0.2151413932442665,
      "rewards/accuracy_reward": 0.5906131863594055,
      "rewards/format_reward": 0.9921875,
      "step": 212
    },
    {
      "completion_length": 88.8515625,
      "epoch": 0.9726027397260274,
      "grad_norm": 2.615966796875,
      "kl": 0.113037109375,
      "learning_rate": 9.027397260273972e-07,
      "loss": 0.0045,
      "reward": 1.5743975639343262,
      "reward_std": 0.24481570720672607,
      "rewards/accuracy_reward": 0.5822100639343262,
      "rewards/format_reward": 0.9921875,
      "step": 213
    },
    {
      "completion_length": 66.3359375,
      "epoch": 0.9771689497716894,
      "grad_norm": 11.469498634338379,
      "kl": 0.106201171875,
      "learning_rate": 9.02283105022831e-07,
      "loss": 0.0042,
      "reward": 1.7420889139175415,
      "reward_std": 0.13549309968948364,
      "rewards/accuracy_reward": 0.7420888245105743,
      "rewards/format_reward": 1.0,
      "step": 214
    },
    {
      "completion_length": 74.0390625,
      "epoch": 0.9817351598173516,
      "grad_norm": 3.873460292816162,
      "kl": 0.0693359375,
      "learning_rate": 9.018264840182648e-07,
      "loss": 0.0028,
      "reward": 1.6421875953674316,
      "reward_std": 0.2022070661187172,
      "rewards/accuracy_reward": 0.6421874761581421,
      "rewards/format_reward": 1.0,
      "step": 215
    },
    {
      "completion_length": 93.375,
      "epoch": 0.9863013698630136,
      "grad_norm": 2.612290143966675,
      "kl": 0.09765625,
      "learning_rate": 9.013698630136985e-07,
      "loss": 0.0039,
      "reward": 1.645312488079071,
      "reward_std": 0.25715554505586624,
      "rewards/accuracy_reward": 0.676562488079071,
      "rewards/format_reward": 0.96875,
      "step": 216
    },
    {
      "completion_length": 83.8515625,
      "epoch": 0.9908675799086758,
      "grad_norm": 1.9043883085250854,
      "kl": 0.077880859375,
      "learning_rate": 9.009132420091324e-07,
      "loss": 0.0031,
      "reward": 1.6007813215255737,
      "reward_std": 0.2224196195602417,
      "rewards/accuracy_reward": 0.6242187321186066,
      "rewards/format_reward": 0.9765625,
      "step": 217
    },
    {
      "completion_length": 70.2421875,
      "epoch": 0.9954337899543378,
      "grad_norm": 9.268891334533691,
      "kl": 0.114013671875,
      "learning_rate": 9.004566210045662e-07,
      "loss": 0.0046,
      "reward": 1.487464964389801,
      "reward_std": 0.3045327961444855,
      "rewards/accuracy_reward": 0.510902464389801,
      "rewards/format_reward": 0.9765625,
      "step": 218
    },
    {
      "completion_length": 138.875,
      "epoch": 1.0,
      "grad_norm": 2.398301839828491,
      "kl": 0.1318359375,
      "learning_rate": 9e-07,
      "loss": 0.0041,
      "reward": 1.899999976158142,
      "reward_std": 0.0,
      "rewards/accuracy_reward": 0.8999999761581421,
      "rewards/format_reward": 1.0,
      "step": 219
    },
    {
      "completion_length": 70.625,
      "epoch": 1.004566210045662,
      "grad_norm": 3.741870880126953,
      "kl": 0.12158203125,
      "learning_rate": 8.995433789954338e-07,
      "loss": 0.0049,
      "reward": 1.50390625,
      "reward_std": 0.26016832888126373,
      "rewards/accuracy_reward": 0.5039062350988388,
      "rewards/format_reward": 1.0,
      "step": 220
    },
    {
      "completion_length": 64.859375,
      "epoch": 1.009132420091324,
      "grad_norm": 11.450385093688965,
      "kl": 0.133056640625,
      "learning_rate": 8.990867579908675e-07,
      "loss": 0.0053,
      "reward": 1.7419270873069763,
      "reward_std": 0.11560981348156929,
      "rewards/accuracy_reward": 0.7419269979000092,
      "rewards/format_reward": 1.0,
      "step": 221
    },
    {
      "completion_length": 82.046875,
      "epoch": 1.0136986301369864,
      "grad_norm": 3.4675490856170654,
      "kl": 0.08935546875,
      "learning_rate": 8.986301369863013e-07,
      "loss": 0.0036,
      "reward": 1.6703497171401978,
      "reward_std": 0.2566395699977875,
      "rewards/accuracy_reward": 0.6859746873378754,
      "rewards/format_reward": 0.984375,
      "step": 222
    },
    {
      "completion_length": 88.4375,
      "epoch": 1.0182648401826484,
      "grad_norm": 2.4552440643310547,
      "kl": 0.11328125,
      "learning_rate": 8.981735159817351e-07,
      "loss": 0.0045,
      "reward": 1.646279752254486,
      "reward_std": 0.1442876234650612,
      "rewards/accuracy_reward": 0.6462797522544861,
      "rewards/format_reward": 1.0,
      "step": 223
    },
    {
      "completion_length": 60.6171875,
      "epoch": 1.0228310502283104,
      "grad_norm": 2.843845844268799,
      "kl": 0.144287109375,
      "learning_rate": 8.97716894977169e-07,
      "loss": 0.0058,
      "reward": 1.5498643517494202,
      "reward_std": 0.2656385153532028,
      "rewards/accuracy_reward": 0.5498644113540649,
      "rewards/format_reward": 1.0,
      "step": 224
    },
    {
      "completion_length": 59.640625,
      "epoch": 1.0273972602739727,
      "grad_norm": 6.659507751464844,
      "kl": 0.26123046875,
      "learning_rate": 8.972602739726027e-07,
      "loss": 0.0105,
      "reward": 1.522736370563507,
      "reward_std": 0.23698078095912933,
      "rewards/accuracy_reward": 0.5383614003658295,
      "rewards/format_reward": 0.984375,
      "step": 225
    },
    {
      "completion_length": 76.2578125,
      "epoch": 1.0319634703196348,
      "grad_norm": 7.1479926109313965,
      "kl": 0.11474609375,
      "learning_rate": 8.968036529680365e-07,
      "loss": 0.0046,
      "reward": 1.5441184043884277,
      "reward_std": 0.1862540915608406,
      "rewards/accuracy_reward": 0.5519307851791382,
      "rewards/format_reward": 0.9921875,
      "step": 226
    },
    {
      "completion_length": 52.7109375,
      "epoch": 1.0365296803652968,
      "grad_norm": 5.714673042297363,
      "kl": 0.17578125,
      "learning_rate": 8.963470319634703e-07,
      "loss": 0.007,
      "reward": 1.4614583253860474,
      "reward_std": 0.3212399482727051,
      "rewards/accuracy_reward": 0.47708331048488617,
      "rewards/format_reward": 0.984375,
      "step": 227
    },
    {
      "completion_length": 59.234375,
      "epoch": 1.0410958904109588,
      "grad_norm": 2.191619396209717,
      "kl": 0.143798828125,
      "learning_rate": 8.958904109589041e-07,
      "loss": 0.0058,
      "reward": 1.6684895753860474,
      "reward_std": 0.1687004156410694,
      "rewards/accuracy_reward": 0.6763020753860474,
      "rewards/format_reward": 0.9921875,
      "step": 228
    },
    {
      "completion_length": 64.375,
      "epoch": 1.045662100456621,
      "grad_norm": 7.204253196716309,
      "kl": 0.1484375,
      "learning_rate": 8.954337899543378e-07,
      "loss": 0.0059,
      "reward": 1.6486244797706604,
      "reward_std": 0.20006748288869858,
      "rewards/accuracy_reward": 0.6486244797706604,
      "rewards/format_reward": 1.0,
      "step": 229
    },
    {
      "completion_length": 68.6796875,
      "epoch": 1.0502283105022832,
      "grad_norm": 34.476558685302734,
      "kl": 0.112060546875,
      "learning_rate": 8.949771689497716e-07,
      "loss": 0.0045,
      "reward": 1.484375,
      "reward_std": 0.2940969169139862,
      "rewards/accuracy_reward": 0.5078124701976776,
      "rewards/format_reward": 0.9765625,
      "step": 230
    },
    {
      "completion_length": 60.7421875,
      "epoch": 1.0547945205479452,
      "grad_norm": 1.8852620124816895,
      "kl": 0.15234375,
      "learning_rate": 8.945205479452055e-07,
      "loss": 0.0061,
      "reward": 1.641055941581726,
      "reward_std": 0.19024673104286194,
      "rewards/accuracy_reward": 0.6488684415817261,
      "rewards/format_reward": 0.9921875,
      "step": 231
    },
    {
      "completion_length": 57.875,
      "epoch": 1.0593607305936072,
      "grad_norm": 5.363572597503662,
      "kl": 0.19677734375,
      "learning_rate": 8.940639269406392e-07,
      "loss": 0.0079,
      "reward": 1.5763392448425293,
      "reward_std": 0.3014761507511139,
      "rewards/accuracy_reward": 0.5997768044471741,
      "rewards/format_reward": 0.9765625,
      "step": 232
    },
    {
      "completion_length": 58.1484375,
      "epoch": 1.0639269406392695,
      "grad_norm": 2.532191276550293,
      "kl": 0.22900390625,
      "learning_rate": 8.93607305936073e-07,
      "loss": 0.0092,
      "reward": 1.5628709197044373,
      "reward_std": 0.1907019466161728,
      "rewards/accuracy_reward": 0.5706833899021149,
      "rewards/format_reward": 0.9921875,
      "step": 233
    },
    {
      "completion_length": 66.8046875,
      "epoch": 1.0684931506849316,
      "grad_norm": 5.099847793579102,
      "kl": 0.16259765625,
      "learning_rate": 8.931506849315068e-07,
      "loss": 0.0065,
      "reward": 1.6207798719406128,
      "reward_std": 0.2609640061855316,
      "rewards/accuracy_reward": 0.659842312335968,
      "rewards/format_reward": 0.9609375,
      "step": 234
    },
    {
      "completion_length": 60.0078125,
      "epoch": 1.0730593607305936,
      "grad_norm": 4.480944633483887,
      "kl": 0.22509765625,
      "learning_rate": 8.926940639269406e-07,
      "loss": 0.009,
      "reward": 1.514843761920929,
      "reward_std": 0.3193802535533905,
      "rewards/accuracy_reward": 0.5460937023162842,
      "rewards/format_reward": 0.96875,
      "step": 235
    },
    {
      "completion_length": 59.7109375,
      "epoch": 1.0776255707762556,
      "grad_norm": 3.332169532775879,
      "kl": 0.203125,
      "learning_rate": 8.922374429223744e-07,
      "loss": 0.0081,
      "reward": 1.5903645753860474,
      "reward_std": 0.2678108364343643,
      "rewards/accuracy_reward": 0.6216145753860474,
      "rewards/format_reward": 0.96875,
      "step": 236
    },
    {
      "completion_length": 60.46875,
      "epoch": 1.0821917808219177,
      "grad_norm": 2.0340123176574707,
      "kl": 0.14794921875,
      "learning_rate": 8.917808219178081e-07,
      "loss": 0.0059,
      "reward": 1.5674479603767395,
      "reward_std": 0.2526354044675827,
      "rewards/accuracy_reward": 0.5752603709697723,
      "rewards/format_reward": 0.9921875,
      "step": 237
    },
    {
      "completion_length": 57.0859375,
      "epoch": 1.08675799086758,
      "grad_norm": 2.9721603393554688,
      "kl": 0.18408203125,
      "learning_rate": 8.91324200913242e-07,
      "loss": 0.0074,
      "reward": 1.5576340556144714,
      "reward_std": 0.17624534666538239,
      "rewards/accuracy_reward": 0.5576339960098267,
      "rewards/format_reward": 1.0,
      "step": 238
    },
    {
      "completion_length": 90.328125,
      "epoch": 1.091324200913242,
      "grad_norm": 2.5770187377929688,
      "kl": 0.18798828125,
      "learning_rate": 8.908675799086758e-07,
      "loss": 0.0075,
      "reward": 1.7246136665344238,
      "reward_std": 0.16096369177103043,
      "rewards/accuracy_reward": 0.7480511367321014,
      "rewards/format_reward": 0.9765625,
      "step": 239
    },
    {
      "completion_length": 42.3828125,
      "epoch": 1.095890410958904,
      "grad_norm": 1.3982386589050293,
      "kl": 0.20263671875,
      "learning_rate": 8.904109589041095e-07,
      "loss": 0.0081,
      "reward": 1.5574405193328857,
      "reward_std": 0.1670553982257843,
      "rewards/accuracy_reward": 0.5574404895305634,
      "rewards/format_reward": 1.0,
      "step": 240
    },
    {
      "completion_length": 71.734375,
      "epoch": 1.1004566210045663,
      "grad_norm": 3.7010498046875,
      "kl": 0.16162109375,
      "learning_rate": 8.899543378995433e-07,
      "loss": 0.0065,
      "reward": 1.8125000596046448,
      "reward_std": 0.12704820185899734,
      "rewards/accuracy_reward": 0.8124999701976776,
      "rewards/format_reward": 1.0,
      "step": 241
    },
    {
      "completion_length": 65.9609375,
      "epoch": 1.1050228310502284,
      "grad_norm": 2.443540096282959,
      "kl": 0.1474609375,
      "learning_rate": 8.894977168949771e-07,
      "loss": 0.0059,
      "reward": 1.6270038485527039,
      "reward_std": 0.2314547374844551,
      "rewards/accuracy_reward": 0.6426288187503815,
      "rewards/format_reward": 0.984375,
      "step": 242
    },
    {
      "completion_length": 63.46875,
      "epoch": 1.1095890410958904,
      "grad_norm": 2.410088062286377,
      "kl": 0.1611328125,
      "learning_rate": 8.890410958904109e-07,
      "loss": 0.0064,
      "reward": 1.5,
      "reward_std": 0.2109457477927208,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 1.0,
      "step": 243
    },
    {
      "completion_length": 61.3671875,
      "epoch": 1.1141552511415524,
      "grad_norm": 3.672989845275879,
      "kl": 0.21826171875,
      "learning_rate": 8.885844748858448e-07,
      "loss": 0.0087,
      "reward": 1.6392003893852234,
      "reward_std": 0.21675443649291992,
      "rewards/accuracy_reward": 0.6626378297805786,
      "rewards/format_reward": 0.9765625,
      "step": 244
    },
    {
      "completion_length": 67.1171875,
      "epoch": 1.1187214611872145,
      "grad_norm": 2.5747616291046143,
      "kl": 0.1923828125,
      "learning_rate": 8.881278538812785e-07,
      "loss": 0.0077,
      "reward": 1.5857422351837158,
      "reward_std": 0.18916229903697968,
      "rewards/accuracy_reward": 0.5935547053813934,
      "rewards/format_reward": 0.9921875,
      "step": 245
    },
    {
      "completion_length": 51.1640625,
      "epoch": 1.1232876712328768,
      "grad_norm": 2.2640976905822754,
      "kl": 0.18017578125,
      "learning_rate": 8.876712328767123e-07,
      "loss": 0.0072,
      "reward": 1.5431283712387085,
      "reward_std": 0.19479237496852875,
      "rewards/accuracy_reward": 0.5431284010410309,
      "rewards/format_reward": 1.0,
      "step": 246
    },
    {
      "completion_length": 70.78125,
      "epoch": 1.1278538812785388,
      "grad_norm": 6.029892444610596,
      "kl": 0.1064453125,
      "learning_rate": 8.872146118721461e-07,
      "loss": 0.0042,
      "reward": 1.6893415451049805,
      "reward_std": 0.1636102795600891,
      "rewards/accuracy_reward": 0.6971540153026581,
      "rewards/format_reward": 0.9921875,
      "step": 247
    },
    {
      "completion_length": 67.9296875,
      "epoch": 1.1324200913242009,
      "grad_norm": 9.093193054199219,
      "kl": 0.1455078125,
      "learning_rate": 8.867579908675798e-07,
      "loss": 0.0058,
      "reward": 1.6471974849700928,
      "reward_std": 0.1802346110343933,
      "rewards/accuracy_reward": 0.647197425365448,
      "rewards/format_reward": 1.0,
      "step": 248
    },
    {
      "completion_length": 86.5234375,
      "epoch": 1.1369863013698631,
      "grad_norm": 3.213435411453247,
      "kl": 0.114501953125,
      "learning_rate": 8.863013698630136e-07,
      "loss": 0.0046,
      "reward": 1.550682783126831,
      "reward_std": 0.20900916308164597,
      "rewards/accuracy_reward": 0.5663077533245087,
      "rewards/format_reward": 0.984375,
      "step": 249
    },
    {
      "completion_length": 79.015625,
      "epoch": 1.1415525114155252,
      "grad_norm": 3.8062937259674072,
      "kl": 0.124267578125,
      "learning_rate": 8.858447488584474e-07,
      "loss": 0.005,
      "reward": 1.5841034650802612,
      "reward_std": 0.17966121435165405,
      "rewards/accuracy_reward": 0.5997284352779388,
      "rewards/format_reward": 0.984375,
      "step": 250
    },
    {
      "completion_length": 67.578125,
      "epoch": 1.1461187214611872,
      "grad_norm": 2.4963302612304688,
      "kl": 0.147705078125,
      "learning_rate": 8.853881278538813e-07,
      "loss": 0.0059,
      "reward": 1.712117850780487,
      "reward_std": 0.15082692354917526,
      "rewards/accuracy_reward": 0.7199303209781647,
      "rewards/format_reward": 0.9921875,
      "step": 251
    },
    {
      "completion_length": 66.7109375,
      "epoch": 1.1506849315068493,
      "grad_norm": 4.159963607788086,
      "kl": 0.132080078125,
      "learning_rate": 8.849315068493151e-07,
      "loss": 0.0053,
      "reward": 1.4834584593772888,
      "reward_std": 0.2506244257092476,
      "rewards/accuracy_reward": 0.4834584891796112,
      "rewards/format_reward": 1.0,
      "step": 252
    },
    {
      "completion_length": 65.9140625,
      "epoch": 1.1552511415525113,
      "grad_norm": 3.3972127437591553,
      "kl": 0.11669921875,
      "learning_rate": 8.844748858447488e-07,
      "loss": 0.0047,
      "reward": 1.7484084367752075,
      "reward_std": 0.17951766774058342,
      "rewards/accuracy_reward": 0.748408317565918,
      "rewards/format_reward": 1.0,
      "step": 253
    },
    {
      "completion_length": 70.4921875,
      "epoch": 1.1598173515981736,
      "grad_norm": 5.158849239349365,
      "kl": 0.105712890625,
      "learning_rate": 8.840182648401826e-07,
      "loss": 0.0042,
      "reward": 1.6956559419631958,
      "reward_std": 0.19831054285168648,
      "rewards/accuracy_reward": 0.7034684121608734,
      "rewards/format_reward": 0.9921875,
      "step": 254
    },
    {
      "completion_length": 65.328125,
      "epoch": 1.1643835616438356,
      "grad_norm": 4.302172660827637,
      "kl": 0.1591796875,
      "learning_rate": 8.835616438356164e-07,
      "loss": 0.0064,
      "reward": 1.5410139560699463,
      "reward_std": 0.24922513961791992,
      "rewards/accuracy_reward": 0.5488264262676239,
      "rewards/format_reward": 0.9921875,
      "step": 255
    },
    {
      "completion_length": 90.1015625,
      "epoch": 1.1689497716894977,
      "grad_norm": 4.157711029052734,
      "kl": 0.07763671875,
      "learning_rate": 8.831050228310501e-07,
      "loss": 0.0031,
      "reward": 1.8432291746139526,
      "reward_std": 0.1643235646188259,
      "rewards/accuracy_reward": 0.8588541150093079,
      "rewards/format_reward": 0.984375,
      "step": 256
    },
    {
      "completion_length": 76.1171875,
      "epoch": 1.17351598173516,
      "grad_norm": 8.10622787475586,
      "kl": 0.111328125,
      "learning_rate": 8.826484018264839e-07,
      "loss": 0.0045,
      "reward": 1.6758702397346497,
      "reward_std": 0.2606264054775238,
      "rewards/accuracy_reward": 0.699307769536972,
      "rewards/format_reward": 0.9765625,
      "step": 257
    },
    {
      "completion_length": 79.9609375,
      "epoch": 1.178082191780822,
      "grad_norm": 9.346076011657715,
      "kl": 0.105224609375,
      "learning_rate": 8.821917808219178e-07,
      "loss": 0.0042,
      "reward": 1.7317607402801514,
      "reward_std": 0.1373404860496521,
      "rewards/accuracy_reward": 0.7395730912685394,
      "rewards/format_reward": 0.9921875,
      "step": 258
    },
    {
      "completion_length": 52.734375,
      "epoch": 1.182648401826484,
      "grad_norm": 2.3975396156311035,
      "kl": 0.13427734375,
      "learning_rate": 8.817351598173516e-07,
      "loss": 0.0054,
      "reward": 1.545721709728241,
      "reward_std": 0.26958315074443817,
      "rewards/accuracy_reward": 0.561346709728241,
      "rewards/format_reward": 0.984375,
      "step": 259
    },
    {
      "completion_length": 59.125,
      "epoch": 1.187214611872146,
      "grad_norm": 4.990502834320068,
      "kl": 0.130126953125,
      "learning_rate": 8.812785388127854e-07,
      "loss": 0.0052,
      "reward": 1.5396197438240051,
      "reward_std": 0.338210791349411,
      "rewards/accuracy_reward": 0.5552447736263275,
      "rewards/format_reward": 0.984375,
      "step": 260
    },
    {
      "completion_length": 92.8984375,
      "epoch": 1.191780821917808,
      "grad_norm": 1.891558051109314,
      "kl": 0.083251953125,
      "learning_rate": 8.808219178082191e-07,
      "loss": 0.0033,
      "reward": 1.6179687976837158,
      "reward_std": 0.19345303252339363,
      "rewards/accuracy_reward": 0.6414062082767487,
      "rewards/format_reward": 0.9765625,
      "step": 261
    },
    {
      "completion_length": 37.671875,
      "epoch": 1.1963470319634704,
      "grad_norm": 3.5058436393737793,
      "kl": 0.12646484375,
      "learning_rate": 8.803652968036529e-07,
      "loss": 0.0051,
      "reward": 1.3005208373069763,
      "reward_std": 0.35989323258399963,
      "rewards/accuracy_reward": 0.3317708224058151,
      "rewards/format_reward": 0.96875,
      "step": 262
    },
    {
      "completion_length": 67.3203125,
      "epoch": 1.2009132420091324,
      "grad_norm": 3.314962387084961,
      "kl": 0.121826171875,
      "learning_rate": 8.799086757990867e-07,
      "loss": 0.0049,
      "reward": 1.6558881998062134,
      "reward_std": 0.2189551442861557,
      "rewards/accuracy_reward": 0.655888170003891,
      "rewards/format_reward": 1.0,
      "step": 263
    },
    {
      "completion_length": 69.8203125,
      "epoch": 1.2054794520547945,
      "grad_norm": 3.078131675720215,
      "kl": 0.14013671875,
      "learning_rate": 8.794520547945205e-07,
      "loss": 0.0056,
      "reward": 1.6604167222976685,
      "reward_std": 0.2180890440940857,
      "rewards/accuracy_reward": 0.6760416328907013,
      "rewards/format_reward": 0.984375,
      "step": 264
    },
    {
      "completion_length": 79.875,
      "epoch": 1.2100456621004567,
      "grad_norm": 2.9452457427978516,
      "kl": 0.086669921875,
      "learning_rate": 8.789954337899543e-07,
      "loss": 0.0035,
      "reward": 1.616637647151947,
      "reward_std": 0.1805976927280426,
      "rewards/accuracy_reward": 0.6244500577449799,
      "rewards/format_reward": 0.9921875,
      "step": 265
    },
    {
      "completion_length": 68.3984375,
      "epoch": 1.2146118721461188,
      "grad_norm": 3.070889711380005,
      "kl": 0.1455078125,
      "learning_rate": 8.785388127853881e-07,
      "loss": 0.0058,
      "reward": 1.6356534361839294,
      "reward_std": 0.20898383855819702,
      "rewards/accuracy_reward": 0.6356533765792847,
      "rewards/format_reward": 1.0,
      "step": 266
    },
    {
      "completion_length": 84.015625,
      "epoch": 1.2191780821917808,
      "grad_norm": 4.916512489318848,
      "kl": 0.090087890625,
      "learning_rate": 8.780821917808219e-07,
      "loss": 0.0036,
      "reward": 1.6764622926712036,
      "reward_std": 0.22947455942630768,
      "rewards/accuracy_reward": 0.6998997330665588,
      "rewards/format_reward": 0.9765625,
      "step": 267
    },
    {
      "completion_length": 56.9296875,
      "epoch": 1.2237442922374429,
      "grad_norm": 3.320021629333496,
      "kl": 0.1376953125,
      "learning_rate": 8.776255707762557e-07,
      "loss": 0.0055,
      "reward": 1.5441592335700989,
      "reward_std": 0.2436549812555313,
      "rewards/accuracy_reward": 0.5519717484712601,
      "rewards/format_reward": 0.9921875,
      "step": 268
    },
    {
      "completion_length": 55.03125,
      "epoch": 1.228310502283105,
      "grad_norm": 2.28918194770813,
      "kl": 0.099609375,
      "learning_rate": 8.771689497716894e-07,
      "loss": 0.004,
      "reward": 1.567187488079071,
      "reward_std": 0.24649156630039215,
      "rewards/accuracy_reward": 0.582812488079071,
      "rewards/format_reward": 0.984375,
      "step": 269
    },
    {
      "completion_length": 65.3203125,
      "epoch": 1.2328767123287672,
      "grad_norm": 4.561470985412598,
      "kl": 0.09423828125,
      "learning_rate": 8.767123287671232e-07,
      "loss": 0.0038,
      "reward": 1.5898438096046448,
      "reward_std": 0.22158773988485336,
      "rewards/accuracy_reward": 0.59765625,
      "rewards/format_reward": 0.9921875,
      "step": 270
    },
    {
      "completion_length": 71.8984375,
      "epoch": 1.2374429223744292,
      "grad_norm": 2.199537992477417,
      "kl": 0.130126953125,
      "learning_rate": 8.762557077625571e-07,
      "loss": 0.0052,
      "reward": 1.6058040857315063,
      "reward_std": 0.20108795166015625,
      "rewards/accuracy_reward": 0.6058041155338287,
      "rewards/format_reward": 1.0,
      "step": 271
    },
    {
      "completion_length": 76.6015625,
      "epoch": 1.2420091324200913,
      "grad_norm": 1.9431648254394531,
      "kl": 0.116455078125,
      "learning_rate": 8.757990867579908e-07,
      "loss": 0.0047,
      "reward": 1.7972594499588013,
      "reward_std": 0.12338224425911903,
      "rewards/accuracy_reward": 0.7972594499588013,
      "rewards/format_reward": 1.0,
      "step": 272
    },
    {
      "completion_length": 69.6015625,
      "epoch": 1.2465753424657535,
      "grad_norm": 1.7630584239959717,
      "kl": 0.103515625,
      "learning_rate": 8.753424657534246e-07,
      "loss": 0.0041,
      "reward": 1.6484375,
      "reward_std": 0.1649293452501297,
      "rewards/accuracy_reward": 0.65625,
      "rewards/format_reward": 0.9921875,
      "step": 273
    },
    {
      "completion_length": 98.6640625,
      "epoch": 1.2511415525114156,
      "grad_norm": 3.8383920192718506,
      "kl": 0.103271484375,
      "learning_rate": 8.748858447488584e-07,
      "loss": 0.0041,
      "reward": 1.765743374824524,
      "reward_std": 0.13689537346363068,
      "rewards/accuracy_reward": 0.7813683450222015,
      "rewards/format_reward": 0.984375,
      "step": 274
    },
    {
      "completion_length": 69.1328125,
      "epoch": 1.2557077625570776,
      "grad_norm": 3.83646297454834,
      "kl": 0.107666015625,
      "learning_rate": 8.744292237442922e-07,
      "loss": 0.0043,
      "reward": 1.5489718914031982,
      "reward_std": 0.2690936550498009,
      "rewards/accuracy_reward": 0.5645968914031982,
      "rewards/format_reward": 0.984375,
      "step": 275
    },
    {
      "completion_length": 74.171875,
      "epoch": 1.2602739726027397,
      "grad_norm": 1.6941070556640625,
      "kl": 0.103759765625,
      "learning_rate": 8.73972602739726e-07,
      "loss": 0.0042,
      "reward": 1.6818824410438538,
      "reward_std": 0.10927051305770874,
      "rewards/accuracy_reward": 0.6818824410438538,
      "rewards/format_reward": 1.0,
      "step": 276
    },
    {
      "completion_length": 68.9375,
      "epoch": 1.2648401826484017,
      "grad_norm": 5.96689510345459,
      "kl": 0.126220703125,
      "learning_rate": 8.735159817351597e-07,
      "loss": 0.0051,
      "reward": 1.4658617973327637,
      "reward_std": 0.19346562027931213,
      "rewards/accuracy_reward": 0.4736742377281189,
      "rewards/format_reward": 0.9921875,
      "step": 277
    },
    {
      "completion_length": 73.265625,
      "epoch": 1.269406392694064,
      "grad_norm": 2.1737794876098633,
      "kl": 0.124267578125,
      "learning_rate": 8.730593607305936e-07,
      "loss": 0.005,
      "reward": 1.6984375715255737,
      "reward_std": 0.13310657069087029,
      "rewards/accuracy_reward": 0.7062499523162842,
      "rewards/format_reward": 0.9921875,
      "step": 278
    },
    {
      "completion_length": 65.4609375,
      "epoch": 1.273972602739726,
      "grad_norm": 2.1923165321350098,
      "kl": 0.1279296875,
      "learning_rate": 8.726027397260274e-07,
      "loss": 0.0051,
      "reward": 1.7382813096046448,
      "reward_std": 0.1450316607952118,
      "rewards/accuracy_reward": 0.7460937201976776,
      "rewards/format_reward": 0.9921875,
      "step": 279
    },
    {
      "completion_length": 69.6328125,
      "epoch": 1.278538812785388,
      "grad_norm": 2.4596610069274902,
      "kl": 0.11279296875,
      "learning_rate": 8.721461187214611e-07,
      "loss": 0.0045,
      "reward": 1.7476563453674316,
      "reward_std": 0.15961872786283493,
      "rewards/accuracy_reward": 0.7476562261581421,
      "rewards/format_reward": 1.0,
      "step": 280
    },
    {
      "completion_length": 74.1640625,
      "epoch": 1.2831050228310503,
      "grad_norm": 1.8160244226455688,
      "kl": 0.08837890625,
      "learning_rate": 8.716894977168949e-07,
      "loss": 0.0035,
      "reward": 1.625745713710785,
      "reward_std": 0.17970408126711845,
      "rewards/accuracy_reward": 0.6257456988096237,
      "rewards/format_reward": 1.0,
      "step": 281
    },
    {
      "completion_length": 56.1875,
      "epoch": 1.2876712328767124,
      "grad_norm": 2.5006446838378906,
      "kl": 0.14306640625,
      "learning_rate": 8.712328767123287e-07,
      "loss": 0.0057,
      "reward": 1.571769893169403,
      "reward_std": 0.2276684269309044,
      "rewards/accuracy_reward": 0.5717698335647583,
      "rewards/format_reward": 1.0,
      "step": 282
    },
    {
      "completion_length": 55.5234375,
      "epoch": 1.2922374429223744,
      "grad_norm": 6.536046504974365,
      "kl": 0.1826171875,
      "learning_rate": 8.707762557077625e-07,
      "loss": 0.0073,
      "reward": 1.6325520277023315,
      "reward_std": 0.21870127320289612,
      "rewards/accuracy_reward": 0.6325520575046539,
      "rewards/format_reward": 1.0,
      "step": 283
    },
    {
      "completion_length": 93.65625,
      "epoch": 1.2968036529680365,
      "grad_norm": 10.913942337036133,
      "kl": 0.0833740234375,
      "learning_rate": 8.703196347031964e-07,
      "loss": 0.0033,
      "reward": 1.6755207777023315,
      "reward_std": 0.17495759949088097,
      "rewards/accuracy_reward": 0.6833332777023315,
      "rewards/format_reward": 0.9921875,
      "step": 284
    },
    {
      "completion_length": 67.125,
      "epoch": 1.3013698630136985,
      "grad_norm": 2.0972490310668945,
      "kl": 0.146484375,
      "learning_rate": 8.698630136986301e-07,
      "loss": 0.0059,
      "reward": 1.4372395873069763,
      "reward_std": 0.24129608273506165,
      "rewards/accuracy_reward": 0.4372395873069763,
      "rewards/format_reward": 1.0,
      "step": 285
    },
    {
      "completion_length": 83.25,
      "epoch": 1.3059360730593608,
      "grad_norm": 2.7569093704223633,
      "kl": 0.1005859375,
      "learning_rate": 8.694063926940639e-07,
      "loss": 0.004,
      "reward": 1.6421875953674316,
      "reward_std": 0.2471975013613701,
      "rewards/accuracy_reward": 0.6734374463558197,
      "rewards/format_reward": 0.96875,
      "step": 286
    },
    {
      "completion_length": 60.7578125,
      "epoch": 1.3105022831050228,
      "grad_norm": 4.288987636566162,
      "kl": 0.10888671875,
      "learning_rate": 8.689497716894977e-07,
      "loss": 0.0044,
      "reward": 1.562853455543518,
      "reward_std": 0.25185681879520416,
      "rewards/accuracy_reward": 0.5706659406423569,
      "rewards/format_reward": 0.9921875,
      "step": 287
    },
    {
      "completion_length": 72.6484375,
      "epoch": 1.3150684931506849,
      "grad_norm": 5.571771621704102,
      "kl": 0.128662109375,
      "learning_rate": 8.684931506849314e-07,
      "loss": 0.0052,
      "reward": 1.57805997133255,
      "reward_std": 0.23417949676513672,
      "rewards/accuracy_reward": 0.5858723521232605,
      "rewards/format_reward": 0.9921875,
      "step": 288
    },
    {
      "completion_length": 84.0390625,
      "epoch": 1.3196347031963471,
      "grad_norm": 1.9545791149139404,
      "kl": 0.085693359375,
      "learning_rate": 8.680365296803652e-07,
      "loss": 0.0034,
      "reward": 1.768750011920929,
      "reward_std": 0.1173202283680439,
      "rewards/accuracy_reward": 0.7765624225139618,
      "rewards/format_reward": 0.9921875,
      "step": 289
    },
    {
      "completion_length": 79.34375,
      "epoch": 1.3242009132420092,
      "grad_norm": 2.6059577465057373,
      "kl": 0.100830078125,
      "learning_rate": 8.67579908675799e-07,
      "loss": 0.004,
      "reward": 1.5458333492279053,
      "reward_std": 0.31738781929016113,
      "rewards/accuracy_reward": 0.5848958194255829,
      "rewards/format_reward": 0.9609375,
      "step": 290
    },
    {
      "completion_length": 77.8828125,
      "epoch": 1.3287671232876712,
      "grad_norm": 2.8494393825531006,
      "kl": 0.126953125,
      "learning_rate": 8.671232876712329e-07,
      "loss": 0.0051,
      "reward": 1.6368772983551025,
      "reward_std": 0.22899606823921204,
      "rewards/accuracy_reward": 0.6603147983551025,
      "rewards/format_reward": 0.9765625,
      "step": 291
    },
    {
      "completion_length": 91.6484375,
      "epoch": 1.3333333333333333,
      "grad_norm": 2.535183906555176,
      "kl": 0.08740234375,
      "learning_rate": 8.666666666666667e-07,
      "loss": 0.0035,
      "reward": 1.6453125476837158,
      "reward_std": 0.15308689326047897,
      "rewards/accuracy_reward": 0.6687499582767487,
      "rewards/format_reward": 0.9765625,
      "step": 292
    },
    {
      "completion_length": 65.921875,
      "epoch": 1.3378995433789953,
      "grad_norm": 3.120173692703247,
      "kl": 0.135009765625,
      "learning_rate": 8.662100456621004e-07,
      "loss": 0.0054,
      "reward": 1.541332721710205,
      "reward_std": 0.2652948349714279,
      "rewards/accuracy_reward": 0.5647702217102051,
      "rewards/format_reward": 0.9765625,
      "step": 293
    },
    {
      "completion_length": 64.9609375,
      "epoch": 1.3424657534246576,
      "grad_norm": 3.71287202835083,
      "kl": 0.096923828125,
      "learning_rate": 8.657534246575342e-07,
      "loss": 0.0039,
      "reward": 1.375390648841858,
      "reward_std": 0.3063650578260422,
      "rewards/accuracy_reward": 0.3988281339406967,
      "rewards/format_reward": 0.9765625,
      "step": 294
    },
    {
      "completion_length": 86.9296875,
      "epoch": 1.3470319634703196,
      "grad_norm": 4.879149913787842,
      "kl": 0.112548828125,
      "learning_rate": 8.65296803652968e-07,
      "loss": 0.0045,
      "reward": 1.7951704263687134,
      "reward_std": 0.1273602545261383,
      "rewards/accuracy_reward": 0.795170396566391,
      "rewards/format_reward": 1.0,
      "step": 295
    },
    {
      "completion_length": 84.3203125,
      "epoch": 1.3515981735159817,
      "grad_norm": 2.7764885425567627,
      "kl": 0.102783203125,
      "learning_rate": 8.648401826484017e-07,
      "loss": 0.0041,
      "reward": 1.7486329078674316,
      "reward_std": 0.2014031484723091,
      "rewards/accuracy_reward": 0.7564452290534973,
      "rewards/format_reward": 0.9921875,
      "step": 296
    },
    {
      "completion_length": 87.6875,
      "epoch": 1.356164383561644,
      "grad_norm": 3.286853551864624,
      "kl": 0.1103515625,
      "learning_rate": 8.643835616438355e-07,
      "loss": 0.0044,
      "reward": 1.5868489742279053,
      "reward_std": 0.21391719579696655,
      "rewards/accuracy_reward": 0.6024739146232605,
      "rewards/format_reward": 0.984375,
      "step": 297
    },
    {
      "completion_length": 72.0,
      "epoch": 1.360730593607306,
      "grad_norm": 3.1161341667175293,
      "kl": 0.130615234375,
      "learning_rate": 8.639269406392694e-07,
      "loss": 0.0052,
      "reward": 1.551562488079071,
      "reward_std": 0.2892322689294815,
      "rewards/accuracy_reward": 0.5750000178813934,
      "rewards/format_reward": 0.9765625,
      "step": 298
    },
    {
      "completion_length": 95.8828125,
      "epoch": 1.365296803652968,
      "grad_norm": 2.0366954803466797,
      "kl": 0.091796875,
      "learning_rate": 8.634703196347032e-07,
      "loss": 0.0037,
      "reward": 1.8364962935447693,
      "reward_std": 0.09190401062369347,
      "rewards/accuracy_reward": 0.8443087041378021,
      "rewards/format_reward": 0.9921875,
      "step": 299
    },
    {
      "completion_length": 101.921875,
      "epoch": 1.36986301369863,
      "grad_norm": 9.95484447479248,
      "kl": 0.080322265625,
      "learning_rate": 8.63013698630137e-07,
      "loss": 0.0032,
      "reward": 1.629079818725586,
      "reward_std": 0.13987341336905956,
      "rewards/accuracy_reward": 0.6290798187255859,
      "rewards/format_reward": 1.0,
      "step": 300
    },
    {
      "completion_length": 79.34375,
      "epoch": 1.374429223744292,
      "grad_norm": 19.206026077270508,
      "kl": 0.117919921875,
      "learning_rate": 8.625570776255707e-07,
      "loss": 0.0047,
      "reward": 1.6438615918159485,
      "reward_std": 0.21473538875579834,
      "rewards/accuracy_reward": 0.6516740918159485,
      "rewards/format_reward": 0.9921875,
      "step": 301
    },
    {
      "completion_length": 86.4609375,
      "epoch": 1.3789954337899544,
      "grad_norm": 9.368078231811523,
      "kl": 0.09228515625,
      "learning_rate": 8.621004566210045e-07,
      "loss": 0.0037,
      "reward": 1.6070312857627869,
      "reward_std": 0.1797475889325142,
      "rewards/accuracy_reward": 0.6226562261581421,
      "rewards/format_reward": 0.984375,
      "step": 302
    },
    {
      "completion_length": 79.0625,
      "epoch": 1.3835616438356164,
      "grad_norm": 2.789032220840454,
      "kl": 0.101806640625,
      "learning_rate": 8.616438356164383e-07,
      "loss": 0.0041,
      "reward": 1.62343031167984,
      "reward_std": 0.2234785482287407,
      "rewards/accuracy_reward": 0.6234302222728729,
      "rewards/format_reward": 1.0,
      "step": 303
    },
    {
      "completion_length": 68.3984375,
      "epoch": 1.3881278538812785,
      "grad_norm": 2.6755340099334717,
      "kl": 0.10009765625,
      "learning_rate": 8.611872146118721e-07,
      "loss": 0.004,
      "reward": 1.5965625047683716,
      "reward_std": 0.1608435958623886,
      "rewards/accuracy_reward": 0.6121874451637268,
      "rewards/format_reward": 0.984375,
      "step": 304
    },
    {
      "completion_length": 79.3203125,
      "epoch": 1.3926940639269407,
      "grad_norm": 3.4425978660583496,
      "kl": 0.1259765625,
      "learning_rate": 8.607305936073059e-07,
      "loss": 0.005,
      "reward": 1.7191716432571411,
      "reward_std": 0.11866222321987152,
      "rewards/accuracy_reward": 0.7269841134548187,
      "rewards/format_reward": 0.9921875,
      "step": 305
    },
    {
      "completion_length": 105.5546875,
      "epoch": 1.3972602739726028,
      "grad_norm": 4.509172439575195,
      "kl": 0.063720703125,
      "learning_rate": 8.602739726027397e-07,
      "loss": 0.0025,
      "reward": 1.7390625476837158,
      "reward_std": 0.14966704696416855,
      "rewards/accuracy_reward": 0.739062488079071,
      "rewards/format_reward": 1.0,
      "step": 306
    },
    {
      "completion_length": 87.1953125,
      "epoch": 1.4018264840182648,
      "grad_norm": 2.9753623008728027,
      "kl": 0.13232421875,
      "learning_rate": 8.598173515981735e-07,
      "loss": 0.0053,
      "reward": 1.6669872403144836,
      "reward_std": 0.14199939370155334,
      "rewards/accuracy_reward": 0.6669871509075165,
      "rewards/format_reward": 1.0,
      "step": 307
    },
    {
      "completion_length": 68.671875,
      "epoch": 1.4063926940639269,
      "grad_norm": 3.5725512504577637,
      "kl": 0.105712890625,
      "learning_rate": 8.593607305936073e-07,
      "loss": 0.0042,
      "reward": 1.5044271349906921,
      "reward_std": 0.2842589318752289,
      "rewards/accuracy_reward": 0.5044270902872086,
      "rewards/format_reward": 1.0,
      "step": 308
    },
    {
      "completion_length": 84.8671875,
      "epoch": 1.410958904109589,
      "grad_norm": 2.2308547496795654,
      "kl": 0.091796875,
      "learning_rate": 8.58904109589041e-07,
      "loss": 0.0037,
      "reward": 1.5451836585998535,
      "reward_std": 0.23898552358150482,
      "rewards/accuracy_reward": 0.5608087480068207,
      "rewards/format_reward": 0.984375,
      "step": 309
    },
    {
      "completion_length": 74.3984375,
      "epoch": 1.4155251141552512,
      "grad_norm": 5.253906726837158,
      "kl": 0.139892578125,
      "learning_rate": 8.584474885844748e-07,
      "loss": 0.0056,
      "reward": 1.5776489973068237,
      "reward_std": 0.20516303181648254,
      "rewards/accuracy_reward": 0.5854615569114685,
      "rewards/format_reward": 0.9921875,
      "step": 310
    },
    {
      "completion_length": 75.515625,
      "epoch": 1.4200913242009132,
      "grad_norm": 3.5134174823760986,
      "kl": 0.0849609375,
      "learning_rate": 8.579908675799087e-07,
      "loss": 0.0034,
      "reward": 1.4553078413009644,
      "reward_std": 0.26812630146741867,
      "rewards/accuracy_reward": 0.47093285620212555,
      "rewards/format_reward": 0.984375,
      "step": 311
    },
    {
      "completion_length": 100.5,
      "epoch": 1.4246575342465753,
      "grad_norm": 2.315570116043091,
      "kl": 0.090087890625,
      "learning_rate": 8.575342465753424e-07,
      "loss": 0.0036,
      "reward": 1.597842276096344,
      "reward_std": 0.192316435277462,
      "rewards/accuracy_reward": 0.5978422611951828,
      "rewards/format_reward": 1.0,
      "step": 312
    },
    {
      "completion_length": 85.46875,
      "epoch": 1.4292237442922375,
      "grad_norm": 4.950066566467285,
      "kl": 0.09375,
      "learning_rate": 8.570776255707762e-07,
      "loss": 0.0037,
      "reward": 1.6005207896232605,
      "reward_std": 0.23914001137018204,
      "rewards/accuracy_reward": 0.6083333194255829,
      "rewards/format_reward": 0.9921875,
      "step": 313
    },
    {
      "completion_length": 73.1328125,
      "epoch": 1.4337899543378996,
      "grad_norm": 3.737804651260376,
      "kl": 0.13037109375,
      "learning_rate": 8.5662100456621e-07,
      "loss": 0.0052,
      "reward": 1.542373538017273,
      "reward_std": 0.24436407536268234,
      "rewards/accuracy_reward": 0.542373538017273,
      "rewards/format_reward": 1.0,
      "step": 314
    },
    {
      "completion_length": 81.2265625,
      "epoch": 1.4383561643835616,
      "grad_norm": 2.7811484336853027,
      "kl": 0.10693359375,
      "learning_rate": 8.561643835616438e-07,
      "loss": 0.0043,
      "reward": 1.7172211408615112,
      "reward_std": 0.15319041907787323,
      "rewards/accuracy_reward": 0.7172211408615112,
      "rewards/format_reward": 1.0,
      "step": 315
    },
    {
      "completion_length": 82.328125,
      "epoch": 1.4429223744292237,
      "grad_norm": 5.399558067321777,
      "kl": 0.10986328125,
      "learning_rate": 8.557077625570776e-07,
      "loss": 0.0044,
      "reward": 1.5121857523918152,
      "reward_std": 0.28201115131378174,
      "rewards/accuracy_reward": 0.5199982225894928,
      "rewards/format_reward": 0.9921875,
      "step": 316
    },
    {
      "completion_length": 103.671875,
      "epoch": 1.4474885844748857,
      "grad_norm": 1.9453126192092896,
      "kl": 0.070068359375,
      "learning_rate": 8.552511415525113e-07,
      "loss": 0.0028,
      "reward": 1.7234273552894592,
      "reward_std": 0.12736555561423302,
      "rewards/accuracy_reward": 0.7234272956848145,
      "rewards/format_reward": 1.0,
      "step": 317
    },
    {
      "completion_length": 91.109375,
      "epoch": 1.452054794520548,
      "grad_norm": 4.417510032653809,
      "kl": 0.09716796875,
      "learning_rate": 8.547945205479452e-07,
      "loss": 0.0039,
      "reward": 1.6735481023788452,
      "reward_std": 0.17602262273430824,
      "rewards/accuracy_reward": 0.68136066198349,
      "rewards/format_reward": 0.9921875,
      "step": 318
    },
    {
      "completion_length": 86.109375,
      "epoch": 1.45662100456621,
      "grad_norm": 3.274066925048828,
      "kl": 0.105712890625,
      "learning_rate": 8.54337899543379e-07,
      "loss": 0.0042,
      "reward": 1.6061198115348816,
      "reward_std": 0.2346249595284462,
      "rewards/accuracy_reward": 0.6061197817325592,
      "rewards/format_reward": 1.0,
      "step": 319
    },
    {
      "completion_length": 103.8828125,
      "epoch": 1.461187214611872,
      "grad_norm": 3.0639874935150146,
      "kl": 0.09375,
      "learning_rate": 8.538812785388127e-07,
      "loss": 0.0037,
      "reward": 1.5671589970588684,
      "reward_std": 0.2653961777687073,
      "rewards/accuracy_reward": 0.5984089076519012,
      "rewards/format_reward": 0.96875,
      "step": 320
    },
    {
      "completion_length": 91.578125,
      "epoch": 1.4657534246575343,
      "grad_norm": 3.1881332397460938,
      "kl": 0.11669921875,
      "learning_rate": 8.534246575342465e-07,
      "loss": 0.0047,
      "reward": 1.7369791269302368,
      "reward_std": 0.24841733276844025,
      "rewards/accuracy_reward": 0.7369791567325592,
      "rewards/format_reward": 1.0,
      "step": 321
    },
    {
      "completion_length": 90.1328125,
      "epoch": 1.4703196347031964,
      "grad_norm": 7.73578405380249,
      "kl": 0.12841796875,
      "learning_rate": 8.529680365296803e-07,
      "loss": 0.0051,
      "reward": 1.5886787176132202,
      "reward_std": 0.27549922466278076,
      "rewards/accuracy_reward": 0.6043036431074142,
      "rewards/format_reward": 0.984375,
      "step": 322
    },
    {
      "completion_length": 105.2578125,
      "epoch": 1.4748858447488584,
      "grad_norm": 2.833080768585205,
      "kl": 0.063720703125,
      "learning_rate": 8.52511415525114e-07,
      "loss": 0.0026,
      "reward": 1.6982238292694092,
      "reward_std": 0.09853163920342922,
      "rewards/accuracy_reward": 0.7060362696647644,
      "rewards/format_reward": 0.9921875,
      "step": 323
    },
    {
      "completion_length": 101.9375,
      "epoch": 1.4794520547945205,
      "grad_norm": 1.9237995147705078,
      "kl": 0.0869140625,
      "learning_rate": 8.52054794520548e-07,
      "loss": 0.0035,
      "reward": 1.6929687857627869,
      "reward_std": 0.14721976220607758,
      "rewards/accuracy_reward": 0.7085936963558197,
      "rewards/format_reward": 0.984375,
      "step": 324
    },
    {
      "completion_length": 93.171875,
      "epoch": 1.4840182648401825,
      "grad_norm": 21.809825897216797,
      "kl": 0.100830078125,
      "learning_rate": 8.515981735159817e-07,
      "loss": 0.004,
      "reward": 1.464756965637207,
      "reward_std": 0.26955385506153107,
      "rewards/accuracy_reward": 0.503819465637207,
      "rewards/format_reward": 0.9609375,
      "step": 325
    },
    {
      "completion_length": 102.03125,
      "epoch": 1.4885844748858448,
      "grad_norm": 2.1946544647216797,
      "kl": 0.075439453125,
      "learning_rate": 8.511415525114155e-07,
      "loss": 0.003,
      "reward": 1.7023438215255737,
      "reward_std": 0.15592241287231445,
      "rewards/accuracy_reward": 0.7179687023162842,
      "rewards/format_reward": 0.984375,
      "step": 326
    },
    {
      "completion_length": 78.6875,
      "epoch": 1.4931506849315068,
      "grad_norm": 5.698726654052734,
      "kl": 0.11376953125,
      "learning_rate": 8.506849315068493e-07,
      "loss": 0.0046,
      "reward": 1.5953141450881958,
      "reward_std": 0.2819615304470062,
      "rewards/accuracy_reward": 0.618751734495163,
      "rewards/format_reward": 0.9765625,
      "step": 327
    },
    {
      "completion_length": 89.71875,
      "epoch": 1.4977168949771689,
      "grad_norm": 4.1140007972717285,
      "kl": 0.080078125,
      "learning_rate": 8.50228310502283e-07,
      "loss": 0.0032,
      "reward": 1.6470133662223816,
      "reward_std": 0.18788425624370575,
      "rewards/accuracy_reward": 0.6626383662223816,
      "rewards/format_reward": 0.984375,
      "step": 328
    },
    {
      "completion_length": 116.9375,
      "epoch": 1.5022831050228311,
      "grad_norm": 2.189347743988037,
      "kl": 0.0545654296875,
      "learning_rate": 8.497716894977168e-07,
      "loss": 0.0022,
      "reward": 1.71875,
      "reward_std": 0.12179600074887276,
      "rewards/accuracy_reward": 0.7343749701976776,
      "rewards/format_reward": 0.984375,
      "step": 329
    },
    {
      "completion_length": 88.703125,
      "epoch": 1.5068493150684932,
      "grad_norm": 2.298283815383911,
      "kl": 0.105712890625,
      "learning_rate": 8.493150684931506e-07,
      "loss": 0.0042,
      "reward": 1.686813473701477,
      "reward_std": 0.1672440692782402,
      "rewards/accuracy_reward": 0.6868133842945099,
      "rewards/format_reward": 1.0,
      "step": 330
    },
    {
      "completion_length": 69.828125,
      "epoch": 1.5114155251141552,
      "grad_norm": 6.333926200866699,
      "kl": 0.13134765625,
      "learning_rate": 8.488584474885845e-07,
      "loss": 0.0053,
      "reward": 1.7103299498558044,
      "reward_std": 0.1922176629304886,
      "rewards/accuracy_reward": 0.7103298306465149,
      "rewards/format_reward": 1.0,
      "step": 331
    },
    {
      "completion_length": 95.6484375,
      "epoch": 1.5159817351598175,
      "grad_norm": 1.9884992837905884,
      "kl": 0.109375,
      "learning_rate": 8.484018264840183e-07,
      "loss": 0.0044,
      "reward": 1.6320313215255737,
      "reward_std": 0.2171119600534439,
      "rewards/accuracy_reward": 0.6398437321186066,
      "rewards/format_reward": 0.9921875,
      "step": 332
    },
    {
      "completion_length": 80.28125,
      "epoch": 1.5205479452054793,
      "grad_norm": 3.4122371673583984,
      "kl": 0.15234375,
      "learning_rate": 8.47945205479452e-07,
      "loss": 0.0061,
      "reward": 1.6989583373069763,
      "reward_std": 0.22790630161762238,
      "rewards/accuracy_reward": 0.7067708373069763,
      "rewards/format_reward": 0.9921875,
      "step": 333
    },
    {
      "completion_length": 82.734375,
      "epoch": 1.5251141552511416,
      "grad_norm": 16.75699234008789,
      "kl": 0.084228515625,
      "learning_rate": 8.474885844748858e-07,
      "loss": 0.0034,
      "reward": 1.6330461502075195,
      "reward_std": 0.1892491653561592,
      "rewards/accuracy_reward": 0.6408586502075195,
      "rewards/format_reward": 0.9921875,
      "step": 334
    },
    {
      "completion_length": 81.421875,
      "epoch": 1.5296803652968036,
      "grad_norm": 2.46637225151062,
      "kl": 0.095703125,
      "learning_rate": 8.470319634703196e-07,
      "loss": 0.0038,
      "reward": 1.6607667207717896,
      "reward_std": 0.23428593575954437,
      "rewards/accuracy_reward": 0.6763917207717896,
      "rewards/format_reward": 0.984375,
      "step": 335
    },
    {
      "completion_length": 87.4375,
      "epoch": 1.5342465753424657,
      "grad_norm": 4.196310043334961,
      "kl": 0.09423828125,
      "learning_rate": 8.465753424657533e-07,
      "loss": 0.0038,
      "reward": 1.5312500596046448,
      "reward_std": 0.2585080787539482,
      "rewards/accuracy_reward": 0.5390624850988388,
      "rewards/format_reward": 0.9921875,
      "step": 336
    },
    {
      "completion_length": 73.046875,
      "epoch": 1.538812785388128,
      "grad_norm": 2.5180585384368896,
      "kl": 0.13671875,
      "learning_rate": 8.461187214611871e-07,
      "loss": 0.0055,
      "reward": 1.630428671836853,
      "reward_std": 0.20693185180425644,
      "rewards/accuracy_reward": 0.6538661420345306,
      "rewards/format_reward": 0.9765625,
      "step": 337
    },
    {
      "completion_length": 83.09375,
      "epoch": 1.54337899543379,
      "grad_norm": 4.625184059143066,
      "kl": 0.107666015625,
      "learning_rate": 8.45662100456621e-07,
      "loss": 0.0043,
      "reward": 1.7304688096046448,
      "reward_std": 0.16690129786729813,
      "rewards/accuracy_reward": 0.7304687201976776,
      "rewards/format_reward": 1.0,
      "step": 338
    },
    {
      "completion_length": 79.3515625,
      "epoch": 1.547945205479452,
      "grad_norm": 7.074364185333252,
      "kl": 0.2109375,
      "learning_rate": 8.452054794520548e-07,
      "loss": 0.0084,
      "reward": 1.6570312976837158,
      "reward_std": 0.24604861438274384,
      "rewards/accuracy_reward": 0.680468738079071,
      "rewards/format_reward": 0.9765625,
      "step": 339
    },
    {
      "completion_length": 79.8125,
      "epoch": 1.5525114155251143,
      "grad_norm": 3.3716442584991455,
      "kl": 0.086669921875,
      "learning_rate": 8.447488584474886e-07,
      "loss": 0.0035,
      "reward": 1.5007859468460083,
      "reward_std": 0.2506624162197113,
      "rewards/accuracy_reward": 0.5242233872413635,
      "rewards/format_reward": 0.9765625,
      "step": 340
    },
    {
      "completion_length": 75.390625,
      "epoch": 1.5570776255707761,
      "grad_norm": 3.0893497467041016,
      "kl": 0.099853515625,
      "learning_rate": 8.442922374429223e-07,
      "loss": 0.004,
      "reward": 1.546875,
      "reward_std": 0.2782461494207382,
      "rewards/accuracy_reward": 0.578125,
      "rewards/format_reward": 0.96875,
      "step": 341
    },
    {
      "completion_length": 80.75,
      "epoch": 1.5616438356164384,
      "grad_norm": 1.9481297731399536,
      "kl": 0.0885009765625,
      "learning_rate": 8.438356164383561e-07,
      "loss": 0.0035,
      "reward": 1.5662733316421509,
      "reward_std": 0.20162209123373032,
      "rewards/accuracy_reward": 0.5897107124328613,
      "rewards/format_reward": 0.9765625,
      "step": 342
    },
    {
      "completion_length": 72.890625,
      "epoch": 1.5662100456621004,
      "grad_norm": 2.927980899810791,
      "kl": 0.112548828125,
      "learning_rate": 8.433789954337899e-07,
      "loss": 0.0045,
      "reward": 1.4613777995109558,
      "reward_std": 0.34207557141780853,
      "rewards/accuracy_reward": 0.5473152101039886,
      "rewards/format_reward": 0.9140625,
      "step": 343
    },
    {
      "completion_length": 56.953125,
      "epoch": 1.5707762557077625,
      "grad_norm": 3.1426687240600586,
      "kl": 0.155517578125,
      "learning_rate": 8.429223744292237e-07,
      "loss": 0.0062,
      "reward": 1.3815755248069763,
      "reward_std": 0.3664630800485611,
      "rewards/accuracy_reward": 0.44407549500465393,
      "rewards/format_reward": 0.9375,
      "step": 344
    },
    {
      "completion_length": 82.2265625,
      "epoch": 1.5753424657534247,
      "grad_norm": 2.98221492767334,
      "kl": 0.126220703125,
      "learning_rate": 8.424657534246576e-07,
      "loss": 0.005,
      "reward": 1.6400888562202454,
      "reward_std": 0.24476776085793972,
      "rewards/accuracy_reward": 0.6947763860225677,
      "rewards/format_reward": 0.9453125,
      "step": 345
    },
    {
      "completion_length": 67.859375,
      "epoch": 1.5799086757990868,
      "grad_norm": 2.4362642765045166,
      "kl": 0.122314453125,
      "learning_rate": 8.420091324200913e-07,
      "loss": 0.0049,
      "reward": 1.4768601059913635,
      "reward_std": 0.258517824113369,
      "rewards/accuracy_reward": 0.5159225761890411,
      "rewards/format_reward": 0.9609375,
      "step": 346
    },
    {
      "completion_length": 70.0859375,
      "epoch": 1.5844748858447488,
      "grad_norm": 3.0861399173736572,
      "kl": 0.166015625,
      "learning_rate": 8.415525114155251e-07,
      "loss": 0.0066,
      "reward": 1.6798083782196045,
      "reward_std": 0.2793383300304413,
      "rewards/accuracy_reward": 0.7032458782196045,
      "rewards/format_reward": 0.9765625,
      "step": 347
    },
    {
      "completion_length": 50.015625,
      "epoch": 1.589041095890411,
      "grad_norm": 3.495870351791382,
      "kl": 0.162109375,
      "learning_rate": 8.410958904109589e-07,
      "loss": 0.0065,
      "reward": 1.582118034362793,
      "reward_std": 0.21435417234897614,
      "rewards/accuracy_reward": 0.5821180641651154,
      "rewards/format_reward": 1.0,
      "step": 348
    },
    {
      "completion_length": 50.2890625,
      "epoch": 1.593607305936073,
      "grad_norm": 4.53682804107666,
      "kl": 0.146240234375,
      "learning_rate": 8.406392694063926e-07,
      "loss": 0.0059,
      "reward": 1.6024181842803955,
      "reward_std": 0.2329491451382637,
      "rewards/accuracy_reward": 0.6258556544780731,
      "rewards/format_reward": 0.9765625,
      "step": 349
    },
    {
      "completion_length": 54.34375,
      "epoch": 1.5981735159817352,
      "grad_norm": 2.6551053524017334,
      "kl": 0.127685546875,
      "learning_rate": 8.401826484018264e-07,
      "loss": 0.0051,
      "reward": 1.5606771111488342,
      "reward_std": 0.1882794126868248,
      "rewards/accuracy_reward": 0.568489596247673,
      "rewards/format_reward": 0.9921875,
      "step": 350
    },
    {
      "completion_length": 66.4140625,
      "epoch": 1.6027397260273972,
      "grad_norm": 6.037808418273926,
      "kl": 0.146484375,
      "learning_rate": 8.397260273972603e-07,
      "loss": 0.0059,
      "reward": 1.6453644037246704,
      "reward_std": 0.18358591571450233,
      "rewards/accuracy_reward": 0.660989373922348,
      "rewards/format_reward": 0.984375,
      "step": 351
    },
    {
      "completion_length": 49.375,
      "epoch": 1.6073059360730593,
      "grad_norm": 2.8399972915649414,
      "kl": 0.1689453125,
      "learning_rate": 8.39269406392694e-07,
      "loss": 0.0068,
      "reward": 1.6010416746139526,
      "reward_std": 0.2630682438611984,
      "rewards/accuracy_reward": 0.6088541746139526,
      "rewards/format_reward": 0.9921875,
      "step": 352
    },
    {
      "completion_length": 58.4921875,
      "epoch": 1.6118721461187215,
      "grad_norm": 1.744779109954834,
      "kl": 0.1396484375,
      "learning_rate": 8.388127853881279e-07,
      "loss": 0.0056,
      "reward": 1.675000011920929,
      "reward_std": 0.16849348694086075,
      "rewards/accuracy_reward": 0.675000011920929,
      "rewards/format_reward": 1.0,
      "step": 353
    },
    {
      "completion_length": 63.8046875,
      "epoch": 1.6164383561643836,
      "grad_norm": 4.681619167327881,
      "kl": 0.14404296875,
      "learning_rate": 8.383561643835616e-07,
      "loss": 0.0058,
      "reward": 1.7920387387275696,
      "reward_std": 0.14393481612205505,
      "rewards/accuracy_reward": 0.7920385599136353,
      "rewards/format_reward": 1.0,
      "step": 354
    },
    {
      "completion_length": 67.0234375,
      "epoch": 1.6210045662100456,
      "grad_norm": 3.167783498764038,
      "kl": 0.1318359375,
      "learning_rate": 8.378995433789954e-07,
      "loss": 0.0053,
      "reward": 1.728416085243225,
      "reward_std": 0.18405602872371674,
      "rewards/accuracy_reward": 0.7518534958362579,
      "rewards/format_reward": 0.9765625,
      "step": 355
    },
    {
      "completion_length": 63.390625,
      "epoch": 1.625570776255708,
      "grad_norm": 3.923906087875366,
      "kl": 0.13330078125,
      "learning_rate": 8.374429223744292e-07,
      "loss": 0.0053,
      "reward": 1.6279487609863281,
      "reward_std": 0.12273544818162918,
      "rewards/accuracy_reward": 0.6279487460851669,
      "rewards/format_reward": 1.0,
      "step": 356
    },
    {
      "completion_length": 61.1796875,
      "epoch": 1.6301369863013697,
      "grad_norm": 3.410731077194214,
      "kl": 0.13427734375,
      "learning_rate": 8.369863013698629e-07,
      "loss": 0.0054,
      "reward": 1.6447545289993286,
      "reward_std": 0.22194860875606537,
      "rewards/accuracy_reward": 0.6603794991970062,
      "rewards/format_reward": 0.984375,
      "step": 357
    },
    {
      "completion_length": 68.53125,
      "epoch": 1.634703196347032,
      "grad_norm": 4.652464389801025,
      "kl": 0.1181640625,
      "learning_rate": 8.365296803652968e-07,
      "loss": 0.0047,
      "reward": 1.6500000953674316,
      "reward_std": 0.17417245358228683,
      "rewards/accuracy_reward": 0.6499999761581421,
      "rewards/format_reward": 1.0,
      "step": 358
    },
    {
      "completion_length": 55.984375,
      "epoch": 1.639269406392694,
      "grad_norm": 1.8993587493896484,
      "kl": 0.1298828125,
      "learning_rate": 8.360730593607306e-07,
      "loss": 0.0052,
      "reward": 1.7149627804756165,
      "reward_std": 0.11658288538455963,
      "rewards/accuracy_reward": 0.7149626910686493,
      "rewards/format_reward": 1.0,
      "step": 359
    },
    {
      "completion_length": 81.7578125,
      "epoch": 1.643835616438356,
      "grad_norm": 1.6242355108261108,
      "kl": 0.145751953125,
      "learning_rate": 8.356164383561643e-07,
      "loss": 0.0058,
      "reward": 1.7650251388549805,
      "reward_std": 0.07628657668828964,
      "rewards/accuracy_reward": 0.7650250494480133,
      "rewards/format_reward": 1.0,
      "step": 360
    },
    {
      "completion_length": 92.7265625,
      "epoch": 1.6484018264840183,
      "grad_norm": 1.6242085695266724,
      "kl": 0.06298828125,
      "learning_rate": 8.351598173515981e-07,
      "loss": 0.0025,
      "reward": 1.75,
      "reward_std": 0.1409430019557476,
      "rewards/accuracy_reward": 0.7656249701976776,
      "rewards/format_reward": 0.984375,
      "step": 361
    },
    {
      "completion_length": 71.1328125,
      "epoch": 1.6529680365296804,
      "grad_norm": 3.2496068477630615,
      "kl": 0.109130859375,
      "learning_rate": 8.347031963470319e-07,
      "loss": 0.0044,
      "reward": 1.614074468612671,
      "reward_std": 0.19926752150058746,
      "rewards/accuracy_reward": 0.6296994388103485,
      "rewards/format_reward": 0.984375,
      "step": 362
    },
    {
      "completion_length": 68.6953125,
      "epoch": 1.6575342465753424,
      "grad_norm": 1.9562387466430664,
      "kl": 0.100830078125,
      "learning_rate": 8.342465753424657e-07,
      "loss": 0.004,
      "reward": 1.7125000953674316,
      "reward_std": 0.13402405753731728,
      "rewards/accuracy_reward": 0.7124999761581421,
      "rewards/format_reward": 1.0,
      "step": 363
    },
    {
      "completion_length": 65.21875,
      "epoch": 1.6621004566210047,
      "grad_norm": 2.6374332904815674,
      "kl": 0.15673828125,
      "learning_rate": 8.337899543378996e-07,
      "loss": 0.0063,
      "reward": 1.653124988079071,
      "reward_std": 0.1894800141453743,
      "rewards/accuracy_reward": 0.668749988079071,
      "rewards/format_reward": 0.984375,
      "step": 364
    },
    {
      "completion_length": 50.03125,
      "epoch": 1.6666666666666665,
      "grad_norm": 3.6687989234924316,
      "kl": 0.1171875,
      "learning_rate": 8.333333333333333e-07,
      "loss": 0.0047,
      "reward": 1.5525545477867126,
      "reward_std": 0.18541519343852997,
      "rewards/accuracy_reward": 0.5525545328855515,
      "rewards/format_reward": 1.0,
      "step": 365
    },
    {
      "completion_length": 50.265625,
      "epoch": 1.6712328767123288,
      "grad_norm": 27.93817710876465,
      "kl": 0.8662109375,
      "learning_rate": 8.328767123287671e-07,
      "loss": 0.0347,
      "reward": 1.6250372529029846,
      "reward_std": 0.26529867947101593,
      "rewards/accuracy_reward": 0.6250371932983398,
      "rewards/format_reward": 1.0,
      "step": 366
    },
    {
      "completion_length": 55.046875,
      "epoch": 1.6757990867579908,
      "grad_norm": 1.959330439567566,
      "kl": 0.128662109375,
      "learning_rate": 8.324200913242009e-07,
      "loss": 0.0051,
      "reward": 1.611718773841858,
      "reward_std": 0.20705362409353256,
      "rewards/accuracy_reward": 0.6117187142372131,
      "rewards/format_reward": 1.0,
      "step": 367
    },
    {
      "completion_length": 73.1796875,
      "epoch": 1.6803652968036529,
      "grad_norm": 4.830105781555176,
      "kl": 0.11083984375,
      "learning_rate": 8.319634703196346e-07,
      "loss": 0.0044,
      "reward": 1.7028512358665466,
      "reward_std": 0.15199671685695648,
      "rewards/accuracy_reward": 0.7106637060642242,
      "rewards/format_reward": 0.9921875,
      "step": 368
    },
    {
      "completion_length": 44.71875,
      "epoch": 1.6849315068493151,
      "grad_norm": 2.5528032779693604,
      "kl": 0.1474609375,
      "learning_rate": 8.315068493150684e-07,
      "loss": 0.0059,
      "reward": 1.5036830306053162,
      "reward_std": 0.2237851321697235,
      "rewards/accuracy_reward": 0.5036830604076385,
      "rewards/format_reward": 1.0,
      "step": 369
    },
    {
      "completion_length": 64.7578125,
      "epoch": 1.6894977168949772,
      "grad_norm": 3.1371445655822754,
      "kl": 0.10693359375,
      "learning_rate": 8.310502283105022e-07,
      "loss": 0.0043,
      "reward": 1.7922433018684387,
      "reward_std": 0.1861533522605896,
      "rewards/accuracy_reward": 0.8000558018684387,
      "rewards/format_reward": 0.9921875,
      "step": 370
    },
    {
      "completion_length": 81.265625,
      "epoch": 1.6940639269406392,
      "grad_norm": 3.3165316581726074,
      "kl": 0.094482421875,
      "learning_rate": 8.305936073059361e-07,
      "loss": 0.0038,
      "reward": 1.5851722359657288,
      "reward_std": 0.17648599669337273,
      "rewards/accuracy_reward": 0.5851722061634064,
      "rewards/format_reward": 1.0,
      "step": 371
    },
    {
      "completion_length": 84.890625,
      "epoch": 1.6986301369863015,
      "grad_norm": 3.0761749744415283,
      "kl": 0.10693359375,
      "learning_rate": 8.301369863013699e-07,
      "loss": 0.0043,
      "reward": 1.7469556331634521,
      "reward_std": 0.1787155643105507,
      "rewards/accuracy_reward": 0.770393043756485,
      "rewards/format_reward": 0.9765625,
      "step": 372
    },
    {
      "completion_length": 50.25,
      "epoch": 1.7031963470319633,
      "grad_norm": 3.6592459678649902,
      "kl": 0.15380859375,
      "learning_rate": 8.296803652968036e-07,
      "loss": 0.0062,
      "reward": 1.6217397451400757,
      "reward_std": 0.2100653052330017,
      "rewards/accuracy_reward": 0.6217397451400757,
      "rewards/format_reward": 1.0,
      "step": 373
    },
    {
      "completion_length": 69.7734375,
      "epoch": 1.7077625570776256,
      "grad_norm": 4.063467502593994,
      "kl": 0.1123046875,
      "learning_rate": 8.292237442922374e-07,
      "loss": 0.0045,
      "reward": 1.6185640096664429,
      "reward_std": 0.17512290179729462,
      "rewards/accuracy_reward": 0.6185639351606369,
      "rewards/format_reward": 1.0,
      "step": 374
    },
    {
      "completion_length": 44.3125,
      "epoch": 1.7123287671232876,
      "grad_norm": 2.8335328102111816,
      "kl": 0.1572265625,
      "learning_rate": 8.287671232876712e-07,
      "loss": 0.0063,
      "reward": 1.528906226158142,
      "reward_std": 0.238134503364563,
      "rewards/accuracy_reward": 0.5289062559604645,
      "rewards/format_reward": 1.0,
      "step": 375
    },
    {
      "completion_length": 47.7578125,
      "epoch": 1.7168949771689497,
      "grad_norm": 2.0584352016448975,
      "kl": 0.14306640625,
      "learning_rate": 8.283105022831049e-07,
      "loss": 0.0057,
      "reward": 1.5130573511123657,
      "reward_std": 0.23582037538290024,
      "rewards/accuracy_reward": 0.5130573809146881,
      "rewards/format_reward": 1.0,
      "step": 376
    },
    {
      "completion_length": 81.203125,
      "epoch": 1.721461187214612,
      "grad_norm": 2.3844358921051025,
      "kl": 0.11669921875,
      "learning_rate": 8.278538812785387e-07,
      "loss": 0.0047,
      "reward": 1.616406261920929,
      "reward_std": 0.22177018970251083,
      "rewards/accuracy_reward": 0.6320312321186066,
      "rewards/format_reward": 0.984375,
      "step": 377
    },
    {
      "completion_length": 70.109375,
      "epoch": 1.726027397260274,
      "grad_norm": 4.480605602264404,
      "kl": 0.093994140625,
      "learning_rate": 8.273972602739726e-07,
      "loss": 0.0038,
      "reward": 1.7281250357627869,
      "reward_std": 0.13258251547813416,
      "rewards/accuracy_reward": 0.7359374761581421,
      "rewards/format_reward": 0.9921875,
      "step": 378
    },
    {
      "completion_length": 57.6015625,
      "epoch": 1.730593607305936,
      "grad_norm": 3.5465571880340576,
      "kl": 0.12109375,
      "learning_rate": 8.269406392694064e-07,
      "loss": 0.0048,
      "reward": 1.6325520873069763,
      "reward_std": 0.20021257549524307,
      "rewards/accuracy_reward": 0.6325520575046539,
      "rewards/format_reward": 1.0,
      "step": 379
    },
    {
      "completion_length": 85.328125,
      "epoch": 1.7351598173515983,
      "grad_norm": 2.483445882797241,
      "kl": 0.0963134765625,
      "learning_rate": 8.264840182648402e-07,
      "loss": 0.0038,
      "reward": 1.5804979801177979,
      "reward_std": 0.16095227003097534,
      "rewards/accuracy_reward": 0.5804979428648949,
      "rewards/format_reward": 1.0,
      "step": 380
    },
    {
      "completion_length": 64.453125,
      "epoch": 1.7397260273972601,
      "grad_norm": 2.9765961170196533,
      "kl": 0.16650390625,
      "learning_rate": 8.260273972602739e-07,
      "loss": 0.0067,
      "reward": 1.6214489340782166,
      "reward_std": 0.20274285972118378,
      "rewards/accuracy_reward": 0.6214488744735718,
      "rewards/format_reward": 1.0,
      "step": 381
    },
    {
      "completion_length": 76.828125,
      "epoch": 1.7442922374429224,
      "grad_norm": 3.132500410079956,
      "kl": 0.1435546875,
      "learning_rate": 8.255707762557077e-07,
      "loss": 0.0058,
      "reward": 1.5574839115142822,
      "reward_std": 0.22679631412029266,
      "rewards/accuracy_reward": 0.5652963519096375,
      "rewards/format_reward": 0.9921875,
      "step": 382
    },
    {
      "completion_length": 70.46875,
      "epoch": 1.7488584474885844,
      "grad_norm": 2.3076906204223633,
      "kl": 0.114501953125,
      "learning_rate": 8.251141552511415e-07,
      "loss": 0.0046,
      "reward": 1.688330888748169,
      "reward_std": 0.13296211138367653,
      "rewards/accuracy_reward": 0.6883308291435242,
      "rewards/format_reward": 1.0,
      "step": 383
    },
    {
      "completion_length": 73.0078125,
      "epoch": 1.7534246575342465,
      "grad_norm": 4.062695026397705,
      "kl": 0.114501953125,
      "learning_rate": 8.246575342465753e-07,
      "loss": 0.0046,
      "reward": 1.5453124642372131,
      "reward_std": 0.2351284772157669,
      "rewards/accuracy_reward": 0.5609375238418579,
      "rewards/format_reward": 0.984375,
      "step": 384
    },
    {
      "completion_length": 48.984375,
      "epoch": 1.7579908675799087,
      "grad_norm": 6.5755510330200195,
      "kl": 0.20068359375,
      "learning_rate": 8.242009132420092e-07,
      "loss": 0.008,
      "reward": 1.5984994769096375,
      "reward_std": 0.3222559839487076,
      "rewards/accuracy_reward": 0.5984995067119598,
      "rewards/format_reward": 1.0,
      "step": 385
    },
    {
      "completion_length": 71.703125,
      "epoch": 1.7625570776255708,
      "grad_norm": 13.117231369018555,
      "kl": 0.083251953125,
      "learning_rate": 8.237442922374429e-07,
      "loss": 0.0033,
      "reward": 1.5032986402511597,
      "reward_std": 0.21686269342899323,
      "rewards/accuracy_reward": 0.5032986104488373,
      "rewards/format_reward": 1.0,
      "step": 386
    },
    {
      "completion_length": 87.375,
      "epoch": 1.7671232876712328,
      "grad_norm": 2.233595371246338,
      "kl": 0.107177734375,
      "learning_rate": 8.232876712328767e-07,
      "loss": 0.0043,
      "reward": 1.701716125011444,
      "reward_std": 0.11170049756765366,
      "rewards/accuracy_reward": 0.7017160058021545,
      "rewards/format_reward": 1.0,
      "step": 387
    },
    {
      "completion_length": 77.515625,
      "epoch": 1.771689497716895,
      "grad_norm": 4.436239719390869,
      "kl": 0.1083984375,
      "learning_rate": 8.228310502283105e-07,
      "loss": 0.0043,
      "reward": 1.6824799180030823,
      "reward_std": 0.18607579916715622,
      "rewards/accuracy_reward": 0.6902924478054047,
      "rewards/format_reward": 0.9921875,
      "step": 388
    },
    {
      "completion_length": 60.2265625,
      "epoch": 1.776255707762557,
      "grad_norm": 2.3695435523986816,
      "kl": 0.1494140625,
      "learning_rate": 8.223744292237442e-07,
      "loss": 0.006,
      "reward": 1.7026600241661072,
      "reward_std": 0.13800616562366486,
      "rewards/accuracy_reward": 0.7104724645614624,
      "rewards/format_reward": 0.9921875,
      "step": 389
    },
    {
      "completion_length": 71.140625,
      "epoch": 1.7808219178082192,
      "grad_norm": 4.455196380615234,
      "kl": 0.15966796875,
      "learning_rate": 8.21917808219178e-07,
      "loss": 0.0064,
      "reward": 1.6663504838943481,
      "reward_std": 0.16872704774141312,
      "rewards/accuracy_reward": 0.6663504242897034,
      "rewards/format_reward": 1.0,
      "step": 390
    },
    {
      "completion_length": 65.8359375,
      "epoch": 1.7853881278538812,
      "grad_norm": 15.527393341064453,
      "kl": 0.148681640625,
      "learning_rate": 8.214611872146119e-07,
      "loss": 0.0059,
      "reward": 1.553125023841858,
      "reward_std": 0.21595830470323563,
      "rewards/accuracy_reward": 0.5531250089406967,
      "rewards/format_reward": 1.0,
      "step": 391
    },
    {
      "completion_length": 73.0625,
      "epoch": 1.7899543378995433,
      "grad_norm": 4.076765537261963,
      "kl": 0.101806640625,
      "learning_rate": 8.210045662100456e-07,
      "loss": 0.0041,
      "reward": 1.6578125357627869,
      "reward_std": 0.17282497137784958,
      "rewards/accuracy_reward": 0.6578125059604645,
      "rewards/format_reward": 1.0,
      "step": 392
    },
    {
      "completion_length": 64.3515625,
      "epoch": 1.7945205479452055,
      "grad_norm": 4.910162448883057,
      "kl": 0.24609375,
      "learning_rate": 8.205479452054795e-07,
      "loss": 0.0099,
      "reward": 1.5686192512512207,
      "reward_std": 0.20994295924901962,
      "rewards/accuracy_reward": 0.5764318108558655,
      "rewards/format_reward": 0.9921875,
      "step": 393
    },
    {
      "completion_length": 71.3359375,
      "epoch": 1.7990867579908676,
      "grad_norm": 3.380282402038574,
      "kl": 0.098876953125,
      "learning_rate": 8.200913242009132e-07,
      "loss": 0.004,
      "reward": 1.6727213859558105,
      "reward_std": 0.09240220487117767,
      "rewards/accuracy_reward": 0.672721341252327,
      "rewards/format_reward": 1.0,
      "step": 394
    },
    {
      "completion_length": 70.625,
      "epoch": 1.8036529680365296,
      "grad_norm": 15.939676284790039,
      "kl": 0.1748046875,
      "learning_rate": 8.19634703196347e-07,
      "loss": 0.007,
      "reward": 1.4804688096046448,
      "reward_std": 0.22750268876552582,
      "rewards/accuracy_reward": 0.4882812350988388,
      "rewards/format_reward": 0.9921875,
      "step": 395
    },
    {
      "completion_length": 76.9453125,
      "epoch": 1.808219178082192,
      "grad_norm": 8.642084121704102,
      "kl": 0.10009765625,
      "learning_rate": 8.191780821917808e-07,
      "loss": 0.004,
      "reward": 1.6074219346046448,
      "reward_std": 0.21606218069791794,
      "rewards/accuracy_reward": 0.615234375,
      "rewards/format_reward": 0.9921875,
      "step": 396
    },
    {
      "completion_length": 71.0546875,
      "epoch": 1.8127853881278537,
      "grad_norm": 3.1132235527038574,
      "kl": 0.1259765625,
      "learning_rate": 8.187214611872145e-07,
      "loss": 0.005,
      "reward": 1.6144831776618958,
      "reward_std": 0.22148973494768143,
      "rewards/accuracy_reward": 0.6144831776618958,
      "rewards/format_reward": 1.0,
      "step": 397
    },
    {
      "completion_length": 64.6484375,
      "epoch": 1.817351598173516,
      "grad_norm": 5.543883800506592,
      "kl": 0.135498046875,
      "learning_rate": 8.182648401826484e-07,
      "loss": 0.0054,
      "reward": 1.5686756372451782,
      "reward_std": 0.1985800489783287,
      "rewards/accuracy_reward": 0.5764880925416946,
      "rewards/format_reward": 0.9921875,
      "step": 398
    },
    {
      "completion_length": 59.421875,
      "epoch": 1.821917808219178,
      "grad_norm": 6.449603080749512,
      "kl": 0.14892578125,
      "learning_rate": 8.178082191780822e-07,
      "loss": 0.006,
      "reward": 1.6453125476837158,
      "reward_std": 0.1793966293334961,
      "rewards/accuracy_reward": 0.6531250476837158,
      "rewards/format_reward": 0.9921875,
      "step": 399
    },
    {
      "completion_length": 65.0859375,
      "epoch": 1.82648401826484,
      "grad_norm": 3.558039665222168,
      "kl": 0.121826171875,
      "learning_rate": 8.173515981735159e-07,
      "loss": 0.0049,
      "reward": 1.5231584310531616,
      "reward_std": 0.27039487659931183,
      "rewards/accuracy_reward": 0.5309710204601288,
      "rewards/format_reward": 0.9921875,
      "step": 400
    },
    {
      "completion_length": 74.875,
      "epoch": 1.8310502283105023,
      "grad_norm": 2.6240084171295166,
      "kl": 0.138427734375,
      "learning_rate": 8.168949771689498e-07,
      "loss": 0.0055,
      "reward": 1.5302269458770752,
      "reward_std": 0.19186384975910187,
      "rewards/accuracy_reward": 0.5302269384264946,
      "rewards/format_reward": 1.0,
      "step": 401
    },
    {
      "completion_length": 72.109375,
      "epoch": 1.8356164383561644,
      "grad_norm": 2.054145574569702,
      "kl": 0.11572265625,
      "learning_rate": 8.164383561643835e-07,
      "loss": 0.0046,
      "reward": 1.7239583730697632,
      "reward_std": 0.13660814613103867,
      "rewards/accuracy_reward": 0.723958283662796,
      "rewards/format_reward": 1.0,
      "step": 402
    },
    {
      "completion_length": 93.1484375,
      "epoch": 1.8401826484018264,
      "grad_norm": 2.950429916381836,
      "kl": 0.108642578125,
      "learning_rate": 8.159817351598172e-07,
      "loss": 0.0043,
      "reward": 1.734375,
      "reward_std": 0.11295716743916273,
      "rewards/accuracy_reward": 0.7343749403953552,
      "rewards/format_reward": 1.0,
      "step": 403
    },
    {
      "completion_length": 68.5546875,
      "epoch": 1.8447488584474887,
      "grad_norm": 2.904849052429199,
      "kl": 0.154541015625,
      "learning_rate": 8.155251141552512e-07,
      "loss": 0.0062,
      "reward": 1.6253038048744202,
      "reward_std": 0.26257922500371933,
      "rewards/accuracy_reward": 0.6409288048744202,
      "rewards/format_reward": 0.984375,
      "step": 404
    },
    {
      "completion_length": 80.0390625,
      "epoch": 1.8493150684931505,
      "grad_norm": 2.698310136795044,
      "kl": 0.10546875,
      "learning_rate": 8.150684931506849e-07,
      "loss": 0.0042,
      "reward": 1.6187500357627869,
      "reward_std": 0.1505398042500019,
      "rewards/accuracy_reward": 0.6343750059604645,
      "rewards/format_reward": 0.984375,
      "step": 405
    },
    {
      "completion_length": 88.8046875,
      "epoch": 1.8538812785388128,
      "grad_norm": 3.7120022773742676,
      "kl": 0.091064453125,
      "learning_rate": 8.146118721461187e-07,
      "loss": 0.0036,
      "reward": 1.6648437976837158,
      "reward_std": 0.16381803154945374,
      "rewards/accuracy_reward": 0.6882811784744263,
      "rewards/format_reward": 0.9765625,
      "step": 406
    },
    {
      "completion_length": 82.1171875,
      "epoch": 1.8584474885844748,
      "grad_norm": 17.568010330200195,
      "kl": 0.09521484375,
      "learning_rate": 8.141552511415525e-07,
      "loss": 0.0038,
      "reward": 1.5686570405960083,
      "reward_std": 0.27213824540376663,
      "rewards/accuracy_reward": 0.5920945107936859,
      "rewards/format_reward": 0.9765625,
      "step": 407
    },
    {
      "completion_length": 72.0390625,
      "epoch": 1.8630136986301369,
      "grad_norm": 5.123001575469971,
      "kl": 0.108642578125,
      "learning_rate": 8.136986301369862e-07,
      "loss": 0.0044,
      "reward": 1.6002225279808044,
      "reward_std": 0.1988915428519249,
      "rewards/accuracy_reward": 0.6002225577831268,
      "rewards/format_reward": 1.0,
      "step": 408
    },
    {
      "completion_length": 65.7890625,
      "epoch": 1.8675799086757991,
      "grad_norm": 3.825585126876831,
      "kl": 0.12646484375,
      "learning_rate": 8.132420091324201e-07,
      "loss": 0.0051,
      "reward": 1.5363582372665405,
      "reward_std": 0.22431423515081406,
      "rewards/accuracy_reward": 0.5363581627607346,
      "rewards/format_reward": 1.0,
      "step": 409
    },
    {
      "completion_length": 84.0859375,
      "epoch": 1.8721461187214612,
      "grad_norm": 2.313400983810425,
      "kl": 0.123779296875,
      "learning_rate": 8.127853881278538e-07,
      "loss": 0.005,
      "reward": 1.6218750476837158,
      "reward_std": 0.28108011931180954,
      "rewards/accuracy_reward": 0.7078124582767487,
      "rewards/format_reward": 0.9140625,
      "step": 410
    },
    {
      "completion_length": 88.03125,
      "epoch": 1.8767123287671232,
      "grad_norm": 4.612303256988525,
      "kl": 0.132568359375,
      "learning_rate": 8.123287671232877e-07,
      "loss": 0.0053,
      "reward": 1.662500023841858,
      "reward_std": 0.18329600244760513,
      "rewards/accuracy_reward": 0.6781249642372131,
      "rewards/format_reward": 0.984375,
      "step": 411
    },
    {
      "completion_length": 60.7265625,
      "epoch": 1.8812785388127855,
      "grad_norm": 2.7391302585601807,
      "kl": 0.14794921875,
      "learning_rate": 8.118721461187215e-07,
      "loss": 0.0059,
      "reward": 1.648708462715149,
      "reward_std": 0.15135541558265686,
      "rewards/accuracy_reward": 0.6487084329128265,
      "rewards/format_reward": 1.0,
      "step": 412
    },
    {
      "completion_length": 103.84375,
      "epoch": 1.8858447488584473,
      "grad_norm": 1.170571208000183,
      "kl": 0.0670166015625,
      "learning_rate": 8.114155251141552e-07,
      "loss": 0.0027,
      "reward": 1.811079502105713,
      "reward_std": 0.0749700665473938,
      "rewards/accuracy_reward": 0.8267044425010681,
      "rewards/format_reward": 0.984375,
      "step": 413
    },
    {
      "completion_length": 70.9921875,
      "epoch": 1.8904109589041096,
      "grad_norm": 3.7362253665924072,
      "kl": 0.14404296875,
      "learning_rate": 8.10958904109589e-07,
      "loss": 0.0058,
      "reward": 1.6984003186225891,
      "reward_std": 0.19452574849128723,
      "rewards/accuracy_reward": 0.7062127590179443,
      "rewards/format_reward": 0.9921875,
      "step": 414
    },
    {
      "completion_length": 66.953125,
      "epoch": 1.8949771689497716,
      "grad_norm": 4.104921340942383,
      "kl": 0.133056640625,
      "learning_rate": 8.105022831050228e-07,
      "loss": 0.0053,
      "reward": 1.5964038372039795,
      "reward_std": 0.2661294490098953,
      "rewards/accuracy_reward": 0.6042163074016571,
      "rewards/format_reward": 0.9921875,
      "step": 415
    },
    {
      "completion_length": 67.984375,
      "epoch": 1.8995433789954337,
      "grad_norm": 5.06357479095459,
      "kl": 0.11669921875,
      "learning_rate": 8.100456621004565e-07,
      "loss": 0.0047,
      "reward": 1.6529513597488403,
      "reward_std": 0.16535750776529312,
      "rewards/accuracy_reward": 0.6529513597488403,
      "rewards/format_reward": 1.0,
      "step": 416
    },
    {
      "completion_length": 72.1640625,
      "epoch": 1.904109589041096,
      "grad_norm": 9.87835693359375,
      "kl": 0.137451171875,
      "learning_rate": 8.095890410958903e-07,
      "loss": 0.0055,
      "reward": 1.6302083134651184,
      "reward_std": 0.21513652801513672,
      "rewards/accuracy_reward": 0.6458333134651184,
      "rewards/format_reward": 0.984375,
      "step": 417
    },
    {
      "completion_length": 73.6015625,
      "epoch": 1.908675799086758,
      "grad_norm": 10.113238334655762,
      "kl": 0.142333984375,
      "learning_rate": 8.091324200913242e-07,
      "loss": 0.0057,
      "reward": 1.6189236044883728,
      "reward_std": 0.23221635073423386,
      "rewards/accuracy_reward": 0.6345485746860504,
      "rewards/format_reward": 0.984375,
      "step": 418
    },
    {
      "completion_length": 62.078125,
      "epoch": 1.91324200913242,
      "grad_norm": 3.4071831703186035,
      "kl": 0.1513671875,
      "learning_rate": 8.08675799086758e-07,
      "loss": 0.0061,
      "reward": 1.356249988079071,
      "reward_std": 0.3108007460832596,
      "rewards/accuracy_reward": 0.36406250298023224,
      "rewards/format_reward": 0.9921875,
      "step": 419
    },
    {
      "completion_length": 83.203125,
      "epoch": 1.9178082191780823,
      "grad_norm": 3.302225351333618,
      "kl": 0.2509765625,
      "learning_rate": 8.082191780821918e-07,
      "loss": 0.0101,
      "reward": 1.6587789058685303,
      "reward_std": 0.1820889264345169,
      "rewards/accuracy_reward": 0.6665914356708527,
      "rewards/format_reward": 0.9921875,
      "step": 420
    },
    {
      "completion_length": 76.1015625,
      "epoch": 1.9223744292237441,
      "grad_norm": 4.783244609832764,
      "kl": 0.1474609375,
      "learning_rate": 8.077625570776255e-07,
      "loss": 0.0059,
      "reward": 1.7332961559295654,
      "reward_std": 0.11496374011039734,
      "rewards/accuracy_reward": 0.733296126127243,
      "rewards/format_reward": 1.0,
      "step": 421
    },
    {
      "completion_length": 94.6484375,
      "epoch": 1.9269406392694064,
      "grad_norm": 2.7827274799346924,
      "kl": 0.09375,
      "learning_rate": 8.073059360730593e-07,
      "loss": 0.0038,
      "reward": 1.8006826043128967,
      "reward_std": 0.09438200853765011,
      "rewards/accuracy_reward": 0.800682544708252,
      "rewards/format_reward": 1.0,
      "step": 422
    },
    {
      "completion_length": 70.4375,
      "epoch": 1.9315068493150684,
      "grad_norm": 4.186135768890381,
      "kl": 0.16455078125,
      "learning_rate": 8.068493150684931e-07,
      "loss": 0.0066,
      "reward": 1.6078130006790161,
      "reward_std": 0.21306797862052917,
      "rewards/accuracy_reward": 0.6078130900859833,
      "rewards/format_reward": 1.0,
      "step": 423
    },
    {
      "completion_length": 77.6953125,
      "epoch": 1.9360730593607305,
      "grad_norm": 2.545297861099243,
      "kl": 0.111083984375,
      "learning_rate": 8.063926940639269e-07,
      "loss": 0.0045,
      "reward": 1.6678841710090637,
      "reward_std": 0.11376722529530525,
      "rewards/accuracy_reward": 0.6756967306137085,
      "rewards/format_reward": 0.9921875,
      "step": 424
    },
    {
      "completion_length": 87.1171875,
      "epoch": 1.9406392694063928,
      "grad_norm": 2.4129421710968018,
      "kl": 0.080078125,
      "learning_rate": 8.059360730593608e-07,
      "loss": 0.0032,
      "reward": 1.6949777007102966,
      "reward_std": 0.14730913192033768,
      "rewards/accuracy_reward": 0.7027901709079742,
      "rewards/format_reward": 0.9921875,
      "step": 425
    },
    {
      "completion_length": 78.53125,
      "epoch": 1.9452054794520548,
      "grad_norm": 2.179104804992676,
      "kl": 0.120361328125,
      "learning_rate": 8.054794520547945e-07,
      "loss": 0.0048,
      "reward": 1.7349414825439453,
      "reward_std": 0.20185434818267822,
      "rewards/accuracy_reward": 0.7505663931369781,
      "rewards/format_reward": 0.984375,
      "step": 426
    },
    {
      "completion_length": 70.2578125,
      "epoch": 1.9497716894977168,
      "grad_norm": 3.5189919471740723,
      "kl": 0.15869140625,
      "learning_rate": 8.050228310502283e-07,
      "loss": 0.0064,
      "reward": 1.6359375715255737,
      "reward_std": 0.1778659224510193,
      "rewards/accuracy_reward": 0.6359374523162842,
      "rewards/format_reward": 1.0,
      "step": 427
    },
    {
      "completion_length": 91.140625,
      "epoch": 1.954337899543379,
      "grad_norm": 4.333745002746582,
      "kl": 0.07861328125,
      "learning_rate": 8.045662100456621e-07,
      "loss": 0.0031,
      "reward": 1.6484509706497192,
      "reward_std": 0.23022788017988205,
      "rewards/accuracy_reward": 0.6797009706497192,
      "rewards/format_reward": 0.96875,
      "step": 428
    },
    {
      "completion_length": 75.453125,
      "epoch": 1.958904109589041,
      "grad_norm": 3.482632637023926,
      "kl": 0.12841796875,
      "learning_rate": 8.041095890410958e-07,
      "loss": 0.0051,
      "reward": 1.6106771230697632,
      "reward_std": 0.23327118158340454,
      "rewards/accuracy_reward": 0.626302033662796,
      "rewards/format_reward": 0.984375,
      "step": 429
    },
    {
      "completion_length": 87.984375,
      "epoch": 1.9634703196347032,
      "grad_norm": 1.5310007333755493,
      "kl": 0.12451171875,
      "learning_rate": 8.036529680365296e-07,
      "loss": 0.005,
      "reward": 1.7687500715255737,
      "reward_std": 0.1117947231978178,
      "rewards/accuracy_reward": 0.7765624523162842,
      "rewards/format_reward": 0.9921875,
      "step": 430
    },
    {
      "completion_length": 74.0625,
      "epoch": 1.9680365296803652,
      "grad_norm": 2.0865707397460938,
      "kl": 0.120849609375,
      "learning_rate": 8.031963470319635e-07,
      "loss": 0.0048,
      "reward": 1.7373016476631165,
      "reward_std": 0.1270090974867344,
      "rewards/accuracy_reward": 0.7373015582561493,
      "rewards/format_reward": 1.0,
      "step": 431
    },
    {
      "completion_length": 74.6953125,
      "epoch": 1.9726027397260273,
      "grad_norm": 2.8522982597351074,
      "kl": 0.104248046875,
      "learning_rate": 8.027397260273972e-07,
      "loss": 0.0042,
      "reward": 1.6342076063156128,
      "reward_std": 0.18551241606473923,
      "rewards/accuracy_reward": 0.6498326063156128,
      "rewards/format_reward": 0.984375,
      "step": 432
    },
    {
      "completion_length": 86.6484375,
      "epoch": 1.9771689497716896,
      "grad_norm": 2.4067180156707764,
      "kl": 0.10009765625,
      "learning_rate": 8.022831050228311e-07,
      "loss": 0.004,
      "reward": 1.6945313215255737,
      "reward_std": 0.11847387999296188,
      "rewards/accuracy_reward": 0.694531261920929,
      "rewards/format_reward": 1.0,
      "step": 433
    },
    {
      "completion_length": 80.296875,
      "epoch": 1.9817351598173516,
      "grad_norm": 2.204301357269287,
      "kl": 0.12890625,
      "learning_rate": 8.018264840182648e-07,
      "loss": 0.0052,
      "reward": 1.6941722631454468,
      "reward_std": 0.16894984245300293,
      "rewards/accuracy_reward": 0.7176096439361572,
      "rewards/format_reward": 0.9765625,
      "step": 434
    },
    {
      "completion_length": 69.578125,
      "epoch": 1.9863013698630136,
      "grad_norm": 3.0245296955108643,
      "kl": 0.127197265625,
      "learning_rate": 8.013698630136985e-07,
      "loss": 0.0051,
      "reward": 1.627698838710785,
      "reward_std": 0.1703873947262764,
      "rewards/accuracy_reward": 0.6276988387107849,
      "rewards/format_reward": 1.0,
      "step": 435
    },
    {
      "completion_length": 59.9140625,
      "epoch": 1.990867579908676,
      "grad_norm": 3.451076030731201,
      "kl": 0.14404296875,
      "learning_rate": 8.009132420091324e-07,
      "loss": 0.0058,
      "reward": 1.5428819060325623,
      "reward_std": 0.2504672184586525,
      "rewards/accuracy_reward": 0.542881965637207,
      "rewards/format_reward": 1.0,
      "step": 436
    },
    {
      "completion_length": 84.5625,
      "epoch": 1.9954337899543377,
      "grad_norm": 2.452648639678955,
      "kl": 0.08056640625,
      "learning_rate": 8.004566210045661e-07,
      "loss": 0.0032,
      "reward": 1.6554688215255737,
      "reward_std": 0.2031169831752777,
      "rewards/accuracy_reward": 0.6710937917232513,
      "rewards/format_reward": 0.984375,
      "step": 437
    },
    {
      "completion_length": 45.5,
      "epoch": 2.0,
      "grad_norm": 1.8669296503067017,
      "kl": 0.126953125,
      "learning_rate": 8e-07,
      "loss": 0.0047,
      "reward": 1.875,
      "reward_std": 0.43671509623527527,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 1.0,
      "step": 438
    },
    {
      "completion_length": 79.21875,
      "epoch": 2.0045662100456623,
      "grad_norm": 1.4508730173110962,
      "kl": 0.127685546875,
      "learning_rate": 7.995433789954338e-07,
      "loss": 0.0051,
      "reward": 1.7248697876930237,
      "reward_std": 0.1105603277683258,
      "rewards/accuracy_reward": 0.7404947876930237,
      "rewards/format_reward": 0.984375,
      "step": 439
    },
    {
      "completion_length": 62.09375,
      "epoch": 2.009132420091324,
      "grad_norm": 2.7377336025238037,
      "kl": 0.1376953125,
      "learning_rate": 7.990867579908675e-07,
      "loss": 0.0055,
      "reward": 1.3338541984558105,
      "reward_std": 0.3162979334592819,
      "rewards/accuracy_reward": 0.34947916865348816,
      "rewards/format_reward": 0.984375,
      "step": 440
    },
    {
      "completion_length": 87.34375,
      "epoch": 2.0136986301369864,
      "grad_norm": 2.2570443153381348,
      "kl": 0.12060546875,
      "learning_rate": 7.986301369863014e-07,
      "loss": 0.0048,
      "reward": 1.7250688076019287,
      "reward_std": 0.08361868560314178,
      "rewards/accuracy_reward": 0.7250687181949615,
      "rewards/format_reward": 1.0,
      "step": 441
    },
    {
      "completion_length": 94.140625,
      "epoch": 2.018264840182648,
      "grad_norm": 8.999687194824219,
      "kl": 0.10595703125,
      "learning_rate": 7.981735159817351e-07,
      "loss": 0.0042,
      "reward": 1.6828125715255737,
      "reward_std": 0.16398613899946213,
      "rewards/accuracy_reward": 0.7062499523162842,
      "rewards/format_reward": 0.9765625,
      "step": 442
    },
    {
      "completion_length": 69.984375,
      "epoch": 2.0228310502283104,
      "grad_norm": 2.834467649459839,
      "kl": 0.096923828125,
      "learning_rate": 7.977168949771688e-07,
      "loss": 0.0039,
      "reward": 1.5455728769302368,
      "reward_std": 0.25657252967357635,
      "rewards/accuracy_reward": 0.5533854067325592,
      "rewards/format_reward": 0.9921875,
      "step": 443
    },
    {
      "completion_length": 98.921875,
      "epoch": 2.0273972602739727,
      "grad_norm": 2.6678521633148193,
      "kl": 0.082275390625,
      "learning_rate": 7.972602739726027e-07,
      "loss": 0.0033,
      "reward": 1.663573145866394,
      "reward_std": 0.1522715613245964,
      "rewards/accuracy_reward": 0.6635731160640717,
      "rewards/format_reward": 1.0,
      "step": 444
    },
    {
      "completion_length": 80.84375,
      "epoch": 2.0319634703196345,
      "grad_norm": 1.9825059175491333,
      "kl": 0.09326171875,
      "learning_rate": 7.968036529680365e-07,
      "loss": 0.0037,
      "reward": 1.603124976158142,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.6109374761581421,
      "rewards/format_reward": 0.9921875,
      "step": 445
    },
    {
      "completion_length": 78.09375,
      "epoch": 2.036529680365297,
      "grad_norm": 5.509555816650391,
      "kl": 0.1341552734375,
      "learning_rate": 7.963470319634703e-07,
      "loss": 0.0054,
      "reward": 1.6062500476837158,
      "reward_std": 0.24452467262744904,
      "rewards/accuracy_reward": 0.621874988079071,
      "rewards/format_reward": 0.984375,
      "step": 446
    },
    {
      "completion_length": 59.6484375,
      "epoch": 2.041095890410959,
      "grad_norm": 4.631182670593262,
      "kl": 0.154296875,
      "learning_rate": 7.958904109589041e-07,
      "loss": 0.0062,
      "reward": 1.6522735357284546,
      "reward_std": 0.24951402842998505,
      "rewards/accuracy_reward": 0.667898565530777,
      "rewards/format_reward": 0.984375,
      "step": 447
    },
    {
      "completion_length": 69.25,
      "epoch": 2.045662100456621,
      "grad_norm": 3.0625455379486084,
      "kl": 0.152099609375,
      "learning_rate": 7.954337899543378e-07,
      "loss": 0.0061,
      "reward": 1.6024305820465088,
      "reward_std": 0.21730707585811615,
      "rewards/accuracy_reward": 0.6102430522441864,
      "rewards/format_reward": 0.9921875,
      "step": 448
    },
    {
      "completion_length": 66.5859375,
      "epoch": 2.050228310502283,
      "grad_norm": 7.600069522857666,
      "kl": 0.118408203125,
      "learning_rate": 7.949771689497717e-07,
      "loss": 0.0047,
      "reward": 1.7647135853767395,
      "reward_std": 0.16752880066633224,
      "rewards/accuracy_reward": 0.7725259959697723,
      "rewards/format_reward": 0.9921875,
      "step": 449
    },
    {
      "completion_length": 83.8828125,
      "epoch": 2.0547945205479454,
      "grad_norm": 4.596118450164795,
      "kl": 0.149658203125,
      "learning_rate": 7.945205479452054e-07,
      "loss": 0.006,
      "reward": 1.6769480109214783,
      "reward_std": 0.1898738443851471,
      "rewards/accuracy_reward": 0.684760570526123,
      "rewards/format_reward": 0.9921875,
      "step": 450
    },
    {
      "completion_length": 48.546875,
      "epoch": 2.0593607305936072,
      "grad_norm": 3.093118906021118,
      "kl": 0.171875,
      "learning_rate": 7.940639269406393e-07,
      "loss": 0.0069,
      "reward": 1.6568829417228699,
      "reward_std": 0.20424779504537582,
      "rewards/accuracy_reward": 0.656883031129837,
      "rewards/format_reward": 1.0,
      "step": 451
    },
    {
      "completion_length": 73.203125,
      "epoch": 2.0639269406392695,
      "grad_norm": 3.5054690837860107,
      "kl": 0.135498046875,
      "learning_rate": 7.936073059360731e-07,
      "loss": 0.0054,
      "reward": 1.6708519458770752,
      "reward_std": 0.18046213686466217,
      "rewards/accuracy_reward": 0.6708519458770752,
      "rewards/format_reward": 1.0,
      "step": 452
    },
    {
      "completion_length": 78.1953125,
      "epoch": 2.0684931506849313,
      "grad_norm": 1.87969172000885,
      "kl": 0.096923828125,
      "learning_rate": 7.931506849315068e-07,
      "loss": 0.0039,
      "reward": 1.6500000953674316,
      "reward_std": 0.1841355338692665,
      "rewards/accuracy_reward": 0.6656249761581421,
      "rewards/format_reward": 0.984375,
      "step": 453
    },
    {
      "completion_length": 66.203125,
      "epoch": 2.0730593607305936,
      "grad_norm": 3.03586483001709,
      "kl": 0.116455078125,
      "learning_rate": 7.926940639269406e-07,
      "loss": 0.0047,
      "reward": 1.5984020233154297,
      "reward_std": 0.2402767539024353,
      "rewards/accuracy_reward": 0.6218394637107849,
      "rewards/format_reward": 0.9765625,
      "step": 454
    },
    {
      "completion_length": 60.4453125,
      "epoch": 2.077625570776256,
      "grad_norm": 5.758031368255615,
      "kl": 0.15185546875,
      "learning_rate": 7.922374429223744e-07,
      "loss": 0.0061,
      "reward": 1.6396695375442505,
      "reward_std": 0.18095022439956665,
      "rewards/accuracy_reward": 0.6396694481372833,
      "rewards/format_reward": 1.0,
      "step": 455
    },
    {
      "completion_length": 80.0703125,
      "epoch": 2.0821917808219177,
      "grad_norm": 2.098630905151367,
      "kl": 0.105712890625,
      "learning_rate": 7.917808219178081e-07,
      "loss": 0.0042,
      "reward": 1.6691096425056458,
      "reward_std": 0.1658947691321373,
      "rewards/accuracy_reward": 0.684734582901001,
      "rewards/format_reward": 0.984375,
      "step": 456
    },
    {
      "completion_length": 72.2734375,
      "epoch": 2.08675799086758,
      "grad_norm": 2.7200841903686523,
      "kl": 0.11083984375,
      "learning_rate": 7.91324200913242e-07,
      "loss": 0.0044,
      "reward": 1.5380051136016846,
      "reward_std": 0.2077661082148552,
      "rewards/accuracy_reward": 0.5458175092935562,
      "rewards/format_reward": 0.9921875,
      "step": 457
    },
    {
      "completion_length": 85.140625,
      "epoch": 2.091324200913242,
      "grad_norm": 12.362757682800293,
      "kl": 0.091552734375,
      "learning_rate": 7.908675799086758e-07,
      "loss": 0.0037,
      "reward": 1.6828125715255737,
      "reward_std": 0.19778337329626083,
      "rewards/accuracy_reward": 0.6984374821186066,
      "rewards/format_reward": 0.984375,
      "step": 458
    },
    {
      "completion_length": 84.1875,
      "epoch": 2.095890410958904,
      "grad_norm": 2.426860809326172,
      "kl": 0.10595703125,
      "learning_rate": 7.904109589041096e-07,
      "loss": 0.0042,
      "reward": 1.6260236501693726,
      "reward_std": 0.18746323138475418,
      "rewards/accuracy_reward": 0.6338361203670502,
      "rewards/format_reward": 0.9921875,
      "step": 459
    },
    {
      "completion_length": 91.34375,
      "epoch": 2.1004566210045663,
      "grad_norm": 9.778014183044434,
      "kl": 0.09765625,
      "learning_rate": 7.899543378995434e-07,
      "loss": 0.0039,
      "reward": 1.660937488079071,
      "reward_std": 0.19728107750415802,
      "rewards/accuracy_reward": 0.6921874582767487,
      "rewards/format_reward": 0.96875,
      "step": 460
    },
    {
      "completion_length": 60.7265625,
      "epoch": 2.105022831050228,
      "grad_norm": 12.834918975830078,
      "kl": 0.32470703125,
      "learning_rate": 7.894977168949771e-07,
      "loss": 0.0129,
      "reward": 1.7225513458251953,
      "reward_std": 0.1815098226070404,
      "rewards/accuracy_reward": 0.7225514352321625,
      "rewards/format_reward": 1.0,
      "step": 461
    },
    {
      "completion_length": 65.0546875,
      "epoch": 2.1095890410958904,
      "grad_norm": 3.4787039756774902,
      "kl": 0.1533203125,
      "learning_rate": 7.890410958904109e-07,
      "loss": 0.0061,
      "reward": 1.6404947638511658,
      "reward_std": 0.19807539880275726,
      "rewards/accuracy_reward": 0.6483072936534882,
      "rewards/format_reward": 0.9921875,
      "step": 462
    },
    {
      "completion_length": 82.46875,
      "epoch": 2.1141552511415527,
      "grad_norm": 2.7784857749938965,
      "kl": 0.1136474609375,
      "learning_rate": 7.885844748858447e-07,
      "loss": 0.0045,
      "reward": 1.6932291984558105,
      "reward_std": 0.14191660657525063,
      "rewards/accuracy_reward": 0.6932291388511658,
      "rewards/format_reward": 1.0,
      "step": 463
    },
    {
      "completion_length": 78.5,
      "epoch": 2.1187214611872145,
      "grad_norm": 2.705900192260742,
      "kl": 0.098388671875,
      "learning_rate": 7.881278538812784e-07,
      "loss": 0.0039,
      "reward": 1.6399182081222534,
      "reward_std": 0.18248122185468674,
      "rewards/accuracy_reward": 0.6477306485176086,
      "rewards/format_reward": 0.9921875,
      "step": 464
    },
    {
      "completion_length": 88.0703125,
      "epoch": 2.1232876712328768,
      "grad_norm": 2.8577768802642822,
      "kl": 0.099609375,
      "learning_rate": 7.876712328767124e-07,
      "loss": 0.004,
      "reward": 1.688281238079071,
      "reward_std": 0.19159993529319763,
      "rewards/accuracy_reward": 0.6960937082767487,
      "rewards/format_reward": 0.9921875,
      "step": 465
    },
    {
      "completion_length": 66.6015625,
      "epoch": 2.127853881278539,
      "grad_norm": 2.5234570503234863,
      "kl": 0.12158203125,
      "learning_rate": 7.872146118721461e-07,
      "loss": 0.0049,
      "reward": 1.5919778943061829,
      "reward_std": 0.24576038867235184,
      "rewards/accuracy_reward": 0.6076028943061829,
      "rewards/format_reward": 0.984375,
      "step": 466
    },
    {
      "completion_length": 75.390625,
      "epoch": 2.132420091324201,
      "grad_norm": 3.172330141067505,
      "kl": 0.1142578125,
      "learning_rate": 7.867579908675798e-07,
      "loss": 0.0046,
      "reward": 1.4447365403175354,
      "reward_std": 0.22060814499855042,
      "rewards/accuracy_reward": 0.452549085021019,
      "rewards/format_reward": 0.9921875,
      "step": 467
    },
    {
      "completion_length": 91.1640625,
      "epoch": 2.136986301369863,
      "grad_norm": 2.385173797607422,
      "kl": 0.0966796875,
      "learning_rate": 7.863013698630137e-07,
      "loss": 0.0039,
      "reward": 1.7272321581840515,
      "reward_std": 0.13396714627742767,
      "rewards/accuracy_reward": 0.7350445687770844,
      "rewards/format_reward": 0.9921875,
      "step": 468
    },
    {
      "completion_length": 75.5234375,
      "epoch": 2.141552511415525,
      "grad_norm": 2.3278234004974365,
      "kl": 0.130126953125,
      "learning_rate": 7.858447488584474e-07,
      "loss": 0.0052,
      "reward": 1.643117606639862,
      "reward_std": 0.1804744228720665,
      "rewards/accuracy_reward": 0.6509300172328949,
      "rewards/format_reward": 0.9921875,
      "step": 469
    },
    {
      "completion_length": 63.9453125,
      "epoch": 2.146118721461187,
      "grad_norm": 5.3924360275268555,
      "kl": 0.16552734375,
      "learning_rate": 7.853881278538812e-07,
      "loss": 0.0066,
      "reward": 1.5500783324241638,
      "reward_std": 0.25609923899173737,
      "rewards/accuracy_reward": 0.5735158026218414,
      "rewards/format_reward": 0.9765625,
      "step": 470
    },
    {
      "completion_length": 86.0625,
      "epoch": 2.1506849315068495,
      "grad_norm": 2.030338764190674,
      "kl": 0.1171875,
      "learning_rate": 7.849315068493151e-07,
      "loss": 0.0047,
      "reward": 1.6851562857627869,
      "reward_std": 0.14745555073022842,
      "rewards/accuracy_reward": 0.6929687559604645,
      "rewards/format_reward": 0.9921875,
      "step": 471
    },
    {
      "completion_length": 66.0,
      "epoch": 2.1552511415525113,
      "grad_norm": 18.199661254882812,
      "kl": 0.13330078125,
      "learning_rate": 7.844748858447488e-07,
      "loss": 0.0053,
      "reward": 1.486718773841858,
      "reward_std": 0.32600878179073334,
      "rewards/accuracy_reward": 0.5101562440395355,
      "rewards/format_reward": 0.9765625,
      "step": 472
    },
    {
      "completion_length": 93.5625,
      "epoch": 2.1598173515981736,
      "grad_norm": 3.143573760986328,
      "kl": 0.07470703125,
      "learning_rate": 7.840182648401827e-07,
      "loss": 0.003,
      "reward": 1.7031250596046448,
      "reward_std": 0.18717344850301743,
      "rewards/accuracy_reward": 0.7109374701976776,
      "rewards/format_reward": 0.9921875,
      "step": 473
    },
    {
      "completion_length": 68.9453125,
      "epoch": 2.1643835616438354,
      "grad_norm": 3.1616029739379883,
      "kl": 0.121337890625,
      "learning_rate": 7.835616438356164e-07,
      "loss": 0.0049,
      "reward": 1.6342764496803284,
      "reward_std": 0.23029568046331406,
      "rewards/accuracy_reward": 0.6499014496803284,
      "rewards/format_reward": 0.984375,
      "step": 474
    },
    {
      "completion_length": 80.5859375,
      "epoch": 2.1689497716894977,
      "grad_norm": 2.141812324523926,
      "kl": 0.11572265625,
      "learning_rate": 7.831050228310501e-07,
      "loss": 0.0046,
      "reward": 1.7305381298065186,
      "reward_std": 0.19372030347585678,
      "rewards/accuracy_reward": 0.7383506596088409,
      "rewards/format_reward": 0.9921875,
      "step": 475
    },
    {
      "completion_length": 89.4921875,
      "epoch": 2.17351598173516,
      "grad_norm": 3.011232852935791,
      "kl": 0.098876953125,
      "learning_rate": 7.82648401826484e-07,
      "loss": 0.004,
      "reward": 1.60247403383255,
      "reward_std": 0.28417903184890747,
      "rewards/accuracy_reward": 0.6415364444255829,
      "rewards/format_reward": 0.9609375,
      "step": 476
    },
    {
      "completion_length": 88.90625,
      "epoch": 2.1780821917808217,
      "grad_norm": 1.868679165840149,
      "kl": 0.101806640625,
      "learning_rate": 7.821917808219177e-07,
      "loss": 0.0041,
      "reward": 1.5997712016105652,
      "reward_std": 0.1929171234369278,
      "rewards/accuracy_reward": 0.5997711420059204,
      "rewards/format_reward": 1.0,
      "step": 477
    },
    {
      "completion_length": 102.1640625,
      "epoch": 2.182648401826484,
      "grad_norm": 3.3878626823425293,
      "kl": 0.11962890625,
      "learning_rate": 7.817351598173516e-07,
      "loss": 0.0048,
      "reward": 1.7378038167953491,
      "reward_std": 0.13828756287693977,
      "rewards/accuracy_reward": 0.7456162571907043,
      "rewards/format_reward": 0.9921875,
      "step": 478
    },
    {
      "completion_length": 73.65625,
      "epoch": 2.1872146118721463,
      "grad_norm": 3.133633613586426,
      "kl": 0.1357421875,
      "learning_rate": 7.812785388127854e-07,
      "loss": 0.0054,
      "reward": 1.7677083611488342,
      "reward_std": 0.18527808785438538,
      "rewards/accuracy_reward": 0.7911458313465118,
      "rewards/format_reward": 0.9765625,
      "step": 479
    },
    {
      "completion_length": 65.0703125,
      "epoch": 2.191780821917808,
      "grad_norm": 2.9012463092803955,
      "kl": 0.14794921875,
      "learning_rate": 7.808219178082191e-07,
      "loss": 0.0059,
      "reward": 1.5422247648239136,
      "reward_std": 0.2771962434053421,
      "rewards/accuracy_reward": 0.5500372052192688,
      "rewards/format_reward": 0.9921875,
      "step": 480
    },
    {
      "completion_length": 80.53125,
      "epoch": 2.1963470319634704,
      "grad_norm": 7.046329498291016,
      "kl": 0.103271484375,
      "learning_rate": 7.80365296803653e-07,
      "loss": 0.0041,
      "reward": 1.66796875,
      "reward_std": 0.1984855979681015,
      "rewards/accuracy_reward": 0.6679687798023224,
      "rewards/format_reward": 1.0,
      "step": 481
    },
    {
      "completion_length": 81.921875,
      "epoch": 2.2009132420091326,
      "grad_norm": 2.9581196308135986,
      "kl": 0.108642578125,
      "learning_rate": 7.799086757990867e-07,
      "loss": 0.0043,
      "reward": 1.6520833373069763,
      "reward_std": 0.23736542463302612,
      "rewards/accuracy_reward": 0.6598958075046539,
      "rewards/format_reward": 0.9921875,
      "step": 482
    },
    {
      "completion_length": 79.78125,
      "epoch": 2.2054794520547945,
      "grad_norm": 8.972982406616211,
      "kl": 0.10693359375,
      "learning_rate": 7.794520547945204e-07,
      "loss": 0.0043,
      "reward": 1.5802912712097168,
      "reward_std": 0.21900298446416855,
      "rewards/accuracy_reward": 0.580291211605072,
      "rewards/format_reward": 1.0,
      "step": 483
    },
    {
      "completion_length": 83.5703125,
      "epoch": 2.2100456621004567,
      "grad_norm": 2.1765618324279785,
      "kl": 0.12548828125,
      "learning_rate": 7.789954337899543e-07,
      "loss": 0.005,
      "reward": 1.6838542222976685,
      "reward_std": 0.21648824214935303,
      "rewards/accuracy_reward": 0.6838541924953461,
      "rewards/format_reward": 1.0,
      "step": 484
    },
    {
      "completion_length": 82.1171875,
      "epoch": 2.2146118721461185,
      "grad_norm": 6.6184563636779785,
      "kl": 0.13916015625,
      "learning_rate": 7.785388127853881e-07,
      "loss": 0.0056,
      "reward": 1.576785683631897,
      "reward_std": 0.2656140699982643,
      "rewards/accuracy_reward": 0.615848183631897,
      "rewards/format_reward": 0.9609375,
      "step": 485
    },
    {
      "completion_length": 79.90625,
      "epoch": 2.219178082191781,
      "grad_norm": 3.397468328475952,
      "kl": 0.12109375,
      "learning_rate": 7.780821917808219e-07,
      "loss": 0.0048,
      "reward": 1.6415550708770752,
      "reward_std": 0.1549607552587986,
      "rewards/accuracy_reward": 0.6571800261735916,
      "rewards/format_reward": 0.984375,
      "step": 486
    },
    {
      "completion_length": 84.609375,
      "epoch": 2.223744292237443,
      "grad_norm": 2.6514155864715576,
      "kl": 0.127685546875,
      "learning_rate": 7.776255707762557e-07,
      "loss": 0.0051,
      "reward": 1.6367188096046448,
      "reward_std": 0.24639248847961426,
      "rewards/accuracy_reward": 0.64453125,
      "rewards/format_reward": 0.9921875,
      "step": 487
    },
    {
      "completion_length": 67.40625,
      "epoch": 2.228310502283105,
      "grad_norm": 2.794630289077759,
      "kl": 0.130859375,
      "learning_rate": 7.771689497716894e-07,
      "loss": 0.0052,
      "reward": 1.7186384201049805,
      "reward_std": 0.18474777042865753,
      "rewards/accuracy_reward": 0.7264508605003357,
      "rewards/format_reward": 0.9921875,
      "step": 488
    },
    {
      "completion_length": 89.7578125,
      "epoch": 2.232876712328767,
      "grad_norm": 3.7008538246154785,
      "kl": 0.115478515625,
      "learning_rate": 7.767123287671233e-07,
      "loss": 0.0046,
      "reward": 1.6265625953674316,
      "reward_std": 0.28076815605163574,
      "rewards/accuracy_reward": 0.6656249761581421,
      "rewards/format_reward": 0.9609375,
      "step": 489
    },
    {
      "completion_length": 66.0,
      "epoch": 2.237442922374429,
      "grad_norm": 5.58867883682251,
      "kl": 0.19970703125,
      "learning_rate": 7.76255707762557e-07,
      "loss": 0.008,
      "reward": 1.714453101158142,
      "reward_std": 0.2889961302280426,
      "rewards/accuracy_reward": 0.7378906309604645,
      "rewards/format_reward": 0.9765625,
      "step": 490
    },
    {
      "completion_length": 71.7265625,
      "epoch": 2.2420091324200913,
      "grad_norm": 2.161491870880127,
      "kl": 0.137451171875,
      "learning_rate": 7.757990867579909e-07,
      "loss": 0.0055,
      "reward": 1.5816146731376648,
      "reward_std": 0.19107923656702042,
      "rewards/accuracy_reward": 0.5894270539283752,
      "rewards/format_reward": 0.9921875,
      "step": 491
    },
    {
      "completion_length": 79.7734375,
      "epoch": 2.2465753424657535,
      "grad_norm": 5.536935329437256,
      "kl": 0.09423828125,
      "learning_rate": 7.753424657534247e-07,
      "loss": 0.0038,
      "reward": 1.5820313096046448,
      "reward_std": 0.23220208287239075,
      "rewards/accuracy_reward": 0.5976562649011612,
      "rewards/format_reward": 0.984375,
      "step": 492
    },
    {
      "completion_length": 105.1640625,
      "epoch": 2.2511415525114153,
      "grad_norm": 2.601624011993408,
      "kl": 0.102294921875,
      "learning_rate": 7.748858447488584e-07,
      "loss": 0.0041,
      "reward": 1.7345969676971436,
      "reward_std": 0.1583983302116394,
      "rewards/accuracy_reward": 0.7424094080924988,
      "rewards/format_reward": 0.9921875,
      "step": 493
    },
    {
      "completion_length": 76.2578125,
      "epoch": 2.2557077625570776,
      "grad_norm": 3.9378445148468018,
      "kl": 0.14208984375,
      "learning_rate": 7.744292237442922e-07,
      "loss": 0.0057,
      "reward": 1.6885417103767395,
      "reward_std": 0.24918173253536224,
      "rewards/accuracy_reward": 0.7197916507720947,
      "rewards/format_reward": 0.96875,
      "step": 494
    },
    {
      "completion_length": 69.578125,
      "epoch": 2.26027397260274,
      "grad_norm": 9.057132720947266,
      "kl": 0.1142578125,
      "learning_rate": 7.73972602739726e-07,
      "loss": 0.0046,
      "reward": 1.7063058018684387,
      "reward_std": 0.18320050090551376,
      "rewards/accuracy_reward": 0.7141183018684387,
      "rewards/format_reward": 0.9921875,
      "step": 495
    },
    {
      "completion_length": 58.1640625,
      "epoch": 2.2648401826484017,
      "grad_norm": 3.844257116317749,
      "kl": 0.14892578125,
      "learning_rate": 7.735159817351597e-07,
      "loss": 0.006,
      "reward": 1.5278646349906921,
      "reward_std": 0.2481004297733307,
      "rewards/accuracy_reward": 0.5278645753860474,
      "rewards/format_reward": 1.0,
      "step": 496
    },
    {
      "completion_length": 88.9140625,
      "epoch": 2.269406392694064,
      "grad_norm": 8.18749713897705,
      "kl": 0.098876953125,
      "learning_rate": 7.730593607305936e-07,
      "loss": 0.004,
      "reward": 1.749678373336792,
      "reward_std": 0.16354048997163773,
      "rewards/accuracy_reward": 0.7496782541275024,
      "rewards/format_reward": 1.0,
      "step": 497
    },
    {
      "completion_length": 93.5078125,
      "epoch": 2.2739726027397262,
      "grad_norm": 2.6278417110443115,
      "kl": 0.077392578125,
      "learning_rate": 7.726027397260274e-07,
      "loss": 0.0031,
      "reward": 1.765897810459137,
      "reward_std": 0.21192234754562378,
      "rewards/accuracy_reward": 0.8049602508544922,
      "rewards/format_reward": 0.9609375,
      "step": 498
    },
    {
      "completion_length": 72.2421875,
      "epoch": 2.278538812785388,
      "grad_norm": 2.370776414871216,
      "kl": 0.112548828125,
      "learning_rate": 7.721461187214611e-07,
      "loss": 0.0045,
      "reward": 1.6037201881408691,
      "reward_std": 0.22106194496154785,
      "rewards/accuracy_reward": 0.6193452179431915,
      "rewards/format_reward": 0.984375,
      "step": 499
    },
    {
      "completion_length": 57.5703125,
      "epoch": 2.2831050228310503,
      "grad_norm": 27.69339942932129,
      "kl": 0.10791015625,
      "learning_rate": 7.71689497716895e-07,
      "loss": 0.0043,
      "reward": 1.6528646349906921,
      "reward_std": 0.2068563476204872,
      "rewards/accuracy_reward": 0.652864545583725,
      "rewards/format_reward": 1.0,
      "step": 500
    },
    {
      "completion_length": 74.9296875,
      "epoch": 2.287671232876712,
      "grad_norm": 3.454657793045044,
      "kl": 0.101318359375,
      "learning_rate": 7.712328767123287e-07,
      "loss": 0.0041,
      "reward": 1.5783380270004272,
      "reward_std": 0.300783634185791,
      "rewards/accuracy_reward": 0.6017755568027496,
      "rewards/format_reward": 0.9765625,
      "step": 501
    },
    {
      "completion_length": 73.5,
      "epoch": 2.2922374429223744,
      "grad_norm": 2.893542766571045,
      "kl": 0.117919921875,
      "learning_rate": 7.707762557077625e-07,
      "loss": 0.0047,
      "reward": 1.6714488863945007,
      "reward_std": 0.23431023210287094,
      "rewards/accuracy_reward": 0.6948863565921783,
      "rewards/format_reward": 0.9765625,
      "step": 502
    },
    {
      "completion_length": 90.9921875,
      "epoch": 2.2968036529680367,
      "grad_norm": 4.334531307220459,
      "kl": 0.075439453125,
      "learning_rate": 7.703196347031963e-07,
      "loss": 0.003,
      "reward": 1.6945313215255737,
      "reward_std": 0.22123288363218307,
      "rewards/accuracy_reward": 0.7179687321186066,
      "rewards/format_reward": 0.9765625,
      "step": 503
    },
    {
      "completion_length": 89.3984375,
      "epoch": 2.3013698630136985,
      "grad_norm": 8.32867431640625,
      "kl": 0.140869140625,
      "learning_rate": 7.6986301369863e-07,
      "loss": 0.0056,
      "reward": 1.6440104246139526,
      "reward_std": 0.12583958357572556,
      "rewards/accuracy_reward": 0.644010454416275,
      "rewards/format_reward": 1.0,
      "step": 504
    },
    {
      "completion_length": 78.046875,
      "epoch": 2.3059360730593608,
      "grad_norm": 3.375293493270874,
      "kl": 0.1474609375,
      "learning_rate": 7.69406392694064e-07,
      "loss": 0.0059,
      "reward": 1.7735260128974915,
      "reward_std": 0.1295642852783203,
      "rewards/accuracy_reward": 0.7813384830951691,
      "rewards/format_reward": 0.9921875,
      "step": 505
    },
    {
      "completion_length": 73.734375,
      "epoch": 2.3105022831050226,
      "grad_norm": 4.356123924255371,
      "kl": 0.114501953125,
      "learning_rate": 7.689497716894977e-07,
      "loss": 0.0046,
      "reward": 1.7072916626930237,
      "reward_std": 0.1626647561788559,
      "rewards/accuracy_reward": 0.7072916328907013,
      "rewards/format_reward": 1.0,
      "step": 506
    },
    {
      "completion_length": 66.375,
      "epoch": 2.315068493150685,
      "grad_norm": 2.643831253051758,
      "kl": 0.1552734375,
      "learning_rate": 7.684931506849314e-07,
      "loss": 0.0062,
      "reward": 1.7758206725120544,
      "reward_std": 0.17595528066158295,
      "rewards/accuracy_reward": 0.7914457023143768,
      "rewards/format_reward": 0.984375,
      "step": 507
    },
    {
      "completion_length": 65.109375,
      "epoch": 2.319634703196347,
      "grad_norm": 6.30082368850708,
      "kl": 0.15869140625,
      "learning_rate": 7.680365296803653e-07,
      "loss": 0.0064,
      "reward": 1.5708190202713013,
      "reward_std": 0.1858999952673912,
      "rewards/accuracy_reward": 0.5708190500736237,
      "rewards/format_reward": 1.0,
      "step": 508
    },
    {
      "completion_length": 76.015625,
      "epoch": 2.324200913242009,
      "grad_norm": 3.34980845451355,
      "kl": 0.1240234375,
      "learning_rate": 7.67579908675799e-07,
      "loss": 0.005,
      "reward": 1.6061203479766846,
      "reward_std": 0.2387639731168747,
      "rewards/accuracy_reward": 0.6295577883720398,
      "rewards/format_reward": 0.9765625,
      "step": 509
    },
    {
      "completion_length": 69.5546875,
      "epoch": 2.328767123287671,
      "grad_norm": 2.7794432640075684,
      "kl": 0.108642578125,
      "learning_rate": 7.671232876712328e-07,
      "loss": 0.0043,
      "reward": 1.628348171710968,
      "reward_std": 0.22845705598592758,
      "rewards/accuracy_reward": 0.6361607015132904,
      "rewards/format_reward": 0.9921875,
      "step": 510
    },
    {
      "completion_length": 77.1171875,
      "epoch": 2.3333333333333335,
      "grad_norm": 1.9683984518051147,
      "kl": 0.10400390625,
      "learning_rate": 7.666666666666667e-07,
      "loss": 0.0041,
      "reward": 1.7567708492279053,
      "reward_std": 0.1648455262184143,
      "rewards/accuracy_reward": 0.7723957598209381,
      "rewards/format_reward": 0.984375,
      "step": 511
    },
    {
      "completion_length": 91.765625,
      "epoch": 2.3378995433789953,
      "grad_norm": 7.019916534423828,
      "kl": 0.120849609375,
      "learning_rate": 7.662100456621004e-07,
      "loss": 0.0048,
      "reward": 1.7004202008247375,
      "reward_std": 0.13108576089143753,
      "rewards/accuracy_reward": 0.708232581615448,
      "rewards/format_reward": 0.9921875,
      "step": 512
    },
    {
      "completion_length": 80.7421875,
      "epoch": 2.3424657534246576,
      "grad_norm": 2.83266282081604,
      "kl": 0.13134765625,
      "learning_rate": 7.657534246575343e-07,
      "loss": 0.0053,
      "reward": 1.658984363079071,
      "reward_std": 0.20922444760799408,
      "rewards/accuracy_reward": 0.658984363079071,
      "rewards/format_reward": 1.0,
      "step": 513
    },
    {
      "completion_length": 71.234375,
      "epoch": 2.34703196347032,
      "grad_norm": 1.9871636629104614,
      "kl": 0.12646484375,
      "learning_rate": 7.65296803652968e-07,
      "loss": 0.005,
      "reward": 1.725781261920929,
      "reward_std": 0.10922157764434814,
      "rewards/accuracy_reward": 0.7257812023162842,
      "rewards/format_reward": 1.0,
      "step": 514
    },
    {
      "completion_length": 73.7265625,
      "epoch": 2.3515981735159817,
      "grad_norm": 2.359335422515869,
      "kl": 0.11279296875,
      "learning_rate": 7.648401826484017e-07,
      "loss": 0.0045,
      "reward": 1.772805094718933,
      "reward_std": 0.16348526254296303,
      "rewards/accuracy_reward": 0.7806175947189331,
      "rewards/format_reward": 0.9921875,
      "step": 515
    },
    {
      "completion_length": 60.3828125,
      "epoch": 2.356164383561644,
      "grad_norm": 5.7438812255859375,
      "kl": 0.14208984375,
      "learning_rate": 7.643835616438356e-07,
      "loss": 0.0057,
      "reward": 1.751901626586914,
      "reward_std": 0.17701375484466553,
      "rewards/accuracy_reward": 0.7597140967845917,
      "rewards/format_reward": 0.9921875,
      "step": 516
    },
    {
      "completion_length": 89.5703125,
      "epoch": 2.3607305936073057,
      "grad_norm": 4.794968128204346,
      "kl": 0.0986328125,
      "learning_rate": 7.639269406392693e-07,
      "loss": 0.0039,
      "reward": 1.7277343273162842,
      "reward_std": 0.1568085253238678,
      "rewards/accuracy_reward": 0.7355467975139618,
      "rewards/format_reward": 0.9921875,
      "step": 517
    },
    {
      "completion_length": 72.1015625,
      "epoch": 2.365296803652968,
      "grad_norm": 3.052950143814087,
      "kl": 0.10009765625,
      "learning_rate": 7.634703196347032e-07,
      "loss": 0.004,
      "reward": 1.7309381365776062,
      "reward_std": 0.13279738277196884,
      "rewards/accuracy_reward": 0.7387505769729614,
      "rewards/format_reward": 0.9921875,
      "step": 518
    },
    {
      "completion_length": 96.421875,
      "epoch": 2.3698630136986303,
      "grad_norm": 1.318949818611145,
      "kl": 0.08251953125,
      "learning_rate": 7.63013698630137e-07,
      "loss": 0.0033,
      "reward": 1.78125,
      "reward_std": 0.18937532603740692,
      "rewards/accuracy_reward": 0.8203124403953552,
      "rewards/format_reward": 0.9609375,
      "step": 519
    },
    {
      "completion_length": 76.15625,
      "epoch": 2.374429223744292,
      "grad_norm": 2.7549917697906494,
      "kl": 0.11669921875,
      "learning_rate": 7.625570776255707e-07,
      "loss": 0.0047,
      "reward": 1.7448863983154297,
      "reward_std": 0.23332231491804123,
      "rewards/accuracy_reward": 0.7683238387107849,
      "rewards/format_reward": 0.9765625,
      "step": 520
    },
    {
      "completion_length": 58.6328125,
      "epoch": 2.3789954337899544,
      "grad_norm": 3.747992753982544,
      "kl": 0.1669921875,
      "learning_rate": 7.621004566210046e-07,
      "loss": 0.0067,
      "reward": 1.6135417222976685,
      "reward_std": 0.16909091174602509,
      "rewards/accuracy_reward": 0.6135416924953461,
      "rewards/format_reward": 1.0,
      "step": 521
    },
    {
      "completion_length": 59.4296875,
      "epoch": 2.383561643835616,
      "grad_norm": 5.62481164932251,
      "kl": 0.14990234375,
      "learning_rate": 7.616438356164383e-07,
      "loss": 0.006,
      "reward": 1.5095030069351196,
      "reward_std": 0.3347364068031311,
      "rewards/accuracy_reward": 0.5407529026269913,
      "rewards/format_reward": 0.96875,
      "step": 522
    },
    {
      "completion_length": 73.7109375,
      "epoch": 2.3881278538812785,
      "grad_norm": 2.828101634979248,
      "kl": 0.15869140625,
      "learning_rate": 7.61187214611872e-07,
      "loss": 0.0063,
      "reward": 1.6775281429290771,
      "reward_std": 0.15744981169700623,
      "rewards/accuracy_reward": 0.6775281727313995,
      "rewards/format_reward": 1.0,
      "step": 523
    },
    {
      "completion_length": 79.359375,
      "epoch": 2.3926940639269407,
      "grad_norm": 1.9861787557601929,
      "kl": 0.1162109375,
      "learning_rate": 7.607305936073059e-07,
      "loss": 0.0046,
      "reward": 1.645312488079071,
      "reward_std": 0.12342093884944916,
      "rewards/accuracy_reward": 0.6453125178813934,
      "rewards/format_reward": 1.0,
      "step": 524
    },
    {
      "completion_length": 78.734375,
      "epoch": 2.3972602739726026,
      "grad_norm": 2.850097179412842,
      "kl": 0.112548828125,
      "learning_rate": 7.602739726027397e-07,
      "loss": 0.0045,
      "reward": 1.6301960349082947,
      "reward_std": 0.18981467187404633,
      "rewards/accuracy_reward": 0.6458209455013275,
      "rewards/format_reward": 0.984375,
      "step": 525
    },
    {
      "completion_length": 62.109375,
      "epoch": 2.401826484018265,
      "grad_norm": 2.852185010910034,
      "kl": 0.16552734375,
      "learning_rate": 7.598173515981735e-07,
      "loss": 0.0066,
      "reward": 1.5752604007720947,
      "reward_std": 0.20407412946224213,
      "rewards/accuracy_reward": 0.5908854007720947,
      "rewards/format_reward": 0.984375,
      "step": 526
    },
    {
      "completion_length": 68.4765625,
      "epoch": 2.406392694063927,
      "grad_norm": 2.7563745975494385,
      "kl": 0.1240234375,
      "learning_rate": 7.593607305936073e-07,
      "loss": 0.005,
      "reward": 1.646093726158142,
      "reward_std": 0.255212739109993,
      "rewards/accuracy_reward": 0.6695312261581421,
      "rewards/format_reward": 0.9765625,
      "step": 527
    },
    {
      "completion_length": 83.1953125,
      "epoch": 2.410958904109589,
      "grad_norm": 3.005842924118042,
      "kl": 0.11328125,
      "learning_rate": 7.58904109589041e-07,
      "loss": 0.0045,
      "reward": 1.4694010615348816,
      "reward_std": 0.34128230810165405,
      "rewards/accuracy_reward": 0.516276016831398,
      "rewards/format_reward": 0.953125,
      "step": 528
    },
    {
      "completion_length": 79.171875,
      "epoch": 2.415525114155251,
      "grad_norm": 10.589370727539062,
      "kl": 0.40234375,
      "learning_rate": 7.584474885844749e-07,
      "loss": 0.0161,
      "reward": 1.7232915163040161,
      "reward_std": 0.16085164994001389,
      "rewards/accuracy_reward": 0.7311040163040161,
      "rewards/format_reward": 0.9921875,
      "step": 529
    },
    {
      "completion_length": 75.6484375,
      "epoch": 2.4200913242009134,
      "grad_norm": 2.8637847900390625,
      "kl": 0.1416015625,
      "learning_rate": 7.579908675799086e-07,
      "loss": 0.0057,
      "reward": 1.5583333373069763,
      "reward_std": 0.24188697338104248,
      "rewards/accuracy_reward": 0.5661458522081375,
      "rewards/format_reward": 0.9921875,
      "step": 530
    },
    {
      "completion_length": 81.7421875,
      "epoch": 2.4246575342465753,
      "grad_norm": 2.9744412899017334,
      "kl": 0.1015625,
      "learning_rate": 7.575342465753424e-07,
      "loss": 0.0041,
      "reward": 1.6184896230697632,
      "reward_std": 0.22770510613918304,
      "rewards/accuracy_reward": 0.6341145932674408,
      "rewards/format_reward": 0.984375,
      "step": 531
    },
    {
      "completion_length": 68.96875,
      "epoch": 2.4292237442922375,
      "grad_norm": 5.560794830322266,
      "kl": 0.1513671875,
      "learning_rate": 7.570776255707763e-07,
      "loss": 0.006,
      "reward": 1.6268364787101746,
      "reward_std": 0.25046712160110474,
      "rewards/accuracy_reward": 0.6346489191055298,
      "rewards/format_reward": 0.9921875,
      "step": 532
    },
    {
      "completion_length": 83.734375,
      "epoch": 2.4337899543378994,
      "grad_norm": 1.856094241142273,
      "kl": 0.103271484375,
      "learning_rate": 7.5662100456621e-07,
      "loss": 0.0041,
      "reward": 1.719410002231598,
      "reward_std": 0.18102595210075378,
      "rewards/accuracy_reward": 0.7272224426269531,
      "rewards/format_reward": 0.9921875,
      "step": 533
    },
    {
      "completion_length": 82.9765625,
      "epoch": 2.4383561643835616,
      "grad_norm": 1.8449293375015259,
      "kl": 0.12255859375,
      "learning_rate": 7.561643835616438e-07,
      "loss": 0.0049,
      "reward": 1.7729809880256653,
      "reward_std": 0.10142140835523605,
      "rewards/accuracy_reward": 0.7807934284210205,
      "rewards/format_reward": 0.9921875,
      "step": 534
    },
    {
      "completion_length": 76.6953125,
      "epoch": 2.442922374429224,
      "grad_norm": 4.276492595672607,
      "kl": 0.13427734375,
      "learning_rate": 7.557077625570776e-07,
      "loss": 0.0054,
      "reward": 1.584375023841858,
      "reward_std": 0.2730569392442703,
      "rewards/accuracy_reward": 0.6078124940395355,
      "rewards/format_reward": 0.9765625,
      "step": 535
    },
    {
      "completion_length": 84.6171875,
      "epoch": 2.4474885844748857,
      "grad_norm": 7.224926471710205,
      "kl": 0.120361328125,
      "learning_rate": 7.552511415525113e-07,
      "loss": 0.0048,
      "reward": 1.7129571437835693,
      "reward_std": 0.16479767858982086,
      "rewards/accuracy_reward": 0.7207695543766022,
      "rewards/format_reward": 0.9921875,
      "step": 536
    },
    {
      "completion_length": 81.859375,
      "epoch": 2.452054794520548,
      "grad_norm": 2.366180181503296,
      "kl": 0.10302734375,
      "learning_rate": 7.547945205479452e-07,
      "loss": 0.0041,
      "reward": 1.7357105612754822,
      "reward_std": 0.1823057383298874,
      "rewards/accuracy_reward": 0.7747730314731598,
      "rewards/format_reward": 0.9609375,
      "step": 537
    },
    {
      "completion_length": 92.140625,
      "epoch": 2.45662100456621,
      "grad_norm": 2.225709915161133,
      "kl": 0.17041015625,
      "learning_rate": 7.54337899543379e-07,
      "loss": 0.0068,
      "reward": 1.7205729484558105,
      "reward_std": 0.136265367269516,
      "rewards/accuracy_reward": 0.7205729186534882,
      "rewards/format_reward": 1.0,
      "step": 538
    },
    {
      "completion_length": 53.6796875,
      "epoch": 2.461187214611872,
      "grad_norm": 2.991917133331299,
      "kl": 0.181640625,
      "learning_rate": 7.538812785388127e-07,
      "loss": 0.0073,
      "reward": 1.531180500984192,
      "reward_std": 0.2481382116675377,
      "rewards/accuracy_reward": 0.5389930307865143,
      "rewards/format_reward": 0.9921875,
      "step": 539
    },
    {
      "completion_length": 72.734375,
      "epoch": 2.4657534246575343,
      "grad_norm": 5.530900001525879,
      "kl": 0.10205078125,
      "learning_rate": 7.534246575342466e-07,
      "loss": 0.0041,
      "reward": 1.705208420753479,
      "reward_std": 0.24594328552484512,
      "rewards/accuracy_reward": 0.7286458015441895,
      "rewards/format_reward": 0.9765625,
      "step": 540
    },
    {
      "completion_length": 69.7734375,
      "epoch": 2.470319634703196,
      "grad_norm": 2.618407964706421,
      "kl": 0.156982421875,
      "learning_rate": 7.529680365296803e-07,
      "loss": 0.0063,
      "reward": 1.7212890982627869,
      "reward_std": 0.19607724994421005,
      "rewards/accuracy_reward": 0.7369140684604645,
      "rewards/format_reward": 0.984375,
      "step": 541
    },
    {
      "completion_length": 66.2109375,
      "epoch": 2.4748858447488584,
      "grad_norm": 2.0617170333862305,
      "kl": 0.14208984375,
      "learning_rate": 7.525114155251141e-07,
      "loss": 0.0057,
      "reward": 1.6648437976837158,
      "reward_std": 0.1707550622522831,
      "rewards/accuracy_reward": 0.672656238079071,
      "rewards/format_reward": 0.9921875,
      "step": 542
    },
    {
      "completion_length": 77.1171875,
      "epoch": 2.4794520547945207,
      "grad_norm": 3.923844814300537,
      "kl": 0.188232421875,
      "learning_rate": 7.520547945205479e-07,
      "loss": 0.0076,
      "reward": 1.7651843428611755,
      "reward_std": 0.18746302276849747,
      "rewards/accuracy_reward": 0.7729967534542084,
      "rewards/format_reward": 0.9921875,
      "step": 543
    },
    {
      "completion_length": 83.75,
      "epoch": 2.4840182648401825,
      "grad_norm": 12.723535537719727,
      "kl": 0.096923828125,
      "learning_rate": 7.515981735159816e-07,
      "loss": 0.0039,
      "reward": 1.7339038252830505,
      "reward_std": 0.16746822372078896,
      "rewards/accuracy_reward": 0.7495287358760834,
      "rewards/format_reward": 0.984375,
      "step": 544
    },
    {
      "completion_length": 63.9921875,
      "epoch": 2.4885844748858448,
      "grad_norm": 2.9950413703918457,
      "kl": 0.130615234375,
      "learning_rate": 7.511415525114156e-07,
      "loss": 0.0052,
      "reward": 1.6598958373069763,
      "reward_std": 0.2278856635093689,
      "rewards/accuracy_reward": 0.6833333075046539,
      "rewards/format_reward": 0.9765625,
      "step": 545
    },
    {
      "completion_length": 82.7890625,
      "epoch": 2.493150684931507,
      "grad_norm": 2.3096249103546143,
      "kl": 0.090576171875,
      "learning_rate": 7.506849315068493e-07,
      "loss": 0.0036,
      "reward": 1.7203125953674316,
      "reward_std": 0.20905159413814545,
      "rewards/accuracy_reward": 0.7281249761581421,
      "rewards/format_reward": 0.9921875,
      "step": 546
    },
    {
      "completion_length": 80.84375,
      "epoch": 2.497716894977169,
      "grad_norm": 3.3282742500305176,
      "kl": 0.093994140625,
      "learning_rate": 7.50228310502283e-07,
      "loss": 0.0038,
      "reward": 1.543817937374115,
      "reward_std": 0.22718056291341782,
      "rewards/accuracy_reward": 0.5594429075717926,
      "rewards/format_reward": 0.984375,
      "step": 547
    },
    {
      "completion_length": 106.3046875,
      "epoch": 2.502283105022831,
      "grad_norm": 5.616523742675781,
      "kl": 0.08251953125,
      "learning_rate": 7.497716894977169e-07,
      "loss": 0.0033,
      "reward": 1.6008946895599365,
      "reward_std": 0.2659634053707123,
      "rewards/accuracy_reward": 0.6321446299552917,
      "rewards/format_reward": 0.96875,
      "step": 548
    },
    {
      "completion_length": 90.3203125,
      "epoch": 2.506849315068493,
      "grad_norm": 5.4013166427612305,
      "kl": 0.0994873046875,
      "learning_rate": 7.493150684931506e-07,
      "loss": 0.004,
      "reward": 1.7019531726837158,
      "reward_std": 0.17125242203474045,
      "rewards/accuracy_reward": 0.7175780534744263,
      "rewards/format_reward": 0.984375,
      "step": 549
    },
    {
      "completion_length": 86.5859375,
      "epoch": 2.5114155251141552,
      "grad_norm": 2.318875789642334,
      "kl": 0.134765625,
      "learning_rate": 7.488584474885844e-07,
      "loss": 0.0054,
      "reward": 1.6225537061691284,
      "reward_std": 0.2404860332608223,
      "rewards/accuracy_reward": 0.645991176366806,
      "rewards/format_reward": 0.9765625,
      "step": 550
    },
    {
      "completion_length": 92.1015625,
      "epoch": 2.5159817351598175,
      "grad_norm": 6.498478412628174,
      "kl": 0.092041015625,
      "learning_rate": 7.484018264840183e-07,
      "loss": 0.0037,
      "reward": 1.6539062857627869,
      "reward_std": 0.20888085663318634,
      "rewards/accuracy_reward": 0.6695312559604645,
      "rewards/format_reward": 0.984375,
      "step": 551
    },
    {
      "completion_length": 98.4375,
      "epoch": 2.5205479452054793,
      "grad_norm": 3.217737913131714,
      "kl": 0.0743408203125,
      "learning_rate": 7.47945205479452e-07,
      "loss": 0.003,
      "reward": 1.8604166507720947,
      "reward_std": 0.10430474206805229,
      "rewards/accuracy_reward": 0.8682291209697723,
      "rewards/format_reward": 0.9921875,
      "step": 552
    },
    {
      "completion_length": 89.34375,
      "epoch": 2.5251141552511416,
      "grad_norm": 1.3329360485076904,
      "kl": 0.12451171875,
      "learning_rate": 7.474885844748859e-07,
      "loss": 0.005,
      "reward": 1.842187523841858,
      "reward_std": 0.09198738168925047,
      "rewards/accuracy_reward": 0.8421874642372131,
      "rewards/format_reward": 1.0,
      "step": 553
    },
    {
      "completion_length": 89.234375,
      "epoch": 2.5296803652968034,
      "grad_norm": 3.4690394401550293,
      "kl": 0.1015625,
      "learning_rate": 7.470319634703196e-07,
      "loss": 0.0041,
      "reward": 1.7106770873069763,
      "reward_std": 0.19660182297229767,
      "rewards/accuracy_reward": 0.7184895873069763,
      "rewards/format_reward": 0.9921875,
      "step": 554
    },
    {
      "completion_length": 82.5859375,
      "epoch": 2.5342465753424657,
      "grad_norm": 3.373579740524292,
      "kl": 0.113037109375,
      "learning_rate": 7.465753424657533e-07,
      "loss": 0.0045,
      "reward": 1.5816163420677185,
      "reward_std": 0.20616846531629562,
      "rewards/accuracy_reward": 0.5816163718700409,
      "rewards/format_reward": 1.0,
      "step": 555
    },
    {
      "completion_length": 93.5078125,
      "epoch": 2.538812785388128,
      "grad_norm": 3.093045473098755,
      "kl": 0.1044921875,
      "learning_rate": 7.461187214611872e-07,
      "loss": 0.0042,
      "reward": 1.608422875404358,
      "reward_std": 0.21706774830818176,
      "rewards/accuracy_reward": 0.6240477561950684,
      "rewards/format_reward": 0.984375,
      "step": 556
    },
    {
      "completion_length": 97.0703125,
      "epoch": 2.54337899543379,
      "grad_norm": 2.3710570335388184,
      "kl": 0.095703125,
      "learning_rate": 7.456621004566209e-07,
      "loss": 0.0038,
      "reward": 1.800067663192749,
      "reward_std": 0.09500321745872498,
      "rewards/accuracy_reward": 0.8000677227973938,
      "rewards/format_reward": 1.0,
      "step": 557
    },
    {
      "completion_length": 71.671875,
      "epoch": 2.547945205479452,
      "grad_norm": 2.5310826301574707,
      "kl": 0.108154296875,
      "learning_rate": 7.452054794520548e-07,
      "loss": 0.0043,
      "reward": 1.5969815254211426,
      "reward_std": 0.2227228805422783,
      "rewards/accuracy_reward": 0.5969814956188202,
      "rewards/format_reward": 1.0,
      "step": 558
    },
    {
      "completion_length": 78.15625,
      "epoch": 2.5525114155251143,
      "grad_norm": 2.1528046131134033,
      "kl": 0.12451171875,
      "learning_rate": 7.447488584474886e-07,
      "loss": 0.005,
      "reward": 1.7443639636039734,
      "reward_std": 0.10288457944989204,
      "rewards/accuracy_reward": 0.752176433801651,
      "rewards/format_reward": 0.9921875,
      "step": 559
    },
    {
      "completion_length": 84.546875,
      "epoch": 2.557077625570776,
      "grad_norm": 2.656460762023926,
      "kl": 0.1171875,
      "learning_rate": 7.442922374429223e-07,
      "loss": 0.0047,
      "reward": 1.7289806604385376,
      "reward_std": 0.17229026556015015,
      "rewards/accuracy_reward": 0.7289806008338928,
      "rewards/format_reward": 1.0,
      "step": 560
    },
    {
      "completion_length": 81.8984375,
      "epoch": 2.5616438356164384,
      "grad_norm": 3.7874462604522705,
      "kl": 0.12548828125,
      "learning_rate": 7.438356164383562e-07,
      "loss": 0.005,
      "reward": 1.7026662230491638,
      "reward_std": 0.2061307728290558,
      "rewards/accuracy_reward": 0.7104786336421967,
      "rewards/format_reward": 0.9921875,
      "step": 561
    },
    {
      "completion_length": 69.5546875,
      "epoch": 2.5662100456621006,
      "grad_norm": 2.563668966293335,
      "kl": 0.1201171875,
      "learning_rate": 7.433789954337899e-07,
      "loss": 0.0048,
      "reward": 1.5755208134651184,
      "reward_std": 0.207001730799675,
      "rewards/accuracy_reward": 0.5833333730697632,
      "rewards/format_reward": 0.9921875,
      "step": 562
    },
    {
      "completion_length": 76.96875,
      "epoch": 2.5707762557077625,
      "grad_norm": 2.503657817840576,
      "kl": 0.098876953125,
      "learning_rate": 7.429223744292236e-07,
      "loss": 0.004,
      "reward": 1.7036305665969849,
      "reward_std": 0.16530893370509148,
      "rewards/accuracy_reward": 0.7036304771900177,
      "rewards/format_reward": 1.0,
      "step": 563
    },
    {
      "completion_length": 86.0078125,
      "epoch": 2.5753424657534247,
      "grad_norm": 4.6145734786987305,
      "kl": 0.087646484375,
      "learning_rate": 7.424657534246575e-07,
      "loss": 0.0035,
      "reward": 1.5936384201049805,
      "reward_std": 0.24451126903295517,
      "rewards/accuracy_reward": 0.6248884201049805,
      "rewards/format_reward": 0.96875,
      "step": 564
    },
    {
      "completion_length": 91.1171875,
      "epoch": 2.5799086757990866,
      "grad_norm": 3.2497313022613525,
      "kl": 0.116455078125,
      "learning_rate": 7.420091324200913e-07,
      "loss": 0.0047,
      "reward": 1.555757462978363,
      "reward_std": 0.2567907050251961,
      "rewards/accuracy_reward": 0.5870074331760406,
      "rewards/format_reward": 0.96875,
      "step": 565
    },
    {
      "completion_length": 89.8984375,
      "epoch": 2.584474885844749,
      "grad_norm": 2.177954912185669,
      "kl": 0.073974609375,
      "learning_rate": 7.415525114155251e-07,
      "loss": 0.003,
      "reward": 1.7282168865203857,
      "reward_std": 0.18810292333364487,
      "rewards/accuracy_reward": 0.7516542971134186,
      "rewards/format_reward": 0.9765625,
      "step": 566
    },
    {
      "completion_length": 59.671875,
      "epoch": 2.589041095890411,
      "grad_norm": 2.342167854309082,
      "kl": 0.12646484375,
      "learning_rate": 7.410958904109589e-07,
      "loss": 0.0051,
      "reward": 1.5524925589561462,
      "reward_std": 0.19321630895137787,
      "rewards/accuracy_reward": 0.5524925589561462,
      "rewards/format_reward": 1.0,
      "step": 567
    },
    {
      "completion_length": 81.0390625,
      "epoch": 2.593607305936073,
      "grad_norm": 25.185625076293945,
      "kl": 0.099609375,
      "learning_rate": 7.406392694063926e-07,
      "loss": 0.004,
      "reward": 1.685937523841858,
      "reward_std": 0.2477683424949646,
      "rewards/accuracy_reward": 0.6937499642372131,
      "rewards/format_reward": 0.9921875,
      "step": 568
    },
    {
      "completion_length": 86.671875,
      "epoch": 2.598173515981735,
      "grad_norm": 3.6345832347869873,
      "kl": 0.095458984375,
      "learning_rate": 7.401826484018265e-07,
      "loss": 0.0038,
      "reward": 1.6493847966194153,
      "reward_std": 0.20508087426424026,
      "rewards/accuracy_reward": 0.6493848264217377,
      "rewards/format_reward": 1.0,
      "step": 569
    },
    {
      "completion_length": 111.0,
      "epoch": 2.602739726027397,
      "grad_norm": 1.393759846687317,
      "kl": 0.06494140625,
      "learning_rate": 7.397260273972602e-07,
      "loss": 0.0026,
      "reward": 1.872656226158142,
      "reward_std": 0.10857155546545982,
      "rewards/accuracy_reward": 0.8960936367511749,
      "rewards/format_reward": 0.9765625,
      "step": 570
    },
    {
      "completion_length": 96.6875,
      "epoch": 2.6073059360730593,
      "grad_norm": 2.2273528575897217,
      "kl": 0.077880859375,
      "learning_rate": 7.39269406392694e-07,
      "loss": 0.0031,
      "reward": 1.8092340230941772,
      "reward_std": 0.10701124370098114,
      "rewards/accuracy_reward": 0.8170464634895325,
      "rewards/format_reward": 0.9921875,
      "step": 571
    },
    {
      "completion_length": 70.09375,
      "epoch": 2.6118721461187215,
      "grad_norm": 2.2020483016967773,
      "kl": 0.108154296875,
      "learning_rate": 7.388127853881279e-07,
      "loss": 0.0043,
      "reward": 1.5493839979171753,
      "reward_std": 0.19577539712190628,
      "rewards/accuracy_reward": 0.5493840277194977,
      "rewards/format_reward": 1.0,
      "step": 572
    },
    {
      "completion_length": 56.6875,
      "epoch": 2.616438356164384,
      "grad_norm": 8.005638122558594,
      "kl": 0.111572265625,
      "learning_rate": 7.383561643835616e-07,
      "loss": 0.0045,
      "reward": 1.457552194595337,
      "reward_std": 0.28487062454223633,
      "rewards/accuracy_reward": 0.45755207538604736,
      "rewards/format_reward": 1.0,
      "step": 573
    },
    {
      "completion_length": 67.3984375,
      "epoch": 2.6210045662100456,
      "grad_norm": 7.093777179718018,
      "kl": 0.1669921875,
      "learning_rate": 7.378995433789954e-07,
      "loss": 0.0067,
      "reward": 1.7374799847602844,
      "reward_std": 0.1953911855816841,
      "rewards/accuracy_reward": 0.745292454957962,
      "rewards/format_reward": 0.9921875,
      "step": 574
    },
    {
      "completion_length": 70.953125,
      "epoch": 2.625570776255708,
      "grad_norm": 2.916884183883667,
      "kl": 0.129150390625,
      "learning_rate": 7.374429223744292e-07,
      "loss": 0.0052,
      "reward": 1.5901844501495361,
      "reward_std": 0.2463785707950592,
      "rewards/accuracy_reward": 0.5901843905448914,
      "rewards/format_reward": 1.0,
      "step": 575
    },
    {
      "completion_length": 79.0859375,
      "epoch": 2.6301369863013697,
      "grad_norm": 7.461641311645508,
      "kl": 0.12890625,
      "learning_rate": 7.369863013698629e-07,
      "loss": 0.0052,
      "reward": 1.666406273841858,
      "reward_std": 0.24838291853666306,
      "rewards/accuracy_reward": 0.6742187440395355,
      "rewards/format_reward": 0.9921875,
      "step": 576
    },
    {
      "completion_length": 71.1796875,
      "epoch": 2.634703196347032,
      "grad_norm": 2.474581718444824,
      "kl": 0.125244140625,
      "learning_rate": 7.365296803652968e-07,
      "loss": 0.005,
      "reward": 1.6498884558677673,
      "reward_std": 0.15810929238796234,
      "rewards/accuracy_reward": 0.6498883664608002,
      "rewards/format_reward": 1.0,
      "step": 577
    },
    {
      "completion_length": 93.484375,
      "epoch": 2.6392694063926943,
      "grad_norm": 2.4049859046936035,
      "kl": 0.068359375,
      "learning_rate": 7.360730593607306e-07,
      "loss": 0.0027,
      "reward": 1.7192708849906921,
      "reward_std": 0.1887570172548294,
      "rewards/accuracy_reward": 0.742708295583725,
      "rewards/format_reward": 0.9765625,
      "step": 578
    },
    {
      "completion_length": 78.71875,
      "epoch": 2.643835616438356,
      "grad_norm": 12.206317901611328,
      "kl": 0.11962890625,
      "learning_rate": 7.356164383561643e-07,
      "loss": 0.0048,
      "reward": 1.6132813096046448,
      "reward_std": 0.19830159842967987,
      "rewards/accuracy_reward": 0.62890625,
      "rewards/format_reward": 0.984375,
      "step": 579
    },
    {
      "completion_length": 84.28125,
      "epoch": 2.6484018264840183,
      "grad_norm": 3.1371257305145264,
      "kl": 0.112060546875,
      "learning_rate": 7.351598173515982e-07,
      "loss": 0.0045,
      "reward": 1.7513157725334167,
      "reward_std": 0.13612205535173416,
      "rewards/accuracy_reward": 0.7513157725334167,
      "rewards/format_reward": 1.0,
      "step": 580
    },
    {
      "completion_length": 74.203125,
      "epoch": 2.65296803652968,
      "grad_norm": 4.9632792472839355,
      "kl": 0.21728515625,
      "learning_rate": 7.347031963470319e-07,
      "loss": 0.0087,
      "reward": 1.5989612340927124,
      "reward_std": 0.21190468221902847,
      "rewards/accuracy_reward": 0.5989611893892288,
      "rewards/format_reward": 1.0,
      "step": 581
    },
    {
      "completion_length": 56.5234375,
      "epoch": 2.6575342465753424,
      "grad_norm": 3.400489091873169,
      "kl": 0.1005859375,
      "learning_rate": 7.342465753424657e-07,
      "loss": 0.004,
      "reward": 1.5473958253860474,
      "reward_std": 0.194392129778862,
      "rewards/accuracy_reward": 0.5473958253860474,
      "rewards/format_reward": 1.0,
      "step": 582
    },
    {
      "completion_length": 56.2421875,
      "epoch": 2.6621004566210047,
      "grad_norm": 2.7040834426879883,
      "kl": 0.19140625,
      "learning_rate": 7.337899543378995e-07,
      "loss": 0.0077,
      "reward": 1.5410130023956299,
      "reward_std": 0.23740805685520172,
      "rewards/accuracy_reward": 0.5488254725933075,
      "rewards/format_reward": 0.9921875,
      "step": 583
    },
    {
      "completion_length": 103.8515625,
      "epoch": 2.6666666666666665,
      "grad_norm": 1.5765724182128906,
      "kl": 0.0721435546875,
      "learning_rate": 7.333333333333332e-07,
      "loss": 0.0029,
      "reward": 1.6796875,
      "reward_std": 0.134404756128788,
      "rewards/accuracy_reward": 0.6874999403953552,
      "rewards/format_reward": 0.9921875,
      "step": 584
    },
    {
      "completion_length": 74.1875,
      "epoch": 2.671232876712329,
      "grad_norm": 7.368083477020264,
      "kl": 0.115234375,
      "learning_rate": 7.328767123287672e-07,
      "loss": 0.0046,
      "reward": 1.6952009201049805,
      "reward_std": 0.23021817207336426,
      "rewards/accuracy_reward": 0.7030133903026581,
      "rewards/format_reward": 0.9921875,
      "step": 585
    },
    {
      "completion_length": 74.0703125,
      "epoch": 2.6757990867579906,
      "grad_norm": 4.19453239440918,
      "kl": 0.112548828125,
      "learning_rate": 7.324200913242009e-07,
      "loss": 0.0045,
      "reward": 1.7542436122894287,
      "reward_std": 0.16448176465928555,
      "rewards/accuracy_reward": 0.7620560824871063,
      "rewards/format_reward": 0.9921875,
      "step": 586
    },
    {
      "completion_length": 83.4453125,
      "epoch": 2.680365296803653,
      "grad_norm": 2.116567373275757,
      "kl": 0.1123046875,
      "learning_rate": 7.319634703196346e-07,
      "loss": 0.0045,
      "reward": 1.8114583492279053,
      "reward_std": 0.10706461034715176,
      "rewards/accuracy_reward": 0.8114582598209381,
      "rewards/format_reward": 1.0,
      "step": 587
    },
    {
      "completion_length": 74.1171875,
      "epoch": 2.684931506849315,
      "grad_norm": 3.840576648712158,
      "kl": 0.106689453125,
      "learning_rate": 7.315068493150685e-07,
      "loss": 0.0043,
      "reward": 1.5600694417953491,
      "reward_std": 0.20616772770881653,
      "rewards/accuracy_reward": 0.5600694417953491,
      "rewards/format_reward": 1.0,
      "step": 588
    },
    {
      "completion_length": 80.515625,
      "epoch": 2.6894977168949774,
      "grad_norm": 5.284098148345947,
      "kl": 0.10986328125,
      "learning_rate": 7.310502283105022e-07,
      "loss": 0.0044,
      "reward": 1.6452972888946533,
      "reward_std": 0.1739073097705841,
      "rewards/accuracy_reward": 0.6452972292900085,
      "rewards/format_reward": 1.0,
      "step": 589
    },
    {
      "completion_length": 80.0078125,
      "epoch": 2.6940639269406392,
      "grad_norm": 2.7230966091156006,
      "kl": 0.1328125,
      "learning_rate": 7.30593607305936e-07,
      "loss": 0.0053,
      "reward": 1.7089230418205261,
      "reward_std": 0.18773558735847473,
      "rewards/accuracy_reward": 0.7167355120182037,
      "rewards/format_reward": 0.9921875,
      "step": 590
    },
    {
      "completion_length": 65.5625,
      "epoch": 2.6986301369863015,
      "grad_norm": 3.291010618209839,
      "kl": 0.1162109375,
      "learning_rate": 7.301369863013699e-07,
      "loss": 0.0047,
      "reward": 1.646875023841858,
      "reward_std": 0.2957848533987999,
      "rewards/accuracy_reward": 0.6624999642372131,
      "rewards/format_reward": 0.984375,
      "step": 591
    },
    {
      "completion_length": 78.953125,
      "epoch": 2.7031963470319633,
      "grad_norm": 3.601032257080078,
      "kl": 0.126953125,
      "learning_rate": 7.296803652968036e-07,
      "loss": 0.0051,
      "reward": 1.6582031846046448,
      "reward_std": 0.2431168407201767,
      "rewards/accuracy_reward": 0.6738280951976776,
      "rewards/format_reward": 0.984375,
      "step": 592
    },
    {
      "completion_length": 75.515625,
      "epoch": 2.7077625570776256,
      "grad_norm": 3.7182750701904297,
      "kl": 0.111572265625,
      "learning_rate": 7.292237442922375e-07,
      "loss": 0.0045,
      "reward": 1.8291015625,
      "reward_std": 0.15765050053596497,
      "rewards/accuracy_reward": 0.8291015625,
      "rewards/format_reward": 1.0,
      "step": 593
    },
    {
      "completion_length": 65.7890625,
      "epoch": 2.712328767123288,
      "grad_norm": 3.52632737159729,
      "kl": 0.15966796875,
      "learning_rate": 7.287671232876712e-07,
      "loss": 0.0064,
      "reward": 1.6682049632072449,
      "reward_std": 0.22022631764411926,
      "rewards/accuracy_reward": 0.6682049036026001,
      "rewards/format_reward": 1.0,
      "step": 594
    },
    {
      "completion_length": 94.90625,
      "epoch": 2.7168949771689497,
      "grad_norm": 1.899457335472107,
      "kl": 0.091796875,
      "learning_rate": 7.283105022831049e-07,
      "loss": 0.0037,
      "reward": 1.6791666746139526,
      "reward_std": 0.18102534115314484,
      "rewards/accuracy_reward": 0.7104166448116302,
      "rewards/format_reward": 0.96875,
      "step": 595
    },
    {
      "completion_length": 80.5390625,
      "epoch": 2.721461187214612,
      "grad_norm": 2.4807851314544678,
      "kl": 0.113525390625,
      "learning_rate": 7.278538812785388e-07,
      "loss": 0.0045,
      "reward": 1.8067708611488342,
      "reward_std": 0.12291676551103592,
      "rewards/accuracy_reward": 0.8067708313465118,
      "rewards/format_reward": 1.0,
      "step": 596
    },
    {
      "completion_length": 83.34375,
      "epoch": 2.7260273972602738,
      "grad_norm": 2.2889952659606934,
      "kl": 0.11669921875,
      "learning_rate": 7.273972602739725e-07,
      "loss": 0.0047,
      "reward": 1.7340867519378662,
      "reward_std": 0.2301034778356552,
      "rewards/accuracy_reward": 0.7731491327285767,
      "rewards/format_reward": 0.9609375,
      "step": 597
    },
    {
      "completion_length": 76.90625,
      "epoch": 2.730593607305936,
      "grad_norm": 2.431467056274414,
      "kl": 0.141845703125,
      "learning_rate": 7.269406392694064e-07,
      "loss": 0.0057,
      "reward": 1.702616572380066,
      "reward_std": 0.16628245636820793,
      "rewards/accuracy_reward": 0.7026165425777435,
      "rewards/format_reward": 1.0,
      "step": 598
    },
    {
      "completion_length": 75.203125,
      "epoch": 2.7351598173515983,
      "grad_norm": 2.5654890537261963,
      "kl": 0.1119384765625,
      "learning_rate": 7.264840182648402e-07,
      "loss": 0.0045,
      "reward": 1.7242187857627869,
      "reward_std": 0.13443218544125557,
      "rewards/accuracy_reward": 0.7320312559604645,
      "rewards/format_reward": 0.9921875,
      "step": 599
    },
    {
      "completion_length": 77.8515625,
      "epoch": 2.73972602739726,
      "grad_norm": 2.4375691413879395,
      "kl": 0.097900390625,
      "learning_rate": 7.260273972602739e-07,
      "loss": 0.0039,
      "reward": 1.7069196701049805,
      "reward_std": 0.24784404039382935,
      "rewards/accuracy_reward": 0.7225446105003357,
      "rewards/format_reward": 0.984375,
      "step": 600
    },
    {
      "completion_length": 73.1953125,
      "epoch": 2.7442922374429224,
      "grad_norm": 2.0869359970092773,
      "kl": 0.1171875,
      "learning_rate": 7.255707762557078e-07,
      "loss": 0.0047,
      "reward": 1.6764508485794067,
      "reward_std": 0.11909351497888565,
      "rewards/accuracy_reward": 0.6764508485794067,
      "rewards/format_reward": 1.0,
      "step": 601
    },
    {
      "completion_length": 59.4765625,
      "epoch": 2.748858447488584,
      "grad_norm": 3.2937729358673096,
      "kl": 0.138427734375,
      "learning_rate": 7.251141552511415e-07,
      "loss": 0.0055,
      "reward": 1.538671851158142,
      "reward_std": 0.36779990792274475,
      "rewards/accuracy_reward": 0.5542968809604645,
      "rewards/format_reward": 0.984375,
      "step": 602
    },
    {
      "completion_length": 92.5703125,
      "epoch": 2.7534246575342465,
      "grad_norm": 3.537177085876465,
      "kl": 0.090087890625,
      "learning_rate": 7.246575342465752e-07,
      "loss": 0.0036,
      "reward": 1.722743034362793,
      "reward_std": 0.18279560655355453,
      "rewards/accuracy_reward": 0.730555534362793,
      "rewards/format_reward": 0.9921875,
      "step": 603
    },
    {
      "completion_length": 93.8046875,
      "epoch": 2.7579908675799087,
      "grad_norm": 2.128953218460083,
      "kl": 0.08056640625,
      "learning_rate": 7.242009132420091e-07,
      "loss": 0.0032,
      "reward": 1.745312511920929,
      "reward_std": 0.1521657407283783,
      "rewards/accuracy_reward": 0.7453123927116394,
      "rewards/format_reward": 1.0,
      "step": 604
    },
    {
      "completion_length": 65.9375,
      "epoch": 2.762557077625571,
      "grad_norm": 17.820329666137695,
      "kl": 0.10205078125,
      "learning_rate": 7.237442922374429e-07,
      "loss": 0.0041,
      "reward": 1.6083807349205017,
      "reward_std": 0.20848889648914337,
      "rewards/accuracy_reward": 0.6161931753158569,
      "rewards/format_reward": 0.9921875,
      "step": 605
    },
    {
      "completion_length": 71.671875,
      "epoch": 2.767123287671233,
      "grad_norm": 2.2198660373687744,
      "kl": 0.10009765625,
      "learning_rate": 7.232876712328767e-07,
      "loss": 0.004,
      "reward": 1.7286458611488342,
      "reward_std": 0.16050894185900688,
      "rewards/accuracy_reward": 0.7286458313465118,
      "rewards/format_reward": 1.0,
      "step": 606
    },
    {
      "completion_length": 84.546875,
      "epoch": 2.771689497716895,
      "grad_norm": 2.738107681274414,
      "kl": 0.11376953125,
      "learning_rate": 7.228310502283105e-07,
      "loss": 0.0046,
      "reward": 1.6790487170219421,
      "reward_std": 0.19785276055335999,
      "rewards/accuracy_reward": 0.6946736574172974,
      "rewards/format_reward": 0.984375,
      "step": 607
    },
    {
      "completion_length": 71.125,
      "epoch": 2.776255707762557,
      "grad_norm": 5.066852569580078,
      "kl": 0.151123046875,
      "learning_rate": 7.223744292237442e-07,
      "loss": 0.006,
      "reward": 1.5531622171401978,
      "reward_std": 0.3353729024529457,
      "rewards/accuracy_reward": 0.5609746873378754,
      "rewards/format_reward": 0.9921875,
      "step": 608
    },
    {
      "completion_length": 83.7890625,
      "epoch": 2.780821917808219,
      "grad_norm": 4.337226867675781,
      "kl": 0.117919921875,
      "learning_rate": 7.219178082191781e-07,
      "loss": 0.0047,
      "reward": 1.7048035860061646,
      "reward_std": 0.13883494585752487,
      "rewards/accuracy_reward": 0.7048035860061646,
      "rewards/format_reward": 1.0,
      "step": 609
    },
    {
      "completion_length": 87.3984375,
      "epoch": 2.7853881278538815,
      "grad_norm": 2.790894031524658,
      "kl": 0.09716796875,
      "learning_rate": 7.214611872146118e-07,
      "loss": 0.0039,
      "reward": 1.6720238327980042,
      "reward_std": 0.16603849083185196,
      "rewards/accuracy_reward": 0.6798363327980042,
      "rewards/format_reward": 0.9921875,
      "step": 610
    },
    {
      "completion_length": 79.4140625,
      "epoch": 2.7899543378995433,
      "grad_norm": 4.233171463012695,
      "kl": 0.1103515625,
      "learning_rate": 7.210045662100456e-07,
      "loss": 0.0044,
      "reward": 1.6883246898651123,
      "reward_std": 0.21825328469276428,
      "rewards/accuracy_reward": 0.6961371600627899,
      "rewards/format_reward": 0.9921875,
      "step": 611
    },
    {
      "completion_length": 63.875,
      "epoch": 2.7945205479452055,
      "grad_norm": 4.263243198394775,
      "kl": 0.138671875,
      "learning_rate": 7.205479452054795e-07,
      "loss": 0.0055,
      "reward": 1.6089910864830017,
      "reward_std": 0.2958259731531143,
      "rewards/accuracy_reward": 0.6246160566806793,
      "rewards/format_reward": 0.984375,
      "step": 612
    },
    {
      "completion_length": 72.21875,
      "epoch": 2.7990867579908674,
      "grad_norm": 2.3356006145477295,
      "kl": 0.110595703125,
      "learning_rate": 7.200913242009132e-07,
      "loss": 0.0044,
      "reward": 1.5526910424232483,
      "reward_std": 0.24793513119220734,
      "rewards/accuracy_reward": 0.5526909977197647,
      "rewards/format_reward": 1.0,
      "step": 613
    },
    {
      "completion_length": 100.8203125,
      "epoch": 2.8036529680365296,
      "grad_norm": 1.9555134773254395,
      "kl": 0.098388671875,
      "learning_rate": 7.19634703196347e-07,
      "loss": 0.0039,
      "reward": 1.7364583611488342,
      "reward_std": 0.15703274309635162,
      "rewards/accuracy_reward": 0.7442708015441895,
      "rewards/format_reward": 0.9921875,
      "step": 614
    },
    {
      "completion_length": 71.8984375,
      "epoch": 2.808219178082192,
      "grad_norm": 5.0422539710998535,
      "kl": 0.16455078125,
      "learning_rate": 7.191780821917808e-07,
      "loss": 0.0066,
      "reward": 1.4983445405960083,
      "reward_std": 0.2954416871070862,
      "rewards/accuracy_reward": 0.5061569809913635,
      "rewards/format_reward": 0.9921875,
      "step": 615
    },
    {
      "completion_length": 81.109375,
      "epoch": 2.8127853881278537,
      "grad_norm": 2.615528106689453,
      "kl": 0.123779296875,
      "learning_rate": 7.187214611872145e-07,
      "loss": 0.005,
      "reward": 1.685937523841858,
      "reward_std": 0.16778654977679253,
      "rewards/accuracy_reward": 0.6859375238418579,
      "rewards/format_reward": 1.0,
      "step": 616
    },
    {
      "completion_length": 78.1171875,
      "epoch": 2.817351598173516,
      "grad_norm": 3.2558462619781494,
      "kl": 0.099609375,
      "learning_rate": 7.182648401826484e-07,
      "loss": 0.004,
      "reward": 1.7515625953674316,
      "reward_std": 0.1005905494093895,
      "rewards/accuracy_reward": 0.7593749463558197,
      "rewards/format_reward": 0.9921875,
      "step": 617
    },
    {
      "completion_length": 85.3671875,
      "epoch": 2.821917808219178,
      "grad_norm": 1.791414499282837,
      "kl": 0.127197265625,
      "learning_rate": 7.178082191780822e-07,
      "loss": 0.0051,
      "reward": 1.6640625,
      "reward_std": 0.16250330954790115,
      "rewards/accuracy_reward": 0.6718749403953552,
      "rewards/format_reward": 0.9921875,
      "step": 618
    },
    {
      "completion_length": 88.2578125,
      "epoch": 2.82648401826484,
      "grad_norm": 5.740970134735107,
      "kl": 0.077392578125,
      "learning_rate": 7.173515981735159e-07,
      "loss": 0.0031,
      "reward": 1.652303397655487,
      "reward_std": 0.17974259704351425,
      "rewards/accuracy_reward": 0.6523034274578094,
      "rewards/format_reward": 1.0,
      "step": 619
    },
    {
      "completion_length": 95.4609375,
      "epoch": 2.8310502283105023,
      "grad_norm": 2.585156202316284,
      "kl": 0.115234375,
      "learning_rate": 7.168949771689498e-07,
      "loss": 0.0046,
      "reward": 1.6767844557762146,
      "reward_std": 0.14024699479341507,
      "rewards/accuracy_reward": 0.6845969557762146,
      "rewards/format_reward": 0.9921875,
      "step": 620
    },
    {
      "completion_length": 80.734375,
      "epoch": 2.8356164383561646,
      "grad_norm": 6.572816848754883,
      "kl": 0.138671875,
      "learning_rate": 7.164383561643835e-07,
      "loss": 0.0055,
      "reward": 1.6025669574737549,
      "reward_std": 0.22662456333637238,
      "rewards/accuracy_reward": 0.6181919574737549,
      "rewards/format_reward": 0.984375,
      "step": 621
    },
    {
      "completion_length": 62.4921875,
      "epoch": 2.8401826484018264,
      "grad_norm": 12.138286590576172,
      "kl": 0.16650390625,
      "learning_rate": 7.159817351598173e-07,
      "loss": 0.0067,
      "reward": 1.7107762694358826,
      "reward_std": 0.16279632598161697,
      "rewards/accuracy_reward": 0.7107762694358826,
      "rewards/format_reward": 1.0,
      "step": 622
    },
    {
      "completion_length": 82.625,
      "epoch": 2.8447488584474887,
      "grad_norm": 4.130621910095215,
      "kl": 0.110595703125,
      "learning_rate": 7.155251141552511e-07,
      "loss": 0.0044,
      "reward": 1.792187511920929,
      "reward_std": 0.12151552736759186,
      "rewards/accuracy_reward": 0.7921874523162842,
      "rewards/format_reward": 1.0,
      "step": 623
    },
    {
      "completion_length": 66.8203125,
      "epoch": 2.8493150684931505,
      "grad_norm": 7.18456506729126,
      "kl": 0.12841796875,
      "learning_rate": 7.150684931506848e-07,
      "loss": 0.0051,
      "reward": 1.5980710983276367,
      "reward_std": 0.21035503596067429,
      "rewards/accuracy_reward": 0.6058836281299591,
      "rewards/format_reward": 0.9921875,
      "step": 624
    },
    {
      "completion_length": 92.1171875,
      "epoch": 2.853881278538813,
      "grad_norm": 5.2410759925842285,
      "kl": 0.119140625,
      "learning_rate": 7.146118721461188e-07,
      "loss": 0.0048,
      "reward": 1.8228118419647217,
      "reward_std": 0.17064978182315826,
      "rewards/accuracy_reward": 0.8306242823600769,
      "rewards/format_reward": 0.9921875,
      "step": 625
    },
    {
      "completion_length": 100.2265625,
      "epoch": 2.858447488584475,
      "grad_norm": 12.312285423278809,
      "kl": 0.086669921875,
      "learning_rate": 7.141552511415525e-07,
      "loss": 0.0035,
      "reward": 1.5916666984558105,
      "reward_std": 0.2057085707783699,
      "rewards/accuracy_reward": 0.5994791686534882,
      "rewards/format_reward": 0.9921875,
      "step": 626
    },
    {
      "completion_length": 94.7109375,
      "epoch": 2.863013698630137,
      "grad_norm": 2.4348156452178955,
      "kl": 0.09423828125,
      "learning_rate": 7.136986301369862e-07,
      "loss": 0.0038,
      "reward": 1.71042400598526,
      "reward_std": 0.15129226446151733,
      "rewards/accuracy_reward": 0.7104238867759705,
      "rewards/format_reward": 1.0,
      "step": 627
    },
    {
      "completion_length": 76.09375,
      "epoch": 2.867579908675799,
      "grad_norm": 2.56550931930542,
      "kl": 0.15283203125,
      "learning_rate": 7.132420091324201e-07,
      "loss": 0.0061,
      "reward": 1.6796875,
      "reward_std": 0.1692301705479622,
      "rewards/accuracy_reward": 0.6796875298023224,
      "rewards/format_reward": 1.0,
      "step": 628
    },
    {
      "completion_length": 63.71875,
      "epoch": 2.872146118721461,
      "grad_norm": 3.254136562347412,
      "kl": 0.1396484375,
      "learning_rate": 7.127853881278538e-07,
      "loss": 0.0056,
      "reward": 1.5118472576141357,
      "reward_std": 0.22680865228176117,
      "rewards/accuracy_reward": 0.5118472576141357,
      "rewards/format_reward": 1.0,
      "step": 629
    },
    {
      "completion_length": 91.984375,
      "epoch": 2.8767123287671232,
      "grad_norm": 1.9678096771240234,
      "kl": 0.118896484375,
      "learning_rate": 7.123287671232876e-07,
      "loss": 0.0047,
      "reward": 1.7608563899993896,
      "reward_std": 0.10874464362859726,
      "rewards/accuracy_reward": 0.7608563899993896,
      "rewards/format_reward": 1.0,
      "step": 630
    },
    {
      "completion_length": 80.421875,
      "epoch": 2.8812785388127855,
      "grad_norm": 3.203822374343872,
      "kl": 0.14111328125,
      "learning_rate": 7.118721461187215e-07,
      "loss": 0.0056,
      "reward": 1.5107174515724182,
      "reward_std": 0.2804914563894272,
      "rewards/accuracy_reward": 0.526342436671257,
      "rewards/format_reward": 0.984375,
      "step": 631
    },
    {
      "completion_length": 69.8671875,
      "epoch": 2.8858447488584473,
      "grad_norm": 2.1312613487243652,
      "kl": 0.1513671875,
      "learning_rate": 7.114155251141552e-07,
      "loss": 0.0061,
      "reward": 1.783835530281067,
      "reward_std": 0.09552156459540129,
      "rewards/accuracy_reward": 0.7838355302810669,
      "rewards/format_reward": 1.0,
      "step": 632
    },
    {
      "completion_length": 83.6796875,
      "epoch": 2.8904109589041096,
      "grad_norm": 2.534517765045166,
      "kl": 0.129638671875,
      "learning_rate": 7.109589041095891e-07,
      "loss": 0.0052,
      "reward": 1.820498526096344,
      "reward_std": 0.14453133195638657,
      "rewards/accuracy_reward": 0.8204984366893768,
      "rewards/format_reward": 1.0,
      "step": 633
    },
    {
      "completion_length": 100.578125,
      "epoch": 2.8949771689497714,
      "grad_norm": 4.368940353393555,
      "kl": 0.137451171875,
      "learning_rate": 7.105022831050228e-07,
      "loss": 0.0055,
      "reward": 1.6578125357627869,
      "reward_std": 0.2853652313351631,
      "rewards/accuracy_reward": 0.6890624761581421,
      "rewards/format_reward": 0.96875,
      "step": 634
    },
    {
      "completion_length": 91.734375,
      "epoch": 2.8995433789954337,
      "grad_norm": 2.2129926681518555,
      "kl": 0.1015625,
      "learning_rate": 7.100456621004565e-07,
      "loss": 0.0041,
      "reward": 1.60442715883255,
      "reward_std": 0.17540115863084793,
      "rewards/accuracy_reward": 0.6200520992279053,
      "rewards/format_reward": 0.984375,
      "step": 635
    },
    {
      "completion_length": 71.65625,
      "epoch": 2.904109589041096,
      "grad_norm": 14.674283981323242,
      "kl": 0.1591796875,
      "learning_rate": 7.095890410958904e-07,
      "loss": 0.0064,
      "reward": 1.5524739027023315,
      "reward_std": 0.24804671853780746,
      "rewards/accuracy_reward": 0.5602864325046539,
      "rewards/format_reward": 0.9921875,
      "step": 636
    },
    {
      "completion_length": 83.1484375,
      "epoch": 2.908675799086758,
      "grad_norm": 4.1890034675598145,
      "kl": 0.119140625,
      "learning_rate": 7.091324200913241e-07,
      "loss": 0.0048,
      "reward": 1.7505208253860474,
      "reward_std": 0.1413591168820858,
      "rewards/accuracy_reward": 0.7583333253860474,
      "rewards/format_reward": 0.9921875,
      "step": 637
    },
    {
      "completion_length": 72.359375,
      "epoch": 2.91324200913242,
      "grad_norm": 4.4610443115234375,
      "kl": 0.1669921875,
      "learning_rate": 7.08675799086758e-07,
      "loss": 0.0067,
      "reward": 1.6082217693328857,
      "reward_std": 0.3458182215690613,
      "rewards/accuracy_reward": 0.647284209728241,
      "rewards/format_reward": 0.9609375,
      "step": 638
    },
    {
      "completion_length": 86.0390625,
      "epoch": 2.9178082191780823,
      "grad_norm": 4.549093723297119,
      "kl": 0.1357421875,
      "learning_rate": 7.082191780821918e-07,
      "loss": 0.0054,
      "reward": 1.6874799728393555,
      "reward_std": 0.2353600338101387,
      "rewards/accuracy_reward": 0.6874799132347107,
      "rewards/format_reward": 1.0,
      "step": 639
    },
    {
      "completion_length": 97.1484375,
      "epoch": 2.922374429223744,
      "grad_norm": 2.3335134983062744,
      "kl": 0.12646484375,
      "learning_rate": 7.077625570776255e-07,
      "loss": 0.0051,
      "reward": 1.7588170766830444,
      "reward_std": 0.14432461559772491,
      "rewards/accuracy_reward": 0.7666293978691101,
      "rewards/format_reward": 0.9921875,
      "step": 640
    },
    {
      "completion_length": 77.0078125,
      "epoch": 2.9269406392694064,
      "grad_norm": 3.070441961288452,
      "kl": 0.126220703125,
      "learning_rate": 7.073059360730594e-07,
      "loss": 0.005,
      "reward": 1.4704504013061523,
      "reward_std": 0.28115857392549515,
      "rewards/accuracy_reward": 0.47826285660266876,
      "rewards/format_reward": 0.9921875,
      "step": 641
    },
    {
      "completion_length": 92.9765625,
      "epoch": 2.9315068493150687,
      "grad_norm": 3.367562770843506,
      "kl": 0.101806640625,
      "learning_rate": 7.068493150684931e-07,
      "loss": 0.0041,
      "reward": 1.729687511920929,
      "reward_std": 0.12863079458475113,
      "rewards/accuracy_reward": 0.7374999523162842,
      "rewards/format_reward": 0.9921875,
      "step": 642
    },
    {
      "completion_length": 96.390625,
      "epoch": 2.9360730593607305,
      "grad_norm": 2.8609812259674072,
      "kl": 0.10302734375,
      "learning_rate": 7.063926940639268e-07,
      "loss": 0.0041,
      "reward": 1.749218761920929,
      "reward_std": 0.18227346241474152,
      "rewards/accuracy_reward": 0.7648437321186066,
      "rewards/format_reward": 0.984375,
      "step": 643
    },
    {
      "completion_length": 89.7734375,
      "epoch": 2.9406392694063928,
      "grad_norm": 1.9020359516143799,
      "kl": 0.090087890625,
      "learning_rate": 7.059360730593607e-07,
      "loss": 0.0036,
      "reward": 1.7804688215255737,
      "reward_std": 0.11121231690049171,
      "rewards/accuracy_reward": 0.7804686725139618,
      "rewards/format_reward": 1.0,
      "step": 644
    },
    {
      "completion_length": 99.03125,
      "epoch": 2.9452054794520546,
      "grad_norm": 9.120920181274414,
      "kl": 0.10595703125,
      "learning_rate": 7.054794520547945e-07,
      "loss": 0.0042,
      "reward": 1.7126150131225586,
      "reward_std": 0.20574645698070526,
      "rewards/accuracy_reward": 0.7282399535179138,
      "rewards/format_reward": 0.984375,
      "step": 645
    },
    {
      "completion_length": 74.2265625,
      "epoch": 2.949771689497717,
      "grad_norm": 5.297418594360352,
      "kl": 0.133056640625,
      "learning_rate": 7.050228310502283e-07,
      "loss": 0.0053,
      "reward": 1.6359771490097046,
      "reward_std": 0.23064683377742767,
      "rewards/accuracy_reward": 0.6672270596027374,
      "rewards/format_reward": 0.96875,
      "step": 646
    },
    {
      "completion_length": 102.6796875,
      "epoch": 2.954337899543379,
      "grad_norm": 1.925435185432434,
      "kl": 0.0947265625,
      "learning_rate": 7.045662100456621e-07,
      "loss": 0.0038,
      "reward": 1.8184895515441895,
      "reward_std": 0.14697792008519173,
      "rewards/accuracy_reward": 0.8341145217418671,
      "rewards/format_reward": 0.984375,
      "step": 647
    },
    {
      "completion_length": 93.6796875,
      "epoch": 2.958904109589041,
      "grad_norm": 8.05284309387207,
      "kl": 0.119140625,
      "learning_rate": 7.041095890410958e-07,
      "loss": 0.0048,
      "reward": 1.7860276699066162,
      "reward_std": 0.13688677921891212,
      "rewards/accuracy_reward": 0.793840080499649,
      "rewards/format_reward": 0.9921875,
      "step": 648
    },
    {
      "completion_length": 89.421875,
      "epoch": 2.963470319634703,
      "grad_norm": 2.366201162338257,
      "kl": 0.150390625,
      "learning_rate": 7.036529680365297e-07,
      "loss": 0.006,
      "reward": 1.6971353888511658,
      "reward_std": 0.16348732262849808,
      "rewards/accuracy_reward": 0.7049478888511658,
      "rewards/format_reward": 0.9921875,
      "step": 649
    },
    {
      "completion_length": 76.46875,
      "epoch": 2.968036529680365,
      "grad_norm": 2.8056042194366455,
      "kl": 0.1259765625,
      "learning_rate": 7.031963470319634e-07,
      "loss": 0.0051,
      "reward": 1.7359544038772583,
      "reward_std": 0.1912137269973755,
      "rewards/accuracy_reward": 0.7437668442726135,
      "rewards/format_reward": 0.9921875,
      "step": 650
    },
    {
      "completion_length": 74.015625,
      "epoch": 2.9726027397260273,
      "grad_norm": 2.3145570755004883,
      "kl": 0.11767578125,
      "learning_rate": 7.027397260273972e-07,
      "loss": 0.0047,
      "reward": 1.5831072330474854,
      "reward_std": 0.25112421810626984,
      "rewards/accuracy_reward": 0.5987322330474854,
      "rewards/format_reward": 0.984375,
      "step": 651
    },
    {
      "completion_length": 93.734375,
      "epoch": 2.9771689497716896,
      "grad_norm": 3.0146915912628174,
      "kl": 0.117431640625,
      "learning_rate": 7.022831050228311e-07,
      "loss": 0.0047,
      "reward": 1.6713745594024658,
      "reward_std": 0.21526063233613968,
      "rewards/accuracy_reward": 0.6791869699954987,
      "rewards/format_reward": 0.9921875,
      "step": 652
    },
    {
      "completion_length": 65.5078125,
      "epoch": 2.981735159817352,
      "grad_norm": 2.2300150394439697,
      "kl": 0.123046875,
      "learning_rate": 7.018264840182648e-07,
      "loss": 0.0049,
      "reward": 1.690625011920929,
      "reward_std": 0.16151440143585205,
      "rewards/accuracy_reward": 0.690625011920929,
      "rewards/format_reward": 1.0,
      "step": 653
    },
    {
      "completion_length": 78.2890625,
      "epoch": 2.9863013698630136,
      "grad_norm": 4.648037433624268,
      "kl": 0.126708984375,
      "learning_rate": 7.013698630136986e-07,
      "loss": 0.0051,
      "reward": 1.744691550731659,
      "reward_std": 0.14016081020236015,
      "rewards/accuracy_reward": 0.7603164613246918,
      "rewards/format_reward": 0.984375,
      "step": 654
    },
    {
      "completion_length": 67.203125,
      "epoch": 2.990867579908676,
      "grad_norm": 2.895372152328491,
      "kl": 0.139404296875,
      "learning_rate": 7.009132420091324e-07,
      "loss": 0.0056,
      "reward": 1.6064826250076294,
      "reward_std": 0.2562439739704132,
      "rewards/accuracy_reward": 0.6142950057983398,
      "rewards/format_reward": 0.9921875,
      "step": 655
    },
    {
      "completion_length": 78.765625,
      "epoch": 2.9954337899543377,
      "grad_norm": 3.7892091274261475,
      "kl": 0.093017578125,
      "learning_rate": 7.004566210045661e-07,
      "loss": 0.0037,
      "reward": 1.754687488079071,
      "reward_std": 0.1408282183110714,
      "rewards/accuracy_reward": 0.754687488079071,
      "rewards/format_reward": 1.0,
      "step": 656
    },
    {
      "completion_length": 58.5,
      "epoch": 3.0,
      "grad_norm": 3.69757080078125,
      "kl": 0.18359375,
      "learning_rate": 7e-07,
      "loss": 0.0056,
      "reward": 1.5,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 1.0,
      "step": 657
    },
    {
      "completion_length": 73.21875,
      "epoch": 3.0045662100456623,
      "grad_norm": 2.8494434356689453,
      "kl": 0.12744140625,
      "learning_rate": 6.995433789954338e-07,
      "loss": 0.0051,
      "reward": 1.7073844075202942,
      "reward_std": 0.18213152885437012,
      "rewards/accuracy_reward": 0.7151968777179718,
      "rewards/format_reward": 0.9921875,
      "step": 658
    },
    {
      "completion_length": 73.71875,
      "epoch": 3.009132420091324,
      "grad_norm": 2.561201810836792,
      "kl": 0.099609375,
      "learning_rate": 6.990867579908675e-07,
      "loss": 0.004,
      "reward": 1.6927083730697632,
      "reward_std": 0.17407646402716637,
      "rewards/accuracy_reward": 0.6927083134651184,
      "rewards/format_reward": 1.0,
      "step": 659
    },
    {
      "completion_length": 98.328125,
      "epoch": 3.0136986301369864,
      "grad_norm": 3.664125442504883,
      "kl": 0.138671875,
      "learning_rate": 6.986301369863014e-07,
      "loss": 0.0055,
      "reward": 1.6324912905693054,
      "reward_std": 0.21703079342842102,
      "rewards/accuracy_reward": 0.6403037309646606,
      "rewards/format_reward": 0.9921875,
      "step": 660
    },
    {
      "completion_length": 94.46875,
      "epoch": 3.018264840182648,
      "grad_norm": 2.527571439743042,
      "kl": 0.1181640625,
      "learning_rate": 6.981735159817351e-07,
      "loss": 0.0047,
      "reward": 1.7589489221572876,
      "reward_std": 0.16745695658028126,
      "rewards/accuracy_reward": 0.7745738625526428,
      "rewards/format_reward": 0.984375,
      "step": 661
    },
    {
      "completion_length": 82.1953125,
      "epoch": 3.0228310502283104,
      "grad_norm": 2.676431894302368,
      "kl": 0.14599609375,
      "learning_rate": 6.977168949771689e-07,
      "loss": 0.0058,
      "reward": 1.681512713432312,
      "reward_std": 0.18309402465820312,
      "rewards/accuracy_reward": 0.6971376836299896,
      "rewards/format_reward": 0.984375,
      "step": 662
    },
    {
      "completion_length": 97.671875,
      "epoch": 3.0273972602739727,
      "grad_norm": 2.549351453781128,
      "kl": 0.078125,
      "learning_rate": 6.972602739726027e-07,
      "loss": 0.0031,
      "reward": 1.6531250476837158,
      "reward_std": 0.24489019811153412,
      "rewards/accuracy_reward": 0.6687500178813934,
      "rewards/format_reward": 0.984375,
      "step": 663
    },
    {
      "completion_length": 76.59375,
      "epoch": 3.0319634703196345,
      "grad_norm": 1.8996092081069946,
      "kl": 0.147216796875,
      "learning_rate": 6.968036529680364e-07,
      "loss": 0.0059,
      "reward": 1.734002947807312,
      "reward_std": 0.1543407365679741,
      "rewards/accuracy_reward": 0.7496279180049896,
      "rewards/format_reward": 0.984375,
      "step": 664
    },
    {
      "completion_length": 62.4453125,
      "epoch": 3.036529680365297,
      "grad_norm": 2.1863508224487305,
      "kl": 0.1217041015625,
      "learning_rate": 6.963470319634704e-07,
      "loss": 0.0049,
      "reward": 1.771484375,
      "reward_std": 0.08306973986327648,
      "rewards/accuracy_reward": 0.7714843153953552,
      "rewards/format_reward": 1.0,
      "step": 665
    },
    {
      "completion_length": 79.71875,
      "epoch": 3.041095890410959,
      "grad_norm": 2.9316976070404053,
      "kl": 0.129150390625,
      "learning_rate": 6.958904109589041e-07,
      "loss": 0.0052,
      "reward": 1.6214410066604614,
      "reward_std": 0.2264525145292282,
      "rewards/accuracy_reward": 0.6370659470558167,
      "rewards/format_reward": 0.984375,
      "step": 666
    },
    {
      "completion_length": 90.0546875,
      "epoch": 3.045662100456621,
      "grad_norm": 2.6280527114868164,
      "kl": 0.101806640625,
      "learning_rate": 6.954337899543378e-07,
      "loss": 0.0041,
      "reward": 1.7546875476837158,
      "reward_std": 0.1054728776216507,
      "rewards/accuracy_reward": 0.7546874284744263,
      "rewards/format_reward": 1.0,
      "step": 667
    },
    {
      "completion_length": 79.453125,
      "epoch": 3.050228310502283,
      "grad_norm": 5.255128383636475,
      "kl": 0.10546875,
      "learning_rate": 6.949771689497717e-07,
      "loss": 0.0042,
      "reward": 1.5757898688316345,
      "reward_std": 0.19272886961698532,
      "rewards/accuracy_reward": 0.5836023390293121,
      "rewards/format_reward": 0.9921875,
      "step": 668
    },
    {
      "completion_length": 96.90625,
      "epoch": 3.0547945205479454,
      "grad_norm": 2.3194217681884766,
      "kl": 0.077392578125,
      "learning_rate": 6.945205479452054e-07,
      "loss": 0.0031,
      "reward": 1.770312488079071,
      "reward_std": 0.11202363669872284,
      "rewards/accuracy_reward": 0.7703124582767487,
      "rewards/format_reward": 1.0,
      "step": 669
    },
    {
      "completion_length": 80.734375,
      "epoch": 3.0593607305936072,
      "grad_norm": 3.282161235809326,
      "kl": 0.163330078125,
      "learning_rate": 6.940639269406392e-07,
      "loss": 0.0065,
      "reward": 1.7177083492279053,
      "reward_std": 0.16301878169178963,
      "rewards/accuracy_reward": 0.7177082896232605,
      "rewards/format_reward": 1.0,
      "step": 670
    },
    {
      "completion_length": 83.0,
      "epoch": 3.0639269406392695,
      "grad_norm": 3.282697916030884,
      "kl": 0.12744140625,
      "learning_rate": 6.93607305936073e-07,
      "loss": 0.0051,
      "reward": 1.8043155074119568,
      "reward_std": 0.11800633184611797,
      "rewards/accuracy_reward": 0.8043154180049896,
      "rewards/format_reward": 1.0,
      "step": 671
    },
    {
      "completion_length": 80.8046875,
      "epoch": 3.0684931506849313,
      "grad_norm": 2.545616388320923,
      "kl": 0.09814453125,
      "learning_rate": 6.931506849315068e-07,
      "loss": 0.0039,
      "reward": 1.7645833492279053,
      "reward_std": 0.14983439445495605,
      "rewards/accuracy_reward": 0.7802082598209381,
      "rewards/format_reward": 0.984375,
      "step": 672
    },
    {
      "completion_length": 75.546875,
      "epoch": 3.0730593607305936,
      "grad_norm": 2.007420778274536,
      "kl": 0.1240234375,
      "learning_rate": 6.926940639269407e-07,
      "loss": 0.005,
      "reward": 1.5703125596046448,
      "reward_std": 0.21254336833953857,
      "rewards/accuracy_reward": 0.5703125,
      "rewards/format_reward": 1.0,
      "step": 673
    },
    {
      "completion_length": 74.5234375,
      "epoch": 3.077625570776256,
      "grad_norm": 3.1193904876708984,
      "kl": 0.119873046875,
      "learning_rate": 6.922374429223744e-07,
      "loss": 0.0048,
      "reward": 1.6553664803504944,
      "reward_std": 0.17046189308166504,
      "rewards/accuracy_reward": 0.6631789207458496,
      "rewards/format_reward": 0.9921875,
      "step": 674
    },
    {
      "completion_length": 85.1484375,
      "epoch": 3.0821917808219177,
      "grad_norm": 2.9060676097869873,
      "kl": 0.105712890625,
      "learning_rate": 6.917808219178081e-07,
      "loss": 0.0042,
      "reward": 1.5601562857627869,
      "reward_std": 0.2604144960641861,
      "rewards/accuracy_reward": 0.5835937559604645,
      "rewards/format_reward": 0.9765625,
      "step": 675
    },
    {
      "completion_length": 92.7109375,
      "epoch": 3.08675799086758,
      "grad_norm": 3.6577494144439697,
      "kl": 0.0640869140625,
      "learning_rate": 6.91324200913242e-07,
      "loss": 0.0026,
      "reward": 1.6146825551986694,
      "reward_std": 0.10726364329457283,
      "rewards/accuracy_reward": 0.614682525396347,
      "rewards/format_reward": 1.0,
      "step": 676
    },
    {
      "completion_length": 76.6875,
      "epoch": 3.091324200913242,
      "grad_norm": 9.413636207580566,
      "kl": 0.13818359375,
      "learning_rate": 6.908675799086757e-07,
      "loss": 0.0055,
      "reward": 1.6631696820259094,
      "reward_std": 0.2051173821091652,
      "rewards/accuracy_reward": 0.6709820926189423,
      "rewards/format_reward": 0.9921875,
      "step": 677
    },
    {
      "completion_length": 75.2734375,
      "epoch": 3.095890410958904,
      "grad_norm": 5.669606685638428,
      "kl": 0.119140625,
      "learning_rate": 6.904109589041097e-07,
      "loss": 0.0048,
      "reward": 1.5031325817108154,
      "reward_std": 0.2111596167087555,
      "rewards/accuracy_reward": 0.5109450221061707,
      "rewards/format_reward": 0.9921875,
      "step": 678
    },
    {
      "completion_length": 74.1796875,
      "epoch": 3.1004566210045663,
      "grad_norm": 2.975776433944702,
      "kl": 0.1220703125,
      "learning_rate": 6.899543378995434e-07,
      "loss": 0.0049,
      "reward": 1.567187488079071,
      "reward_std": 0.26324766874313354,
      "rewards/accuracy_reward": 0.582812488079071,
      "rewards/format_reward": 0.984375,
      "step": 679
    },
    {
      "completion_length": 84.171875,
      "epoch": 3.105022831050228,
      "grad_norm": 2.021631956100464,
      "kl": 0.09765625,
      "learning_rate": 6.894977168949771e-07,
      "loss": 0.0039,
      "reward": 1.723046898841858,
      "reward_std": 0.14361856132745743,
      "rewards/accuracy_reward": 0.7230468392372131,
      "rewards/format_reward": 1.0,
      "step": 680
    },
    {
      "completion_length": 82.0390625,
      "epoch": 3.1095890410958904,
      "grad_norm": 4.33884334564209,
      "kl": 0.108642578125,
      "learning_rate": 6.89041095890411e-07,
      "loss": 0.0043,
      "reward": 1.6755682229995728,
      "reward_std": 0.11651190742850304,
      "rewards/accuracy_reward": 0.6755681037902832,
      "rewards/format_reward": 1.0,
      "step": 681
    },
    {
      "completion_length": 86.203125,
      "epoch": 3.1141552511415527,
      "grad_norm": 1.7685151100158691,
      "kl": 0.099853515625,
      "learning_rate": 6.885844748858447e-07,
      "loss": 0.004,
      "reward": 1.71484375,
      "reward_std": 0.15246989578008652,
      "rewards/accuracy_reward": 0.72265625,
      "rewards/format_reward": 0.9921875,
      "step": 682
    },
    {
      "completion_length": 74.515625,
      "epoch": 3.1187214611872145,
      "grad_norm": 3.724857807159424,
      "kl": 0.136962890625,
      "learning_rate": 6.881278538812784e-07,
      "loss": 0.0055,
      "reward": 1.6691706776618958,
      "reward_std": 0.20630130916833878,
      "rewards/accuracy_reward": 0.6691707074642181,
      "rewards/format_reward": 1.0,
      "step": 683
    },
    {
      "completion_length": 88.515625,
      "epoch": 3.1232876712328768,
      "grad_norm": 1.960091471672058,
      "kl": 0.08740234375,
      "learning_rate": 6.876712328767123e-07,
      "loss": 0.0035,
      "reward": 1.7096437811851501,
      "reward_std": 0.1030731052160263,
      "rewards/accuracy_reward": 0.709643691778183,
      "rewards/format_reward": 1.0,
      "step": 684
    },
    {
      "completion_length": 69.921875,
      "epoch": 3.127853881278539,
      "grad_norm": 4.404357433319092,
      "kl": 0.11376953125,
      "learning_rate": 6.872146118721461e-07,
      "loss": 0.0046,
      "reward": 1.7164434790611267,
      "reward_std": 0.16906945407390594,
      "rewards/accuracy_reward": 0.7164434790611267,
      "rewards/format_reward": 1.0,
      "step": 685
    },
    {
      "completion_length": 106.78125,
      "epoch": 3.132420091324201,
      "grad_norm": 1.5101971626281738,
      "kl": 0.073974609375,
      "learning_rate": 6.867579908675799e-07,
      "loss": 0.003,
      "reward": 1.798065423965454,
      "reward_std": 0.0902152806520462,
      "rewards/accuracy_reward": 0.8058778941631317,
      "rewards/format_reward": 0.9921875,
      "step": 686
    },
    {
      "completion_length": 76.1953125,
      "epoch": 3.136986301369863,
      "grad_norm": 2.8091230392456055,
      "kl": 0.16064453125,
      "learning_rate": 6.863013698630137e-07,
      "loss": 0.0064,
      "reward": 1.5283854007720947,
      "reward_std": 0.21707772463560104,
      "rewards/accuracy_reward": 0.5283854007720947,
      "rewards/format_reward": 1.0,
      "step": 687
    },
    {
      "completion_length": 87.125,
      "epoch": 3.141552511415525,
      "grad_norm": 6.1914591789245605,
      "kl": 0.078857421875,
      "learning_rate": 6.858447488584474e-07,
      "loss": 0.0032,
      "reward": 1.7078726291656494,
      "reward_std": 0.15342054888606071,
      "rewards/accuracy_reward": 0.7156850397586823,
      "rewards/format_reward": 0.9921875,
      "step": 688
    },
    {
      "completion_length": 101.140625,
      "epoch": 3.146118721461187,
      "grad_norm": 2.486093521118164,
      "kl": 0.076904296875,
      "learning_rate": 6.853881278538813e-07,
      "loss": 0.0031,
      "reward": 1.8033654689788818,
      "reward_std": 0.08035113476216793,
      "rewards/accuracy_reward": 0.8033653199672699,
      "rewards/format_reward": 1.0,
      "step": 689
    },
    {
      "completion_length": 72.546875,
      "epoch": 3.1506849315068495,
      "grad_norm": 4.52701473236084,
      "kl": 0.13427734375,
      "learning_rate": 6.84931506849315e-07,
      "loss": 0.0054,
      "reward": 1.9150173664093018,
      "reward_std": 0.12145426124334335,
      "rewards/accuracy_reward": 0.9228298366069794,
      "rewards/format_reward": 0.9921875,
      "step": 690
    },
    {
      "completion_length": 77.0625,
      "epoch": 3.1552511415525113,
      "grad_norm": 2.9900949001312256,
      "kl": 0.124755859375,
      "learning_rate": 6.844748858447487e-07,
      "loss": 0.005,
      "reward": 1.6368862390518188,
      "reward_std": 0.17363014817237854,
      "rewards/accuracy_reward": 0.6446986198425293,
      "rewards/format_reward": 0.9921875,
      "step": 691
    },
    {
      "completion_length": 91.0859375,
      "epoch": 3.1598173515981736,
      "grad_norm": 2.327423095703125,
      "kl": 0.099609375,
      "learning_rate": 6.840182648401827e-07,
      "loss": 0.004,
      "reward": 1.682031273841858,
      "reward_std": 0.17397383973002434,
      "rewards/accuracy_reward": 0.7054687440395355,
      "rewards/format_reward": 0.9765625,
      "step": 692
    },
    {
      "completion_length": 67.2578125,
      "epoch": 3.1643835616438354,
      "grad_norm": 2.4474422931671143,
      "kl": 0.133544921875,
      "learning_rate": 6.835616438356164e-07,
      "loss": 0.0053,
      "reward": 1.6786458492279053,
      "reward_std": 0.16388440132141113,
      "rewards/accuracy_reward": 0.6786458194255829,
      "rewards/format_reward": 1.0,
      "step": 693
    },
    {
      "completion_length": 83.015625,
      "epoch": 3.1689497716894977,
      "grad_norm": 12.002235412597656,
      "kl": 0.1435546875,
      "learning_rate": 6.831050228310502e-07,
      "loss": 0.0057,
      "reward": 1.6534380912780762,
      "reward_std": 0.19905856251716614,
      "rewards/accuracy_reward": 0.6534381806850433,
      "rewards/format_reward": 1.0,
      "step": 694
    },
    {
      "completion_length": 72.671875,
      "epoch": 3.17351598173516,
      "grad_norm": 4.494046211242676,
      "kl": 0.127197265625,
      "learning_rate": 6.82648401826484e-07,
      "loss": 0.0051,
      "reward": 1.8556300401687622,
      "reward_std": 0.14538883790373802,
      "rewards/accuracy_reward": 0.8556298911571503,
      "rewards/format_reward": 1.0,
      "step": 695
    },
    {
      "completion_length": 86.609375,
      "epoch": 3.1780821917808217,
      "grad_norm": 1.9116896390914917,
      "kl": 0.103759765625,
      "learning_rate": 6.821917808219177e-07,
      "loss": 0.0041,
      "reward": 1.7070313096046448,
      "reward_std": 0.19209937751293182,
      "rewards/accuracy_reward": 0.7148437201976776,
      "rewards/format_reward": 0.9921875,
      "step": 696
    },
    {
      "completion_length": 87.53125,
      "epoch": 3.182648401826484,
      "grad_norm": 6.58709192276001,
      "kl": 0.1005859375,
      "learning_rate": 6.817351598173516e-07,
      "loss": 0.004,
      "reward": 1.7153645753860474,
      "reward_std": 0.13286828622221947,
      "rewards/accuracy_reward": 0.7153644859790802,
      "rewards/format_reward": 1.0,
      "step": 697
    },
    {
      "completion_length": 70.234375,
      "epoch": 3.1872146118721463,
      "grad_norm": 2.434600591659546,
      "kl": 0.145751953125,
      "learning_rate": 6.812785388127854e-07,
      "loss": 0.0058,
      "reward": 1.7907050848007202,
      "reward_std": 0.12088606879115105,
      "rewards/accuracy_reward": 0.7907051146030426,
      "rewards/format_reward": 1.0,
      "step": 698
    },
    {
      "completion_length": 68.5,
      "epoch": 3.191780821917808,
      "grad_norm": 26.128162384033203,
      "kl": 0.126953125,
      "learning_rate": 6.808219178082191e-07,
      "loss": 0.0051,
      "reward": 1.715624988079071,
      "reward_std": 0.2467075139284134,
      "rewards/accuracy_reward": 0.715624988079071,
      "rewards/format_reward": 1.0,
      "step": 699
    },
    {
      "completion_length": 87.078125,
      "epoch": 3.1963470319634704,
      "grad_norm": 3.1288247108459473,
      "kl": 0.14404296875,
      "learning_rate": 6.80365296803653e-07,
      "loss": 0.0057,
      "reward": 1.860937476158142,
      "reward_std": 0.08299508690834045,
      "rewards/accuracy_reward": 0.8687499463558197,
      "rewards/format_reward": 0.9921875,
      "step": 700
    },
    {
      "completion_length": 67.5,
      "epoch": 3.2009132420091326,
      "grad_norm": 7.457474708557129,
      "kl": 0.1416015625,
      "learning_rate": 6.799086757990867e-07,
      "loss": 0.0057,
      "reward": 1.6931147575378418,
      "reward_std": 0.18138662725687027,
      "rewards/accuracy_reward": 0.6931147575378418,
      "rewards/format_reward": 1.0,
      "step": 701
    },
    {
      "completion_length": 78.4921875,
      "epoch": 3.2054794520547945,
      "grad_norm": 3.00228214263916,
      "kl": 0.182861328125,
      "learning_rate": 6.794520547945205e-07,
      "loss": 0.0073,
      "reward": 1.7272436618804932,
      "reward_std": 0.18743212521076202,
      "rewards/accuracy_reward": 0.727243572473526,
      "rewards/format_reward": 1.0,
      "step": 702
    },
    {
      "completion_length": 83.515625,
      "epoch": 3.2100456621004567,
      "grad_norm": 2.2364189624786377,
      "kl": 0.10693359375,
      "learning_rate": 6.789954337899543e-07,
      "loss": 0.0043,
      "reward": 1.7500601410865784,
      "reward_std": 0.10543964058160782,
      "rewards/accuracy_reward": 0.7500600218772888,
      "rewards/format_reward": 1.0,
      "step": 703
    },
    {
      "completion_length": 59.78125,
      "epoch": 3.2146118721461185,
      "grad_norm": 15.17089557647705,
      "kl": 0.64404296875,
      "learning_rate": 6.78538812785388e-07,
      "loss": 0.0258,
      "reward": 1.5231274366378784,
      "reward_std": 0.3583277612924576,
      "rewards/accuracy_reward": 0.5309399664402008,
      "rewards/format_reward": 0.9921875,
      "step": 704
    },
    {
      "completion_length": 85.2734375,
      "epoch": 3.219178082191781,
      "grad_norm": 1.857921838760376,
      "kl": 0.121826171875,
      "learning_rate": 6.78082191780822e-07,
      "loss": 0.0049,
      "reward": 1.8244792222976685,
      "reward_std": 0.08049174584448338,
      "rewards/accuracy_reward": 0.8244791626930237,
      "rewards/format_reward": 1.0,
      "step": 705
    },
    {
      "completion_length": 81.7578125,
      "epoch": 3.223744292237443,
      "grad_norm": 6.063529014587402,
      "kl": 0.11328125,
      "learning_rate": 6.776255707762557e-07,
      "loss": 0.0045,
      "reward": 1.7606770992279053,
      "reward_std": 0.21317215263843536,
      "rewards/accuracy_reward": 0.7684895098209381,
      "rewards/format_reward": 0.9921875,
      "step": 706
    },
    {
      "completion_length": 86.984375,
      "epoch": 3.228310502283105,
      "grad_norm": 1.6526107788085938,
      "kl": 0.1376953125,
      "learning_rate": 6.771689497716894e-07,
      "loss": 0.0055,
      "reward": 1.6617188453674316,
      "reward_std": 0.175977885723114,
      "rewards/accuracy_reward": 0.6695312261581421,
      "rewards/format_reward": 0.9921875,
      "step": 707
    },
    {
      "completion_length": 90.5234375,
      "epoch": 3.232876712328767,
      "grad_norm": 3.4753360748291016,
      "kl": 0.090087890625,
      "learning_rate": 6.767123287671233e-07,
      "loss": 0.0036,
      "reward": 1.7570313215255737,
      "reward_std": 0.09644587151706219,
      "rewards/accuracy_reward": 0.7570312023162842,
      "rewards/format_reward": 1.0,
      "step": 708
    },
    {
      "completion_length": 78.34375,
      "epoch": 3.237442922374429,
      "grad_norm": 5.928980350494385,
      "kl": 0.1376953125,
      "learning_rate": 6.76255707762557e-07,
      "loss": 0.0055,
      "reward": 1.7825521230697632,
      "reward_std": 0.19306360930204391,
      "rewards/accuracy_reward": 0.7825520932674408,
      "rewards/format_reward": 1.0,
      "step": 709
    },
    {
      "completion_length": 78.015625,
      "epoch": 3.2420091324200913,
      "grad_norm": 5.787877559661865,
      "kl": 0.098876953125,
      "learning_rate": 6.757990867579907e-07,
      "loss": 0.004,
      "reward": 1.6989798545837402,
      "reward_std": 0.22939839959144592,
      "rewards/accuracy_reward": 0.7146047651767731,
      "rewards/format_reward": 0.984375,
      "step": 710
    },
    {
      "completion_length": 73.984375,
      "epoch": 3.2465753424657535,
      "grad_norm": 1.9611449241638184,
      "kl": 0.135986328125,
      "learning_rate": 6.753424657534246e-07,
      "loss": 0.0054,
      "reward": 1.7150809168815613,
      "reward_std": 0.12537125870585442,
      "rewards/accuracy_reward": 0.7228934466838837,
      "rewards/format_reward": 0.9921875,
      "step": 711
    },
    {
      "completion_length": 76.640625,
      "epoch": 3.2511415525114153,
      "grad_norm": 2.202395439147949,
      "kl": 0.15673828125,
      "learning_rate": 6.748858447488584e-07,
      "loss": 0.0063,
      "reward": 1.7408854365348816,
      "reward_std": 0.18467864021658897,
      "rewards/accuracy_reward": 0.7486979365348816,
      "rewards/format_reward": 0.9921875,
      "step": 712
    },
    {
      "completion_length": 91.578125,
      "epoch": 3.2557077625570776,
      "grad_norm": 2.922081232070923,
      "kl": 0.075439453125,
      "learning_rate": 6.744292237442923e-07,
      "loss": 0.003,
      "reward": 1.6480501890182495,
      "reward_std": 0.08936248533427715,
      "rewards/accuracy_reward": 0.6480501890182495,
      "rewards/format_reward": 1.0,
      "step": 713
    },
    {
      "completion_length": 97.6953125,
      "epoch": 3.26027397260274,
      "grad_norm": 1.8138573169708252,
      "kl": 0.0810546875,
      "learning_rate": 6.73972602739726e-07,
      "loss": 0.0032,
      "reward": 1.678125023841858,
      "reward_std": 0.17582245916128159,
      "rewards/accuracy_reward": 0.7015624642372131,
      "rewards/format_reward": 0.9765625,
      "step": 714
    },
    {
      "completion_length": 81.609375,
      "epoch": 3.2648401826484017,
      "grad_norm": 2.29533052444458,
      "kl": 0.117919921875,
      "learning_rate": 6.735159817351597e-07,
      "loss": 0.0047,
      "reward": 1.5667868852615356,
      "reward_std": 0.23160236328840256,
      "rewards/accuracy_reward": 0.5902243554592133,
      "rewards/format_reward": 0.9765625,
      "step": 715
    },
    {
      "completion_length": 79.046875,
      "epoch": 3.269406392694064,
      "grad_norm": 3.5871529579162598,
      "kl": 0.1748046875,
      "learning_rate": 6.730593607305936e-07,
      "loss": 0.007,
      "reward": 1.5300781726837158,
      "reward_std": 0.381600484251976,
      "rewards/accuracy_reward": 0.592578113079071,
      "rewards/format_reward": 0.9375,
      "step": 716
    },
    {
      "completion_length": 81.21875,
      "epoch": 3.2739726027397262,
      "grad_norm": 7.025841236114502,
      "kl": 0.177734375,
      "learning_rate": 6.726027397260273e-07,
      "loss": 0.0071,
      "reward": 1.5895833373069763,
      "reward_std": 0.3286707103252411,
      "rewards/accuracy_reward": 0.6442708373069763,
      "rewards/format_reward": 0.9453125,
      "step": 717
    },
    {
      "completion_length": 79.859375,
      "epoch": 3.278538812785388,
      "grad_norm": 3.0471503734588623,
      "kl": 0.099609375,
      "learning_rate": 6.721461187214613e-07,
      "loss": 0.004,
      "reward": 1.6017058491706848,
      "reward_std": 0.18348699063062668,
      "rewards/accuracy_reward": 0.6095183193683624,
      "rewards/format_reward": 0.9921875,
      "step": 718
    },
    {
      "completion_length": 78.8515625,
      "epoch": 3.2831050228310503,
      "grad_norm": 2.4444265365600586,
      "kl": 0.09912109375,
      "learning_rate": 6.71689497716895e-07,
      "loss": 0.004,
      "reward": 1.5035117268562317,
      "reward_std": 0.2922291085124016,
      "rewards/accuracy_reward": 0.5191366672515869,
      "rewards/format_reward": 0.984375,
      "step": 719
    },
    {
      "completion_length": 80.515625,
      "epoch": 3.287671232876712,
      "grad_norm": 4.203682899475098,
      "kl": 0.11962890625,
      "learning_rate": 6.712328767123287e-07,
      "loss": 0.0048,
      "reward": 1.5486140251159668,
      "reward_std": 0.18936936557292938,
      "rewards/accuracy_reward": 0.5564264357089996,
      "rewards/format_reward": 0.9921875,
      "step": 720
    },
    {
      "completion_length": 90.1015625,
      "epoch": 3.2922374429223744,
      "grad_norm": 3.4824612140655518,
      "kl": 0.103759765625,
      "learning_rate": 6.707762557077626e-07,
      "loss": 0.0042,
      "reward": 1.7547819018363953,
      "reward_std": 0.14520251005887985,
      "rewards/accuracy_reward": 0.7782192826271057,
      "rewards/format_reward": 0.9765625,
      "step": 721
    },
    {
      "completion_length": 71.7734375,
      "epoch": 3.2968036529680367,
      "grad_norm": 1.9310115575790405,
      "kl": 0.18310546875,
      "learning_rate": 6.703196347031963e-07,
      "loss": 0.0073,
      "reward": 1.494028091430664,
      "reward_std": 0.24336419254541397,
      "rewards/accuracy_reward": 0.5174656212329865,
      "rewards/format_reward": 0.9765625,
      "step": 722
    },
    {
      "completion_length": 87.65625,
      "epoch": 3.3013698630136985,
      "grad_norm": 2.703258991241455,
      "kl": 0.1064453125,
      "learning_rate": 6.6986301369863e-07,
      "loss": 0.0043,
      "reward": 1.6925916075706482,
      "reward_std": 0.22123625874519348,
      "rewards/accuracy_reward": 0.7238415777683258,
      "rewards/format_reward": 0.96875,
      "step": 723
    },
    {
      "completion_length": 86.796875,
      "epoch": 3.3059360730593608,
      "grad_norm": 1.1868500709533691,
      "kl": 0.10107421875,
      "learning_rate": 6.694063926940639e-07,
      "loss": 0.004,
      "reward": 1.868213951587677,
      "reward_std": 0.16314804006833583,
      "rewards/accuracy_reward": 0.8916513323783875,
      "rewards/format_reward": 0.9765625,
      "step": 724
    },
    {
      "completion_length": 89.765625,
      "epoch": 3.3105022831050226,
      "grad_norm": 1.9978954792022705,
      "kl": 0.09814453125,
      "learning_rate": 6.689497716894977e-07,
      "loss": 0.0039,
      "reward": 1.6983258724212646,
      "reward_std": 0.1776369959115982,
      "rewards/accuracy_reward": 0.7139508724212646,
      "rewards/format_reward": 0.984375,
      "step": 725
    },
    {
      "completion_length": 82.3515625,
      "epoch": 3.315068493150685,
      "grad_norm": 2.7241666316986084,
      "kl": 0.0947265625,
      "learning_rate": 6.684931506849316e-07,
      "loss": 0.0038,
      "reward": 1.6841517686843872,
      "reward_std": 0.20553293824195862,
      "rewards/accuracy_reward": 0.6997767686843872,
      "rewards/format_reward": 0.984375,
      "step": 726
    },
    {
      "completion_length": 66.3515625,
      "epoch": 3.319634703196347,
      "grad_norm": 2.9195847511291504,
      "kl": 0.1630859375,
      "learning_rate": 6.680365296803653e-07,
      "loss": 0.0065,
      "reward": 1.7539063096046448,
      "reward_std": 0.18009010702371597,
      "rewards/accuracy_reward": 0.7617186903953552,
      "rewards/format_reward": 0.9921875,
      "step": 727
    },
    {
      "completion_length": 75.4765625,
      "epoch": 3.324200913242009,
      "grad_norm": 2.223017930984497,
      "kl": 0.091064453125,
      "learning_rate": 6.67579908675799e-07,
      "loss": 0.0036,
      "reward": 1.535528302192688,
      "reward_std": 0.24083293601870537,
      "rewards/accuracy_reward": 0.5667782425880432,
      "rewards/format_reward": 0.96875,
      "step": 728
    },
    {
      "completion_length": 78.046875,
      "epoch": 3.328767123287671,
      "grad_norm": 2.5224952697753906,
      "kl": 0.102783203125,
      "learning_rate": 6.671232876712329e-07,
      "loss": 0.0041,
      "reward": 1.732812523841858,
      "reward_std": 0.21059568226337433,
      "rewards/accuracy_reward": 0.7640624642372131,
      "rewards/format_reward": 0.96875,
      "step": 729
    },
    {
      "completion_length": 71.8203125,
      "epoch": 3.3333333333333335,
      "grad_norm": 2.83807110786438,
      "kl": 0.10498046875,
      "learning_rate": 6.666666666666666e-07,
      "loss": 0.0042,
      "reward": 1.579541265964508,
      "reward_std": 0.26198120415210724,
      "rewards/accuracy_reward": 0.5951661765575409,
      "rewards/format_reward": 0.984375,
      "step": 730
    },
    {
      "completion_length": 63.3125,
      "epoch": 3.3378995433789953,
      "grad_norm": 3.6584157943725586,
      "kl": 0.16015625,
      "learning_rate": 6.662100456621003e-07,
      "loss": 0.0064,
      "reward": 1.7500744462013245,
      "reward_std": 0.19155671447515488,
      "rewards/accuracy_reward": 0.7500744163990021,
      "rewards/format_reward": 1.0,
      "step": 731
    },
    {
      "completion_length": 56.4765625,
      "epoch": 3.3424657534246576,
      "grad_norm": 2.3448262214660645,
      "kl": 0.13818359375,
      "learning_rate": 6.657534246575343e-07,
      "loss": 0.0055,
      "reward": 1.6341642141342163,
      "reward_std": 0.20584679394960403,
      "rewards/accuracy_reward": 0.6576017141342163,
      "rewards/format_reward": 0.9765625,
      "step": 732
    },
    {
      "completion_length": 56.953125,
      "epoch": 3.34703196347032,
      "grad_norm": 2.6592612266540527,
      "kl": 0.1484375,
      "learning_rate": 6.65296803652968e-07,
      "loss": 0.0059,
      "reward": 1.694618046283722,
      "reward_std": 0.17367403209209442,
      "rewards/accuracy_reward": 0.6946180760860443,
      "rewards/format_reward": 1.0,
      "step": 733
    },
    {
      "completion_length": 86.1640625,
      "epoch": 3.3515981735159817,
      "grad_norm": 1.7500590085983276,
      "kl": 0.128662109375,
      "learning_rate": 6.648401826484019e-07,
      "loss": 0.0051,
      "reward": 1.7081771492958069,
      "reward_std": 0.11748043447732925,
      "rewards/accuracy_reward": 0.7159895300865173,
      "rewards/format_reward": 0.9921875,
      "step": 734
    },
    {
      "completion_length": 73.1171875,
      "epoch": 3.356164383561644,
      "grad_norm": 2.0890016555786133,
      "kl": 0.111328125,
      "learning_rate": 6.643835616438356e-07,
      "loss": 0.0045,
      "reward": 1.669720709323883,
      "reward_std": 0.14257927983999252,
      "rewards/accuracy_reward": 0.6775331199169159,
      "rewards/format_reward": 0.9921875,
      "step": 735
    },
    {
      "completion_length": 75.4453125,
      "epoch": 3.3607305936073057,
      "grad_norm": 1.7575141191482544,
      "kl": 0.10302734375,
      "learning_rate": 6.639269406392693e-07,
      "loss": 0.0041,
      "reward": 1.760156273841858,
      "reward_std": 0.1406225487589836,
      "rewards/accuracy_reward": 0.7679686844348907,
      "rewards/format_reward": 0.9921875,
      "step": 736
    },
    {
      "completion_length": 79.0625,
      "epoch": 3.365296803652968,
      "grad_norm": 1.9857916831970215,
      "kl": 0.0750732421875,
      "learning_rate": 6.634703196347032e-07,
      "loss": 0.003,
      "reward": 1.8437398672103882,
      "reward_std": 0.10023375414311886,
      "rewards/accuracy_reward": 0.8515522480010986,
      "rewards/format_reward": 0.9921875,
      "step": 737
    },
    {
      "completion_length": 43.953125,
      "epoch": 3.3698630136986303,
      "grad_norm": 1.8931981325149536,
      "kl": 0.23876953125,
      "learning_rate": 6.63013698630137e-07,
      "loss": 0.0096,
      "reward": 1.777430534362793,
      "reward_std": 0.2545855790376663,
      "rewards/accuracy_reward": 0.808680534362793,
      "rewards/format_reward": 0.96875,
      "step": 738
    },
    {
      "completion_length": 66.671875,
      "epoch": 3.374429223744292,
      "grad_norm": 4.250927925109863,
      "kl": 0.11474609375,
      "learning_rate": 6.625570776255707e-07,
      "loss": 0.0046,
      "reward": 1.5928385257720947,
      "reward_std": 0.13594963401556015,
      "rewards/accuracy_reward": 0.5928385257720947,
      "rewards/format_reward": 1.0,
      "step": 739
    },
    {
      "completion_length": 59.28125,
      "epoch": 3.3789954337899544,
      "grad_norm": 1.9088664054870605,
      "kl": 0.114990234375,
      "learning_rate": 6.621004566210046e-07,
      "loss": 0.0046,
      "reward": 1.5951822996139526,
      "reward_std": 0.1505616046488285,
      "rewards/accuracy_reward": 0.5951822698116302,
      "rewards/format_reward": 1.0,
      "step": 740
    },
    {
      "completion_length": 74.8984375,
      "epoch": 3.383561643835616,
      "grad_norm": 2.5536303520202637,
      "kl": 0.126953125,
      "learning_rate": 6.616438356164383e-07,
      "loss": 0.0051,
      "reward": 1.6819568276405334,
      "reward_std": 0.1260900031775236,
      "rewards/accuracy_reward": 0.6819568276405334,
      "rewards/format_reward": 1.0,
      "step": 741
    },
    {
      "completion_length": 81.7421875,
      "epoch": 3.3881278538812785,
      "grad_norm": 1.9142627716064453,
      "kl": 0.091064453125,
      "learning_rate": 6.61187214611872e-07,
      "loss": 0.0036,
      "reward": 1.7414063215255737,
      "reward_std": 0.15430963411927223,
      "rewards/accuracy_reward": 0.7492187023162842,
      "rewards/format_reward": 0.9921875,
      "step": 742
    },
    {
      "completion_length": 63.8828125,
      "epoch": 3.3926940639269407,
      "grad_norm": 2.298887252807617,
      "kl": 0.15625,
      "learning_rate": 6.607305936073059e-07,
      "loss": 0.0063,
      "reward": 1.6809749007225037,
      "reward_std": 0.13766025006771088,
      "rewards/accuracy_reward": 0.6809749007225037,
      "rewards/format_reward": 1.0,
      "step": 743
    },
    {
      "completion_length": 62.3046875,
      "epoch": 3.3972602739726026,
      "grad_norm": 2.9588382244110107,
      "kl": 0.13916015625,
      "learning_rate": 6.602739726027396e-07,
      "loss": 0.0056,
      "reward": 1.668749988079071,
      "reward_std": 0.28636179864406586,
      "rewards/accuracy_reward": 0.692187488079071,
      "rewards/format_reward": 0.9765625,
      "step": 744
    },
    {
      "completion_length": 58.328125,
      "epoch": 3.401826484018265,
      "grad_norm": 2.7415127754211426,
      "kl": 0.130126953125,
      "learning_rate": 6.598173515981736e-07,
      "loss": 0.0052,
      "reward": 1.755094826221466,
      "reward_std": 0.2010849490761757,
      "rewards/accuracy_reward": 0.7707198262214661,
      "rewards/format_reward": 0.984375,
      "step": 745
    },
    {
      "completion_length": 61.15625,
      "epoch": 3.406392694063927,
      "grad_norm": 2.1522672176361084,
      "kl": 0.138427734375,
      "learning_rate": 6.593607305936073e-07,
      "loss": 0.0055,
      "reward": 1.779687523841858,
      "reward_std": 0.1337989792227745,
      "rewards/accuracy_reward": 0.7796874344348907,
      "rewards/format_reward": 1.0,
      "step": 746
    },
    {
      "completion_length": 93.4296875,
      "epoch": 3.410958904109589,
      "grad_norm": 2.299729824066162,
      "kl": 0.127197265625,
      "learning_rate": 6.58904109589041e-07,
      "loss": 0.0051,
      "reward": 1.787500023841858,
      "reward_std": 0.12756995856761932,
      "rewards/accuracy_reward": 0.8031249642372131,
      "rewards/format_reward": 0.984375,
      "step": 747
    },
    {
      "completion_length": 72.625,
      "epoch": 3.415525114155251,
      "grad_norm": 3.2953529357910156,
      "kl": 0.13623046875,
      "learning_rate": 6.584474885844749e-07,
      "loss": 0.0055,
      "reward": 1.5497395992279053,
      "reward_std": 0.2357780486345291,
      "rewards/accuracy_reward": 0.5575520694255829,
      "rewards/format_reward": 0.9921875,
      "step": 748
    },
    {
      "completion_length": 66.1015625,
      "epoch": 3.4200913242009134,
      "grad_norm": 3.811732530593872,
      "kl": 0.15625,
      "learning_rate": 6.579908675799086e-07,
      "loss": 0.0063,
      "reward": 1.465334177017212,
      "reward_std": 0.281493678689003,
      "rewards/accuracy_reward": 0.48095911741256714,
      "rewards/format_reward": 0.984375,
      "step": 749
    },
    {
      "completion_length": 78.1171875,
      "epoch": 3.4246575342465753,
      "grad_norm": 6.028533458709717,
      "kl": 0.14501953125,
      "learning_rate": 6.575342465753423e-07,
      "loss": 0.0058,
      "reward": 1.612395167350769,
      "reward_std": 0.1732819825410843,
      "rewards/accuracy_reward": 0.6202076524496078,
      "rewards/format_reward": 0.9921875,
      "step": 750
    },
    {
      "completion_length": 80.921875,
      "epoch": 3.4292237442922375,
      "grad_norm": 2.363414764404297,
      "kl": 0.10107421875,
      "learning_rate": 6.570776255707762e-07,
      "loss": 0.004,
      "reward": 1.7257813215255737,
      "reward_std": 0.15576134249567986,
      "rewards/accuracy_reward": 0.7257812023162842,
      "rewards/format_reward": 1.0,
      "step": 751
    },
    {
      "completion_length": 60.625,
      "epoch": 3.4337899543378994,
      "grad_norm": 2.295811176300049,
      "kl": 0.1298828125,
      "learning_rate": 6.5662100456621e-07,
      "loss": 0.0052,
      "reward": 1.8014204502105713,
      "reward_std": 0.14195573329925537,
      "rewards/accuracy_reward": 0.8092329502105713,
      "rewards/format_reward": 0.9921875,
      "step": 752
    },
    {
      "completion_length": 79.6875,
      "epoch": 3.4383561643835616,
      "grad_norm": 1.9623810052871704,
      "kl": 0.09326171875,
      "learning_rate": 6.561643835616439e-07,
      "loss": 0.0037,
      "reward": 1.654836356639862,
      "reward_std": 0.16391075402498245,
      "rewards/accuracy_reward": 0.6704612970352173,
      "rewards/format_reward": 0.984375,
      "step": 753
    },
    {
      "completion_length": 80.84375,
      "epoch": 3.442922374429224,
      "grad_norm": 2.4038403034210205,
      "kl": 0.117919921875,
      "learning_rate": 6.557077625570776e-07,
      "loss": 0.0047,
      "reward": 1.7968750596046448,
      "reward_std": 0.10437997803092003,
      "rewards/accuracy_reward": 0.7968749105930328,
      "rewards/format_reward": 1.0,
      "step": 754
    },
    {
      "completion_length": 72.078125,
      "epoch": 3.4474885844748857,
      "grad_norm": 1.999448299407959,
      "kl": 0.11962890625,
      "learning_rate": 6.552511415525113e-07,
      "loss": 0.0048,
      "reward": 1.829807698726654,
      "reward_std": 0.07720155641436577,
      "rewards/accuracy_reward": 0.8298076391220093,
      "rewards/format_reward": 1.0,
      "step": 755
    },
    {
      "completion_length": 75.8828125,
      "epoch": 3.452054794520548,
      "grad_norm": 1.9827911853790283,
      "kl": 0.0849609375,
      "learning_rate": 6.547945205479452e-07,
      "loss": 0.0034,
      "reward": 1.7932292222976685,
      "reward_std": 0.14317410439252853,
      "rewards/accuracy_reward": 0.8088541030883789,
      "rewards/format_reward": 0.984375,
      "step": 756
    },
    {
      "completion_length": 77.359375,
      "epoch": 3.45662100456621,
      "grad_norm": 2.601567268371582,
      "kl": 0.118896484375,
      "learning_rate": 6.543378995433789e-07,
      "loss": 0.0048,
      "reward": 1.7324219346046448,
      "reward_std": 0.15861116349697113,
      "rewards/accuracy_reward": 0.740234375,
      "rewards/format_reward": 0.9921875,
      "step": 757
    },
    {
      "completion_length": 83.2421875,
      "epoch": 3.461187214611872,
      "grad_norm": 2.38321852684021,
      "kl": 0.133056640625,
      "learning_rate": 6.538812785388129e-07,
      "loss": 0.0053,
      "reward": 1.6742457747459412,
      "reward_std": 0.18774619698524475,
      "rewards/accuracy_reward": 0.6820583343505859,
      "rewards/format_reward": 0.9921875,
      "step": 758
    },
    {
      "completion_length": 80.3671875,
      "epoch": 3.4657534246575343,
      "grad_norm": 1.9422838687896729,
      "kl": 0.128662109375,
      "learning_rate": 6.534246575342466e-07,
      "loss": 0.0051,
      "reward": 1.7861049175262451,
      "reward_std": 0.09223857149481773,
      "rewards/accuracy_reward": 0.7861048579216003,
      "rewards/format_reward": 1.0,
      "step": 759
    },
    {
      "completion_length": 77.8125,
      "epoch": 3.470319634703196,
      "grad_norm": 2.641580820083618,
      "kl": 0.119873046875,
      "learning_rate": 6.529680365296803e-07,
      "loss": 0.0048,
      "reward": 1.6968767046928406,
      "reward_std": 0.2295239269733429,
      "rewards/accuracy_reward": 0.7046891748905182,
      "rewards/format_reward": 0.9921875,
      "step": 760
    },
    {
      "completion_length": 73.4375,
      "epoch": 3.4748858447488584,
      "grad_norm": 2.904656171798706,
      "kl": 0.12744140625,
      "learning_rate": 6.525114155251142e-07,
      "loss": 0.0051,
      "reward": 1.5930989980697632,
      "reward_std": 0.1885790079832077,
      "rewards/accuracy_reward": 0.5930989682674408,
      "rewards/format_reward": 1.0,
      "step": 761
    },
    {
      "completion_length": 78.75,
      "epoch": 3.4794520547945207,
      "grad_norm": 2.564729690551758,
      "kl": 0.1591796875,
      "learning_rate": 6.520547945205479e-07,
      "loss": 0.0064,
      "reward": 1.7561274766921997,
      "reward_std": 0.1594456396996975,
      "rewards/accuracy_reward": 0.7639399170875549,
      "rewards/format_reward": 0.9921875,
      "step": 762
    },
    {
      "completion_length": 85.6953125,
      "epoch": 3.4840182648401825,
      "grad_norm": 4.693384170532227,
      "kl": 0.117919921875,
      "learning_rate": 6.515981735159816e-07,
      "loss": 0.0047,
      "reward": 1.4937500357627869,
      "reward_std": 0.26740533858537674,
      "rewards/accuracy_reward": 0.5093749910593033,
      "rewards/format_reward": 0.984375,
      "step": 763
    },
    {
      "completion_length": 74.28125,
      "epoch": 3.4885844748858448,
      "grad_norm": 3.116814136505127,
      "kl": 0.13037109375,
      "learning_rate": 6.511415525114155e-07,
      "loss": 0.0052,
      "reward": 1.760156273841858,
      "reward_std": 0.17397862672805786,
      "rewards/accuracy_reward": 0.7679687440395355,
      "rewards/format_reward": 0.9921875,
      "step": 764
    },
    {
      "completion_length": 72.8671875,
      "epoch": 3.493150684931507,
      "grad_norm": 4.106734752655029,
      "kl": 0.319091796875,
      "learning_rate": 6.506849315068493e-07,
      "loss": 0.0128,
      "reward": 1.6917535066604614,
      "reward_std": 0.1726042479276657,
      "rewards/accuracy_reward": 0.7073784470558167,
      "rewards/format_reward": 0.984375,
      "step": 765
    },
    {
      "completion_length": 72.4296875,
      "epoch": 3.497716894977169,
      "grad_norm": 2.7671029567718506,
      "kl": 0.11962890625,
      "learning_rate": 6.502283105022832e-07,
      "loss": 0.0048,
      "reward": 1.7573699951171875,
      "reward_std": 0.12242420390248299,
      "rewards/accuracy_reward": 0.7573699355125427,
      "rewards/format_reward": 1.0,
      "step": 766
    },
    {
      "completion_length": 91.5078125,
      "epoch": 3.502283105022831,
      "grad_norm": 2.353253126144409,
      "kl": 0.12158203125,
      "learning_rate": 6.497716894977169e-07,
      "loss": 0.0049,
      "reward": 1.5520833730697632,
      "reward_std": 0.21783916652202606,
      "rewards/accuracy_reward": 0.5677083134651184,
      "rewards/format_reward": 0.984375,
      "step": 767
    },
    {
      "completion_length": 86.0390625,
      "epoch": 3.506849315068493,
      "grad_norm": 2.568850040435791,
      "kl": 0.108642578125,
      "learning_rate": 6.493150684931506e-07,
      "loss": 0.0043,
      "reward": 1.8059896230697632,
      "reward_std": 0.1456034928560257,
      "rewards/accuracy_reward": 0.8138020634651184,
      "rewards/format_reward": 0.9921875,
      "step": 768
    },
    {
      "completion_length": 97.078125,
      "epoch": 3.5114155251141552,
      "grad_norm": 9.077414512634277,
      "kl": 0.084716796875,
      "learning_rate": 6.488584474885845e-07,
      "loss": 0.0034,
      "reward": 1.6987351775169373,
      "reward_std": 0.2012891098856926,
      "rewards/accuracy_reward": 0.7221725881099701,
      "rewards/format_reward": 0.9765625,
      "step": 769
    },
    {
      "completion_length": 90.3828125,
      "epoch": 3.5159817351598175,
      "grad_norm": 4.764615058898926,
      "kl": 0.107421875,
      "learning_rate": 6.484018264840182e-07,
      "loss": 0.0043,
      "reward": 1.6770833730697632,
      "reward_std": 0.20456601679325104,
      "rewards/accuracy_reward": 0.6927083432674408,
      "rewards/format_reward": 0.984375,
      "step": 770
    },
    {
      "completion_length": 91.53125,
      "epoch": 3.5205479452054793,
      "grad_norm": 2.192065477371216,
      "kl": 0.125244140625,
      "learning_rate": 6.479452054794519e-07,
      "loss": 0.005,
      "reward": 1.822656273841858,
      "reward_std": 0.11972266435623169,
      "rewards/accuracy_reward": 0.8226562142372131,
      "rewards/format_reward": 1.0,
      "step": 771
    },
    {
      "completion_length": 66.875,
      "epoch": 3.5251141552511416,
      "grad_norm": 2.9831643104553223,
      "kl": 0.16259765625,
      "learning_rate": 6.474885844748859e-07,
      "loss": 0.0065,
      "reward": 1.6729073524475098,
      "reward_std": 0.18848184496164322,
      "rewards/accuracy_reward": 0.6807198524475098,
      "rewards/format_reward": 0.9921875,
      "step": 772
    },
    {
      "completion_length": 112.515625,
      "epoch": 3.5296803652968034,
      "grad_norm": 4.550014495849609,
      "kl": 0.107177734375,
      "learning_rate": 6.470319634703196e-07,
      "loss": 0.0043,
      "reward": 1.748437523841858,
      "reward_std": 0.21249166131019592,
      "rewards/accuracy_reward": 0.7796874046325684,
      "rewards/format_reward": 0.96875,
      "step": 773
    },
    {
      "completion_length": 86.4453125,
      "epoch": 3.5342465753424657,
      "grad_norm": 12.177788734436035,
      "kl": 0.134521484375,
      "learning_rate": 6.465753424657535e-07,
      "loss": 0.0054,
      "reward": 1.5592572093009949,
      "reward_std": 0.3570391535758972,
      "rewards/accuracy_reward": 0.6061321794986725,
      "rewards/format_reward": 0.953125,
      "step": 774
    },
    {
      "completion_length": 91.90625,
      "epoch": 3.538812785388128,
      "grad_norm": 2.1934075355529785,
      "kl": 0.09423828125,
      "learning_rate": 6.461187214611872e-07,
      "loss": 0.0038,
      "reward": 1.7695313096046448,
      "reward_std": 0.2720255181193352,
      "rewards/accuracy_reward": 0.7929687201976776,
      "rewards/format_reward": 0.9765625,
      "step": 775
    },
    {
      "completion_length": 93.34375,
      "epoch": 3.54337899543379,
      "grad_norm": 1.7237935066223145,
      "kl": 0.10302734375,
      "learning_rate": 6.456621004566209e-07,
      "loss": 0.0041,
      "reward": 1.839062511920929,
      "reward_std": 0.08532825112342834,
      "rewards/accuracy_reward": 0.8390624821186066,
      "rewards/format_reward": 1.0,
      "step": 776
    },
    {
      "completion_length": 68.5234375,
      "epoch": 3.547945205479452,
      "grad_norm": 9.828876495361328,
      "kl": 0.130859375,
      "learning_rate": 6.452054794520548e-07,
      "loss": 0.0052,
      "reward": 1.699999988079071,
      "reward_std": 0.23911622911691666,
      "rewards/accuracy_reward": 0.7156249582767487,
      "rewards/format_reward": 0.984375,
      "step": 777
    },
    {
      "completion_length": 111.7265625,
      "epoch": 3.5525114155251143,
      "grad_norm": 2.7527942657470703,
      "kl": 0.0751953125,
      "learning_rate": 6.447488584474886e-07,
      "loss": 0.003,
      "reward": 1.6765625476837158,
      "reward_std": 0.17730073630809784,
      "rewards/accuracy_reward": 0.6921874582767487,
      "rewards/format_reward": 0.984375,
      "step": 778
    },
    {
      "completion_length": 86.328125,
      "epoch": 3.557077625570776,
      "grad_norm": 8.820732116699219,
      "kl": 0.127685546875,
      "learning_rate": 6.442922374429223e-07,
      "loss": 0.0051,
      "reward": 1.5761924982070923,
      "reward_std": 0.2583516389131546,
      "rewards/accuracy_reward": 0.5996299386024475,
      "rewards/format_reward": 0.9765625,
      "step": 779
    },
    {
      "completion_length": 83.546875,
      "epoch": 3.5616438356164384,
      "grad_norm": 4.555209636688232,
      "kl": 0.11083984375,
      "learning_rate": 6.438356164383562e-07,
      "loss": 0.0044,
      "reward": 1.6642005443572998,
      "reward_std": 0.2603069022297859,
      "rewards/accuracy_reward": 0.6876380443572998,
      "rewards/format_reward": 0.9765625,
      "step": 780
    },
    {
      "completion_length": 88.859375,
      "epoch": 3.5662100456621006,
      "grad_norm": 3.635876417160034,
      "kl": 0.133056640625,
      "learning_rate": 6.433789954337899e-07,
      "loss": 0.0053,
      "reward": 1.6102182865142822,
      "reward_std": 0.2444395273923874,
      "rewards/accuracy_reward": 0.6414682567119598,
      "rewards/format_reward": 0.96875,
      "step": 781
    },
    {
      "completion_length": 83.7734375,
      "epoch": 3.5707762557077625,
      "grad_norm": 2.809690237045288,
      "kl": 0.118896484375,
      "learning_rate": 6.429223744292238e-07,
      "loss": 0.0048,
      "reward": 1.6714159846305847,
      "reward_std": 0.2755907028913498,
      "rewards/accuracy_reward": 0.6948534548282623,
      "rewards/format_reward": 0.9765625,
      "step": 782
    },
    {
      "completion_length": 66.0,
      "epoch": 3.5753424657534247,
      "grad_norm": 4.247856140136719,
      "kl": 0.1552734375,
      "learning_rate": 6.424657534246575e-07,
      "loss": 0.0062,
      "reward": 1.5200520753860474,
      "reward_std": 0.2615704759955406,
      "rewards/accuracy_reward": 0.5434895753860474,
      "rewards/format_reward": 0.9765625,
      "step": 783
    },
    {
      "completion_length": 86.296875,
      "epoch": 3.5799086757990866,
      "grad_norm": 3.303392171859741,
      "kl": 0.17236328125,
      "learning_rate": 6.420091324200912e-07,
      "loss": 0.0069,
      "reward": 1.7383928894996643,
      "reward_std": 0.20217304676771164,
      "rewards/accuracy_reward": 0.7618303298950195,
      "rewards/format_reward": 0.9765625,
      "step": 784
    },
    {
      "completion_length": 94.890625,
      "epoch": 3.584474885844749,
      "grad_norm": 2.6589694023132324,
      "kl": 0.105712890625,
      "learning_rate": 6.415525114155252e-07,
      "loss": 0.0042,
      "reward": 1.710684597492218,
      "reward_std": 0.20098017156124115,
      "rewards/accuracy_reward": 0.7341220676898956,
      "rewards/format_reward": 0.9765625,
      "step": 785
    },
    {
      "completion_length": 90.21875,
      "epoch": 3.589041095890411,
      "grad_norm": 2.0707075595855713,
      "kl": 0.0810546875,
      "learning_rate": 6.410958904109589e-07,
      "loss": 0.0032,
      "reward": 1.70947265625,
      "reward_std": 0.13764164596796036,
      "rewards/accuracy_reward": 0.7172850966453552,
      "rewards/format_reward": 0.9921875,
      "step": 786
    },
    {
      "completion_length": 69.5078125,
      "epoch": 3.593607305936073,
      "grad_norm": 2.431682586669922,
      "kl": 0.197265625,
      "learning_rate": 6.406392694063926e-07,
      "loss": 0.0079,
      "reward": 1.6103981137275696,
      "reward_std": 0.27858249843120575,
      "rewards/accuracy_reward": 0.6338355839252472,
      "rewards/format_reward": 0.9765625,
      "step": 787
    },
    {
      "completion_length": 73.4453125,
      "epoch": 3.598173515981735,
      "grad_norm": 2.722654342651367,
      "kl": 0.13330078125,
      "learning_rate": 6.401826484018265e-07,
      "loss": 0.0053,
      "reward": 1.7416667342185974,
      "reward_std": 0.18196804821491241,
      "rewards/accuracy_reward": 0.7651041448116302,
      "rewards/format_reward": 0.9765625,
      "step": 788
    },
    {
      "completion_length": 80.890625,
      "epoch": 3.602739726027397,
      "grad_norm": 5.7324018478393555,
      "kl": 0.151123046875,
      "learning_rate": 6.397260273972602e-07,
      "loss": 0.0061,
      "reward": 1.7906250357627869,
      "reward_std": 0.23605135083198547,
      "rewards/accuracy_reward": 0.8062499463558197,
      "rewards/format_reward": 0.984375,
      "step": 789
    },
    {
      "completion_length": 72.625,
      "epoch": 3.6073059360730593,
      "grad_norm": 2.2109618186950684,
      "kl": 0.11767578125,
      "learning_rate": 6.39269406392694e-07,
      "loss": 0.0047,
      "reward": 1.5838541984558105,
      "reward_std": 0.16635090112686157,
      "rewards/accuracy_reward": 0.583854153752327,
      "rewards/format_reward": 1.0,
      "step": 790
    },
    {
      "completion_length": 79.53125,
      "epoch": 3.6118721461187215,
      "grad_norm": 1.7857418060302734,
      "kl": 0.142822265625,
      "learning_rate": 6.388127853881278e-07,
      "loss": 0.0057,
      "reward": 1.801552414894104,
      "reward_std": 0.1699754223227501,
      "rewards/accuracy_reward": 0.8171773254871368,
      "rewards/format_reward": 0.984375,
      "step": 791
    },
    {
      "completion_length": 72.3359375,
      "epoch": 3.616438356164384,
      "grad_norm": 6.849174499511719,
      "kl": 0.111572265625,
      "learning_rate": 6.383561643835616e-07,
      "loss": 0.0045,
      "reward": 1.5342634320259094,
      "reward_std": 0.26482056826353073,
      "rewards/accuracy_reward": 0.534263402223587,
      "rewards/format_reward": 1.0,
      "step": 792
    },
    {
      "completion_length": 98.8828125,
      "epoch": 3.6210045662100456,
      "grad_norm": 7.422957897186279,
      "kl": 0.0927734375,
      "learning_rate": 6.378995433789955e-07,
      "loss": 0.0037,
      "reward": 1.7940475940704346,
      "reward_std": 0.1312719490379095,
      "rewards/accuracy_reward": 0.8096725642681122,
      "rewards/format_reward": 0.984375,
      "step": 793
    },
    {
      "completion_length": 81.0546875,
      "epoch": 3.625570776255708,
      "grad_norm": 1.819212794303894,
      "kl": 0.11572265625,
      "learning_rate": 6.374429223744292e-07,
      "loss": 0.0046,
      "reward": 1.664595365524292,
      "reward_std": 0.1356574185192585,
      "rewards/accuracy_reward": 0.6802203357219696,
      "rewards/format_reward": 0.984375,
      "step": 794
    },
    {
      "completion_length": 80.25,
      "epoch": 3.6301369863013697,
      "grad_norm": 5.144930362701416,
      "kl": 0.15966796875,
      "learning_rate": 6.369863013698629e-07,
      "loss": 0.0064,
      "reward": 1.6710898280143738,
      "reward_std": 0.27560608088970184,
      "rewards/accuracy_reward": 0.710152268409729,
      "rewards/format_reward": 0.9609375,
      "step": 795
    },
    {
      "completion_length": 70.6015625,
      "epoch": 3.634703196347032,
      "grad_norm": 5.67324686050415,
      "kl": 0.119140625,
      "learning_rate": 6.365296803652968e-07,
      "loss": 0.0048,
      "reward": 1.5117551684379578,
      "reward_std": 0.2634401321411133,
      "rewards/accuracy_reward": 0.5195676535367966,
      "rewards/format_reward": 0.9921875,
      "step": 796
    },
    {
      "completion_length": 67.71875,
      "epoch": 3.6392694063926943,
      "grad_norm": 2.1526761054992676,
      "kl": 0.14599609375,
      "learning_rate": 6.360730593607305e-07,
      "loss": 0.0058,
      "reward": 1.7081494331359863,
      "reward_std": 0.15050432085990906,
      "rewards/accuracy_reward": 0.7237744629383087,
      "rewards/format_reward": 0.984375,
      "step": 797
    },
    {
      "completion_length": 90.4609375,
      "epoch": 3.643835616438356,
      "grad_norm": 7.734349250793457,
      "kl": 0.107666015625,
      "learning_rate": 6.356164383561645e-07,
      "loss": 0.0043,
      "reward": 1.572656273841858,
      "reward_std": 0.2658383846282959,
      "rewards/accuracy_reward": 0.5960937142372131,
      "rewards/format_reward": 0.9765625,
      "step": 798
    },
    {
      "completion_length": 76.6640625,
      "epoch": 3.6484018264840183,
      "grad_norm": 3.576162815093994,
      "kl": 0.115966796875,
      "learning_rate": 6.351598173515982e-07,
      "loss": 0.0046,
      "reward": 1.6600198149681091,
      "reward_std": 0.15953533351421356,
      "rewards/accuracy_reward": 0.6678323149681091,
      "rewards/format_reward": 0.9921875,
      "step": 799
    },
    {
      "completion_length": 91.1796875,
      "epoch": 3.65296803652968,
      "grad_norm": 3.9656198024749756,
      "kl": 0.107421875,
      "learning_rate": 6.347031963470319e-07,
      "loss": 0.0043,
      "reward": 1.6131696701049805,
      "reward_std": 0.19991916418075562,
      "rewards/accuracy_reward": 0.6366070806980133,
      "rewards/format_reward": 0.9765625,
      "step": 800
    },
    {
      "completion_length": 91.5,
      "epoch": 3.6575342465753424,
      "grad_norm": 2.609405755996704,
      "kl": 0.1591796875,
      "learning_rate": 6.342465753424658e-07,
      "loss": 0.0064,
      "reward": 1.6630208492279053,
      "reward_std": 0.2102055549621582,
      "rewards/accuracy_reward": 0.6708333194255829,
      "rewards/format_reward": 0.9921875,
      "step": 801
    },
    {
      "completion_length": 83.9609375,
      "epoch": 3.6621004566210047,
      "grad_norm": 2.384371280670166,
      "kl": 0.13818359375,
      "learning_rate": 6.337899543378995e-07,
      "loss": 0.0055,
      "reward": 1.7731770873069763,
      "reward_std": 0.16687272489070892,
      "rewards/accuracy_reward": 0.7888020277023315,
      "rewards/format_reward": 0.984375,
      "step": 802
    },
    {
      "completion_length": 83.25,
      "epoch": 3.6666666666666665,
      "grad_norm": 2.3074464797973633,
      "kl": 0.105224609375,
      "learning_rate": 6.333333333333332e-07,
      "loss": 0.0042,
      "reward": 1.6963542103767395,
      "reward_std": 0.2274407297372818,
      "rewards/accuracy_reward": 0.7354166209697723,
      "rewards/format_reward": 0.9609375,
      "step": 803
    },
    {
      "completion_length": 91.296875,
      "epoch": 3.671232876712329,
      "grad_norm": 4.80971097946167,
      "kl": 0.2685546875,
      "learning_rate": 6.328767123287671e-07,
      "loss": 0.0107,
      "reward": 1.8013640642166138,
      "reward_std": 0.18622903525829315,
      "rewards/accuracy_reward": 0.8326140344142914,
      "rewards/format_reward": 0.96875,
      "step": 804
    },
    {
      "completion_length": 76.4375,
      "epoch": 3.6757990867579906,
      "grad_norm": 2.313264846801758,
      "kl": 0.13623046875,
      "learning_rate": 6.324200913242009e-07,
      "loss": 0.0054,
      "reward": 1.573582112789154,
      "reward_std": 0.28307729959487915,
      "rewards/accuracy_reward": 0.5892071425914764,
      "rewards/format_reward": 0.984375,
      "step": 805
    },
    {
      "completion_length": 71.53125,
      "epoch": 3.680365296803653,
      "grad_norm": 5.89130163192749,
      "kl": 0.135986328125,
      "learning_rate": 6.319634703196348e-07,
      "loss": 0.0054,
      "reward": 1.6453125476837158,
      "reward_std": 0.2522100582718849,
      "rewards/accuracy_reward": 0.6687499582767487,
      "rewards/format_reward": 0.9765625,
      "step": 806
    },
    {
      "completion_length": 84.84375,
      "epoch": 3.684931506849315,
      "grad_norm": 2.150968074798584,
      "kl": 0.103759765625,
      "learning_rate": 6.315068493150685e-07,
      "loss": 0.0042,
      "reward": 1.6665269136428833,
      "reward_std": 0.17805734649300575,
      "rewards/accuracy_reward": 0.6743394434452057,
      "rewards/format_reward": 0.9921875,
      "step": 807
    },
    {
      "completion_length": 118.5390625,
      "epoch": 3.6894977168949774,
      "grad_norm": 1.7179328203201294,
      "kl": 0.0628662109375,
      "learning_rate": 6.310502283105022e-07,
      "loss": 0.0025,
      "reward": 1.6565104722976685,
      "reward_std": 0.31547820568084717,
      "rewards/accuracy_reward": 0.7268228828907013,
      "rewards/format_reward": 0.9296875,
      "step": 808
    },
    {
      "completion_length": 86.203125,
      "epoch": 3.6940639269406392,
      "grad_norm": 4.058598518371582,
      "kl": 0.12353515625,
      "learning_rate": 6.305936073059361e-07,
      "loss": 0.0049,
      "reward": 1.7383049130439758,
      "reward_std": 0.2762962728738785,
      "rewards/accuracy_reward": 0.7617424130439758,
      "rewards/format_reward": 0.9765625,
      "step": 809
    },
    {
      "completion_length": 72.484375,
      "epoch": 3.6986301369863015,
      "grad_norm": 6.153299331665039,
      "kl": 0.1328125,
      "learning_rate": 6.301369863013698e-07,
      "loss": 0.0053,
      "reward": 1.607812523841858,
      "reward_std": 0.26409636437892914,
      "rewards/accuracy_reward": 0.6312500238418579,
      "rewards/format_reward": 0.9765625,
      "step": 810
    },
    {
      "completion_length": 73.5078125,
      "epoch": 3.7031963470319633,
      "grad_norm": 4.83391809463501,
      "kl": 0.115478515625,
      "learning_rate": 6.296803652968035e-07,
      "loss": 0.0046,
      "reward": 1.6350895166397095,
      "reward_std": 0.3858235031366348,
      "rewards/accuracy_reward": 0.6897769868373871,
      "rewards/format_reward": 0.9453125,
      "step": 811
    },
    {
      "completion_length": 79.734375,
      "epoch": 3.7077625570776256,
      "grad_norm": 7.388575553894043,
      "kl": 0.121826171875,
      "learning_rate": 6.292237442922375e-07,
      "loss": 0.0049,
      "reward": 1.6204612851142883,
      "reward_std": 0.2048889473080635,
      "rewards/accuracy_reward": 0.6360863149166107,
      "rewards/format_reward": 0.984375,
      "step": 812
    },
    {
      "completion_length": 87.6875,
      "epoch": 3.712328767123288,
      "grad_norm": 2.1843342781066895,
      "kl": 0.091064453125,
      "learning_rate": 6.287671232876712e-07,
      "loss": 0.0036,
      "reward": 1.7203125357627869,
      "reward_std": 0.14389308542013168,
      "rewards/accuracy_reward": 0.7281249761581421,
      "rewards/format_reward": 0.9921875,
      "step": 813
    },
    {
      "completion_length": 87.6484375,
      "epoch": 3.7168949771689497,
      "grad_norm": 6.269219875335693,
      "kl": 0.100341796875,
      "learning_rate": 6.283105022831051e-07,
      "loss": 0.004,
      "reward": 1.6382812857627869,
      "reward_std": 0.17056220024824142,
      "rewards/accuracy_reward": 0.6460937112569809,
      "rewards/format_reward": 0.9921875,
      "step": 814
    },
    {
      "completion_length": 113.0,
      "epoch": 3.721461187214612,
      "grad_norm": 7.404258728027344,
      "kl": 0.0799560546875,
      "learning_rate": 6.278538812785388e-07,
      "loss": 0.0032,
      "reward": 1.7445932626724243,
      "reward_std": 0.27295994758605957,
      "rewards/accuracy_reward": 0.8149056732654572,
      "rewards/format_reward": 0.9296875,
      "step": 815
    },
    {
      "completion_length": 83.03125,
      "epoch": 3.7260273972602738,
      "grad_norm": 2.5489566326141357,
      "kl": 0.095703125,
      "learning_rate": 6.273972602739725e-07,
      "loss": 0.0038,
      "reward": 1.6650923490524292,
      "reward_std": 0.2622094973921776,
      "rewards/accuracy_reward": 0.7041547894477844,
      "rewards/format_reward": 0.9609375,
      "step": 816
    },
    {
      "completion_length": 82.0859375,
      "epoch": 3.730593607305936,
      "grad_norm": 9.064318656921387,
      "kl": 0.10693359375,
      "learning_rate": 6.269406392694064e-07,
      "loss": 0.0043,
      "reward": 1.565638542175293,
      "reward_std": 0.20368661731481552,
      "rewards/accuracy_reward": 0.589076042175293,
      "rewards/format_reward": 0.9765625,
      "step": 817
    },
    {
      "completion_length": 85.734375,
      "epoch": 3.7351598173515983,
      "grad_norm": 3.814204692840576,
      "kl": 0.0927734375,
      "learning_rate": 6.264840182648402e-07,
      "loss": 0.0037,
      "reward": 1.7833616733551025,
      "reward_std": 0.1723785549402237,
      "rewards/accuracy_reward": 0.7911740839481354,
      "rewards/format_reward": 0.9921875,
      "step": 818
    },
    {
      "completion_length": 59.921875,
      "epoch": 3.73972602739726,
      "grad_norm": 3.6849331855773926,
      "kl": 0.127685546875,
      "learning_rate": 6.260273972602739e-07,
      "loss": 0.0051,
      "reward": 1.6573927998542786,
      "reward_std": 0.23803135752677917,
      "rewards/accuracy_reward": 0.6808302700519562,
      "rewards/format_reward": 0.9765625,
      "step": 819
    },
    {
      "completion_length": 73.921875,
      "epoch": 3.7442922374429224,
      "grad_norm": 2.6054327487945557,
      "kl": 0.10498046875,
      "learning_rate": 6.255707762557078e-07,
      "loss": 0.0042,
      "reward": 1.5656526684761047,
      "reward_std": 0.2571340575814247,
      "rewards/accuracy_reward": 0.5890901386737823,
      "rewards/format_reward": 0.9765625,
      "step": 820
    },
    {
      "completion_length": 74.3046875,
      "epoch": 3.748858447488584,
      "grad_norm": 8.960927963256836,
      "kl": 0.413818359375,
      "learning_rate": 6.251141552511415e-07,
      "loss": 0.0166,
      "reward": 1.6491714119911194,
      "reward_std": 0.190296471118927,
      "rewards/accuracy_reward": 0.664796382188797,
      "rewards/format_reward": 0.984375,
      "step": 821
    },
    {
      "completion_length": 80.5859375,
      "epoch": 3.7534246575342465,
      "grad_norm": 4.059770107269287,
      "kl": 0.15185546875,
      "learning_rate": 6.246575342465754e-07,
      "loss": 0.0061,
      "reward": 1.7381510734558105,
      "reward_std": 0.23838761448860168,
      "rewards/accuracy_reward": 0.7615885138511658,
      "rewards/format_reward": 0.9765625,
      "step": 822
    },
    {
      "completion_length": 81.765625,
      "epoch": 3.7579908675799087,
      "grad_norm": 3.5536246299743652,
      "kl": 0.244140625,
      "learning_rate": 6.242009132420091e-07,
      "loss": 0.0098,
      "reward": 1.6796875,
      "reward_std": 0.19722937047481537,
      "rewards/accuracy_reward": 0.7031249403953552,
      "rewards/format_reward": 0.9765625,
      "step": 823
    },
    {
      "completion_length": 67.1015625,
      "epoch": 3.762557077625571,
      "grad_norm": 2.649221658706665,
      "kl": 0.148681640625,
      "learning_rate": 6.237442922374428e-07,
      "loss": 0.006,
      "reward": 1.5904948711395264,
      "reward_std": 0.26418011635541916,
      "rewards/accuracy_reward": 0.6217447817325592,
      "rewards/format_reward": 0.96875,
      "step": 824
    },
    {
      "completion_length": 67.453125,
      "epoch": 3.767123287671233,
      "grad_norm": 4.655685901641846,
      "kl": 0.13623046875,
      "learning_rate": 6.232876712328768e-07,
      "loss": 0.0054,
      "reward": 1.611718773841858,
      "reward_std": 0.17845439538359642,
      "rewards/accuracy_reward": 0.6195312142372131,
      "rewards/format_reward": 0.9921875,
      "step": 825
    },
    {
      "completion_length": 75.90625,
      "epoch": 3.771689497716895,
      "grad_norm": 1.6092960834503174,
      "kl": 0.127685546875,
      "learning_rate": 6.228310502283105e-07,
      "loss": 0.0051,
      "reward": 1.71875,
      "reward_std": 0.16304787248373032,
      "rewards/accuracy_reward": 0.7343750298023224,
      "rewards/format_reward": 0.984375,
      "step": 826
    },
    {
      "completion_length": 74.7890625,
      "epoch": 3.776255707762557,
      "grad_norm": 1.904418706893921,
      "kl": 0.090087890625,
      "learning_rate": 6.223744292237442e-07,
      "loss": 0.0036,
      "reward": 1.7890625596046448,
      "reward_std": 0.16059691458940506,
      "rewards/accuracy_reward": 0.8125,
      "rewards/format_reward": 0.9765625,
      "step": 827
    },
    {
      "completion_length": 78.609375,
      "epoch": 3.780821917808219,
      "grad_norm": 2.7213988304138184,
      "kl": 0.13037109375,
      "learning_rate": 6.219178082191781e-07,
      "loss": 0.0052,
      "reward": 1.6396701335906982,
      "reward_std": 0.23586497455835342,
      "rewards/accuracy_reward": 0.6787325739860535,
      "rewards/format_reward": 0.9609375,
      "step": 828
    },
    {
      "completion_length": 87.4140625,
      "epoch": 3.7853881278538815,
      "grad_norm": 3.867840051651001,
      "kl": 0.1123046875,
      "learning_rate": 6.214611872146118e-07,
      "loss": 0.0045,
      "reward": 1.6255208849906921,
      "reward_std": 0.19572605937719345,
      "rewards/accuracy_reward": 0.6489582806825638,
      "rewards/format_reward": 0.9765625,
      "step": 829
    },
    {
      "completion_length": 80.6953125,
      "epoch": 3.7899543378995433,
      "grad_norm": 5.212503910064697,
      "kl": 0.177490234375,
      "learning_rate": 6.210045662100457e-07,
      "loss": 0.0071,
      "reward": 1.7282792925834656,
      "reward_std": 0.31921522319316864,
      "rewards/accuracy_reward": 0.7673417627811432,
      "rewards/format_reward": 0.9609375,
      "step": 830
    },
    {
      "completion_length": 87.03125,
      "epoch": 3.7945205479452055,
      "grad_norm": 7.442432880401611,
      "kl": 0.093017578125,
      "learning_rate": 6.205479452054794e-07,
      "loss": 0.0037,
      "reward": 1.6146034002304077,
      "reward_std": 0.196245439350605,
      "rewards/accuracy_reward": 0.6224157810211182,
      "rewards/format_reward": 0.9921875,
      "step": 831
    },
    {
      "completion_length": 77.1640625,
      "epoch": 3.7990867579908674,
      "grad_norm": 4.140781402587891,
      "kl": 0.2216796875,
      "learning_rate": 6.200913242009132e-07,
      "loss": 0.0089,
      "reward": 1.4750909209251404,
      "reward_std": 0.26224930584430695,
      "rewards/accuracy_reward": 0.5141534507274628,
      "rewards/format_reward": 0.9609375,
      "step": 832
    },
    {
      "completion_length": 73.5859375,
      "epoch": 3.8036529680365296,
      "grad_norm": 2.893556594848633,
      "kl": 0.105224609375,
      "learning_rate": 6.196347031963471e-07,
      "loss": 0.0042,
      "reward": 1.6718750596046448,
      "reward_std": 0.25111906230449677,
      "rewards/accuracy_reward": 0.6953124701976776,
      "rewards/format_reward": 0.9765625,
      "step": 833
    },
    {
      "completion_length": 68.46875,
      "epoch": 3.808219178082192,
      "grad_norm": 4.676414489746094,
      "kl": 0.20361328125,
      "learning_rate": 6.191780821917808e-07,
      "loss": 0.0081,
      "reward": 1.682812511920929,
      "reward_std": 0.20322410762310028,
      "rewards/accuracy_reward": 0.6984374821186066,
      "rewards/format_reward": 0.984375,
      "step": 834
    },
    {
      "completion_length": 78.9375,
      "epoch": 3.8127853881278537,
      "grad_norm": 1.6920398473739624,
      "kl": 0.134765625,
      "learning_rate": 6.187214611872145e-07,
      "loss": 0.0054,
      "reward": 1.8067708611488342,
      "reward_std": 0.14878704398870468,
      "rewards/accuracy_reward": 0.8223958313465118,
      "rewards/format_reward": 0.984375,
      "step": 835
    },
    {
      "completion_length": 73.34375,
      "epoch": 3.817351598173516,
      "grad_norm": 3.9718282222747803,
      "kl": 0.1552734375,
      "learning_rate": 6.182648401826484e-07,
      "loss": 0.0062,
      "reward": 1.586328148841858,
      "reward_std": 0.2995697557926178,
      "rewards/accuracy_reward": 0.6410156190395355,
      "rewards/format_reward": 0.9453125,
      "step": 836
    },
    {
      "completion_length": 74.859375,
      "epoch": 3.821917808219178,
      "grad_norm": 2.3701136112213135,
      "kl": 0.146240234375,
      "learning_rate": 6.178082191780821e-07,
      "loss": 0.0059,
      "reward": 1.6562500596046448,
      "reward_std": 0.1812673956155777,
      "rewards/accuracy_reward": 0.6640624701976776,
      "rewards/format_reward": 0.9921875,
      "step": 837
    },
    {
      "completion_length": 63.671875,
      "epoch": 3.82648401826484,
      "grad_norm": 3.7064878940582275,
      "kl": 0.1220703125,
      "learning_rate": 6.173515981735161e-07,
      "loss": 0.0049,
      "reward": 1.574496567249298,
      "reward_std": 0.24655399471521378,
      "rewards/accuracy_reward": 0.5979340374469757,
      "rewards/format_reward": 0.9765625,
      "step": 838
    },
    {
      "completion_length": 69.6171875,
      "epoch": 3.8310502283105023,
      "grad_norm": 2.966919422149658,
      "kl": 0.119384765625,
      "learning_rate": 6.168949771689498e-07,
      "loss": 0.0048,
      "reward": 1.5951017141342163,
      "reward_std": 0.22802505642175674,
      "rewards/accuracy_reward": 0.6263516843318939,
      "rewards/format_reward": 0.96875,
      "step": 839
    },
    {
      "completion_length": 67.4921875,
      "epoch": 3.8356164383561646,
      "grad_norm": 5.754858493804932,
      "kl": 0.125732421875,
      "learning_rate": 6.164383561643835e-07,
      "loss": 0.005,
      "reward": 1.60744047164917,
      "reward_std": 0.2320682480931282,
      "rewards/accuracy_reward": 0.6230654716491699,
      "rewards/format_reward": 0.984375,
      "step": 840
    },
    {
      "completion_length": 94.8828125,
      "epoch": 3.8401826484018264,
      "grad_norm": 3.788088321685791,
      "kl": 0.09912109375,
      "learning_rate": 6.159817351598174e-07,
      "loss": 0.004,
      "reward": 1.65234375,
      "reward_std": 0.1578691005706787,
      "rewards/accuracy_reward": 0.65234375,
      "rewards/format_reward": 1.0,
      "step": 841
    },
    {
      "completion_length": 82.28125,
      "epoch": 3.8447488584474887,
      "grad_norm": 3.032823085784912,
      "kl": 0.103271484375,
      "learning_rate": 6.155251141552511e-07,
      "loss": 0.0041,
      "reward": 1.538119375705719,
      "reward_std": 0.2931046634912491,
      "rewards/accuracy_reward": 0.569369375705719,
      "rewards/format_reward": 0.96875,
      "step": 842
    },
    {
      "completion_length": 88.265625,
      "epoch": 3.8493150684931505,
      "grad_norm": 2.753324031829834,
      "kl": 0.08544921875,
      "learning_rate": 6.150684931506848e-07,
      "loss": 0.0034,
      "reward": 1.6361016035079956,
      "reward_std": 0.2543141394853592,
      "rewards/accuracy_reward": 0.6673516035079956,
      "rewards/format_reward": 0.96875,
      "step": 843
    },
    {
      "completion_length": 88.3203125,
      "epoch": 3.853881278538813,
      "grad_norm": 1.9302741289138794,
      "kl": 0.116455078125,
      "learning_rate": 6.146118721461187e-07,
      "loss": 0.0047,
      "reward": 1.723133623600006,
      "reward_std": 0.12216833233833313,
      "rewards/accuracy_reward": 0.7231336236000061,
      "rewards/format_reward": 1.0,
      "step": 844
    },
    {
      "completion_length": 68.3046875,
      "epoch": 3.858447488584475,
      "grad_norm": 2.5496842861175537,
      "kl": 0.115234375,
      "learning_rate": 6.141552511415525e-07,
      "loss": 0.0046,
      "reward": 1.63571435213089,
      "reward_std": 0.22781573235988617,
      "rewards/accuracy_reward": 0.6513392478227615,
      "rewards/format_reward": 0.984375,
      "step": 845
    },
    {
      "completion_length": 98.5546875,
      "epoch": 3.863013698630137,
      "grad_norm": 1.8447142839431763,
      "kl": 0.098388671875,
      "learning_rate": 6.136986301369864e-07,
      "loss": 0.0039,
      "reward": 1.7750434279441833,
      "reward_std": 0.16683020442724228,
      "rewards/accuracy_reward": 0.798480898141861,
      "rewards/format_reward": 0.9765625,
      "step": 846
    },
    {
      "completion_length": 86.2265625,
      "epoch": 3.867579908675799,
      "grad_norm": 4.997533798217773,
      "kl": 0.123779296875,
      "learning_rate": 6.132420091324201e-07,
      "loss": 0.0049,
      "reward": 1.6639086604118347,
      "reward_std": 0.298796147108078,
      "rewards/accuracy_reward": 0.7185961008071899,
      "rewards/format_reward": 0.9453125,
      "step": 847
    },
    {
      "completion_length": 76.2578125,
      "epoch": 3.872146118721461,
      "grad_norm": 4.215785503387451,
      "kl": 0.130859375,
      "learning_rate": 6.127853881278538e-07,
      "loss": 0.0052,
      "reward": 1.5344713926315308,
      "reward_std": 0.3186282366514206,
      "rewards/accuracy_reward": 0.5735338628292084,
      "rewards/format_reward": 0.9609375,
      "step": 848
    },
    {
      "completion_length": 76.5546875,
      "epoch": 3.8767123287671232,
      "grad_norm": 3.0258474349975586,
      "kl": 0.11962890625,
      "learning_rate": 6.123287671232877e-07,
      "loss": 0.0048,
      "reward": 1.5302554368972778,
      "reward_std": 0.21649178117513657,
      "rewards/accuracy_reward": 0.5458804070949554,
      "rewards/format_reward": 0.984375,
      "step": 849
    },
    {
      "completion_length": 88.828125,
      "epoch": 3.8812785388127855,
      "grad_norm": 3.9653830528259277,
      "kl": 0.134765625,
      "learning_rate": 6.118721461187214e-07,
      "loss": 0.0054,
      "reward": 1.7723276019096375,
      "reward_std": 0.27654044330120087,
      "rewards/accuracy_reward": 0.8035775721073151,
      "rewards/format_reward": 0.96875,
      "step": 850
    },
    {
      "completion_length": 70.6796875,
      "epoch": 3.8858447488584473,
      "grad_norm": 13.911476135253906,
      "kl": 0.1279296875,
      "learning_rate": 6.114155251141551e-07,
      "loss": 0.0051,
      "reward": 1.5849445462226868,
      "reward_std": 0.2708437442779541,
      "rewards/accuracy_reward": 0.5849444717168808,
      "rewards/format_reward": 1.0,
      "step": 851
    },
    {
      "completion_length": 84.46875,
      "epoch": 3.8904109589041096,
      "grad_norm": 2.8287229537963867,
      "kl": 0.122802734375,
      "learning_rate": 6.109589041095891e-07,
      "loss": 0.0049,
      "reward": 1.6496233344078064,
      "reward_std": 0.1977412924170494,
      "rewards/accuracy_reward": 0.6730607748031616,
      "rewards/format_reward": 0.9765625,
      "step": 852
    },
    {
      "completion_length": 84.328125,
      "epoch": 3.8949771689497714,
      "grad_norm": 3.292459487915039,
      "kl": 0.110107421875,
      "learning_rate": 6.105022831050228e-07,
      "loss": 0.0044,
      "reward": 1.560290813446045,
      "reward_std": 0.2603389769792557,
      "rewards/accuracy_reward": 0.5681032538414001,
      "rewards/format_reward": 0.9921875,
      "step": 853
    },
    {
      "completion_length": 70.046875,
      "epoch": 3.8995433789954337,
      "grad_norm": 3.479142189025879,
      "kl": 0.13818359375,
      "learning_rate": 6.100456621004567e-07,
      "loss": 0.0055,
      "reward": 1.5651041865348816,
      "reward_std": 0.327400267124176,
      "rewards/accuracy_reward": 0.6041666567325592,
      "rewards/format_reward": 0.9609375,
      "step": 854
    },
    {
      "completion_length": 78.3046875,
      "epoch": 3.904109589041096,
      "grad_norm": 2.2671470642089844,
      "kl": 0.103515625,
      "learning_rate": 6.095890410958904e-07,
      "loss": 0.0041,
      "reward": 1.6139509677886963,
      "reward_std": 0.221808023750782,
      "rewards/accuracy_reward": 0.6608258783817291,
      "rewards/format_reward": 0.953125,
      "step": 855
    },
    {
      "completion_length": 79.015625,
      "epoch": 3.908675799086758,
      "grad_norm": 3.359975814819336,
      "kl": 0.097900390625,
      "learning_rate": 6.091324200913241e-07,
      "loss": 0.0039,
      "reward": 1.7910323739051819,
      "reward_std": 0.16605094820261002,
      "rewards/accuracy_reward": 0.80665722489357,
      "rewards/format_reward": 0.984375,
      "step": 856
    },
    {
      "completion_length": 108.4375,
      "epoch": 3.91324200913242,
      "grad_norm": 2.9608707427978516,
      "kl": 0.064208984375,
      "learning_rate": 6.08675799086758e-07,
      "loss": 0.0026,
      "reward": 1.7960938215255737,
      "reward_std": 0.15246989950537682,
      "rewards/accuracy_reward": 0.8195312321186066,
      "rewards/format_reward": 0.9765625,
      "step": 857
    },
    {
      "completion_length": 70.484375,
      "epoch": 3.9178082191780823,
      "grad_norm": 1.9071540832519531,
      "kl": 0.115234375,
      "learning_rate": 6.082191780821918e-07,
      "loss": 0.0046,
      "reward": 1.8630208373069763,
      "reward_std": 0.18414238840341568,
      "rewards/accuracy_reward": 0.8864583075046539,
      "rewards/format_reward": 0.9765625,
      "step": 858
    },
    {
      "completion_length": 86.25,
      "epoch": 3.922374429223744,
      "grad_norm": 1.7425415515899658,
      "kl": 0.0791015625,
      "learning_rate": 6.077625570776255e-07,
      "loss": 0.0032,
      "reward": 1.7500391602516174,
      "reward_std": 0.1361438985913992,
      "rewards/accuracy_reward": 0.757851630449295,
      "rewards/format_reward": 0.9921875,
      "step": 859
    },
    {
      "completion_length": 83.4765625,
      "epoch": 3.9269406392694064,
      "grad_norm": 2.0878822803497314,
      "kl": 0.132080078125,
      "learning_rate": 6.073059360730594e-07,
      "loss": 0.0053,
      "reward": 1.7227915525436401,
      "reward_std": 0.2049795687198639,
      "rewards/accuracy_reward": 0.7384164929389954,
      "rewards/format_reward": 0.984375,
      "step": 860
    },
    {
      "completion_length": 82.6328125,
      "epoch": 3.9315068493150687,
      "grad_norm": 2.1125905513763428,
      "kl": 0.10400390625,
      "learning_rate": 6.068493150684931e-07,
      "loss": 0.0041,
      "reward": 1.7052381038665771,
      "reward_std": 0.19698219001293182,
      "rewards/accuracy_reward": 0.7208629250526428,
      "rewards/format_reward": 0.984375,
      "step": 861
    },
    {
      "completion_length": 70.6171875,
      "epoch": 3.9360730593607305,
      "grad_norm": 3.4557483196258545,
      "kl": 0.125244140625,
      "learning_rate": 6.06392694063927e-07,
      "loss": 0.005,
      "reward": 1.6386160850524902,
      "reward_std": 0.2512796074151993,
      "rewards/accuracy_reward": 0.6464285254478455,
      "rewards/format_reward": 0.9921875,
      "step": 862
    },
    {
      "completion_length": 60.2890625,
      "epoch": 3.9406392694063928,
      "grad_norm": 2.9261093139648438,
      "kl": 0.197265625,
      "learning_rate": 6.059360730593607e-07,
      "loss": 0.0079,
      "reward": 1.6639323234558105,
      "reward_std": 0.18012882769107819,
      "rewards/accuracy_reward": 0.6717447936534882,
      "rewards/format_reward": 0.9921875,
      "step": 863
    },
    {
      "completion_length": 68.984375,
      "epoch": 3.9452054794520546,
      "grad_norm": 8.949965476989746,
      "kl": 0.16455078125,
      "learning_rate": 6.054794520547944e-07,
      "loss": 0.0066,
      "reward": 1.634996235370636,
      "reward_std": 0.30956215411424637,
      "rewards/accuracy_reward": 0.658433735370636,
      "rewards/format_reward": 0.9765625,
      "step": 864
    },
    {
      "completion_length": 94.5078125,
      "epoch": 3.949771689497717,
      "grad_norm": 1.5857025384902954,
      "kl": 0.07958984375,
      "learning_rate": 6.050228310502284e-07,
      "loss": 0.0032,
      "reward": 1.7937500476837158,
      "reward_std": 0.08982988260686398,
      "rewards/accuracy_reward": 0.793749988079071,
      "rewards/format_reward": 1.0,
      "step": 865
    },
    {
      "completion_length": 72.328125,
      "epoch": 3.954337899543379,
      "grad_norm": 3.955540418624878,
      "kl": 0.144775390625,
      "learning_rate": 6.045662100456621e-07,
      "loss": 0.0058,
      "reward": 1.5606706142425537,
      "reward_std": 0.3175853192806244,
      "rewards/accuracy_reward": 0.6075455844402313,
      "rewards/format_reward": 0.953125,
      "step": 866
    },
    {
      "completion_length": 56.71875,
      "epoch": 3.958904109589041,
      "grad_norm": 3.506572723388672,
      "kl": 0.142822265625,
      "learning_rate": 6.041095890410958e-07,
      "loss": 0.0057,
      "reward": 1.7238853573799133,
      "reward_std": 0.22640568763017654,
      "rewards/accuracy_reward": 0.7316978275775909,
      "rewards/format_reward": 0.9921875,
      "step": 867
    },
    {
      "completion_length": 69.9921875,
      "epoch": 3.963470319634703,
      "grad_norm": 6.0887770652771,
      "kl": 0.0931396484375,
      "learning_rate": 6.036529680365297e-07,
      "loss": 0.0037,
      "reward": 1.851125419139862,
      "reward_std": 0.1146822888404131,
      "rewards/accuracy_reward": 0.8511254191398621,
      "rewards/format_reward": 1.0,
      "step": 868
    },
    {
      "completion_length": 78.6875,
      "epoch": 3.968036529680365,
      "grad_norm": 2.3538548946380615,
      "kl": 0.134765625,
      "learning_rate": 6.031963470319634e-07,
      "loss": 0.0054,
      "reward": 1.6927083730697632,
      "reward_std": 0.20856676995754242,
      "rewards/accuracy_reward": 0.7083333134651184,
      "rewards/format_reward": 0.984375,
      "step": 869
    },
    {
      "completion_length": 67.84375,
      "epoch": 3.9726027397260273,
      "grad_norm": 3.3203752040863037,
      "kl": 0.19482421875,
      "learning_rate": 6.027397260273972e-07,
      "loss": 0.0078,
      "reward": 1.682466745376587,
      "reward_std": 0.20432885736227036,
      "rewards/accuracy_reward": 0.6824667155742645,
      "rewards/format_reward": 1.0,
      "step": 870
    },
    {
      "completion_length": 60.8046875,
      "epoch": 3.9771689497716896,
      "grad_norm": 4.509641647338867,
      "kl": 0.12548828125,
      "learning_rate": 6.02283105022831e-07,
      "loss": 0.005,
      "reward": 1.5753461122512817,
      "reward_std": 0.23474501818418503,
      "rewards/accuracy_reward": 0.5753461122512817,
      "rewards/format_reward": 1.0,
      "step": 871
    },
    {
      "completion_length": 89.40625,
      "epoch": 3.981735159817352,
      "grad_norm": 2.3428871631622314,
      "kl": 0.113037109375,
      "learning_rate": 6.018264840182648e-07,
      "loss": 0.0045,
      "reward": 1.8376488089561462,
      "reward_std": 0.12626906298100948,
      "rewards/accuracy_reward": 0.8454612195491791,
      "rewards/format_reward": 0.9921875,
      "step": 872
    },
    {
      "completion_length": 70.9921875,
      "epoch": 3.9863013698630136,
      "grad_norm": 1.8384939432144165,
      "kl": 0.12890625,
      "learning_rate": 6.013698630136987e-07,
      "loss": 0.0052,
      "reward": 1.6695202589035034,
      "reward_std": 0.13698631152510643,
      "rewards/accuracy_reward": 0.6695202589035034,
      "rewards/format_reward": 1.0,
      "step": 873
    },
    {
      "completion_length": 72.8828125,
      "epoch": 3.990867579908676,
      "grad_norm": 2.790123224258423,
      "kl": 0.132568359375,
      "learning_rate": 6.009132420091324e-07,
      "loss": 0.0053,
      "reward": 1.5515252947807312,
      "reward_std": 0.24770487844944,
      "rewards/accuracy_reward": 0.5749628096818924,
      "rewards/format_reward": 0.9765625,
      "step": 874
    },
    {
      "completion_length": 70.765625,
      "epoch": 3.9954337899543377,
      "grad_norm": 3.470076084136963,
      "kl": 0.09423828125,
      "learning_rate": 6.004566210045661e-07,
      "loss": 0.0038,
      "reward": 1.8020795583724976,
      "reward_std": 0.08546407520771027,
      "rewards/accuracy_reward": 0.8098919987678528,
      "rewards/format_reward": 0.9921875,
      "step": 875
    },
    {
      "completion_length": 37.0,
      "epoch": 4.0,
      "grad_norm": 3.645047903060913,
      "kl": 0.158203125,
      "learning_rate": 6e-07,
      "loss": 0.0061,
      "reward": 1.5,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.5,
      "rewards/format_reward": 1.0,
      "step": 876
    },
    {
      "completion_length": 63.9921875,
      "epoch": 4.004566210045662,
      "grad_norm": 8.070154190063477,
      "kl": 0.143310546875,
      "learning_rate": 5.995433789954337e-07,
      "loss": 0.0057,
      "reward": 1.726302146911621,
      "reward_std": 0.1457817256450653,
      "rewards/accuracy_reward": 0.7263020873069763,
      "rewards/format_reward": 1.0,
      "step": 877
    },
    {
      "completion_length": 80.28125,
      "epoch": 4.0091324200913245,
      "grad_norm": 3.771700620651245,
      "kl": 0.1318359375,
      "learning_rate": 5.990867579908675e-07,
      "loss": 0.0053,
      "reward": 1.7184895873069763,
      "reward_std": 0.14457330107688904,
      "rewards/accuracy_reward": 0.7263020873069763,
      "rewards/format_reward": 0.9921875,
      "step": 878
    },
    {
      "completion_length": 55.3828125,
      "epoch": 4.013698630136986,
      "grad_norm": 3.114535093307495,
      "kl": 0.193359375,
      "learning_rate": 5.986301369863014e-07,
      "loss": 0.0077,
      "reward": 1.699999988079071,
      "reward_std": 0.24036270380020142,
      "rewards/accuracy_reward": 0.715624988079071,
      "rewards/format_reward": 0.984375,
      "step": 879
    },
    {
      "completion_length": 68.9921875,
      "epoch": 4.018264840182648,
      "grad_norm": 5.49289608001709,
      "kl": 0.1513671875,
      "learning_rate": 5.981735159817351e-07,
      "loss": 0.006,
      "reward": 1.7083333730697632,
      "reward_std": 0.14341074973344803,
      "rewards/accuracy_reward": 0.7083333134651184,
      "rewards/format_reward": 1.0,
      "step": 880
    },
    {
      "completion_length": 68.1484375,
      "epoch": 4.0228310502283104,
      "grad_norm": 8.557650566101074,
      "kl": 0.47802734375,
      "learning_rate": 5.97716894977169e-07,
      "loss": 0.0191,
      "reward": 1.7641276121139526,
      "reward_std": 0.13008537888526917,
      "rewards/accuracy_reward": 0.7641275823116302,
      "rewards/format_reward": 1.0,
      "step": 881
    },
    {
      "completion_length": 74.2265625,
      "epoch": 4.027397260273973,
      "grad_norm": 1.5429102182388306,
      "kl": 0.1396484375,
      "learning_rate": 5.972602739726027e-07,
      "loss": 0.0056,
      "reward": 1.7580729722976685,
      "reward_std": 0.1651972383260727,
      "rewards/accuracy_reward": 0.7736978530883789,
      "rewards/format_reward": 0.984375,
      "step": 882
    },
    {
      "completion_length": 60.5703125,
      "epoch": 4.031963470319635,
      "grad_norm": 1.6412632465362549,
      "kl": 0.123779296875,
      "learning_rate": 5.968036529680364e-07,
      "loss": 0.005,
      "reward": 1.7804688215255737,
      "reward_std": 0.1583872102200985,
      "rewards/accuracy_reward": 0.7882812321186066,
      "rewards/format_reward": 0.9921875,
      "step": 883
    },
    {
      "completion_length": 74.703125,
      "epoch": 4.036529680365296,
      "grad_norm": 5.128458499908447,
      "kl": 0.2080078125,
      "learning_rate": 5.963470319634703e-07,
      "loss": 0.0083,
      "reward": 1.706798791885376,
      "reward_std": 0.19644346833229065,
      "rewards/accuracy_reward": 0.7302362024784088,
      "rewards/format_reward": 0.9765625,
      "step": 884
    },
    {
      "completion_length": 60.921875,
      "epoch": 4.041095890410959,
      "grad_norm": 2.6405813694000244,
      "kl": 0.14892578125,
      "learning_rate": 5.958904109589041e-07,
      "loss": 0.006,
      "reward": 1.7202391624450684,
      "reward_std": 0.218642920255661,
      "rewards/accuracy_reward": 0.7202391028404236,
      "rewards/format_reward": 1.0,
      "step": 885
    },
    {
      "completion_length": 66.59375,
      "epoch": 4.045662100456621,
      "grad_norm": 2.522268533706665,
      "kl": 0.1220703125,
      "learning_rate": 5.95433789954338e-07,
      "loss": 0.0049,
      "reward": 1.5917280912399292,
      "reward_std": 0.18826671689748764,
      "rewards/accuracy_reward": 0.6073530614376068,
      "rewards/format_reward": 0.984375,
      "step": 886
    },
    {
      "completion_length": 51.8828125,
      "epoch": 4.050228310502283,
      "grad_norm": 2.2267160415649414,
      "kl": 0.173828125,
      "learning_rate": 5.949771689497717e-07,
      "loss": 0.007,
      "reward": 1.6658853888511658,
      "reward_std": 0.22664503753185272,
      "rewards/accuracy_reward": 0.6815103888511658,
      "rewards/format_reward": 0.984375,
      "step": 887
    },
    {
      "completion_length": 51.3203125,
      "epoch": 4.054794520547945,
      "grad_norm": 1.9868104457855225,
      "kl": 0.1552734375,
      "learning_rate": 5.945205479452054e-07,
      "loss": 0.0062,
      "reward": 1.7429263591766357,
      "reward_std": 0.18497039377689362,
      "rewards/accuracy_reward": 0.7741763293743134,
      "rewards/format_reward": 0.96875,
      "step": 888
    },
    {
      "completion_length": 56.0390625,
      "epoch": 4.059360730593608,
      "grad_norm": 2.1909918785095215,
      "kl": 0.140869140625,
      "learning_rate": 5.940639269406393e-07,
      "loss": 0.0056,
      "reward": 1.726171851158142,
      "reward_std": 0.20413171127438545,
      "rewards/accuracy_reward": 0.7339843213558197,
      "rewards/format_reward": 0.9921875,
      "step": 889
    },
    {
      "completion_length": 69.4609375,
      "epoch": 4.063926940639269,
      "grad_norm": 1.6129564046859741,
      "kl": 0.125732421875,
      "learning_rate": 5.93607305936073e-07,
      "loss": 0.005,
      "reward": 1.7187500596046448,
      "reward_std": 0.1436246931552887,
      "rewards/accuracy_reward": 0.734375,
      "rewards/format_reward": 0.984375,
      "step": 890
    },
    {
      "completion_length": 67.9921875,
      "epoch": 4.068493150684931,
      "grad_norm": 3.2070353031158447,
      "kl": 0.119140625,
      "learning_rate": 5.931506849315067e-07,
      "loss": 0.0048,
      "reward": 1.685523271560669,
      "reward_std": 0.18283560872077942,
      "rewards/accuracy_reward": 0.6933356523513794,
      "rewards/format_reward": 0.9921875,
      "step": 891
    },
    {
      "completion_length": 74.7578125,
      "epoch": 4.073059360730594,
      "grad_norm": 2.196234703063965,
      "kl": 0.092041015625,
      "learning_rate": 5.926940639269407e-07,
      "loss": 0.0037,
      "reward": 1.7648438215255737,
      "reward_std": 0.10456175357103348,
      "rewards/accuracy_reward": 0.7648437023162842,
      "rewards/format_reward": 1.0,
      "step": 892
    },
    {
      "completion_length": 62.0,
      "epoch": 4.077625570776256,
      "grad_norm": 3.306205987930298,
      "kl": 0.12451171875,
      "learning_rate": 5.922374429223744e-07,
      "loss": 0.005,
      "reward": 1.7713721990585327,
      "reward_std": 0.16517749428749084,
      "rewards/accuracy_reward": 0.7713720798492432,
      "rewards/format_reward": 1.0,
      "step": 893
    },
    {
      "completion_length": 59.890625,
      "epoch": 4.082191780821918,
      "grad_norm": 2.852937698364258,
      "kl": 0.13525390625,
      "learning_rate": 5.917808219178083e-07,
      "loss": 0.0054,
      "reward": 1.6403688192367554,
      "reward_std": 0.19380497932434082,
      "rewards/accuracy_reward": 0.648181289434433,
      "rewards/format_reward": 0.9921875,
      "step": 894
    },
    {
      "completion_length": 80.5546875,
      "epoch": 4.0867579908675795,
      "grad_norm": 4.440982818603516,
      "kl": 0.110107421875,
      "learning_rate": 5.91324200913242e-07,
      "loss": 0.0044,
      "reward": 1.7101563215255737,
      "reward_std": 0.15725971013307571,
      "rewards/accuracy_reward": 0.7179687321186066,
      "rewards/format_reward": 0.9921875,
      "step": 895
    },
    {
      "completion_length": 76.8046875,
      "epoch": 4.091324200913242,
      "grad_norm": 1.442449927330017,
      "kl": 0.130859375,
      "learning_rate": 5.908675799086757e-07,
      "loss": 0.0052,
      "reward": 1.831250011920929,
      "reward_std": 0.0731260534375906,
      "rewards/accuracy_reward": 0.8312499225139618,
      "rewards/format_reward": 1.0,
      "step": 896
    },
    {
      "completion_length": 43.0,
      "epoch": 4.095890410958904,
      "grad_norm": 5.812748908996582,
      "kl": 0.21533203125,
      "learning_rate": 5.904109589041096e-07,
      "loss": 0.0086,
      "reward": 1.7768229246139526,
      "reward_std": 0.22991649061441422,
      "rewards/accuracy_reward": 0.7768228948116302,
      "rewards/format_reward": 1.0,
      "step": 897
    },
    {
      "completion_length": 87.3828125,
      "epoch": 4.100456621004566,
      "grad_norm": 2.7623848915100098,
      "kl": 0.093505859375,
      "learning_rate": 5.899543378995433e-07,
      "loss": 0.0037,
      "reward": 1.8036272525787354,
      "reward_std": 0.11107254587113857,
      "rewards/accuracy_reward": 0.8114396631717682,
      "rewards/format_reward": 0.9921875,
      "step": 898
    },
    {
      "completion_length": 91.8125,
      "epoch": 4.105022831050229,
      "grad_norm": 2.0738606452941895,
      "kl": 0.133056640625,
      "learning_rate": 5.894977168949771e-07,
      "loss": 0.0053,
      "reward": 1.7905096411705017,
      "reward_std": 0.12370277941226959,
      "rewards/accuracy_reward": 0.8061346113681793,
      "rewards/format_reward": 0.984375,
      "step": 899
    },
    {
      "completion_length": 76.5546875,
      "epoch": 4.109589041095891,
      "grad_norm": 3.7335944175720215,
      "kl": 0.130859375,
      "learning_rate": 5.89041095890411e-07,
      "loss": 0.0052,
      "reward": 1.6451637148857117,
      "reward_std": 0.1962077133357525,
      "rewards/accuracy_reward": 0.6529761403799057,
      "rewards/format_reward": 0.9921875,
      "step": 900
    },
    {
      "completion_length": 63.0625,
      "epoch": 4.114155251141552,
      "grad_norm": 7.811534881591797,
      "kl": 0.117431640625,
      "learning_rate": 5.885844748858447e-07,
      "loss": 0.0047,
      "reward": 1.6516927480697632,
      "reward_std": 0.20568183064460754,
      "rewards/accuracy_reward": 0.651692658662796,
      "rewards/format_reward": 1.0,
      "step": 901
    },
    {
      "completion_length": 75.40625,
      "epoch": 4.1187214611872145,
      "grad_norm": 4.964561462402344,
      "kl": 0.11474609375,
      "learning_rate": 5.881278538812785e-07,
      "loss": 0.0046,
      "reward": 1.767968773841858,
      "reward_std": 0.17296937853097916,
      "rewards/accuracy_reward": 0.7757811844348907,
      "rewards/format_reward": 0.9921875,
      "step": 902
    },
    {
      "completion_length": 78.671875,
      "epoch": 4.123287671232877,
      "grad_norm": 7.2277679443359375,
      "kl": 0.111083984375,
      "learning_rate": 5.876712328767123e-07,
      "loss": 0.0044,
      "reward": 1.8033853769302368,
      "reward_std": 0.12047793343663216,
      "rewards/accuracy_reward": 0.8111978769302368,
      "rewards/format_reward": 0.9921875,
      "step": 903
    },
    {
      "completion_length": 76.0546875,
      "epoch": 4.127853881278539,
      "grad_norm": 3.666255474090576,
      "kl": 0.14697265625,
      "learning_rate": 5.87214611872146e-07,
      "loss": 0.0059,
      "reward": 1.7567708492279053,
      "reward_std": 0.15414869785308838,
      "rewards/accuracy_reward": 0.7645833492279053,
      "rewards/format_reward": 0.9921875,
      "step": 904
    },
    {
      "completion_length": 64.28125,
      "epoch": 4.132420091324201,
      "grad_norm": 6.144541263580322,
      "kl": 0.15673828125,
      "learning_rate": 5.8675799086758e-07,
      "loss": 0.0063,
      "reward": 1.6296875476837158,
      "reward_std": 0.23411936312913895,
      "rewards/accuracy_reward": 0.637499988079071,
      "rewards/format_reward": 0.9921875,
      "step": 905
    },
    {
      "completion_length": 96.265625,
      "epoch": 4.136986301369863,
      "grad_norm": 4.049592018127441,
      "kl": 0.103759765625,
      "learning_rate": 5.863013698630137e-07,
      "loss": 0.0041,
      "reward": 1.807031273841858,
      "reward_std": 0.16107044368982315,
      "rewards/accuracy_reward": 0.8070311546325684,
      "rewards/format_reward": 1.0,
      "step": 906
    },
    {
      "completion_length": 73.2578125,
      "epoch": 4.141552511415525,
      "grad_norm": 1.5114738941192627,
      "kl": 0.085693359375,
      "learning_rate": 5.858447488584474e-07,
      "loss": 0.0034,
      "reward": 1.6850818991661072,
      "reward_std": 0.16721044853329659,
      "rewards/accuracy_reward": 0.69289430975914,
      "rewards/format_reward": 0.9921875,
      "step": 907
    },
    {
      "completion_length": 56.8515625,
      "epoch": 4.146118721461187,
      "grad_norm": 2.3427228927612305,
      "kl": 0.12890625,
      "learning_rate": 5.853881278538813e-07,
      "loss": 0.0052,
      "reward": 1.756416380405426,
      "reward_std": 0.1625683754682541,
      "rewards/accuracy_reward": 0.7642288506031036,
      "rewards/format_reward": 0.9921875,
      "step": 908
    },
    {
      "completion_length": 85.671875,
      "epoch": 4.1506849315068495,
      "grad_norm": 2.933675765991211,
      "kl": 0.092529296875,
      "learning_rate": 5.84931506849315e-07,
      "loss": 0.0037,
      "reward": 1.723825216293335,
      "reward_std": 0.14209723100066185,
      "rewards/accuracy_reward": 0.7238251268863678,
      "rewards/format_reward": 1.0,
      "step": 909
    },
    {
      "completion_length": 67.4921875,
      "epoch": 4.155251141552512,
      "grad_norm": 2.615705966949463,
      "kl": 0.11328125,
      "learning_rate": 5.844748858447488e-07,
      "loss": 0.0045,
      "reward": 1.6570913791656494,
      "reward_std": 0.1796240657567978,
      "rewards/accuracy_reward": 0.664903849363327,
      "rewards/format_reward": 0.9921875,
      "step": 910
    },
    {
      "completion_length": 76.96875,
      "epoch": 4.159817351598173,
      "grad_norm": 1.9582023620605469,
      "kl": 0.090087890625,
      "learning_rate": 5.840182648401826e-07,
      "loss": 0.0036,
      "reward": 1.7256065011024475,
      "reward_std": 0.17783351242542267,
      "rewards/accuracy_reward": 0.7334189414978027,
      "rewards/format_reward": 0.9921875,
      "step": 911
    },
    {
      "completion_length": 60.40625,
      "epoch": 4.164383561643835,
      "grad_norm": 4.850559711456299,
      "kl": 0.103271484375,
      "learning_rate": 5.835616438356164e-07,
      "loss": 0.0041,
      "reward": 1.524218738079071,
      "reward_std": 0.2731492444872856,
      "rewards/accuracy_reward": 0.5320312678813934,
      "rewards/format_reward": 0.9921875,
      "step": 912
    },
    {
      "completion_length": 50.765625,
      "epoch": 4.168949771689498,
      "grad_norm": 2.815920352935791,
      "kl": 0.18408203125,
      "learning_rate": 5.831050228310503e-07,
      "loss": 0.0074,
      "reward": 1.653542160987854,
      "reward_std": 0.23784886300563812,
      "rewards/accuracy_reward": 0.6691671311855316,
      "rewards/format_reward": 0.984375,
      "step": 913
    },
    {
      "completion_length": 76.1875,
      "epoch": 4.17351598173516,
      "grad_norm": 2.3140416145324707,
      "kl": 0.14208984375,
      "learning_rate": 5.82648401826484e-07,
      "loss": 0.0057,
      "reward": 1.6864583492279053,
      "reward_std": 0.1315075010061264,
      "rewards/accuracy_reward": 0.6864583194255829,
      "rewards/format_reward": 1.0,
      "step": 914
    },
    {
      "completion_length": 73.3515625,
      "epoch": 4.178082191780822,
      "grad_norm": 3.9995479583740234,
      "kl": 0.12451171875,
      "learning_rate": 5.821917808219177e-07,
      "loss": 0.005,
      "reward": 1.7815169095993042,
      "reward_std": 0.08393021672964096,
      "rewards/accuracy_reward": 0.7815168499946594,
      "rewards/format_reward": 1.0,
      "step": 915
    },
    {
      "completion_length": 66.3515625,
      "epoch": 4.182648401826484,
      "grad_norm": 1.871342420578003,
      "kl": 0.12353515625,
      "learning_rate": 5.817351598173516e-07,
      "loss": 0.0049,
      "reward": 1.7460670471191406,
      "reward_std": 0.15686815977096558,
      "rewards/accuracy_reward": 0.7538795471191406,
      "rewards/format_reward": 0.9921875,
      "step": 916
    },
    {
      "completion_length": 77.828125,
      "epoch": 4.187214611872146,
      "grad_norm": 1.733489751815796,
      "kl": 0.095458984375,
      "learning_rate": 5.812785388127853e-07,
      "loss": 0.0038,
      "reward": 1.7660456895828247,
      "reward_std": 0.16200437024235725,
      "rewards/accuracy_reward": 0.7816706299781799,
      "rewards/format_reward": 0.984375,
      "step": 917
    },
    {
      "completion_length": 65.046875,
      "epoch": 4.191780821917808,
      "grad_norm": 3.412703275680542,
      "kl": 0.13232421875,
      "learning_rate": 5.808219178082191e-07,
      "loss": 0.0053,
      "reward": 1.6960819363594055,
      "reward_std": 0.19595444947481155,
      "rewards/accuracy_reward": 0.7117068469524384,
      "rewards/format_reward": 0.984375,
      "step": 918
    },
    {
      "completion_length": 83.4140625,
      "epoch": 4.19634703196347,
      "grad_norm": 1.8262677192687988,
      "kl": 0.120849609375,
      "learning_rate": 5.80365296803653e-07,
      "loss": 0.0048,
      "reward": 1.7575623989105225,
      "reward_std": 0.1326010897755623,
      "rewards/accuracy_reward": 0.7653749287128448,
      "rewards/format_reward": 0.9921875,
      "step": 919
    },
    {
      "completion_length": 86.09375,
      "epoch": 4.200913242009133,
      "grad_norm": 4.315945625305176,
      "kl": 0.10107421875,
      "learning_rate": 5.799086757990867e-07,
      "loss": 0.004,
      "reward": 1.701339304447174,
      "reward_std": 0.14829950034618378,
      "rewards/accuracy_reward": 0.7013393044471741,
      "rewards/format_reward": 1.0,
      "step": 920
    },
    {
      "completion_length": 74.46875,
      "epoch": 4.205479452054795,
      "grad_norm": 1.9493387937545776,
      "kl": 0.08251953125,
      "learning_rate": 5.794520547945206e-07,
      "loss": 0.0033,
      "reward": 1.8070870637893677,
      "reward_std": 0.036666832864284515,
      "rewards/accuracy_reward": 0.8070869743824005,
      "rewards/format_reward": 1.0,
      "step": 921
    },
    {
      "completion_length": 77.328125,
      "epoch": 4.210045662100456,
      "grad_norm": 1.4725087881088257,
      "kl": 0.1162109375,
      "learning_rate": 5.789954337899543e-07,
      "loss": 0.0047,
      "reward": 1.8280134201049805,
      "reward_std": 0.10532564483582973,
      "rewards/accuracy_reward": 0.8358259201049805,
      "rewards/format_reward": 0.9921875,
      "step": 922
    },
    {
      "completion_length": 72.9921875,
      "epoch": 4.2146118721461185,
      "grad_norm": 4.775530815124512,
      "kl": 0.13037109375,
      "learning_rate": 5.78538812785388e-07,
      "loss": 0.0052,
      "reward": 1.7100632786750793,
      "reward_std": 0.17319176718592644,
      "rewards/accuracy_reward": 0.7178757190704346,
      "rewards/format_reward": 0.9921875,
      "step": 923
    },
    {
      "completion_length": 70.015625,
      "epoch": 4.219178082191781,
      "grad_norm": 2.908025026321411,
      "kl": 0.162109375,
      "learning_rate": 5.780821917808219e-07,
      "loss": 0.0065,
      "reward": 1.6276227831840515,
      "reward_std": 0.22536901384592056,
      "rewards/accuracy_reward": 0.6432477831840515,
      "rewards/format_reward": 0.984375,
      "step": 924
    },
    {
      "completion_length": 59.578125,
      "epoch": 4.223744292237443,
      "grad_norm": 5.317495346069336,
      "kl": 0.10595703125,
      "learning_rate": 5.776255707762557e-07,
      "loss": 0.0042,
      "reward": 1.6319011449813843,
      "reward_std": 0.18694934993982315,
      "rewards/accuracy_reward": 0.6397135257720947,
      "rewards/format_reward": 0.9921875,
      "step": 925
    },
    {
      "completion_length": 81.40625,
      "epoch": 4.228310502283105,
      "grad_norm": 2.3484511375427246,
      "kl": 0.091796875,
      "learning_rate": 5.771689497716896e-07,
      "loss": 0.0037,
      "reward": 1.7912667393684387,
      "reward_std": 0.08569350093603134,
      "rewards/accuracy_reward": 0.7912667095661163,
      "rewards/format_reward": 1.0,
      "step": 926
    },
    {
      "completion_length": 68.578125,
      "epoch": 4.232876712328767,
      "grad_norm": 2.1061465740203857,
      "kl": 0.125244140625,
      "learning_rate": 5.767123287671233e-07,
      "loss": 0.005,
      "reward": 1.8626301884651184,
      "reward_std": 0.10864401236176491,
      "rewards/accuracy_reward": 0.8626301884651184,
      "rewards/format_reward": 1.0,
      "step": 927
    },
    {
      "completion_length": 74.9765625,
      "epoch": 4.237442922374429,
      "grad_norm": 3.035919427871704,
      "kl": 0.13134765625,
      "learning_rate": 5.76255707762557e-07,
      "loss": 0.0053,
      "reward": 1.6922495365142822,
      "reward_std": 0.15072567015886307,
      "rewards/accuracy_reward": 0.7156869769096375,
      "rewards/format_reward": 0.9765625,
      "step": 928
    },
    {
      "completion_length": 58.78125,
      "epoch": 4.242009132420091,
      "grad_norm": 2.8555891513824463,
      "kl": 0.16064453125,
      "learning_rate": 5.757990867579909e-07,
      "loss": 0.0064,
      "reward": 1.8009114861488342,
      "reward_std": 0.18833597749471664,
      "rewards/accuracy_reward": 0.8009114265441895,
      "rewards/format_reward": 1.0,
      "step": 929
    },
    {
      "completion_length": 79.625,
      "epoch": 4.2465753424657535,
      "grad_norm": 4.95534086227417,
      "kl": 0.113037109375,
      "learning_rate": 5.753424657534246e-07,
      "loss": 0.0045,
      "reward": 1.668163239955902,
      "reward_std": 0.20845329016447067,
      "rewards/accuracy_reward": 0.6759756505489349,
      "rewards/format_reward": 0.9921875,
      "step": 930
    },
    {
      "completion_length": 57.59375,
      "epoch": 4.251141552511416,
      "grad_norm": 3.5745315551757812,
      "kl": 0.158203125,
      "learning_rate": 5.748858447488583e-07,
      "loss": 0.0063,
      "reward": 1.7814725637435913,
      "reward_std": 0.1947070211172104,
      "rewards/accuracy_reward": 0.7814724445343018,
      "rewards/format_reward": 1.0,
      "step": 931
    },
    {
      "completion_length": 69.296875,
      "epoch": 4.255707762557078,
      "grad_norm": 2.9775948524475098,
      "kl": 0.129638671875,
      "learning_rate": 5.744292237442923e-07,
      "loss": 0.0052,
      "reward": 1.5528324842453003,
      "reward_std": 0.23214496672153473,
      "rewards/accuracy_reward": 0.5684574842453003,
      "rewards/format_reward": 0.984375,
      "step": 932
    },
    {
      "completion_length": 85.0859375,
      "epoch": 4.260273972602739,
      "grad_norm": 2.191046714782715,
      "kl": 0.13232421875,
      "learning_rate": 5.73972602739726e-07,
      "loss": 0.0053,
      "reward": 1.666010558605194,
      "reward_std": 0.16855772212147713,
      "rewards/accuracy_reward": 0.6816355586051941,
      "rewards/format_reward": 0.984375,
      "step": 933
    },
    {
      "completion_length": 84.921875,
      "epoch": 4.264840182648402,
      "grad_norm": 1.6789612770080566,
      "kl": 0.0821533203125,
      "learning_rate": 5.735159817351598e-07,
      "loss": 0.0033,
      "reward": 1.749913215637207,
      "reward_std": 0.07257125526666641,
      "rewards/accuracy_reward": 0.7499131858348846,
      "rewards/format_reward": 1.0,
      "step": 934
    },
    {
      "completion_length": 78.6171875,
      "epoch": 4.269406392694064,
      "grad_norm": 4.655454158782959,
      "kl": 0.116455078125,
      "learning_rate": 5.730593607305936e-07,
      "loss": 0.0047,
      "reward": 1.7203125953674316,
      "reward_std": 0.14768873527646065,
      "rewards/accuracy_reward": 0.7203124463558197,
      "rewards/format_reward": 1.0,
      "step": 935
    },
    {
      "completion_length": 79.765625,
      "epoch": 4.273972602739726,
      "grad_norm": 2.1534454822540283,
      "kl": 0.093017578125,
      "learning_rate": 5.726027397260273e-07,
      "loss": 0.0037,
      "reward": 1.7255195379257202,
      "reward_std": 0.18480905890464783,
      "rewards/accuracy_reward": 0.7489570677280426,
      "rewards/format_reward": 0.9765625,
      "step": 936
    },
    {
      "completion_length": 82.1015625,
      "epoch": 4.2785388127853885,
      "grad_norm": 4.468496799468994,
      "kl": 0.081298828125,
      "learning_rate": 5.721461187214612e-07,
      "loss": 0.0032,
      "reward": 1.8500558733940125,
      "reward_std": 0.13352815061807632,
      "rewards/accuracy_reward": 0.8578682541847229,
      "rewards/format_reward": 0.9921875,
      "step": 937
    },
    {
      "completion_length": 55.5078125,
      "epoch": 4.28310502283105,
      "grad_norm": 2.743320941925049,
      "kl": 0.18212890625,
      "learning_rate": 5.716894977168949e-07,
      "loss": 0.0073,
      "reward": 1.5885499119758606,
      "reward_std": 0.20670751482248306,
      "rewards/accuracy_reward": 0.6041748821735382,
      "rewards/format_reward": 0.984375,
      "step": 938
    },
    {
      "completion_length": 77.984375,
      "epoch": 4.287671232876712,
      "grad_norm": 3.1169564723968506,
      "kl": 0.102783203125,
      "learning_rate": 5.712328767123287e-07,
      "loss": 0.0041,
      "reward": 1.6959820985794067,
      "reward_std": 0.1572049930691719,
      "rewards/accuracy_reward": 0.7037945985794067,
      "rewards/format_reward": 0.9921875,
      "step": 939
    },
    {
      "completion_length": 78.359375,
      "epoch": 4.292237442922374,
      "grad_norm": 1.5095194578170776,
      "kl": 0.11083984375,
      "learning_rate": 5.707762557077626e-07,
      "loss": 0.0044,
      "reward": 1.7596355080604553,
      "reward_std": 0.1378481425344944,
      "rewards/accuracy_reward": 0.7830728888511658,
      "rewards/format_reward": 0.9765625,
      "step": 940
    },
    {
      "completion_length": 61.1171875,
      "epoch": 4.296803652968037,
      "grad_norm": 2.8175227642059326,
      "kl": 0.153076171875,
      "learning_rate": 5.703196347031963e-07,
      "loss": 0.0061,
      "reward": 1.6882672905921936,
      "reward_std": 0.1907612383365631,
      "rewards/accuracy_reward": 0.696079820394516,
      "rewards/format_reward": 0.9921875,
      "step": 941
    },
    {
      "completion_length": 77.0546875,
      "epoch": 4.301369863013699,
      "grad_norm": 2.724752426147461,
      "kl": 0.14306640625,
      "learning_rate": 5.698630136986301e-07,
      "loss": 0.0057,
      "reward": 1.6338477730751038,
      "reward_std": 0.1553211621940136,
      "rewards/accuracy_reward": 0.633847713470459,
      "rewards/format_reward": 1.0,
      "step": 942
    },
    {
      "completion_length": 72.0625,
      "epoch": 4.30593607305936,
      "grad_norm": 1.5134310722351074,
      "kl": 0.109375,
      "learning_rate": 5.694063926940639e-07,
      "loss": 0.0044,
      "reward": 1.8307477235794067,
      "reward_std": 0.10607551783323288,
      "rewards/accuracy_reward": 0.8307477533817291,
      "rewards/format_reward": 1.0,
      "step": 943
    },
    {
      "completion_length": 84.6796875,
      "epoch": 4.310502283105023,
      "grad_norm": 1.9289695024490356,
      "kl": 0.120849609375,
      "learning_rate": 5.689497716894976e-07,
      "loss": 0.0048,
      "reward": 1.7625186443328857,
      "reward_std": 0.14161107502877712,
      "rewards/accuracy_reward": 0.7703310549259186,
      "rewards/format_reward": 0.9921875,
      "step": 944
    },
    {
      "completion_length": 68.6171875,
      "epoch": 4.315068493150685,
      "grad_norm": 2.0734076499938965,
      "kl": 0.149169921875,
      "learning_rate": 5.684931506849316e-07,
      "loss": 0.006,
      "reward": 1.6850000619888306,
      "reward_std": 0.14654473960399628,
      "rewards/accuracy_reward": 0.6928124725818634,
      "rewards/format_reward": 0.9921875,
      "step": 945
    },
    {
      "completion_length": 68.4765625,
      "epoch": 4.319634703196347,
      "grad_norm": 2.146794080734253,
      "kl": 0.11572265625,
      "learning_rate": 5.680365296803653e-07,
      "loss": 0.0046,
      "reward": 1.70016747713089,
      "reward_std": 0.20281969010829926,
      "rewards/accuracy_reward": 0.7157924175262451,
      "rewards/format_reward": 0.984375,
      "step": 946
    },
    {
      "completion_length": 68.03125,
      "epoch": 4.324200913242009,
      "grad_norm": 2.0439021587371826,
      "kl": 0.126953125,
      "learning_rate": 5.67579908675799e-07,
      "loss": 0.0051,
      "reward": 1.8492187857627869,
      "reward_std": 0.09047675505280495,
      "rewards/accuracy_reward": 0.8492186665534973,
      "rewards/format_reward": 1.0,
      "step": 947
    },
    {
      "completion_length": 68.6953125,
      "epoch": 4.328767123287671,
      "grad_norm": 5.651437759399414,
      "kl": 0.129638671875,
      "learning_rate": 5.671232876712329e-07,
      "loss": 0.0052,
      "reward": 1.7400199174880981,
      "reward_std": 0.19576279073953629,
      "rewards/accuracy_reward": 0.7478323876857758,
      "rewards/format_reward": 0.9921875,
      "step": 948
    },
    {
      "completion_length": 92.3828125,
      "epoch": 4.333333333333333,
      "grad_norm": 2.3988163471221924,
      "kl": 0.107666015625,
      "learning_rate": 5.666666666666666e-07,
      "loss": 0.0043,
      "reward": 1.7797867059707642,
      "reward_std": 0.091391421854496,
      "rewards/accuracy_reward": 0.7797866761684418,
      "rewards/format_reward": 1.0,
      "step": 949
    },
    {
      "completion_length": 65.796875,
      "epoch": 4.337899543378995,
      "grad_norm": 4.604662895202637,
      "kl": 0.14306640625,
      "learning_rate": 5.662100456621004e-07,
      "loss": 0.0057,
      "reward": 1.5967974662780762,
      "reward_std": 0.19326772540807724,
      "rewards/accuracy_reward": 0.596797525882721,
      "rewards/format_reward": 1.0,
      "step": 950
    },
    {
      "completion_length": 65.34375,
      "epoch": 4.342465753424658,
      "grad_norm": 2.6304314136505127,
      "kl": 0.194091796875,
      "learning_rate": 5.657534246575342e-07,
      "loss": 0.0078,
      "reward": 1.6250000596046448,
      "reward_std": 0.22805283963680267,
      "rewards/accuracy_reward": 0.6484375,
      "rewards/format_reward": 0.9765625,
      "step": 951
    },
    {
      "completion_length": 79.5390625,
      "epoch": 4.34703196347032,
      "grad_norm": 26.240947723388672,
      "kl": 0.1016845703125,
      "learning_rate": 5.65296803652968e-07,
      "loss": 0.0041,
      "reward": 1.7042073011398315,
      "reward_std": 0.1676829382777214,
      "rewards/accuracy_reward": 0.7120197415351868,
      "rewards/format_reward": 0.9921875,
      "step": 952
    },
    {
      "completion_length": 74.0390625,
      "epoch": 4.351598173515982,
      "grad_norm": 2.3484628200531006,
      "kl": 0.178466796875,
      "learning_rate": 5.648401826484019e-07,
      "loss": 0.0072,
      "reward": 1.7171673774719238,
      "reward_std": 0.16094867885112762,
      "rewards/accuracy_reward": 0.7249797880649567,
      "rewards/format_reward": 0.9921875,
      "step": 953
    },
    {
      "completion_length": 74.578125,
      "epoch": 4.3561643835616435,
      "grad_norm": 42.42313766479492,
      "kl": 0.120361328125,
      "learning_rate": 5.643835616438356e-07,
      "loss": 0.0048,
      "reward": 1.6389508843421936,
      "reward_std": 0.1658247858285904,
      "rewards/accuracy_reward": 0.638950914144516,
      "rewards/format_reward": 1.0,
      "step": 954
    },
    {
      "completion_length": 76.90625,
      "epoch": 4.360730593607306,
      "grad_norm": 12.59420108795166,
      "kl": 0.126220703125,
      "learning_rate": 5.639269406392693e-07,
      "loss": 0.005,
      "reward": 1.704687476158142,
      "reward_std": 0.18361148238182068,
      "rewards/accuracy_reward": 0.7046875059604645,
      "rewards/format_reward": 1.0,
      "step": 955
    },
    {
      "completion_length": 84.125,
      "epoch": 4.365296803652968,
      "grad_norm": 2.1863152980804443,
      "kl": 0.107177734375,
      "learning_rate": 5.634703196347032e-07,
      "loss": 0.0043,
      "reward": 1.6656250953674316,
      "reward_std": 0.1505398079752922,
      "rewards/accuracy_reward": 0.6812499761581421,
      "rewards/format_reward": 0.984375,
      "step": 956
    },
    {
      "completion_length": 66.25,
      "epoch": 4.36986301369863,
      "grad_norm": 5.098891735076904,
      "kl": 0.134033203125,
      "learning_rate": 5.630136986301369e-07,
      "loss": 0.0054,
      "reward": 1.7317472100257874,
      "reward_std": 0.13505896925926208,
      "rewards/accuracy_reward": 0.7395596504211426,
      "rewards/format_reward": 0.9921875,
      "step": 957
    },
    {
      "completion_length": 66.5078125,
      "epoch": 4.3744292237442925,
      "grad_norm": 10.807661056518555,
      "kl": 0.119873046875,
      "learning_rate": 5.625570776255707e-07,
      "loss": 0.0048,
      "reward": 1.7822917103767395,
      "reward_std": 0.11853177845478058,
      "rewards/accuracy_reward": 0.7822916507720947,
      "rewards/format_reward": 1.0,
      "step": 958
    },
    {
      "completion_length": 67.375,
      "epoch": 4.378995433789954,
      "grad_norm": 1.9763253927230835,
      "kl": 0.16064453125,
      "learning_rate": 5.621004566210046e-07,
      "loss": 0.0064,
      "reward": 1.6397321224212646,
      "reward_std": 0.2425994575023651,
      "rewards/accuracy_reward": 0.6631696224212646,
      "rewards/format_reward": 0.9765625,
      "step": 959
    },
    {
      "completion_length": 73.609375,
      "epoch": 4.383561643835616,
      "grad_norm": 11.640439987182617,
      "kl": 0.12841796875,
      "learning_rate": 5.616438356164383e-07,
      "loss": 0.0051,
      "reward": 1.6920552849769592,
      "reward_std": 0.23473482113331556,
      "rewards/accuracy_reward": 0.7154927253723145,
      "rewards/format_reward": 0.9765625,
      "step": 960
    },
    {
      "completion_length": 67.1640625,
      "epoch": 4.3881278538812785,
      "grad_norm": 4.730476379394531,
      "kl": 0.091796875,
      "learning_rate": 5.611872146118722e-07,
      "loss": 0.0037,
      "reward": 1.8379933834075928,
      "reward_std": 0.07210628129541874,
      "rewards/accuracy_reward": 0.8379934132099152,
      "rewards/format_reward": 1.0,
      "step": 961
    },
    {
      "completion_length": 95.25,
      "epoch": 4.392694063926941,
      "grad_norm": 1.63775634765625,
      "kl": 0.069091796875,
      "learning_rate": 5.607305936073059e-07,
      "loss": 0.0028,
      "reward": 1.818750023841858,
      "reward_std": 0.0883883461356163,
      "rewards/accuracy_reward": 0.8343749344348907,
      "rewards/format_reward": 0.984375,
      "step": 962
    },
    {
      "completion_length": 58.0078125,
      "epoch": 4.397260273972603,
      "grad_norm": 2.0879337787628174,
      "kl": 0.162109375,
      "learning_rate": 5.602739726027396e-07,
      "loss": 0.0065,
      "reward": 1.810954213142395,
      "reward_std": 0.18426654487848282,
      "rewards/accuracy_reward": 0.818766713142395,
      "rewards/format_reward": 0.9921875,
      "step": 963
    },
    {
      "completion_length": 60.8125,
      "epoch": 4.401826484018265,
      "grad_norm": 4.00584077835083,
      "kl": 0.1513671875,
      "learning_rate": 5.598173515981735e-07,
      "loss": 0.0061,
      "reward": 1.765178620815277,
      "reward_std": 0.20918096601963043,
      "rewards/accuracy_reward": 0.7729910910129547,
      "rewards/format_reward": 0.9921875,
      "step": 964
    },
    {
      "completion_length": 62.6171875,
      "epoch": 4.406392694063927,
      "grad_norm": 2.43243408203125,
      "kl": 0.134521484375,
      "learning_rate": 5.593607305936073e-07,
      "loss": 0.0054,
      "reward": 1.7512852549552917,
      "reward_std": 0.20323559269309044,
      "rewards/accuracy_reward": 0.7512852847576141,
      "rewards/format_reward": 1.0,
      "step": 965
    },
    {
      "completion_length": 87.6875,
      "epoch": 4.410958904109589,
      "grad_norm": 2.643324375152588,
      "kl": 0.10693359375,
      "learning_rate": 5.589041095890411e-07,
      "loss": 0.0043,
      "reward": 1.7709821462631226,
      "reward_std": 0.12626906856894493,
      "rewards/accuracy_reward": 0.7787946164608002,
      "rewards/format_reward": 0.9921875,
      "step": 966
    },
    {
      "completion_length": 73.7421875,
      "epoch": 4.415525114155251,
      "grad_norm": 3.2244210243225098,
      "kl": 0.10791015625,
      "learning_rate": 5.584474885844749e-07,
      "loss": 0.0043,
      "reward": 1.76171875,
      "reward_std": 0.18859816156327724,
      "rewards/accuracy_reward": 0.7695312201976776,
      "rewards/format_reward": 0.9921875,
      "step": 967
    },
    {
      "completion_length": 77.84375,
      "epoch": 4.420091324200913,
      "grad_norm": 4.4358625411987305,
      "kl": 0.11962890625,
      "learning_rate": 5.579908675799086e-07,
      "loss": 0.0048,
      "reward": 1.8305882215499878,
      "reward_std": 0.09982336312532425,
      "rewards/accuracy_reward": 0.8305881917476654,
      "rewards/format_reward": 1.0,
      "step": 968
    },
    {
      "completion_length": 62.171875,
      "epoch": 4.424657534246576,
      "grad_norm": 4.14423942565918,
      "kl": 0.133056640625,
      "learning_rate": 5.575342465753425e-07,
      "loss": 0.0053,
      "reward": 1.6299851536750793,
      "reward_std": 0.19754792749881744,
      "rewards/accuracy_reward": 0.637797623872757,
      "rewards/format_reward": 0.9921875,
      "step": 969
    },
    {
      "completion_length": 83.1953125,
      "epoch": 4.429223744292237,
      "grad_norm": 1.427676796913147,
      "kl": 0.087158203125,
      "learning_rate": 5.570776255707762e-07,
      "loss": 0.0035,
      "reward": 1.8171875476837158,
      "reward_std": 0.11048543266952038,
      "rewards/accuracy_reward": 0.8249998986721039,
      "rewards/format_reward": 0.9921875,
      "step": 970
    },
    {
      "completion_length": 83.6484375,
      "epoch": 4.433789954337899,
      "grad_norm": 2.3894028663635254,
      "kl": 0.098388671875,
      "learning_rate": 5.566210045662099e-07,
      "loss": 0.0039,
      "reward": 1.7342448234558105,
      "reward_std": 0.24424592405557632,
      "rewards/accuracy_reward": 0.7576822340488434,
      "rewards/format_reward": 0.9765625,
      "step": 971
    },
    {
      "completion_length": 86.203125,
      "epoch": 4.438356164383562,
      "grad_norm": 2.410252571105957,
      "kl": 0.103515625,
      "learning_rate": 5.561643835616439e-07,
      "loss": 0.0041,
      "reward": 1.8436079621315002,
      "reward_std": 0.08567275106906891,
      "rewards/accuracy_reward": 0.8514204323291779,
      "rewards/format_reward": 0.9921875,
      "step": 972
    },
    {
      "completion_length": 67.296875,
      "epoch": 4.442922374429224,
      "grad_norm": 3.1359169483184814,
      "kl": 0.156982421875,
      "learning_rate": 5.557077625570776e-07,
      "loss": 0.0063,
      "reward": 1.6325520277023315,
      "reward_std": 0.1258198693394661,
      "rewards/accuracy_reward": 0.6325520575046539,
      "rewards/format_reward": 1.0,
      "step": 973
    },
    {
      "completion_length": 75.625,
      "epoch": 4.447488584474886,
      "grad_norm": 3.9389309883117676,
      "kl": 0.17529296875,
      "learning_rate": 5.552511415525114e-07,
      "loss": 0.007,
      "reward": 1.6369017958641052,
      "reward_std": 0.21986636519432068,
      "rewards/accuracy_reward": 0.6447142958641052,
      "rewards/format_reward": 0.9921875,
      "step": 974
    },
    {
      "completion_length": 75.4765625,
      "epoch": 4.4520547945205475,
      "grad_norm": 19.899776458740234,
      "kl": 0.638916015625,
      "learning_rate": 5.547945205479452e-07,
      "loss": 0.0255,
      "reward": 1.819531261920929,
      "reward_std": 0.07878133933991194,
      "rewards/accuracy_reward": 0.819531261920929,
      "rewards/format_reward": 1.0,
      "step": 975
    },
    {
      "completion_length": 73.5625,
      "epoch": 4.45662100456621,
      "grad_norm": 3.0659143924713135,
      "kl": 0.095703125,
      "learning_rate": 5.543378995433789e-07,
      "loss": 0.0038,
      "reward": 1.6691592335700989,
      "reward_std": 0.19636988639831543,
      "rewards/accuracy_reward": 0.6847842335700989,
      "rewards/format_reward": 0.984375,
      "step": 976
    },
    {
      "completion_length": 82.0,
      "epoch": 4.461187214611872,
      "grad_norm": 2.006531000137329,
      "kl": 0.101318359375,
      "learning_rate": 5.538812785388128e-07,
      "loss": 0.0041,
      "reward": 1.8178571462631226,
      "reward_std": 0.1673773005604744,
      "rewards/accuracy_reward": 0.8334820866584778,
      "rewards/format_reward": 0.984375,
      "step": 977
    },
    {
      "completion_length": 55.5234375,
      "epoch": 4.465753424657534,
      "grad_norm": 2.1843132972717285,
      "kl": 0.158203125,
      "learning_rate": 5.534246575342465e-07,
      "loss": 0.0063,
      "reward": 1.770518183708191,
      "reward_std": 0.19590065628290176,
      "rewards/accuracy_reward": 0.7783306241035461,
      "rewards/format_reward": 0.9921875,
      "step": 978
    },
    {
      "completion_length": 54.3984375,
      "epoch": 4.470319634703197,
      "grad_norm": 5.7283830642700195,
      "kl": 0.16845703125,
      "learning_rate": 5.529680365296803e-07,
      "loss": 0.0067,
      "reward": 1.692187488079071,
      "reward_std": 0.193861223757267,
      "rewards/accuracy_reward": 0.6999999582767487,
      "rewards/format_reward": 0.9921875,
      "step": 979
    },
    {
      "completion_length": 73.6171875,
      "epoch": 4.474885844748858,
      "grad_norm": 4.302793979644775,
      "kl": 0.139404296875,
      "learning_rate": 5.525114155251142e-07,
      "loss": 0.0056,
      "reward": 1.7293124198913574,
      "reward_std": 0.14934544544667006,
      "rewards/accuracy_reward": 0.7293124198913574,
      "rewards/format_reward": 1.0,
      "step": 980
    },
    {
      "completion_length": 87.0625,
      "epoch": 4.47945205479452,
      "grad_norm": 2.7895452976226807,
      "kl": 0.115966796875,
      "learning_rate": 5.520547945205479e-07,
      "loss": 0.0046,
      "reward": 1.7789062857627869,
      "reward_std": 0.09265873953700066,
      "rewards/accuracy_reward": 0.7789061963558197,
      "rewards/format_reward": 1.0,
      "step": 981
    },
    {
      "completion_length": 78.3984375,
      "epoch": 4.4840182648401825,
      "grad_norm": 3.9894137382507324,
      "kl": 0.1279296875,
      "learning_rate": 5.515981735159817e-07,
      "loss": 0.0051,
      "reward": 1.6846325397491455,
      "reward_std": 0.18365756422281265,
      "rewards/accuracy_reward": 0.7002575993537903,
      "rewards/format_reward": 0.984375,
      "step": 982
    },
    {
      "completion_length": 76.875,
      "epoch": 4.488584474885845,
      "grad_norm": 3.384289503097534,
      "kl": 0.150634765625,
      "learning_rate": 5.511415525114155e-07,
      "loss": 0.006,
      "reward": 1.7085938453674316,
      "reward_std": 0.20954116433858871,
      "rewards/accuracy_reward": 0.7320311665534973,
      "rewards/format_reward": 0.9765625,
      "step": 983
    },
    {
      "completion_length": 49.8671875,
      "epoch": 4.493150684931507,
      "grad_norm": 5.41481876373291,
      "kl": 0.14111328125,
      "learning_rate": 5.506849315068492e-07,
      "loss": 0.0056,
      "reward": 1.5635416507720947,
      "reward_std": 0.23787462711334229,
      "rewards/accuracy_reward": 0.5635416507720947,
      "rewards/format_reward": 1.0,
      "step": 984
    },
    {
      "completion_length": 87.890625,
      "epoch": 4.497716894977169,
      "grad_norm": 1.2277454137802124,
      "kl": 0.0751953125,
      "learning_rate": 5.502283105022832e-07,
      "loss": 0.003,
      "reward": 1.7507812976837158,
      "reward_std": 0.05524272099137306,
      "rewards/accuracy_reward": 0.7507811486721039,
      "rewards/format_reward": 1.0,
      "step": 985
    },
    {
      "completion_length": 61.9453125,
      "epoch": 4.502283105022831,
      "grad_norm": 5.256927490234375,
      "kl": 0.1865234375,
      "learning_rate": 5.497716894977169e-07,
      "loss": 0.0075,
      "reward": 1.610156238079071,
      "reward_std": 0.2878893092274666,
      "rewards/accuracy_reward": 0.625781238079071,
      "rewards/format_reward": 0.984375,
      "step": 986
    },
    {
      "completion_length": 82.71875,
      "epoch": 4.506849315068493,
      "grad_norm": 3.515986919403076,
      "kl": 0.18603515625,
      "learning_rate": 5.493150684931506e-07,
      "loss": 0.0074,
      "reward": 1.7588542103767395,
      "reward_std": 0.1455376148223877,
      "rewards/accuracy_reward": 0.7588541209697723,
      "rewards/format_reward": 1.0,
      "step": 987
    },
    {
      "completion_length": 67.7578125,
      "epoch": 4.511415525114155,
      "grad_norm": 2.741415023803711,
      "kl": 0.14306640625,
      "learning_rate": 5.488584474885845e-07,
      "loss": 0.0057,
      "reward": 1.5901537537574768,
      "reward_std": 0.15988682955503464,
      "rewards/accuracy_reward": 0.5901537537574768,
      "rewards/format_reward": 1.0,
      "step": 988
    },
    {
      "completion_length": 66.84375,
      "epoch": 4.5159817351598175,
      "grad_norm": 2.9208884239196777,
      "kl": 0.12060546875,
      "learning_rate": 5.484018264840182e-07,
      "loss": 0.0048,
      "reward": 1.7678571939468384,
      "reward_std": 0.14058196544647217,
      "rewards/accuracy_reward": 0.7678571343421936,
      "rewards/format_reward": 1.0,
      "step": 989
    },
    {
      "completion_length": 67.2734375,
      "epoch": 4.52054794520548,
      "grad_norm": 3.776416540145874,
      "kl": 0.126708984375,
      "learning_rate": 5.47945205479452e-07,
      "loss": 0.0051,
      "reward": 1.6418346166610718,
      "reward_std": 0.1818918213248253,
      "rewards/accuracy_reward": 0.6418345868587494,
      "rewards/format_reward": 1.0,
      "step": 990
    },
    {
      "completion_length": 72.625,
      "epoch": 4.525114155251142,
      "grad_norm": 2.1685075759887695,
      "kl": 0.138916015625,
      "learning_rate": 5.474885844748858e-07,
      "loss": 0.0056,
      "reward": 1.7085193395614624,
      "reward_std": 0.22063866257667542,
      "rewards/accuracy_reward": 0.72414430975914,
      "rewards/format_reward": 0.984375,
      "step": 991
    },
    {
      "completion_length": 66.28125,
      "epoch": 4.529680365296803,
      "grad_norm": 14.555278778076172,
      "kl": 0.1513671875,
      "learning_rate": 5.470319634703196e-07,
      "loss": 0.006,
      "reward": 1.6950520873069763,
      "reward_std": 0.19789891690015793,
      "rewards/accuracy_reward": 0.6950520873069763,
      "rewards/format_reward": 1.0,
      "step": 992
    },
    {
      "completion_length": 68.71875,
      "epoch": 4.534246575342466,
      "grad_norm": 4.036056995391846,
      "kl": 0.15283203125,
      "learning_rate": 5.465753424657535e-07,
      "loss": 0.0061,
      "reward": 1.5639322996139526,
      "reward_std": 0.2465272918343544,
      "rewards/accuracy_reward": 0.5795572698116302,
      "rewards/format_reward": 0.984375,
      "step": 993
    },
    {
      "completion_length": 72.296875,
      "epoch": 4.538812785388128,
      "grad_norm": 4.413112163543701,
      "kl": 0.2080078125,
      "learning_rate": 5.461187214611872e-07,
      "loss": 0.0083,
      "reward": 1.6929687857627869,
      "reward_std": 0.26582426577806473,
      "rewards/accuracy_reward": 0.7242187261581421,
      "rewards/format_reward": 0.96875,
      "step": 994
    },
    {
      "completion_length": 71.1171875,
      "epoch": 4.54337899543379,
      "grad_norm": 2.5498805046081543,
      "kl": 0.13623046875,
      "learning_rate": 5.456621004566209e-07,
      "loss": 0.0055,
      "reward": 1.696093738079071,
      "reward_std": 0.20334278792142868,
      "rewards/accuracy_reward": 0.7039062678813934,
      "rewards/format_reward": 0.9921875,
      "step": 995
    },
    {
      "completion_length": 80.1796875,
      "epoch": 4.5479452054794525,
      "grad_norm": 1.7603377103805542,
      "kl": 0.117919921875,
      "learning_rate": 5.452054794520548e-07,
      "loss": 0.0047,
      "reward": 1.6623343229293823,
      "reward_std": 0.15708915889263153,
      "rewards/accuracy_reward": 0.6779592037200928,
      "rewards/format_reward": 0.984375,
      "step": 996
    },
    {
      "completion_length": 56.375,
      "epoch": 4.552511415525114,
      "grad_norm": 4.723056316375732,
      "kl": 0.17236328125,
      "learning_rate": 5.447488584474885e-07,
      "loss": 0.0069,
      "reward": 1.7257593870162964,
      "reward_std": 0.21335439383983612,
      "rewards/accuracy_reward": 0.7257594168186188,
      "rewards/format_reward": 1.0,
      "step": 997
    },
    {
      "completion_length": 59.25,
      "epoch": 4.557077625570776,
      "grad_norm": 2.3055455684661865,
      "kl": 0.1591796875,
      "learning_rate": 5.442922374429223e-07,
      "loss": 0.0064,
      "reward": 1.7297247648239136,
      "reward_std": 0.17667409405112267,
      "rewards/accuracy_reward": 0.7375371754169464,
      "rewards/format_reward": 0.9921875,
      "step": 998
    },
    {
      "completion_length": 55.4296875,
      "epoch": 4.561643835616438,
      "grad_norm": 10.311870574951172,
      "kl": 0.135009765625,
      "learning_rate": 5.438356164383562e-07,
      "loss": 0.0054,
      "reward": 1.48444002866745,
      "reward_std": 0.28679582476615906,
      "rewards/accuracy_reward": 0.49225252866744995,
      "rewards/format_reward": 0.9921875,
      "step": 999
    },
    {
      "completion_length": 60.46875,
      "epoch": 4.566210045662101,
      "grad_norm": 3.404649019241333,
      "kl": 0.1982421875,
      "learning_rate": 5.433789954337899e-07,
      "loss": 0.0079,
      "reward": 1.5433160066604614,
      "reward_std": 0.2616465389728546,
      "rewards/accuracy_reward": 0.574565976858139,
      "rewards/format_reward": 0.96875,
      "step": 1000
    },
    {
      "completion_length": 73.15625,
      "epoch": 4.570776255707763,
      "grad_norm": 1.8293238878250122,
      "kl": 0.118408203125,
      "learning_rate": 5.429223744292238e-07,
      "loss": 0.0047,
      "reward": 1.8298460245132446,
      "reward_std": 0.1409841626882553,
      "rewards/accuracy_reward": 0.8454709947109222,
      "rewards/format_reward": 0.984375,
      "step": 1001
    },
    {
      "completion_length": 76.0703125,
      "epoch": 4.575342465753424,
      "grad_norm": 1.9321002960205078,
      "kl": 0.111328125,
      "learning_rate": 5.424657534246575e-07,
      "loss": 0.0044,
      "reward": 1.725000023841858,
      "reward_std": 0.15813970565795898,
      "rewards/accuracy_reward": 0.7328124344348907,
      "rewards/format_reward": 0.9921875,
      "step": 1002
    },
    {
      "completion_length": 65.5703125,
      "epoch": 4.579908675799087,
      "grad_norm": 2.1403932571411133,
      "kl": 0.133056640625,
      "learning_rate": 5.420091324200912e-07,
      "loss": 0.0053,
      "reward": 1.7848958373069763,
      "reward_std": 0.21307425945997238,
      "rewards/accuracy_reward": 0.8083333373069763,
      "rewards/format_reward": 0.9765625,
      "step": 1003
    },
    {
      "completion_length": 108.484375,
      "epoch": 4.584474885844749,
      "grad_norm": 1.755354642868042,
      "kl": 0.10107421875,
      "learning_rate": 5.415525114155251e-07,
      "loss": 0.004,
      "reward": 1.8961884379386902,
      "reward_std": 0.060667259618639946,
      "rewards/accuracy_reward": 0.896188348531723,
      "rewards/format_reward": 1.0,
      "step": 1004
    },
    {
      "completion_length": 48.4140625,
      "epoch": 4.589041095890411,
      "grad_norm": 1.387018084526062,
      "kl": 0.1494140625,
      "learning_rate": 5.410958904109589e-07,
      "loss": 0.006,
      "reward": 1.8950520753860474,
      "reward_std": 0.12889967486262321,
      "rewards/accuracy_reward": 0.9028645753860474,
      "rewards/format_reward": 0.9921875,
      "step": 1005
    },
    {
      "completion_length": 62.9140625,
      "epoch": 4.593607305936073,
      "grad_norm": 3.6895904541015625,
      "kl": 0.20751953125,
      "learning_rate": 5.406392694063927e-07,
      "loss": 0.0083,
      "reward": 1.6447916626930237,
      "reward_std": 0.25781603902578354,
      "rewards/accuracy_reward": 0.6526041626930237,
      "rewards/format_reward": 0.9921875,
      "step": 1006
    },
    {
      "completion_length": 73.59375,
      "epoch": 4.598173515981735,
      "grad_norm": 5.844861030578613,
      "kl": 0.118896484375,
      "learning_rate": 5.401826484018265e-07,
      "loss": 0.0048,
      "reward": 1.6738625168800354,
      "reward_std": 0.16072557866573334,
      "rewards/accuracy_reward": 0.6894874572753906,
      "rewards/format_reward": 0.984375,
      "step": 1007
    },
    {
      "completion_length": 62.390625,
      "epoch": 4.602739726027397,
      "grad_norm": 9.675552368164062,
      "kl": 0.143798828125,
      "learning_rate": 5.397260273972602e-07,
      "loss": 0.0058,
      "reward": 1.6939173936843872,
      "reward_std": 0.1647581309080124,
      "rewards/accuracy_reward": 0.6939173638820648,
      "rewards/format_reward": 1.0,
      "step": 1008
    },
    {
      "completion_length": 67.1015625,
      "epoch": 4.607305936073059,
      "grad_norm": 5.169093132019043,
      "kl": 0.130859375,
      "learning_rate": 5.392694063926941e-07,
      "loss": 0.0052,
      "reward": 1.6269097924232483,
      "reward_std": 0.23041004687547684,
      "rewards/accuracy_reward": 0.6425347030162811,
      "rewards/format_reward": 0.984375,
      "step": 1009
    },
    {
      "completion_length": 66.6328125,
      "epoch": 4.6118721461187215,
      "grad_norm": 1.9441876411437988,
      "kl": 0.106201171875,
      "learning_rate": 5.388127853881278e-07,
      "loss": 0.0043,
      "reward": 1.5781250596046448,
      "reward_std": 0.14283225312829018,
      "rewards/accuracy_reward": 0.578125,
      "rewards/format_reward": 1.0,
      "step": 1010
    },
    {
      "completion_length": 62.9296875,
      "epoch": 4.616438356164384,
      "grad_norm": 3.9866504669189453,
      "kl": 0.20361328125,
      "learning_rate": 5.383561643835615e-07,
      "loss": 0.0081,
      "reward": 1.718553066253662,
      "reward_std": 0.1621505692601204,
      "rewards/accuracy_reward": 0.7185530364513397,
      "rewards/format_reward": 1.0,
      "step": 1011
    },
    {
      "completion_length": 54.625,
      "epoch": 4.621004566210045,
      "grad_norm": 3.620591402053833,
      "kl": 0.1474609375,
      "learning_rate": 5.378995433789955e-07,
      "loss": 0.0059,
      "reward": 1.813281238079071,
      "reward_std": 0.07733980193734169,
      "rewards/accuracy_reward": 0.813281238079071,
      "rewards/format_reward": 1.0,
      "step": 1012
    },
    {
      "completion_length": 71.109375,
      "epoch": 4.6255707762557075,
      "grad_norm": 3.377457857131958,
      "kl": 0.14208984375,
      "learning_rate": 5.374429223744292e-07,
      "loss": 0.0057,
      "reward": 1.74702388048172,
      "reward_std": 0.1983274295926094,
      "rewards/accuracy_reward": 0.7470237612724304,
      "rewards/format_reward": 1.0,
      "step": 1013
    },
    {
      "completion_length": 59.6171875,
      "epoch": 4.63013698630137,
      "grad_norm": 3.1818978786468506,
      "kl": 0.16064453125,
      "learning_rate": 5.36986301369863e-07,
      "loss": 0.0064,
      "reward": 1.667509913444519,
      "reward_std": 0.17691579461097717,
      "rewards/accuracy_reward": 0.6675098836421967,
      "rewards/format_reward": 1.0,
      "step": 1014
    },
    {
      "completion_length": 62.40625,
      "epoch": 4.634703196347032,
      "grad_norm": 2.588076114654541,
      "kl": 0.198974609375,
      "learning_rate": 5.365296803652968e-07,
      "loss": 0.008,
      "reward": 1.7416791319847107,
      "reward_std": 0.19379056990146637,
      "rewards/accuracy_reward": 0.7573040723800659,
      "rewards/format_reward": 0.984375,
      "step": 1015
    },
    {
      "completion_length": 64.6171875,
      "epoch": 4.639269406392694,
      "grad_norm": 2.7476561069488525,
      "kl": 0.1728515625,
      "learning_rate": 5.360730593607305e-07,
      "loss": 0.0069,
      "reward": 1.8008184432983398,
      "reward_std": 0.1534968689084053,
      "rewards/accuracy_reward": 0.8008184134960175,
      "rewards/format_reward": 1.0,
      "step": 1016
    },
    {
      "completion_length": 71.171875,
      "epoch": 4.6438356164383565,
      "grad_norm": 3.0133135318756104,
      "kl": 0.121826171875,
      "learning_rate": 5.356164383561644e-07,
      "loss": 0.0049,
      "reward": 1.820052146911621,
      "reward_std": 0.1514080874621868,
      "rewards/accuracy_reward": 0.8278645575046539,
      "rewards/format_reward": 0.9921875,
      "step": 1017
    },
    {
      "completion_length": 80.9375,
      "epoch": 4.648401826484018,
      "grad_norm": 55.512779235839844,
      "kl": 0.109375,
      "learning_rate": 5.351598173515981e-07,
      "loss": 0.0044,
      "reward": 1.7374999523162842,
      "reward_std": 0.1173202246427536,
      "rewards/accuracy_reward": 0.7374999821186066,
      "rewards/format_reward": 1.0,
      "step": 1018
    },
    {
      "completion_length": 73.75,
      "epoch": 4.65296803652968,
      "grad_norm": 3.2106876373291016,
      "kl": 0.150390625,
      "learning_rate": 5.347031963470319e-07,
      "loss": 0.006,
      "reward": 1.7260440587997437,
      "reward_std": 0.22125811874866486,
      "rewards/accuracy_reward": 0.7338564693927765,
      "rewards/format_reward": 0.9921875,
      "step": 1019
    },
    {
      "completion_length": 53.90625,
      "epoch": 4.657534246575342,
      "grad_norm": 3.0299935340881348,
      "kl": 0.133056640625,
      "learning_rate": 5.342465753424658e-07,
      "loss": 0.0053,
      "reward": 1.7598958611488342,
      "reward_std": 0.14248863141983747,
      "rewards/accuracy_reward": 0.7598957717418671,
      "rewards/format_reward": 1.0,
      "step": 1020
    },
    {
      "completion_length": 44.4296875,
      "epoch": 4.662100456621005,
      "grad_norm": 17.613971710205078,
      "kl": 0.751953125,
      "learning_rate": 5.337899543378995e-07,
      "loss": 0.03,
      "reward": 1.843323826789856,
      "reward_std": 0.22472677379846573,
      "rewards/accuracy_reward": 0.866761326789856,
      "rewards/format_reward": 0.9765625,
      "step": 1021
    },
    {
      "completion_length": 68.1640625,
      "epoch": 4.666666666666667,
      "grad_norm": 3.197855234146118,
      "kl": 0.171875,
      "learning_rate": 5.333333333333333e-07,
      "loss": 0.0069,
      "reward": 1.805405616760254,
      "reward_std": 0.15048640966415405,
      "rewards/accuracy_reward": 0.8054055869579315,
      "rewards/format_reward": 1.0,
      "step": 1022
    },
    {
      "completion_length": 66.9921875,
      "epoch": 4.671232876712329,
      "grad_norm": 3.1693930625915527,
      "kl": 0.126708984375,
      "learning_rate": 5.328767123287671e-07,
      "loss": 0.0051,
      "reward": 1.6846354603767395,
      "reward_std": 0.12612807750701904,
      "rewards/accuracy_reward": 0.6846353709697723,
      "rewards/format_reward": 1.0,
      "step": 1023
    },
    {
      "completion_length": 81.1484375,
      "epoch": 4.675799086757991,
      "grad_norm": 2.464825391769409,
      "kl": 0.106689453125,
      "learning_rate": 5.324200913242008e-07,
      "loss": 0.0043,
      "reward": 1.8835937976837158,
      "reward_std": 0.08758953772485256,
      "rewards/accuracy_reward": 0.8835937082767487,
      "rewards/format_reward": 1.0,
      "step": 1024
    },
    {
      "completion_length": 64.78125,
      "epoch": 4.680365296803653,
      "grad_norm": 3.6387760639190674,
      "kl": 0.17333984375,
      "learning_rate": 5.319634703196348e-07,
      "loss": 0.0069,
      "reward": 1.7731905579566956,
      "reward_std": 0.16570382565259933,
      "rewards/accuracy_reward": 0.7731905579566956,
      "rewards/format_reward": 1.0,
      "step": 1025
    },
    {
      "completion_length": 60.9296875,
      "epoch": 4.684931506849315,
      "grad_norm": 20.164772033691406,
      "kl": 0.13232421875,
      "learning_rate": 5.315068493150685e-07,
      "loss": 0.0053,
      "reward": 1.689843773841858,
      "reward_std": 0.17715102434158325,
      "rewards/accuracy_reward": 0.6976562738418579,
      "rewards/format_reward": 0.9921875,
      "step": 1026
    },
    {
      "completion_length": 75.0,
      "epoch": 4.689497716894977,
      "grad_norm": 2.0931997299194336,
      "kl": 0.1484375,
      "learning_rate": 5.310502283105022e-07,
      "loss": 0.0059,
      "reward": 1.630468726158142,
      "reward_std": 0.16156607866287231,
      "rewards/accuracy_reward": 0.6460937261581421,
      "rewards/format_reward": 0.984375,
      "step": 1027
    },
    {
      "completion_length": 49.234375,
      "epoch": 4.69406392694064,
      "grad_norm": 5.067066192626953,
      "kl": 0.1962890625,
      "learning_rate": 5.305936073059361e-07,
      "loss": 0.0078,
      "reward": 1.8552082777023315,
      "reward_std": 0.15139273926615715,
      "rewards/accuracy_reward": 0.8552083373069763,
      "rewards/format_reward": 1.0,
      "step": 1028
    },
    {
      "completion_length": 80.9453125,
      "epoch": 4.698630136986301,
      "grad_norm": 4.103106498718262,
      "kl": 0.126708984375,
      "learning_rate": 5.301369863013698e-07,
      "loss": 0.0051,
      "reward": 1.678125023841858,
      "reward_std": 0.17818758636713028,
      "rewards/accuracy_reward": 0.6859374642372131,
      "rewards/format_reward": 0.9921875,
      "step": 1029
    },
    {
      "completion_length": 75.2734375,
      "epoch": 4.703196347031963,
      "grad_norm": 4.427400588989258,
      "kl": 0.25732421875,
      "learning_rate": 5.296803652968036e-07,
      "loss": 0.0103,
      "reward": 1.696877896785736,
      "reward_std": 0.16482967138290405,
      "rewards/accuracy_reward": 0.7046903669834137,
      "rewards/format_reward": 0.9921875,
      "step": 1030
    },
    {
      "completion_length": 56.8515625,
      "epoch": 4.707762557077626,
      "grad_norm": 5.5338921546936035,
      "kl": 0.22607421875,
      "learning_rate": 5.292237442922374e-07,
      "loss": 0.0091,
      "reward": 1.6015625,
      "reward_std": 0.25800998508930206,
      "rewards/accuracy_reward": 0.6249999403953552,
      "rewards/format_reward": 0.9765625,
      "step": 1031
    },
    {
      "completion_length": 68.96875,
      "epoch": 4.712328767123288,
      "grad_norm": 2.648360013961792,
      "kl": 0.138427734375,
      "learning_rate": 5.287671232876712e-07,
      "loss": 0.0055,
      "reward": 1.7213541865348816,
      "reward_std": 0.1301564909517765,
      "rewards/accuracy_reward": 0.7291666567325592,
      "rewards/format_reward": 0.9921875,
      "step": 1032
    },
    {
      "completion_length": 76.53125,
      "epoch": 4.71689497716895,
      "grad_norm": 3.0798656940460205,
      "kl": 0.126953125,
      "learning_rate": 5.283105022831051e-07,
      "loss": 0.0051,
      "reward": 1.7125211358070374,
      "reward_std": 0.09494294971227646,
      "rewards/accuracy_reward": 0.7125210464000702,
      "rewards/format_reward": 1.0,
      "step": 1033
    },
    {
      "completion_length": 68.859375,
      "epoch": 4.7214611872146115,
      "grad_norm": 5.309957981109619,
      "kl": 0.154296875,
      "learning_rate": 5.278538812785388e-07,
      "loss": 0.0062,
      "reward": 1.609375,
      "reward_std": 0.20264848321676254,
      "rewards/accuracy_reward": 0.6093749701976776,
      "rewards/format_reward": 1.0,
      "step": 1034
    },
    {
      "completion_length": 77.921875,
      "epoch": 4.726027397260274,
      "grad_norm": 4.624930381774902,
      "kl": 0.13427734375,
      "learning_rate": 5.273972602739725e-07,
      "loss": 0.0054,
      "reward": 1.7170308232307434,
      "reward_std": 0.19927022606134415,
      "rewards/accuracy_reward": 0.7248433232307434,
      "rewards/format_reward": 0.9921875,
      "step": 1035
    },
    {
      "completion_length": 69.5234375,
      "epoch": 4.730593607305936,
      "grad_norm": 3.3374686241149902,
      "kl": 0.116943359375,
      "learning_rate": 5.269406392694064e-07,
      "loss": 0.0047,
      "reward": 1.6639309525489807,
      "reward_std": 0.1389038860797882,
      "rewards/accuracy_reward": 0.6717434823513031,
      "rewards/format_reward": 0.9921875,
      "step": 1036
    },
    {
      "completion_length": 65.140625,
      "epoch": 4.735159817351598,
      "grad_norm": 3.858757495880127,
      "kl": 0.1416015625,
      "learning_rate": 5.264840182648401e-07,
      "loss": 0.0057,
      "reward": 1.6567708253860474,
      "reward_std": 0.16906771063804626,
      "rewards/accuracy_reward": 0.6723958551883698,
      "rewards/format_reward": 0.984375,
      "step": 1037
    },
    {
      "completion_length": 70.0859375,
      "epoch": 4.739726027397261,
      "grad_norm": 3.435810089111328,
      "kl": 0.156005859375,
      "learning_rate": 5.260273972602739e-07,
      "loss": 0.0063,
      "reward": 1.6902902126312256,
      "reward_std": 0.19441108405590057,
      "rewards/accuracy_reward": 0.698102593421936,
      "rewards/format_reward": 0.9921875,
      "step": 1038
    },
    {
      "completion_length": 60.0859375,
      "epoch": 4.744292237442922,
      "grad_norm": 4.343775272369385,
      "kl": 0.20263671875,
      "learning_rate": 5.255707762557078e-07,
      "loss": 0.0081,
      "reward": 1.8015338778495789,
      "reward_std": 0.20425771176815033,
      "rewards/accuracy_reward": 0.8015338182449341,
      "rewards/format_reward": 1.0,
      "step": 1039
    },
    {
      "completion_length": 67.5,
      "epoch": 4.748858447488584,
      "grad_norm": 2.354308605194092,
      "kl": 0.1474609375,
      "learning_rate": 5.251141552511415e-07,
      "loss": 0.0059,
      "reward": 1.8274368047714233,
      "reward_std": 0.13800114393234253,
      "rewards/accuracy_reward": 0.8352491855621338,
      "rewards/format_reward": 0.9921875,
      "step": 1040
    },
    {
      "completion_length": 65.109375,
      "epoch": 4.7534246575342465,
      "grad_norm": 4.133329391479492,
      "kl": 0.119140625,
      "learning_rate": 5.246575342465754e-07,
      "loss": 0.0048,
      "reward": 1.7390583753585815,
      "reward_std": 0.1775147169828415,
      "rewards/accuracy_reward": 0.7390583753585815,
      "rewards/format_reward": 1.0,
      "step": 1041
    },
    {
      "completion_length": 96.8125,
      "epoch": 4.757990867579909,
      "grad_norm": 2.921767473220825,
      "kl": 0.090576171875,
      "learning_rate": 5.242009132420091e-07,
      "loss": 0.0036,
      "reward": 1.8279520869255066,
      "reward_std": 0.11337075009942055,
      "rewards/accuracy_reward": 0.843576967716217,
      "rewards/format_reward": 0.984375,
      "step": 1042
    },
    {
      "completion_length": 76.25,
      "epoch": 4.762557077625571,
      "grad_norm": 1.6646754741668701,
      "kl": 0.0908203125,
      "learning_rate": 5.237442922374428e-07,
      "loss": 0.0036,
      "reward": 1.761111080646515,
      "reward_std": 0.08823190443217754,
      "rewards/accuracy_reward": 0.7611111104488373,
      "rewards/format_reward": 1.0,
      "step": 1043
    },
    {
      "completion_length": 68.90625,
      "epoch": 4.767123287671232,
      "grad_norm": 6.794203758239746,
      "kl": 0.12890625,
      "learning_rate": 5.232876712328767e-07,
      "loss": 0.0052,
      "reward": 1.6395359635353088,
      "reward_std": 0.1700442135334015,
      "rewards/accuracy_reward": 0.6473484635353088,
      "rewards/format_reward": 0.9921875,
      "step": 1044
    },
    {
      "completion_length": 64.5625,
      "epoch": 4.771689497716895,
      "grad_norm": 1.9526876211166382,
      "kl": 0.160888671875,
      "learning_rate": 5.228310502283105e-07,
      "loss": 0.0064,
      "reward": 1.818750023841858,
      "reward_std": 0.19044626876711845,
      "rewards/accuracy_reward": 0.8265624642372131,
      "rewards/format_reward": 0.9921875,
      "step": 1045
    },
    {
      "completion_length": 71.8359375,
      "epoch": 4.776255707762557,
      "grad_norm": 5.581969261169434,
      "kl": 0.121337890625,
      "learning_rate": 5.223744292237443e-07,
      "loss": 0.0049,
      "reward": 1.7421875596046448,
      "reward_std": 0.1561211347579956,
      "rewards/accuracy_reward": 0.7421874701976776,
      "rewards/format_reward": 1.0,
      "step": 1046
    },
    {
      "completion_length": 65.2109375,
      "epoch": 4.780821917808219,
      "grad_norm": 3.522632360458374,
      "kl": 0.139892578125,
      "learning_rate": 5.219178082191781e-07,
      "loss": 0.0056,
      "reward": 1.8170573115348816,
      "reward_std": 0.1570434384047985,
      "rewards/accuracy_reward": 0.840494692325592,
      "rewards/format_reward": 0.9765625,
      "step": 1047
    },
    {
      "completion_length": 82.296875,
      "epoch": 4.7853881278538815,
      "grad_norm": 8.847128868103027,
      "kl": 0.110107421875,
      "learning_rate": 5.214611872146118e-07,
      "loss": 0.0044,
      "reward": 1.8859375715255737,
      "reward_std": 0.0849014800041914,
      "rewards/accuracy_reward": 0.8859374523162842,
      "rewards/format_reward": 1.0,
      "step": 1048
    },
    {
      "completion_length": 73.2578125,
      "epoch": 4.789954337899544,
      "grad_norm": 9.131427764892578,
      "kl": 0.13525390625,
      "learning_rate": 5.210045662100457e-07,
      "loss": 0.0054,
      "reward": 1.6504226922988892,
      "reward_std": 0.18509591370821,
      "rewards/accuracy_reward": 0.6582351326942444,
      "rewards/format_reward": 0.9921875,
      "step": 1049
    },
    {
      "completion_length": 79.2265625,
      "epoch": 4.794520547945205,
      "grad_norm": 3.1302144527435303,
      "kl": 0.123779296875,
      "learning_rate": 5.205479452054794e-07,
      "loss": 0.005,
      "reward": 1.6324219703674316,
      "reward_std": 0.22072409093379974,
      "rewards/accuracy_reward": 0.6480468809604645,
      "rewards/format_reward": 0.984375,
      "step": 1050
    },
    {
      "completion_length": 68.59375,
      "epoch": 4.799086757990867,
      "grad_norm": 1.5062321424484253,
      "kl": 0.13525390625,
      "learning_rate": 5.200913242009131e-07,
      "loss": 0.0054,
      "reward": 1.7230769991874695,
      "reward_std": 0.10901044122874737,
      "rewards/accuracy_reward": 0.7387019097805023,
      "rewards/format_reward": 0.984375,
      "step": 1051
    },
    {
      "completion_length": 57.203125,
      "epoch": 4.80365296803653,
      "grad_norm": 2.9574577808380127,
      "kl": 0.18505859375,
      "learning_rate": 5.196347031963471e-07,
      "loss": 0.0074,
      "reward": 1.6598585844039917,
      "reward_std": 0.14134880900382996,
      "rewards/accuracy_reward": 0.6676710844039917,
      "rewards/format_reward": 0.9921875,
      "step": 1052
    },
    {
      "completion_length": 76.7734375,
      "epoch": 4.808219178082192,
      "grad_norm": 2.4928853511810303,
      "kl": 0.1630859375,
      "learning_rate": 5.191780821917808e-07,
      "loss": 0.0065,
      "reward": 1.7294270992279053,
      "reward_std": 0.2240969017148018,
      "rewards/accuracy_reward": 0.7528644800186157,
      "rewards/format_reward": 0.9765625,
      "step": 1053
    },
    {
      "completion_length": 84.828125,
      "epoch": 4.812785388127854,
      "grad_norm": 4.220333576202393,
      "kl": 0.114013671875,
      "learning_rate": 5.187214611872146e-07,
      "loss": 0.0046,
      "reward": 1.5883237719535828,
      "reward_std": 0.18970628082752228,
      "rewards/accuracy_reward": 0.5961362421512604,
      "rewards/format_reward": 0.9921875,
      "step": 1054
    },
    {
      "completion_length": 80.0546875,
      "epoch": 4.817351598173516,
      "grad_norm": 2.9225761890411377,
      "kl": 0.084716796875,
      "learning_rate": 5.182648401826484e-07,
      "loss": 0.0034,
      "reward": 1.8234375715255737,
      "reward_std": 0.1099486481398344,
      "rewards/accuracy_reward": 0.8234374225139618,
      "rewards/format_reward": 1.0,
      "step": 1055
    },
    {
      "completion_length": 68.84375,
      "epoch": 4.821917808219178,
      "grad_norm": 15.956554412841797,
      "kl": 0.126953125,
      "learning_rate": 5.178082191780821e-07,
      "loss": 0.0051,
      "reward": 1.707698404788971,
      "reward_std": 0.12136031687259674,
      "rewards/accuracy_reward": 0.7155108153820038,
      "rewards/format_reward": 0.9921875,
      "step": 1056
    },
    {
      "completion_length": 61.59375,
      "epoch": 4.82648401826484,
      "grad_norm": 3.7795159816741943,
      "kl": 0.19775390625,
      "learning_rate": 5.17351598173516e-07,
      "loss": 0.0079,
      "reward": 1.7173488140106201,
      "reward_std": 0.15133387595415115,
      "rewards/accuracy_reward": 0.7251611948013306,
      "rewards/format_reward": 0.9921875,
      "step": 1057
    },
    {
      "completion_length": 85.015625,
      "epoch": 4.831050228310502,
      "grad_norm": 4.373653888702393,
      "kl": 0.104736328125,
      "learning_rate": 5.168949771689497e-07,
      "loss": 0.0042,
      "reward": 1.8429688215255737,
      "reward_std": 0.08631845097988844,
      "rewards/accuracy_reward": 0.8507812023162842,
      "rewards/format_reward": 0.9921875,
      "step": 1058
    },
    {
      "completion_length": 80.9296875,
      "epoch": 4.835616438356165,
      "grad_norm": 2.2139017581939697,
      "kl": 0.091552734375,
      "learning_rate": 5.164383561643836e-07,
      "loss": 0.0037,
      "reward": 1.7554687857627869,
      "reward_std": 0.12153397500514984,
      "rewards/accuracy_reward": 0.7632811963558197,
      "rewards/format_reward": 0.9921875,
      "step": 1059
    },
    {
      "completion_length": 69.9296875,
      "epoch": 4.840182648401827,
      "grad_norm": 2.391106128692627,
      "kl": 0.113525390625,
      "learning_rate": 5.159817351598174e-07,
      "loss": 0.0045,
      "reward": 1.6293052434921265,
      "reward_std": 0.18556885421276093,
      "rewards/accuracy_reward": 0.6371176838874817,
      "rewards/format_reward": 0.9921875,
      "step": 1060
    },
    {
      "completion_length": 75.5078125,
      "epoch": 4.844748858447488,
      "grad_norm": 3.7311413288116455,
      "kl": 0.1337890625,
      "learning_rate": 5.155251141552511e-07,
      "loss": 0.0053,
      "reward": 1.8738667964935303,
      "reward_std": 0.09744009375572205,
      "rewards/accuracy_reward": 0.8738666772842407,
      "rewards/format_reward": 1.0,
      "step": 1061
    },
    {
      "completion_length": 88.328125,
      "epoch": 4.8493150684931505,
      "grad_norm": 2.399247884750366,
      "kl": 0.146484375,
      "learning_rate": 5.150684931506849e-07,
      "loss": 0.0058,
      "reward": 1.706250011920929,
      "reward_std": 0.21963772177696228,
      "rewards/accuracy_reward": 0.7140624821186066,
      "rewards/format_reward": 0.9921875,
      "step": 1062
    },
    {
      "completion_length": 73.6171875,
      "epoch": 4.853881278538813,
      "grad_norm": 2.581521511077881,
      "kl": 0.1435546875,
      "learning_rate": 5.146118721461187e-07,
      "loss": 0.0058,
      "reward": 1.7601500153541565,
      "reward_std": 0.14986789226531982,
      "rewards/accuracy_reward": 0.7679624557495117,
      "rewards/format_reward": 0.9921875,
      "step": 1063
    },
    {
      "completion_length": 72.1484375,
      "epoch": 4.858447488584475,
      "grad_norm": 8.218365669250488,
      "kl": 0.1533203125,
      "learning_rate": 5.141552511415524e-07,
      "loss": 0.0061,
      "reward": 1.7895833253860474,
      "reward_std": 0.18269683420658112,
      "rewards/accuracy_reward": 0.8052083253860474,
      "rewards/format_reward": 0.984375,
      "step": 1064
    },
    {
      "completion_length": 75.6875,
      "epoch": 4.863013698630137,
      "grad_norm": 6.358737468719482,
      "kl": 0.130126953125,
      "learning_rate": 5.136986301369864e-07,
      "loss": 0.0052,
      "reward": 1.8111705780029297,
      "reward_std": 0.11504914239048958,
      "rewards/accuracy_reward": 0.8111704587936401,
      "rewards/format_reward": 1.0,
      "step": 1065
    },
    {
      "completion_length": 72.0859375,
      "epoch": 4.867579908675799,
      "grad_norm": 1.565542221069336,
      "kl": 0.123779296875,
      "learning_rate": 5.132420091324201e-07,
      "loss": 0.005,
      "reward": 1.7661458849906921,
      "reward_std": 0.13723178207874298,
      "rewards/accuracy_reward": 0.773958295583725,
      "rewards/format_reward": 0.9921875,
      "step": 1066
    },
    {
      "completion_length": 81.34375,
      "epoch": 4.872146118721461,
      "grad_norm": 1.8809258937835693,
      "kl": 0.123291015625,
      "learning_rate": 5.127853881278538e-07,
      "loss": 0.0049,
      "reward": 1.8238808512687683,
      "reward_std": 0.09778433851897717,
      "rewards/accuracy_reward": 0.8238807916641235,
      "rewards/format_reward": 1.0,
      "step": 1067
    },
    {
      "completion_length": 80.703125,
      "epoch": 4.876712328767123,
      "grad_norm": 1.7312757968902588,
      "kl": 0.0810546875,
      "learning_rate": 5.123287671232877e-07,
      "loss": 0.0032,
      "reward": 1.624392330646515,
      "reward_std": 0.09445247054100037,
      "rewards/accuracy_reward": 0.6243923306465149,
      "rewards/format_reward": 1.0,
      "step": 1068
    },
    {
      "completion_length": 73.265625,
      "epoch": 4.8812785388127855,
      "grad_norm": 2.2705366611480713,
      "kl": 0.134033203125,
      "learning_rate": 5.118721461187214e-07,
      "loss": 0.0054,
      "reward": 1.6665404438972473,
      "reward_std": 0.17886501550674438,
      "rewards/accuracy_reward": 0.6821653842926025,
      "rewards/format_reward": 0.984375,
      "step": 1069
    },
    {
      "completion_length": 74.671875,
      "epoch": 4.885844748858448,
      "grad_norm": 2.8938543796539307,
      "kl": 0.12646484375,
      "learning_rate": 5.114155251141552e-07,
      "loss": 0.0051,
      "reward": 1.8554381728172302,
      "reward_std": 0.18266908079385757,
      "rewards/accuracy_reward": 0.8866880834102631,
      "rewards/format_reward": 0.96875,
      "step": 1070
    },
    {
      "completion_length": 84.8203125,
      "epoch": 4.890410958904109,
      "grad_norm": 6.559085845947266,
      "kl": 0.0888671875,
      "learning_rate": 5.10958904109589e-07,
      "loss": 0.0036,
      "reward": 1.6242188215255737,
      "reward_std": 0.14865797758102417,
      "rewards/accuracy_reward": 0.6320312321186066,
      "rewards/format_reward": 0.9921875,
      "step": 1071
    },
    {
      "completion_length": 72.203125,
      "epoch": 4.894977168949771,
      "grad_norm": 2.1274454593658447,
      "kl": 0.126708984375,
      "learning_rate": 5.105022831050228e-07,
      "loss": 0.0051,
      "reward": 1.7550346851348877,
      "reward_std": 0.15320852398872375,
      "rewards/accuracy_reward": 0.7628472149372101,
      "rewards/format_reward": 0.9921875,
      "step": 1072
    },
    {
      "completion_length": 73.6484375,
      "epoch": 4.899543378995434,
      "grad_norm": 2.215421676635742,
      "kl": 0.134765625,
      "learning_rate": 5.100456621004567e-07,
      "loss": 0.0054,
      "reward": 1.7328497171401978,
      "reward_std": 0.17783771082758904,
      "rewards/accuracy_reward": 0.756287157535553,
      "rewards/format_reward": 0.9765625,
      "step": 1073
    },
    {
      "completion_length": 82.7109375,
      "epoch": 4.904109589041096,
      "grad_norm": 7.915944576263428,
      "kl": 0.1328125,
      "learning_rate": 5.095890410958904e-07,
      "loss": 0.0053,
      "reward": 1.7984375953674316,
      "reward_std": 0.14389308542013168,
      "rewards/accuracy_reward": 0.8062499761581421,
      "rewards/format_reward": 0.9921875,
      "step": 1074
    },
    {
      "completion_length": 56.046875,
      "epoch": 4.908675799086758,
      "grad_norm": 6.48905086517334,
      "kl": 0.142822265625,
      "learning_rate": 5.091324200913241e-07,
      "loss": 0.0057,
      "reward": 1.7908854484558105,
      "reward_std": 0.16380437463521957,
      "rewards/accuracy_reward": 0.7986978888511658,
      "rewards/format_reward": 0.9921875,
      "step": 1075
    },
    {
      "completion_length": 76.328125,
      "epoch": 4.91324200913242,
      "grad_norm": 15.81816577911377,
      "kl": 0.103759765625,
      "learning_rate": 5.08675799086758e-07,
      "loss": 0.0042,
      "reward": 1.774218738079071,
      "reward_std": 0.14345712214708328,
      "rewards/accuracy_reward": 0.782031238079071,
      "rewards/format_reward": 0.9921875,
      "step": 1076
    },
    {
      "completion_length": 65.71875,
      "epoch": 4.917808219178082,
      "grad_norm": 2.252168655395508,
      "kl": 0.138916015625,
      "learning_rate": 5.082191780821917e-07,
      "loss": 0.0056,
      "reward": 1.7870659828186035,
      "reward_std": 0.14891928434371948,
      "rewards/accuracy_reward": 0.7870660126209259,
      "rewards/format_reward": 1.0,
      "step": 1077
    },
    {
      "completion_length": 84.4921875,
      "epoch": 4.922374429223744,
      "grad_norm": 6.789572715759277,
      "kl": 0.1016845703125,
      "learning_rate": 5.077625570776255e-07,
      "loss": 0.0041,
      "reward": 1.780468761920929,
      "reward_std": 0.12099719420075417,
      "rewards/accuracy_reward": 0.7882812023162842,
      "rewards/format_reward": 0.9921875,
      "step": 1078
    },
    {
      "completion_length": 93.4765625,
      "epoch": 4.926940639269406,
      "grad_norm": 1.4697335958480835,
      "kl": 0.0927734375,
      "learning_rate": 5.073059360730594e-07,
      "loss": 0.0037,
      "reward": 1.7980054020881653,
      "reward_std": 0.06987884640693665,
      "rewards/accuracy_reward": 0.7980053424835205,
      "rewards/format_reward": 1.0,
      "step": 1079
    },
    {
      "completion_length": 67.8125,
      "epoch": 4.931506849315069,
      "grad_norm": 1.78167724609375,
      "kl": 0.12353515625,
      "learning_rate": 5.068493150684931e-07,
      "loss": 0.005,
      "reward": 1.8179687857627869,
      "reward_std": 0.09306978806853294,
      "rewards/accuracy_reward": 0.8257812857627869,
      "rewards/format_reward": 0.9921875,
      "step": 1080
    },
    {
      "completion_length": 71.4453125,
      "epoch": 4.936073059360731,
      "grad_norm": 2.6702702045440674,
      "kl": 0.146484375,
      "learning_rate": 5.06392694063927e-07,
      "loss": 0.0059,
      "reward": 1.8533853888511658,
      "reward_std": 0.27950893342494965,
      "rewards/accuracy_reward": 0.884635329246521,
      "rewards/format_reward": 0.96875,
      "step": 1081
    },
    {
      "completion_length": 98.71875,
      "epoch": 4.940639269406392,
      "grad_norm": 2.6240620613098145,
      "kl": 0.06982421875,
      "learning_rate": 5.059360730593607e-07,
      "loss": 0.0028,
      "reward": 1.7433823347091675,
      "reward_std": 0.1912398487329483,
      "rewards/accuracy_reward": 0.7746323049068451,
      "rewards/format_reward": 0.96875,
      "step": 1082
    },
    {
      "completion_length": 66.03125,
      "epoch": 4.945205479452055,
      "grad_norm": 5.760696887969971,
      "kl": 0.15234375,
      "learning_rate": 5.054794520547944e-07,
      "loss": 0.0061,
      "reward": 1.7298898100852966,
      "reward_std": 0.21766673773527145,
      "rewards/accuracy_reward": 0.7377023696899414,
      "rewards/format_reward": 0.9921875,
      "step": 1083
    },
    {
      "completion_length": 68.59375,
      "epoch": 4.949771689497717,
      "grad_norm": 5.8692216873168945,
      "kl": 0.16357421875,
      "learning_rate": 5.050228310502283e-07,
      "loss": 0.0065,
      "reward": 1.7146621942520142,
      "reward_std": 0.10843057557940483,
      "rewards/accuracy_reward": 0.7146620750427246,
      "rewards/format_reward": 1.0,
      "step": 1084
    },
    {
      "completion_length": 92.359375,
      "epoch": 4.954337899543379,
      "grad_norm": 3.167354106903076,
      "kl": 0.1181640625,
      "learning_rate": 5.045662100456621e-07,
      "loss": 0.0047,
      "reward": 1.6780134439468384,
      "reward_std": 0.2429228127002716,
      "rewards/accuracy_reward": 0.6936383247375488,
      "rewards/format_reward": 0.984375,
      "step": 1085
    },
    {
      "completion_length": 79.984375,
      "epoch": 4.958904109589041,
      "grad_norm": 2.3840832710266113,
      "kl": 0.113037109375,
      "learning_rate": 5.041095890410959e-07,
      "loss": 0.0045,
      "reward": 1.7828125357627869,
      "reward_std": 0.19722937047481537,
      "rewards/accuracy_reward": 0.8140624761581421,
      "rewards/format_reward": 0.96875,
      "step": 1086
    },
    {
      "completion_length": 60.84375,
      "epoch": 4.963470319634704,
      "grad_norm": 2.938784599304199,
      "kl": 0.140380859375,
      "learning_rate": 5.036529680365297e-07,
      "loss": 0.0056,
      "reward": 1.6975947618484497,
      "reward_std": 0.22429338097572327,
      "rewards/accuracy_reward": 0.7210322618484497,
      "rewards/format_reward": 0.9765625,
      "step": 1087
    },
    {
      "completion_length": 73.8125,
      "epoch": 4.968036529680365,
      "grad_norm": 3.95701265335083,
      "kl": 0.11474609375,
      "learning_rate": 5.031963470319634e-07,
      "loss": 0.0046,
      "reward": 1.6440104246139526,
      "reward_std": 0.1992247775197029,
      "rewards/accuracy_reward": 0.651822954416275,
      "rewards/format_reward": 0.9921875,
      "step": 1088
    },
    {
      "completion_length": 71.96875,
      "epoch": 4.972602739726027,
      "grad_norm": 7.810092926025391,
      "kl": 0.1279296875,
      "learning_rate": 5.027397260273973e-07,
      "loss": 0.0051,
      "reward": 1.7612414360046387,
      "reward_std": 0.17399008572101593,
      "rewards/accuracy_reward": 0.7690538763999939,
      "rewards/format_reward": 0.9921875,
      "step": 1089
    },
    {
      "completion_length": 95.4765625,
      "epoch": 4.9771689497716896,
      "grad_norm": 2.2332069873809814,
      "kl": 0.106201171875,
      "learning_rate": 5.02283105022831e-07,
      "loss": 0.0042,
      "reward": 1.6498697996139526,
      "reward_std": 0.23457543551921844,
      "rewards/accuracy_reward": 0.6811197698116302,
      "rewards/format_reward": 0.96875,
      "step": 1090
    },
    {
      "completion_length": 67.2421875,
      "epoch": 4.981735159817352,
      "grad_norm": 2.379009485244751,
      "kl": 0.138916015625,
      "learning_rate": 5.018264840182647e-07,
      "loss": 0.0056,
      "reward": 1.758593738079071,
      "reward_std": 0.15046585351228714,
      "rewards/accuracy_reward": 0.766406238079071,
      "rewards/format_reward": 0.9921875,
      "step": 1091
    },
    {
      "completion_length": 80.3515625,
      "epoch": 4.986301369863014,
      "grad_norm": 1.9919111728668213,
      "kl": 0.100830078125,
      "learning_rate": 5.013698630136987e-07,
      "loss": 0.004,
      "reward": 1.7437500357627869,
      "reward_std": 0.16439745388925076,
      "rewards/accuracy_reward": 0.7671874463558197,
      "rewards/format_reward": 0.9765625,
      "step": 1092
    },
    {
      "completion_length": 87.265625,
      "epoch": 4.9908675799086755,
      "grad_norm": 1.678598403930664,
      "kl": 0.108642578125,
      "learning_rate": 5.009132420091324e-07,
      "loss": 0.0043,
      "reward": 1.6603811383247375,
      "reward_std": 0.1863521747291088,
      "rewards/accuracy_reward": 0.6838186085224152,
      "rewards/format_reward": 0.9765625,
      "step": 1093
    },
    {
      "completion_length": 86.515625,
      "epoch": 4.995433789954338,
      "grad_norm": 1.237178921699524,
      "kl": 0.105224609375,
      "learning_rate": 5.004566210045662e-07,
      "loss": 0.0042,
      "reward": 1.8344122171401978,
      "reward_std": 0.09498458355665207,
      "rewards/accuracy_reward": 0.8422246873378754,
      "rewards/format_reward": 0.9921875,
      "step": 1094
    },
    {
      "completion_length": 49.625,
      "epoch": 5.0,
      "grad_norm": 3.034471273422241,
      "kl": 0.10205078125,
      "learning_rate": 5e-07,
      "loss": 0.0039,
      "reward": 1.1875,
      "reward_std": 0.36278264224529266,
      "rewards/accuracy_reward": 0.1875,
      "rewards/format_reward": 1.0,
      "step": 1095
    },
    {
      "completion_length": 73.484375,
      "epoch": 5.004566210045662,
      "grad_norm": 4.169987678527832,
      "kl": 0.15771484375,
      "learning_rate": 4.995433789954337e-07,
      "loss": 0.0063,
      "reward": 1.7950236201286316,
      "reward_std": 0.1954728439450264,
      "rewards/accuracy_reward": 0.818461149930954,
      "rewards/format_reward": 0.9765625,
      "step": 1096
    },
    {
      "completion_length": 53.1484375,
      "epoch": 5.0091324200913245,
      "grad_norm": 2.2528958320617676,
      "kl": 0.18115234375,
      "learning_rate": 4.990867579908676e-07,
      "loss": 0.0073,
      "reward": 1.6565104126930237,
      "reward_std": 0.2434339076280594,
      "rewards/accuracy_reward": 0.6799479424953461,
      "rewards/format_reward": 0.9765625,
      "step": 1097
    },
    {
      "completion_length": 77.0390625,
      "epoch": 5.013698630136986,
      "grad_norm": 2.312629461288452,
      "kl": 0.150634765625,
      "learning_rate": 4.986301369863014e-07,
      "loss": 0.006,
      "reward": 1.6930060386657715,
      "reward_std": 0.15761591121554375,
      "rewards/accuracy_reward": 0.7008183896541595,
      "rewards/format_reward": 0.9921875,
      "step": 1098
    },
    {
      "completion_length": 68.0625,
      "epoch": 5.018264840182648,
      "grad_norm": 4.447121620178223,
      "kl": 0.180908203125,
      "learning_rate": 4.981735159817351e-07,
      "loss": 0.0072,
      "reward": 1.689843773841858,
      "reward_std": 0.19052022695541382,
      "rewards/accuracy_reward": 0.7054686844348907,
      "rewards/format_reward": 0.984375,
      "step": 1099
    },
    {
      "completion_length": 63.734375,
      "epoch": 5.0228310502283104,
      "grad_norm": 5.222453594207764,
      "kl": 0.1005859375,
      "learning_rate": 4.977168949771689e-07,
      "loss": 0.004,
      "reward": 1.8026909828186035,
      "reward_std": 0.11854784563183784,
      "rewards/accuracy_reward": 0.8105034232139587,
      "rewards/format_reward": 0.9921875,
      "step": 1100
    },
    {
      "completion_length": 97.75,
      "epoch": 5.027397260273973,
      "grad_norm": 2.315014362335205,
      "kl": 0.0711669921875,
      "learning_rate": 4.972602739726027e-07,
      "loss": 0.0028,
      "reward": 1.7503806352615356,
      "reward_std": 0.11963648349046707,
      "rewards/accuracy_reward": 0.7503806054592133,
      "rewards/format_reward": 1.0,
      "step": 1101
    },
    {
      "completion_length": 69.828125,
      "epoch": 5.031963470319635,
      "grad_norm": 2.1413135528564453,
      "kl": 0.13818359375,
      "learning_rate": 4.968036529680365e-07,
      "loss": 0.0055,
      "reward": 1.8750000596046448,
      "reward_std": 0.10205792635679245,
      "rewards/accuracy_reward": 0.8749999105930328,
      "rewards/format_reward": 1.0,
      "step": 1102
    },
    {
      "completion_length": 95.1328125,
      "epoch": 5.036529680365296,
      "grad_norm": 1.4922230243682861,
      "kl": 0.1259765625,
      "learning_rate": 4.963470319634703e-07,
      "loss": 0.005,
      "reward": 1.7955846786499023,
      "reward_std": 0.07053190469741821,
      "rewards/accuracy_reward": 0.79558464884758,
      "rewards/format_reward": 1.0,
      "step": 1103
    },
    {
      "completion_length": 75.15625,
      "epoch": 5.041095890410959,
      "grad_norm": 4.347703456878662,
      "kl": 0.125244140625,
      "learning_rate": 4.958904109589041e-07,
      "loss": 0.005,
      "reward": 1.6125000715255737,
      "reward_std": 0.16528953611850739,
      "rewards/accuracy_reward": 0.612500011920929,
      "rewards/format_reward": 1.0,
      "step": 1104
    },
    {
      "completion_length": 83.9453125,
      "epoch": 5.045662100456621,
      "grad_norm": 1.9074816703796387,
      "kl": 0.109375,
      "learning_rate": 4.954337899543379e-07,
      "loss": 0.0044,
      "reward": 1.810937523841858,
      "reward_std": 0.16448542103171349,
      "rewards/accuracy_reward": 0.8265624642372131,
      "rewards/format_reward": 0.984375,
      "step": 1105
    },
    {
      "completion_length": 66.875,
      "epoch": 5.050228310502283,
      "grad_norm": 2.969620943069458,
      "kl": 0.12841796875,
      "learning_rate": 4.949771689497717e-07,
      "loss": 0.0051,
      "reward": 1.6987723112106323,
      "reward_std": 0.19702571630477905,
      "rewards/accuracy_reward": 0.6987722814083099,
      "rewards/format_reward": 1.0,
      "step": 1106
    },
    {
      "completion_length": 97.59375,
      "epoch": 5.054794520547945,
      "grad_norm": 2.6807308197021484,
      "kl": 0.086669921875,
      "learning_rate": 4.945205479452055e-07,
      "loss": 0.0035,
      "reward": 1.841210961341858,
      "reward_std": 0.1313929297029972,
      "rewards/accuracy_reward": 0.8490233421325684,
      "rewards/format_reward": 0.9921875,
      "step": 1107
    },
    {
      "completion_length": 67.1015625,
      "epoch": 5.059360730593608,
      "grad_norm": 1.8944909572601318,
      "kl": 0.135986328125,
      "learning_rate": 4.940639269406393e-07,
      "loss": 0.0054,
      "reward": 1.8903645873069763,
      "reward_std": 0.09690769761800766,
      "rewards/accuracy_reward": 0.8903645575046539,
      "rewards/format_reward": 1.0,
      "step": 1108
    },
    {
      "completion_length": 77.6015625,
      "epoch": 5.063926940639269,
      "grad_norm": 3.560889959335327,
      "kl": 0.17138671875,
      "learning_rate": 4.93607305936073e-07,
      "loss": 0.0069,
      "reward": 1.6163216829299927,
      "reward_std": 0.2479196935892105,
      "rewards/accuracy_reward": 0.6319466531276703,
      "rewards/format_reward": 0.984375,
      "step": 1109
    },
    {
      "completion_length": 64.453125,
      "epoch": 5.068493150684931,
      "grad_norm": 1.8171228170394897,
      "kl": 0.158203125,
      "learning_rate": 4.931506849315068e-07,
      "loss": 0.0063,
      "reward": 1.7542577981948853,
      "reward_std": 0.1399587020277977,
      "rewards/accuracy_reward": 0.7542578279972076,
      "rewards/format_reward": 1.0,
      "step": 1110
    },
    {
      "completion_length": 86.5078125,
      "epoch": 5.073059360730594,
      "grad_norm": 2.2326784133911133,
      "kl": 0.106689453125,
      "learning_rate": 4.926940639269407e-07,
      "loss": 0.0043,
      "reward": 1.7170758247375488,
      "reward_std": 0.146156445145607,
      "rewards/accuracy_reward": 0.7248883247375488,
      "rewards/format_reward": 0.9921875,
      "step": 1111
    },
    {
      "completion_length": 60.7109375,
      "epoch": 5.077625570776256,
      "grad_norm": 2.7463958263397217,
      "kl": 0.16796875,
      "learning_rate": 4.922374429223744e-07,
      "loss": 0.0067,
      "reward": 1.6381410360336304,
      "reward_std": 0.25498080253601074,
      "rewards/accuracy_reward": 0.6615785360336304,
      "rewards/format_reward": 0.9765625,
      "step": 1112
    },
    {
      "completion_length": 72.3046875,
      "epoch": 5.082191780821918,
      "grad_norm": 1.4025739431381226,
      "kl": 0.10888671875,
      "learning_rate": 4.917808219178081e-07,
      "loss": 0.0044,
      "reward": 1.7656250596046448,
      "reward_std": 0.07576144114136696,
      "rewards/accuracy_reward": 0.7734374403953552,
      "rewards/format_reward": 0.9921875,
      "step": 1113
    },
    {
      "completion_length": 53.21875,
      "epoch": 5.0867579908675795,
      "grad_norm": 3.9440958499908447,
      "kl": 0.142822265625,
      "learning_rate": 4.91324200913242e-07,
      "loss": 0.0057,
      "reward": 1.8531250357627869,
      "reward_std": 0.14943039789795876,
      "rewards/accuracy_reward": 0.8609374761581421,
      "rewards/format_reward": 0.9921875,
      "step": 1114
    },
    {
      "completion_length": 71.0859375,
      "epoch": 5.091324200913242,
      "grad_norm": 2.5945069789886475,
      "kl": 0.101318359375,
      "learning_rate": 4.908675799086758e-07,
      "loss": 0.0041,
      "reward": 1.7131696939468384,
      "reward_std": 0.09270836971700191,
      "rewards/accuracy_reward": 0.7131696045398712,
      "rewards/format_reward": 1.0,
      "step": 1115
    },
    {
      "completion_length": 65.1796875,
      "epoch": 5.095890410958904,
      "grad_norm": 1.469778060913086,
      "kl": 0.121337890625,
      "learning_rate": 4.904109589041096e-07,
      "loss": 0.0048,
      "reward": 1.7162946462631226,
      "reward_std": 0.12621085345745087,
      "rewards/accuracy_reward": 0.7241071164608002,
      "rewards/format_reward": 0.9921875,
      "step": 1116
    },
    {
      "completion_length": 80.140625,
      "epoch": 5.100456621004566,
      "grad_norm": 2.706517457962036,
      "kl": 0.099365234375,
      "learning_rate": 4.899543378995434e-07,
      "loss": 0.004,
      "reward": 1.8068211078643799,
      "reward_std": 0.12114018388092518,
      "rewards/accuracy_reward": 0.8224460184574127,
      "rewards/format_reward": 0.984375,
      "step": 1117
    },
    {
      "completion_length": 58.5078125,
      "epoch": 5.105022831050229,
      "grad_norm": 4.029208183288574,
      "kl": 0.17138671875,
      "learning_rate": 4.894977168949771e-07,
      "loss": 0.0069,
      "reward": 1.7942708134651184,
      "reward_std": 0.09960613120347261,
      "rewards/accuracy_reward": 0.7942708432674408,
      "rewards/format_reward": 1.0,
      "step": 1118
    },
    {
      "completion_length": 67.359375,
      "epoch": 5.109589041095891,
      "grad_norm": 2.30452823638916,
      "kl": 0.15185546875,
      "learning_rate": 4.89041095890411e-07,
      "loss": 0.0061,
      "reward": 1.7510417103767395,
      "reward_std": 0.18200254440307617,
      "rewards/accuracy_reward": 0.7666666209697723,
      "rewards/format_reward": 0.984375,
      "step": 1119
    },
    {
      "completion_length": 70.984375,
      "epoch": 5.114155251141552,
      "grad_norm": 1.813476324081421,
      "kl": 0.1337890625,
      "learning_rate": 4.885844748858447e-07,
      "loss": 0.0054,
      "reward": 1.7627604007720947,
      "reward_std": 0.09784993343055248,
      "rewards/accuracy_reward": 0.76276034116745,
      "rewards/format_reward": 1.0,
      "step": 1120
    },
    {
      "completion_length": 67.1171875,
      "epoch": 5.1187214611872145,
      "grad_norm": 2.7639548778533936,
      "kl": 0.158203125,
      "learning_rate": 4.881278538812786e-07,
      "loss": 0.0063,
      "reward": 1.6960286498069763,
      "reward_std": 0.1810276135802269,
      "rewards/accuracy_reward": 0.6960286498069763,
      "rewards/format_reward": 1.0,
      "step": 1121
    },
    {
      "completion_length": 64.8984375,
      "epoch": 5.123287671232877,
      "grad_norm": 1.0785552263259888,
      "kl": 0.116943359375,
      "learning_rate": 4.876712328767123e-07,
      "loss": 0.0047,
      "reward": 1.7366800904273987,
      "reward_std": 0.12058132514357567,
      "rewards/accuracy_reward": 0.7523050308227539,
      "rewards/format_reward": 0.984375,
      "step": 1122
    },
    {
      "completion_length": 70.203125,
      "epoch": 5.127853881278539,
      "grad_norm": 1.4450359344482422,
      "kl": 0.165771484375,
      "learning_rate": 4.872146118721461e-07,
      "loss": 0.0066,
      "reward": 1.8294271230697632,
      "reward_std": 0.1284763477742672,
      "rewards/accuracy_reward": 0.8450520634651184,
      "rewards/format_reward": 0.984375,
      "step": 1123
    },
    {
      "completion_length": 73.5390625,
      "epoch": 5.132420091324201,
      "grad_norm": 1.4733330011367798,
      "kl": 0.126220703125,
      "learning_rate": 4.867579908675799e-07,
      "loss": 0.0051,
      "reward": 1.8328125476837158,
      "reward_std": 0.09086007624864578,
      "rewards/accuracy_reward": 0.8328124284744263,
      "rewards/format_reward": 1.0,
      "step": 1124
    },
    {
      "completion_length": 62.890625,
      "epoch": 5.136986301369863,
      "grad_norm": 5.508318901062012,
      "kl": 0.16796875,
      "learning_rate": 4.863013698630137e-07,
      "loss": 0.0067,
      "reward": 1.7619792222976685,
      "reward_std": 0.18441661447286606,
      "rewards/accuracy_reward": 0.7619791626930237,
      "rewards/format_reward": 1.0,
      "step": 1125
    },
    {
      "completion_length": 68.4453125,
      "epoch": 5.141552511415525,
      "grad_norm": 2.6522018909454346,
      "kl": 0.16748046875,
      "learning_rate": 4.858447488584474e-07,
      "loss": 0.0067,
      "reward": 1.705004334449768,
      "reward_std": 0.19281967729330063,
      "rewards/accuracy_reward": 0.7206293344497681,
      "rewards/format_reward": 0.984375,
      "step": 1126
    },
    {
      "completion_length": 69.0,
      "epoch": 5.146118721461187,
      "grad_norm": 2.304168224334717,
      "kl": 0.143310546875,
      "learning_rate": 4.853881278538813e-07,
      "loss": 0.0057,
      "reward": 1.7598021030426025,
      "reward_std": 0.18482757359743118,
      "rewards/accuracy_reward": 0.7832395136356354,
      "rewards/format_reward": 0.9765625,
      "step": 1127
    },
    {
      "completion_length": 92.8203125,
      "epoch": 5.1506849315068495,
      "grad_norm": 2.8171322345733643,
      "kl": 0.082763671875,
      "learning_rate": 4.84931506849315e-07,
      "loss": 0.0033,
      "reward": 1.7391226291656494,
      "reward_std": 0.10928737744688988,
      "rewards/accuracy_reward": 0.7469350695610046,
      "rewards/format_reward": 0.9921875,
      "step": 1128
    },
    {
      "completion_length": 66.7578125,
      "epoch": 5.155251141552512,
      "grad_norm": 3.0008230209350586,
      "kl": 0.22119140625,
      "learning_rate": 4.844748858447489e-07,
      "loss": 0.0088,
      "reward": 1.7692708373069763,
      "reward_std": 0.20399662107229233,
      "rewards/accuracy_reward": 0.7927083671092987,
      "rewards/format_reward": 0.9765625,
      "step": 1129
    },
    {
      "completion_length": 71.7734375,
      "epoch": 5.159817351598173,
      "grad_norm": 3.2780561447143555,
      "kl": 0.15673828125,
      "learning_rate": 4.840182648401826e-07,
      "loss": 0.0063,
      "reward": 1.8493314981460571,
      "reward_std": 0.12171986699104309,
      "rewards/accuracy_reward": 0.8493313789367676,
      "rewards/format_reward": 1.0,
      "step": 1130
    },
    {
      "completion_length": 72.875,
      "epoch": 5.164383561643835,
      "grad_norm": 2.125765800476074,
      "kl": 0.1103515625,
      "learning_rate": 4.835616438356164e-07,
      "loss": 0.0044,
      "reward": 1.8101563453674316,
      "reward_std": 0.09943688660860062,
      "rewards/accuracy_reward": 0.8179686963558197,
      "rewards/format_reward": 0.9921875,
      "step": 1131
    },
    {
      "completion_length": 70.375,
      "epoch": 5.168949771689498,
      "grad_norm": 8.532958984375,
      "kl": 0.138671875,
      "learning_rate": 4.831050228310502e-07,
      "loss": 0.0055,
      "reward": 1.7585145235061646,
      "reward_std": 0.1326666846871376,
      "rewards/accuracy_reward": 0.7663269340991974,
      "rewards/format_reward": 0.9921875,
      "step": 1132
    },
    {
      "completion_length": 86.296875,
      "epoch": 5.17351598173516,
      "grad_norm": 1.4057005643844604,
      "kl": 0.105224609375,
      "learning_rate": 4.82648401826484e-07,
      "loss": 0.0042,
      "reward": 1.8115254044532776,
      "reward_std": 0.059396788477897644,
      "rewards/accuracy_reward": 0.8115253746509552,
      "rewards/format_reward": 1.0,
      "step": 1133
    },
    {
      "completion_length": 50.203125,
      "epoch": 5.178082191780822,
      "grad_norm": 4.812795162200928,
      "kl": 0.13623046875,
      "learning_rate": 4.821917808219178e-07,
      "loss": 0.0055,
      "reward": 1.5721354484558105,
      "reward_std": 0.19010350108146667,
      "rewards/accuracy_reward": 0.5721354186534882,
      "rewards/format_reward": 1.0,
      "step": 1134
    },
    {
      "completion_length": 62.796875,
      "epoch": 5.182648401826484,
      "grad_norm": 5.795816421508789,
      "kl": 0.14892578125,
      "learning_rate": 4.817351598173516e-07,
      "loss": 0.006,
      "reward": 1.839672565460205,
      "reward_std": 0.16954216361045837,
      "rewards/accuracy_reward": 0.8631100654602051,
      "rewards/format_reward": 0.9765625,
      "step": 1135
    },
    {
      "completion_length": 65.6015625,
      "epoch": 5.187214611872146,
      "grad_norm": 4.185856342315674,
      "kl": 0.14453125,
      "learning_rate": 4.812785388127853e-07,
      "loss": 0.0058,
      "reward": 1.727814257144928,
      "reward_std": 0.19888605177402496,
      "rewards/accuracy_reward": 0.7512516975402832,
      "rewards/format_reward": 0.9765625,
      "step": 1136
    },
    {
      "completion_length": 74.3046875,
      "epoch": 5.191780821917808,
      "grad_norm": 1.525724172592163,
      "kl": 0.113525390625,
      "learning_rate": 4.808219178082192e-07,
      "loss": 0.0045,
      "reward": 1.7773438096046448,
      "reward_std": 0.12639044970273972,
      "rewards/accuracy_reward": 0.77734375,
      "rewards/format_reward": 1.0,
      "step": 1137
    },
    {
      "completion_length": 54.1875,
      "epoch": 5.19634703196347,
      "grad_norm": 2.832690954208374,
      "kl": 0.21484375,
      "learning_rate": 4.80365296803653e-07,
      "loss": 0.0086,
      "reward": 1.666732907295227,
      "reward_std": 0.19027332961559296,
      "rewards/accuracy_reward": 0.6745454370975494,
      "rewards/format_reward": 0.9921875,
      "step": 1138
    },
    {
      "completion_length": 58.1953125,
      "epoch": 5.200913242009133,
      "grad_norm": 2.4609923362731934,
      "kl": 0.17724609375,
      "learning_rate": 4.799086757990867e-07,
      "loss": 0.0071,
      "reward": 1.7091332077980042,
      "reward_std": 0.24910558015108109,
      "rewards/accuracy_reward": 0.7247581779956818,
      "rewards/format_reward": 0.984375,
      "step": 1139
    },
    {
      "completion_length": 94.75,
      "epoch": 5.205479452054795,
      "grad_norm": 11.776135444641113,
      "kl": 0.2222900390625,
      "learning_rate": 4.794520547945205e-07,
      "loss": 0.0089,
      "reward": 1.803125023841858,
      "reward_std": 0.09865947626531124,
      "rewards/accuracy_reward": 0.8109373450279236,
      "rewards/format_reward": 0.9921875,
      "step": 1140
    },
    {
      "completion_length": 75.5546875,
      "epoch": 5.210045662100456,
      "grad_norm": 1.514672875404358,
      "kl": 0.13330078125,
      "learning_rate": 4.789954337899543e-07,
      "loss": 0.0053,
      "reward": 1.8759238719940186,
      "reward_std": 0.10365894064307213,
      "rewards/accuracy_reward": 0.8915487825870514,
      "rewards/format_reward": 0.984375,
      "step": 1141
    },
    {
      "completion_length": 65.953125,
      "epoch": 5.2146118721461185,
      "grad_norm": 2.5782454013824463,
      "kl": 0.110595703125,
      "learning_rate": 4.785388127853881e-07,
      "loss": 0.0044,
      "reward": 1.6188101172447205,
      "reward_std": 0.24680721014738083,
      "rewards/accuracy_reward": 0.6422475874423981,
      "rewards/format_reward": 0.9765625,
      "step": 1142
    },
    {
      "completion_length": 54.1875,
      "epoch": 5.219178082191781,
      "grad_norm": 6.618932723999023,
      "kl": 0.1689453125,
      "learning_rate": 4.780821917808219e-07,
      "loss": 0.0068,
      "reward": 1.6172246932983398,
      "reward_std": 0.23022934794425964,
      "rewards/accuracy_reward": 0.6250371932983398,
      "rewards/format_reward": 0.9921875,
      "step": 1143
    },
    {
      "completion_length": 75.8203125,
      "epoch": 5.223744292237443,
      "grad_norm": 3.0787389278411865,
      "kl": 0.10595703125,
      "learning_rate": 4.776255707762557e-07,
      "loss": 0.0042,
      "reward": 1.806249976158142,
      "reward_std": 0.10205793380737305,
      "rewards/accuracy_reward": 0.8062499165534973,
      "rewards/format_reward": 1.0,
      "step": 1144
    },
    {
      "completion_length": 65.4453125,
      "epoch": 5.228310502283105,
      "grad_norm": 2.9733009338378906,
      "kl": 0.1591796875,
      "learning_rate": 4.771689497716894e-07,
      "loss": 0.0064,
      "reward": 1.7895833849906921,
      "reward_std": 0.13740738481283188,
      "rewards/accuracy_reward": 0.7973958253860474,
      "rewards/format_reward": 0.9921875,
      "step": 1145
    },
    {
      "completion_length": 68.8828125,
      "epoch": 5.232876712328767,
      "grad_norm": 3.3826847076416016,
      "kl": 0.1767578125,
      "learning_rate": 4.7671232876712324e-07,
      "loss": 0.0071,
      "reward": 1.7227678894996643,
      "reward_std": 0.2944817692041397,
      "rewards/accuracy_reward": 0.7540178298950195,
      "rewards/format_reward": 0.96875,
      "step": 1146
    },
    {
      "completion_length": 104.4140625,
      "epoch": 5.237442922374429,
      "grad_norm": 2.469144105911255,
      "kl": 0.07373046875,
      "learning_rate": 4.762557077625571e-07,
      "loss": 0.0029,
      "reward": 1.896093726158142,
      "reward_std": 0.07282309047877789,
      "rewards/accuracy_reward": 0.9039061367511749,
      "rewards/format_reward": 0.9921875,
      "step": 1147
    },
    {
      "completion_length": 73.578125,
      "epoch": 5.242009132420091,
      "grad_norm": 2.969399929046631,
      "kl": 0.139892578125,
      "learning_rate": 4.7579908675799086e-07,
      "loss": 0.0056,
      "reward": 1.8597139716148376,
      "reward_std": 0.13000112399458885,
      "rewards/accuracy_reward": 0.8597138822078705,
      "rewards/format_reward": 1.0,
      "step": 1148
    },
    {
      "completion_length": 77.1953125,
      "epoch": 5.2465753424657535,
      "grad_norm": 1.6237159967422485,
      "kl": 0.115478515625,
      "learning_rate": 4.7534246575342465e-07,
      "loss": 0.0046,
      "reward": 1.7437500953674316,
      "reward_std": 0.1552036553621292,
      "rewards/accuracy_reward": 0.7593749761581421,
      "rewards/format_reward": 0.984375,
      "step": 1149
    },
    {
      "completion_length": 73.2578125,
      "epoch": 5.251141552511416,
      "grad_norm": 3.891481637954712,
      "kl": 0.108154296875,
      "learning_rate": 4.748858447488584e-07,
      "loss": 0.0043,
      "reward": 1.792336344718933,
      "reward_std": 0.23970390856266022,
      "rewards/accuracy_reward": 0.8157738149166107,
      "rewards/format_reward": 0.9765625,
      "step": 1150
    },
    {
      "completion_length": 92.109375,
      "epoch": 5.255707762557078,
      "grad_norm": 2.1832501888275146,
      "kl": 0.074951171875,
      "learning_rate": 4.744292237442922e-07,
      "loss": 0.003,
      "reward": 1.7013021111488342,
      "reward_std": 0.13120698183774948,
      "rewards/accuracy_reward": 0.7091145515441895,
      "rewards/format_reward": 0.9921875,
      "step": 1151
    },
    {
      "completion_length": 79.3359375,
      "epoch": 5.260273972602739,
      "grad_norm": 2.02133846282959,
      "kl": 0.15185546875,
      "learning_rate": 4.73972602739726e-07,
      "loss": 0.0061,
      "reward": 1.7348758578300476,
      "reward_std": 0.2100473716855049,
      "rewards/accuracy_reward": 0.7661258280277252,
      "rewards/format_reward": 0.96875,
      "step": 1152
    },
    {
      "completion_length": 69.28125,
      "epoch": 5.264840182648402,
      "grad_norm": 1.7159175872802734,
      "kl": 0.146484375,
      "learning_rate": 4.735159817351598e-07,
      "loss": 0.0058,
      "reward": 1.6989583373069763,
      "reward_std": 0.17543572187423706,
      "rewards/accuracy_reward": 0.7223958075046539,
      "rewards/format_reward": 0.9765625,
      "step": 1153
    },
    {
      "completion_length": 62.1328125,
      "epoch": 5.269406392694064,
      "grad_norm": 3.155850887298584,
      "kl": 0.1455078125,
      "learning_rate": 4.730593607305936e-07,
      "loss": 0.0058,
      "reward": 1.734188973903656,
      "reward_std": 0.2661764770746231,
      "rewards/accuracy_reward": 0.7576265037059784,
      "rewards/format_reward": 0.9765625,
      "step": 1154
    },
    {
      "completion_length": 64.859375,
      "epoch": 5.273972602739726,
      "grad_norm": 2.906141757965088,
      "kl": 0.18798828125,
      "learning_rate": 4.726027397260274e-07,
      "loss": 0.0075,
      "reward": 1.7283979058265686,
      "reward_std": 0.15935904532670975,
      "rewards/accuracy_reward": 0.7362103760242462,
      "rewards/format_reward": 0.9921875,
      "step": 1155
    },
    {
      "completion_length": 68.3984375,
      "epoch": 5.2785388127853885,
      "grad_norm": 3.6288256645202637,
      "kl": 0.1611328125,
      "learning_rate": 4.7214611872146116e-07,
      "loss": 0.0065,
      "reward": 1.6093750596046448,
      "reward_std": 0.2415616661310196,
      "rewards/accuracy_reward": 0.6328125298023224,
      "rewards/format_reward": 0.9765625,
      "step": 1156
    },
    {
      "completion_length": 65.546875,
      "epoch": 5.28310502283105,
      "grad_norm": 5.221689701080322,
      "kl": 0.21435546875,
      "learning_rate": 4.71689497716895e-07,
      "loss": 0.0086,
      "reward": 1.7025855779647827,
      "reward_std": 0.11607009917497635,
      "rewards/accuracy_reward": 0.7025855779647827,
      "rewards/format_reward": 1.0,
      "step": 1157
    },
    {
      "completion_length": 66.5546875,
      "epoch": 5.287671232876712,
      "grad_norm": 1.9122740030288696,
      "kl": 0.1474609375,
      "learning_rate": 4.7123287671232874e-07,
      "loss": 0.0059,
      "reward": 1.8352110981941223,
      "reward_std": 0.097220653668046,
      "rewards/accuracy_reward": 0.8430235981941223,
      "rewards/format_reward": 0.9921875,
      "step": 1158
    },
    {
      "completion_length": 70.34375,
      "epoch": 5.292237442922374,
      "grad_norm": 10.378006935119629,
      "kl": 0.1640625,
      "learning_rate": 4.707762557077625e-07,
      "loss": 0.0066,
      "reward": 1.8343608975410461,
      "reward_std": 0.10324066504836082,
      "rewards/accuracy_reward": 0.8343608975410461,
      "rewards/format_reward": 1.0,
      "step": 1159
    },
    {
      "completion_length": 67.8828125,
      "epoch": 5.296803652968037,
      "grad_norm": 6.8480119705200195,
      "kl": 0.14599609375,
      "learning_rate": 4.703196347031963e-07,
      "loss": 0.0058,
      "reward": 1.7857979536056519,
      "reward_std": 0.13422124087810516,
      "rewards/accuracy_reward": 0.7936104536056519,
      "rewards/format_reward": 0.9921875,
      "step": 1160
    },
    {
      "completion_length": 100.640625,
      "epoch": 5.301369863013699,
      "grad_norm": 3.1650795936584473,
      "kl": 0.0848388671875,
      "learning_rate": 4.6986301369863015e-07,
      "loss": 0.0034,
      "reward": 1.803321123123169,
      "reward_std": 0.17942239344120026,
      "rewards/accuracy_reward": 0.8345710933208466,
      "rewards/format_reward": 0.96875,
      "step": 1161
    },
    {
      "completion_length": 53.546875,
      "epoch": 5.30593607305936,
      "grad_norm": 4.188801288604736,
      "kl": 0.20947265625,
      "learning_rate": 4.694063926940639e-07,
      "loss": 0.0084,
      "reward": 1.7130786776542664,
      "reward_std": 0.21442808210849762,
      "rewards/accuracy_reward": 0.7287036776542664,
      "rewards/format_reward": 0.984375,
      "step": 1162
    },
    {
      "completion_length": 80.5625,
      "epoch": 5.310502283105023,
      "grad_norm": 2.222104787826538,
      "kl": 0.08837890625,
      "learning_rate": 4.689497716894977e-07,
      "loss": 0.0035,
      "reward": 1.7203125953674316,
      "reward_std": 0.16751586645841599,
      "rewards/accuracy_reward": 0.7437499463558197,
      "rewards/format_reward": 0.9765625,
      "step": 1163
    },
    {
      "completion_length": 54.359375,
      "epoch": 5.315068493150685,
      "grad_norm": 2.051166296005249,
      "kl": 0.15380859375,
      "learning_rate": 4.684931506849315e-07,
      "loss": 0.0062,
      "reward": 1.7621361017227173,
      "reward_std": 0.11842145770788193,
      "rewards/accuracy_reward": 0.7699486315250397,
      "rewards/format_reward": 0.9921875,
      "step": 1164
    },
    {
      "completion_length": 64.0546875,
      "epoch": 5.319634703196347,
      "grad_norm": 3.0098822116851807,
      "kl": 0.177734375,
      "learning_rate": 4.680365296803653e-07,
      "loss": 0.0071,
      "reward": 1.6876825094223022,
      "reward_std": 0.1372067779302597,
      "rewards/accuracy_reward": 0.6876825392246246,
      "rewards/format_reward": 1.0,
      "step": 1165
    },
    {
      "completion_length": 80.75,
      "epoch": 5.324200913242009,
      "grad_norm": 2.739140033721924,
      "kl": 0.0966796875,
      "learning_rate": 4.6757990867579904e-07,
      "loss": 0.0039,
      "reward": 1.740625023841858,
      "reward_std": 0.14044751226902008,
      "rewards/accuracy_reward": 0.7484374046325684,
      "rewards/format_reward": 0.9921875,
      "step": 1166
    },
    {
      "completion_length": 67.6875,
      "epoch": 5.328767123287671,
      "grad_norm": 2.510779857635498,
      "kl": 0.103271484375,
      "learning_rate": 4.671232876712329e-07,
      "loss": 0.0041,
      "reward": 1.607812523841858,
      "reward_std": 0.1767766922712326,
      "rewards/accuracy_reward": 0.6078124940395355,
      "rewards/format_reward": 1.0,
      "step": 1167
    },
    {
      "completion_length": 78.734375,
      "epoch": 5.333333333333333,
      "grad_norm": 2.7751314640045166,
      "kl": 0.126708984375,
      "learning_rate": 4.6666666666666666e-07,
      "loss": 0.0051,
      "reward": 1.853255271911621,
      "reward_std": 0.19070542603731155,
      "rewards/accuracy_reward": 0.8688801825046539,
      "rewards/format_reward": 0.984375,
      "step": 1168
    },
    {
      "completion_length": 67.75,
      "epoch": 5.337899543378995,
      "grad_norm": 1.8172931671142578,
      "kl": 0.146240234375,
      "learning_rate": 4.6621004566210045e-07,
      "loss": 0.0059,
      "reward": 1.7744792103767395,
      "reward_std": 0.09880542568862438,
      "rewards/accuracy_reward": 0.78229159116745,
      "rewards/format_reward": 0.9921875,
      "step": 1169
    },
    {
      "completion_length": 96.1640625,
      "epoch": 5.342465753424658,
      "grad_norm": 2.242799758911133,
      "kl": 0.11083984375,
      "learning_rate": 4.657534246575342e-07,
      "loss": 0.0044,
      "reward": 1.6858011484146118,
      "reward_std": 0.16090433299541473,
      "rewards/accuracy_reward": 0.7014259994029999,
      "rewards/format_reward": 0.984375,
      "step": 1170
    },
    {
      "completion_length": 94.90625,
      "epoch": 5.34703196347032,
      "grad_norm": 1.8737952709197998,
      "kl": 0.102783203125,
      "learning_rate": 4.65296803652968e-07,
      "loss": 0.0041,
      "reward": 1.8359375,
      "reward_std": 0.1631505787372589,
      "rewards/accuracy_reward": 0.8515624403953552,
      "rewards/format_reward": 0.984375,
      "step": 1171
    },
    {
      "completion_length": 108.46875,
      "epoch": 5.351598173515982,
      "grad_norm": 1.0832995176315308,
      "kl": 0.1112060546875,
      "learning_rate": 4.648401826484018e-07,
      "loss": 0.0044,
      "reward": 1.8958333730697632,
      "reward_std": 0.05308555904775858,
      "rewards/accuracy_reward": 0.8958332538604736,
      "rewards/format_reward": 1.0,
      "step": 1172
    },
    {
      "completion_length": 89.1875,
      "epoch": 5.3561643835616435,
      "grad_norm": 2.971842050552368,
      "kl": 0.1611328125,
      "learning_rate": 4.643835616438356e-07,
      "loss": 0.0065,
      "reward": 1.6929687857627869,
      "reward_std": 0.22123289108276367,
      "rewards/accuracy_reward": 0.7164062261581421,
      "rewards/format_reward": 0.9765625,
      "step": 1173
    },
    {
      "completion_length": 79.7109375,
      "epoch": 5.360730593607306,
      "grad_norm": 3.3845016956329346,
      "kl": 0.107177734375,
      "learning_rate": 4.639269406392694e-07,
      "loss": 0.0043,
      "reward": 1.7539063096046448,
      "reward_std": 0.1683424860239029,
      "rewards/accuracy_reward": 0.75390625,
      "rewards/format_reward": 1.0,
      "step": 1174
    },
    {
      "completion_length": 88.7265625,
      "epoch": 5.365296803652968,
      "grad_norm": 1.8644884824752808,
      "kl": 0.115478515625,
      "learning_rate": 4.634703196347032e-07,
      "loss": 0.0046,
      "reward": 1.635702133178711,
      "reward_std": 0.2014181986451149,
      "rewards/accuracy_reward": 0.6591395735740662,
      "rewards/format_reward": 0.9765625,
      "step": 1175
    },
    {
      "completion_length": 79.828125,
      "epoch": 5.36986301369863,
      "grad_norm": 11.902091026306152,
      "kl": 0.126953125,
      "learning_rate": 4.6301369863013696e-07,
      "loss": 0.0051,
      "reward": 1.7371652126312256,
      "reward_std": 0.1940011829137802,
      "rewards/accuracy_reward": 0.7527901232242584,
      "rewards/format_reward": 0.984375,
      "step": 1176
    },
    {
      "completion_length": 66.2734375,
      "epoch": 5.3744292237442925,
      "grad_norm": 2.139374017715454,
      "kl": 0.11083984375,
      "learning_rate": 4.625570776255708e-07,
      "loss": 0.0044,
      "reward": 1.8981026411056519,
      "reward_std": 0.12666313955560327,
      "rewards/accuracy_reward": 0.9137276411056519,
      "rewards/format_reward": 0.984375,
      "step": 1177
    },
    {
      "completion_length": 68.4921875,
      "epoch": 5.378995433789954,
      "grad_norm": 1.7856290340423584,
      "kl": 0.12744140625,
      "learning_rate": 4.6210045662100454e-07,
      "loss": 0.0051,
      "reward": 1.6984019875526428,
      "reward_std": 0.18527702055871487,
      "rewards/accuracy_reward": 0.7140269875526428,
      "rewards/format_reward": 0.984375,
      "step": 1178
    },
    {
      "completion_length": 85.734375,
      "epoch": 5.383561643835616,
      "grad_norm": 3.6187846660614014,
      "kl": 0.090576171875,
      "learning_rate": 4.616438356164383e-07,
      "loss": 0.0036,
      "reward": 1.8394480347633362,
      "reward_std": 0.08028085343539715,
      "rewards/accuracy_reward": 0.8394480347633362,
      "rewards/format_reward": 1.0,
      "step": 1179
    },
    {
      "completion_length": 73.796875,
      "epoch": 5.3881278538812785,
      "grad_norm": 2.5453333854675293,
      "kl": 0.17138671875,
      "learning_rate": 4.611872146118721e-07,
      "loss": 0.0068,
      "reward": 1.810937523841858,
      "reward_std": 0.19586525857448578,
      "rewards/accuracy_reward": 0.8343749344348907,
      "rewards/format_reward": 0.9765625,
      "step": 1180
    },
    {
      "completion_length": 75.5546875,
      "epoch": 5.392694063926941,
      "grad_norm": 2.8857522010803223,
      "kl": 0.107177734375,
      "learning_rate": 4.6073059360730595e-07,
      "loss": 0.0043,
      "reward": 1.685937523841858,
      "reward_std": 0.14966705068945885,
      "rewards/accuracy_reward": 0.6859375238418579,
      "rewards/format_reward": 1.0,
      "step": 1181
    },
    {
      "completion_length": 64.7890625,
      "epoch": 5.397260273972603,
      "grad_norm": 2.9432222843170166,
      "kl": 0.14306640625,
      "learning_rate": 4.602739726027397e-07,
      "loss": 0.0057,
      "reward": 1.8203496932983398,
      "reward_std": 0.20706991106271744,
      "rewards/accuracy_reward": 0.8281622231006622,
      "rewards/format_reward": 0.9921875,
      "step": 1182
    },
    {
      "completion_length": 88.6328125,
      "epoch": 5.401826484018265,
      "grad_norm": 2.689836025238037,
      "kl": 0.0927734375,
      "learning_rate": 4.5981735159817347e-07,
      "loss": 0.0037,
      "reward": 1.772569477558136,
      "reward_std": 0.2539152055978775,
      "rewards/accuracy_reward": 0.8116318881511688,
      "rewards/format_reward": 0.9609375,
      "step": 1183
    },
    {
      "completion_length": 74.28125,
      "epoch": 5.406392694063927,
      "grad_norm": 3.8888282775878906,
      "kl": 0.14453125,
      "learning_rate": 4.593607305936073e-07,
      "loss": 0.0058,
      "reward": 1.6562862396240234,
      "reward_std": 0.22701606899499893,
      "rewards/accuracy_reward": 0.679723709821701,
      "rewards/format_reward": 0.9765625,
      "step": 1184
    },
    {
      "completion_length": 90.3671875,
      "epoch": 5.410958904109589,
      "grad_norm": 5.0972981452941895,
      "kl": 0.13427734375,
      "learning_rate": 4.589041095890411e-07,
      "loss": 0.0054,
      "reward": 1.7076823115348816,
      "reward_std": 0.1942017376422882,
      "rewards/accuracy_reward": 0.7154947817325592,
      "rewards/format_reward": 0.9921875,
      "step": 1185
    },
    {
      "completion_length": 94.515625,
      "epoch": 5.415525114155251,
      "grad_norm": 2.337641716003418,
      "kl": 0.112548828125,
      "learning_rate": 4.5844748858447483e-07,
      "loss": 0.0045,
      "reward": 1.8321614265441895,
      "reward_std": 0.08036978542804718,
      "rewards/accuracy_reward": 0.8321613669395447,
      "rewards/format_reward": 1.0,
      "step": 1186
    },
    {
      "completion_length": 70.9296875,
      "epoch": 5.420091324200913,
      "grad_norm": 1.6933238506317139,
      "kl": 0.099853515625,
      "learning_rate": 4.579908675799087e-07,
      "loss": 0.004,
      "reward": 1.6846354007720947,
      "reward_std": 0.20877984166145325,
      "rewards/accuracy_reward": 0.7002603709697723,
      "rewards/format_reward": 0.984375,
      "step": 1187
    },
    {
      "completion_length": 86.8359375,
      "epoch": 5.424657534246576,
      "grad_norm": 9.663613319396973,
      "kl": 0.094970703125,
      "learning_rate": 4.5753424657534246e-07,
      "loss": 0.0038,
      "reward": 1.784375011920929,
      "reward_std": 0.10888781771063805,
      "rewards/accuracy_reward": 0.7843749225139618,
      "rewards/format_reward": 1.0,
      "step": 1188
    },
    {
      "completion_length": 90.921875,
      "epoch": 5.429223744292237,
      "grad_norm": 3.4232490062713623,
      "kl": 0.1181640625,
      "learning_rate": 4.5707762557077625e-07,
      "loss": 0.0047,
      "reward": 1.7392657399177551,
      "reward_std": 0.12979238480329514,
      "rewards/accuracy_reward": 0.7470781803131104,
      "rewards/format_reward": 0.9921875,
      "step": 1189
    },
    {
      "completion_length": 74.96875,
      "epoch": 5.433789954337899,
      "grad_norm": 2.6465795040130615,
      "kl": 0.132080078125,
      "learning_rate": 4.5662100456621e-07,
      "loss": 0.0053,
      "reward": 1.7466145753860474,
      "reward_std": 0.129929106682539,
      "rewards/accuracy_reward": 0.7544271051883698,
      "rewards/format_reward": 0.9921875,
      "step": 1190
    },
    {
      "completion_length": 77.8828125,
      "epoch": 5.438356164383562,
      "grad_norm": 4.823699951171875,
      "kl": 0.12158203125,
      "learning_rate": 4.561643835616438e-07,
      "loss": 0.0049,
      "reward": 1.679805874824524,
      "reward_std": 0.12897202000021935,
      "rewards/accuracy_reward": 0.6798058748245239,
      "rewards/format_reward": 1.0,
      "step": 1191
    },
    {
      "completion_length": 61.109375,
      "epoch": 5.442922374429224,
      "grad_norm": 1.3143653869628906,
      "kl": 0.13623046875,
      "learning_rate": 4.557077625570776e-07,
      "loss": 0.0055,
      "reward": 1.6895833611488342,
      "reward_std": 0.10995453409850597,
      "rewards/accuracy_reward": 0.6895833015441895,
      "rewards/format_reward": 1.0,
      "step": 1192
    },
    {
      "completion_length": 94.640625,
      "epoch": 5.447488584474886,
      "grad_norm": 26.47618293762207,
      "kl": 0.086181640625,
      "learning_rate": 4.552511415525114e-07,
      "loss": 0.0035,
      "reward": 1.8638021349906921,
      "reward_std": 0.14757593348622322,
      "rewards/accuracy_reward": 0.8872395753860474,
      "rewards/format_reward": 0.9765625,
      "step": 1193
    },
    {
      "completion_length": 52.4609375,
      "epoch": 5.4520547945205475,
      "grad_norm": 5.206370830535889,
      "kl": 0.17138671875,
      "learning_rate": 4.547945205479452e-07,
      "loss": 0.0069,
      "reward": 1.6548038721084595,
      "reward_std": 0.23970913887023926,
      "rewards/accuracy_reward": 0.6704288721084595,
      "rewards/format_reward": 0.984375,
      "step": 1194
    },
    {
      "completion_length": 61.78125,
      "epoch": 5.45662100456621,
      "grad_norm": 4.111104488372803,
      "kl": 0.15966796875,
      "learning_rate": 4.54337899543379e-07,
      "loss": 0.0064,
      "reward": 1.7300130128860474,
      "reward_std": 0.17958877980709076,
      "rewards/accuracy_reward": 0.7456380128860474,
      "rewards/format_reward": 0.984375,
      "step": 1195
    },
    {
      "completion_length": 79.1796875,
      "epoch": 5.461187214611872,
      "grad_norm": 1.4285087585449219,
      "kl": 0.10009765625,
      "learning_rate": 4.5388127853881276e-07,
      "loss": 0.004,
      "reward": 1.8193080425262451,
      "reward_std": 0.11568646691739559,
      "rewards/accuracy_reward": 0.827120453119278,
      "rewards/format_reward": 0.9921875,
      "step": 1196
    },
    {
      "completion_length": 80.234375,
      "epoch": 5.465753424657534,
      "grad_norm": 24.182233810424805,
      "kl": 0.119384765625,
      "learning_rate": 4.534246575342466e-07,
      "loss": 0.0048,
      "reward": 1.6295573711395264,
      "reward_std": 0.16487130522727966,
      "rewards/accuracy_reward": 0.6451822966337204,
      "rewards/format_reward": 0.984375,
      "step": 1197
    },
    {
      "completion_length": 73.296875,
      "epoch": 5.470319634703197,
      "grad_norm": 1.8759607076644897,
      "kl": 0.115234375,
      "learning_rate": 4.5296803652968034e-07,
      "loss": 0.0046,
      "reward": 1.742903709411621,
      "reward_std": 0.14027476869523525,
      "rewards/accuracy_reward": 0.7585286498069763,
      "rewards/format_reward": 0.984375,
      "step": 1198
    },
    {
      "completion_length": 56.65625,
      "epoch": 5.474885844748858,
      "grad_norm": 2.124757766723633,
      "kl": 0.129638671875,
      "learning_rate": 4.525114155251141e-07,
      "loss": 0.0052,
      "reward": 1.766406238079071,
      "reward_std": 0.11525032296776772,
      "rewards/accuracy_reward": 0.7664062678813934,
      "rewards/format_reward": 1.0,
      "step": 1199
    },
    {
      "completion_length": 71.984375,
      "epoch": 5.47945205479452,
      "grad_norm": 3.025282382965088,
      "kl": 0.1591796875,
      "learning_rate": 4.520547945205479e-07,
      "loss": 0.0064,
      "reward": 1.565625011920929,
      "reward_std": 0.25334832072257996,
      "rewards/accuracy_reward": 0.5812499821186066,
      "rewards/format_reward": 0.984375,
      "step": 1200
    },
    {
      "completion_length": 71.421875,
      "epoch": 5.4840182648401825,
      "grad_norm": 3.4319229125976562,
      "kl": 0.141357421875,
      "learning_rate": 4.5159817351598175e-07,
      "loss": 0.0057,
      "reward": 1.8551432490348816,
      "reward_std": 0.10183484852313995,
      "rewards/accuracy_reward": 0.8551432490348816,
      "rewards/format_reward": 1.0,
      "step": 1201
    },
    {
      "completion_length": 85.296875,
      "epoch": 5.488584474885845,
      "grad_norm": 2.546898365020752,
      "kl": 0.114501953125,
      "learning_rate": 4.511415525114155e-07,
      "loss": 0.0046,
      "reward": 1.6248489022254944,
      "reward_std": 0.21220972388982773,
      "rewards/accuracy_reward": 0.6404739022254944,
      "rewards/format_reward": 0.984375,
      "step": 1202
    },
    {
      "completion_length": 107.4453125,
      "epoch": 5.493150684931507,
      "grad_norm": 2.221569299697876,
      "kl": 0.12939453125,
      "learning_rate": 4.5068493150684927e-07,
      "loss": 0.0052,
      "reward": 1.7825521230697632,
      "reward_std": 0.11523282900452614,
      "rewards/accuracy_reward": 0.7903645038604736,
      "rewards/format_reward": 0.9921875,
      "step": 1203
    },
    {
      "completion_length": 68.296875,
      "epoch": 5.497716894977169,
      "grad_norm": 5.981943130493164,
      "kl": 0.162109375,
      "learning_rate": 4.502283105022831e-07,
      "loss": 0.0065,
      "reward": 1.6208333373069763,
      "reward_std": 0.18434154987335205,
      "rewards/accuracy_reward": 0.6208333373069763,
      "rewards/format_reward": 1.0,
      "step": 1204
    },
    {
      "completion_length": 82.796875,
      "epoch": 5.502283105022831,
      "grad_norm": 6.735628128051758,
      "kl": 0.10888671875,
      "learning_rate": 4.497716894977169e-07,
      "loss": 0.0044,
      "reward": 1.7694801092147827,
      "reward_std": 0.08905210345983505,
      "rewards/accuracy_reward": 0.7694801688194275,
      "rewards/format_reward": 1.0,
      "step": 1205
    },
    {
      "completion_length": 48.140625,
      "epoch": 5.506849315068493,
      "grad_norm": 4.568028926849365,
      "kl": 0.12255859375,
      "learning_rate": 4.4931506849315063e-07,
      "loss": 0.0049,
      "reward": 1.8573929071426392,
      "reward_std": 0.12051096558570862,
      "rewards/accuracy_reward": 0.8573929369449615,
      "rewards/format_reward": 1.0,
      "step": 1206
    },
    {
      "completion_length": 79.2265625,
      "epoch": 5.511415525114155,
      "grad_norm": 2.16815447807312,
      "kl": 0.114990234375,
      "learning_rate": 4.488584474885845e-07,
      "loss": 0.0046,
      "reward": 1.8097842931747437,
      "reward_std": 0.13499605283141136,
      "rewards/accuracy_reward": 0.8175966441631317,
      "rewards/format_reward": 0.9921875,
      "step": 1207
    },
    {
      "completion_length": 83.3828125,
      "epoch": 5.5159817351598175,
      "grad_norm": 9.177694320678711,
      "kl": 0.100830078125,
      "learning_rate": 4.4840182648401826e-07,
      "loss": 0.004,
      "reward": 1.8088541626930237,
      "reward_std": 0.14230135083198547,
      "rewards/accuracy_reward": 0.8166666626930237,
      "rewards/format_reward": 0.9921875,
      "step": 1208
    },
    {
      "completion_length": 66.96875,
      "epoch": 5.52054794520548,
      "grad_norm": 5.906530380249023,
      "kl": 0.109375,
      "learning_rate": 4.4794520547945205e-07,
      "loss": 0.0044,
      "reward": 1.6549479365348816,
      "reward_std": 0.16808292269706726,
      "rewards/accuracy_reward": 0.6549479067325592,
      "rewards/format_reward": 1.0,
      "step": 1209
    },
    {
      "completion_length": 71.90625,
      "epoch": 5.525114155251142,
      "grad_norm": 2.249753713607788,
      "kl": 0.14599609375,
      "learning_rate": 4.474885844748858e-07,
      "loss": 0.0058,
      "reward": 1.7879971265792847,
      "reward_std": 0.13272903114557266,
      "rewards/accuracy_reward": 0.8036221563816071,
      "rewards/format_reward": 0.984375,
      "step": 1210
    },
    {
      "completion_length": 71.3359375,
      "epoch": 5.529680365296803,
      "grad_norm": 3.3971409797668457,
      "kl": 0.142578125,
      "learning_rate": 4.470319634703196e-07,
      "loss": 0.0057,
      "reward": 1.7578125596046448,
      "reward_std": 0.14966705441474915,
      "rewards/accuracy_reward": 0.7578124403953552,
      "rewards/format_reward": 1.0,
      "step": 1211
    },
    {
      "completion_length": 82.359375,
      "epoch": 5.534246575342466,
      "grad_norm": 4.350315093994141,
      "kl": 0.116455078125,
      "learning_rate": 4.465753424657534e-07,
      "loss": 0.0047,
      "reward": 1.66940575838089,
      "reward_std": 0.14690347015857697,
      "rewards/accuracy_reward": 0.6850306689739227,
      "rewards/format_reward": 0.984375,
      "step": 1212
    },
    {
      "completion_length": 81.0234375,
      "epoch": 5.538812785388128,
      "grad_norm": 2.2065749168395996,
      "kl": 0.094970703125,
      "learning_rate": 4.461187214611872e-07,
      "loss": 0.0038,
      "reward": 1.7763858437538147,
      "reward_std": 0.07247142866253853,
      "rewards/accuracy_reward": 0.7763858437538147,
      "rewards/format_reward": 1.0,
      "step": 1213
    },
    {
      "completion_length": 50.5703125,
      "epoch": 5.54337899543379,
      "grad_norm": 1.8587976694107056,
      "kl": 0.17333984375,
      "learning_rate": 4.45662100456621e-07,
      "loss": 0.0069,
      "reward": 1.8041483163833618,
      "reward_std": 0.1315086344256997,
      "rewards/accuracy_reward": 0.8119607865810394,
      "rewards/format_reward": 0.9921875,
      "step": 1214
    },
    {
      "completion_length": 72.6953125,
      "epoch": 5.5479452054794525,
      "grad_norm": 2.4681146144866943,
      "kl": 0.10400390625,
      "learning_rate": 4.4520547945205477e-07,
      "loss": 0.0042,
      "reward": 1.7903646230697632,
      "reward_std": 0.08470549248158932,
      "rewards/accuracy_reward": 0.7903645634651184,
      "rewards/format_reward": 1.0,
      "step": 1215
    },
    {
      "completion_length": 85.7890625,
      "epoch": 5.552511415525114,
      "grad_norm": 3.0611839294433594,
      "kl": 0.077392578125,
      "learning_rate": 4.4474885844748856e-07,
      "loss": 0.0031,
      "reward": 1.7632812857627869,
      "reward_std": 0.20296207815408707,
      "rewards/accuracy_reward": 0.7789061665534973,
      "rewards/format_reward": 0.984375,
      "step": 1216
    },
    {
      "completion_length": 68.2421875,
      "epoch": 5.557077625570776,
      "grad_norm": 2.2732932567596436,
      "kl": 0.120849609375,
      "learning_rate": 4.442922374429224e-07,
      "loss": 0.0048,
      "reward": 1.663699746131897,
      "reward_std": 0.1902095228433609,
      "rewards/accuracy_reward": 0.6715123057365417,
      "rewards/format_reward": 0.9921875,
      "step": 1217
    },
    {
      "completion_length": 87.59375,
      "epoch": 5.561643835616438,
      "grad_norm": 1.4389771223068237,
      "kl": 0.117431640625,
      "learning_rate": 4.4383561643835613e-07,
      "loss": 0.0047,
      "reward": 1.8174665570259094,
      "reward_std": 0.0971344392746687,
      "rewards/accuracy_reward": 0.825279027223587,
      "rewards/format_reward": 0.9921875,
      "step": 1218
    },
    {
      "completion_length": 98.3046875,
      "epoch": 5.566210045662101,
      "grad_norm": 1.4271727800369263,
      "kl": 0.12255859375,
      "learning_rate": 4.433789954337899e-07,
      "loss": 0.0049,
      "reward": 1.8162059783935547,
      "reward_std": 0.11401430889964104,
      "rewards/accuracy_reward": 0.8240183591842651,
      "rewards/format_reward": 0.9921875,
      "step": 1219
    },
    {
      "completion_length": 73.3046875,
      "epoch": 5.570776255707763,
      "grad_norm": 2.519207239151001,
      "kl": 0.117919921875,
      "learning_rate": 4.429223744292237e-07,
      "loss": 0.0047,
      "reward": 1.7786458730697632,
      "reward_std": 0.13994821161031723,
      "rewards/accuracy_reward": 0.7786457538604736,
      "rewards/format_reward": 1.0,
      "step": 1220
    },
    {
      "completion_length": 68.9453125,
      "epoch": 5.575342465753424,
      "grad_norm": 2.901245594024658,
      "kl": 0.16357421875,
      "learning_rate": 4.4246575342465755e-07,
      "loss": 0.0065,
      "reward": 1.639657735824585,
      "reward_std": 0.24836767464876175,
      "rewards/accuracy_reward": 0.655282735824585,
      "rewards/format_reward": 0.984375,
      "step": 1221
    },
    {
      "completion_length": 96.234375,
      "epoch": 5.579908675799087,
      "grad_norm": 1.3380907773971558,
      "kl": 0.1065673828125,
      "learning_rate": 4.420091324200913e-07,
      "loss": 0.0043,
      "reward": 1.8123116493225098,
      "reward_std": 0.06037373095750809,
      "rewards/accuracy_reward": 0.8123115301132202,
      "rewards/format_reward": 1.0,
      "step": 1222
    },
    {
      "completion_length": 78.203125,
      "epoch": 5.584474885844749,
      "grad_norm": 2.1692612171173096,
      "kl": 0.11328125,
      "learning_rate": 4.4155251141552507e-07,
      "loss": 0.0045,
      "reward": 1.852519154548645,
      "reward_std": 0.08084761165082455,
      "rewards/accuracy_reward": 0.8603315949440002,
      "rewards/format_reward": 0.9921875,
      "step": 1223
    },
    {
      "completion_length": 76.1015625,
      "epoch": 5.589041095890411,
      "grad_norm": 1.5047615766525269,
      "kl": 0.13916015625,
      "learning_rate": 4.410958904109589e-07,
      "loss": 0.0056,
      "reward": 1.8007813096046448,
      "reward_std": 0.13284454122185707,
      "rewards/accuracy_reward": 0.8164061903953552,
      "rewards/format_reward": 0.984375,
      "step": 1224
    },
    {
      "completion_length": 95.3671875,
      "epoch": 5.593607305936073,
      "grad_norm": 5.50695276260376,
      "kl": 0.0765380859375,
      "learning_rate": 4.406392694063927e-07,
      "loss": 0.0031,
      "reward": 1.821587860584259,
      "reward_std": 0.11784346960484982,
      "rewards/accuracy_reward": 0.8294003307819366,
      "rewards/format_reward": 0.9921875,
      "step": 1225
    },
    {
      "completion_length": 79.359375,
      "epoch": 5.598173515981735,
      "grad_norm": 1.8532605171203613,
      "kl": 0.123046875,
      "learning_rate": 4.4018264840182643e-07,
      "loss": 0.0049,
      "reward": 1.672764003276825,
      "reward_std": 0.1454983726143837,
      "rewards/accuracy_reward": 0.672764003276825,
      "rewards/format_reward": 1.0,
      "step": 1226
    },
    {
      "completion_length": 49.9296875,
      "epoch": 5.602739726027397,
      "grad_norm": 4.534260272979736,
      "kl": 0.18798828125,
      "learning_rate": 4.397260273972603e-07,
      "loss": 0.0075,
      "reward": 1.7486504912376404,
      "reward_std": 0.23138123750686646,
      "rewards/accuracy_reward": 0.7642754018306732,
      "rewards/format_reward": 0.984375,
      "step": 1227
    },
    {
      "completion_length": 79.5390625,
      "epoch": 5.607305936073059,
      "grad_norm": 2.4813179969787598,
      "kl": 0.113525390625,
      "learning_rate": 4.3926940639269406e-07,
      "loss": 0.0045,
      "reward": 1.854801058769226,
      "reward_std": 0.15426605194807053,
      "rewards/accuracy_reward": 0.8782385289669037,
      "rewards/format_reward": 0.9765625,
      "step": 1228
    },
    {
      "completion_length": 65.8984375,
      "epoch": 5.6118721461187215,
      "grad_norm": 1.7068575620651245,
      "kl": 0.123291015625,
      "learning_rate": 4.3881278538812785e-07,
      "loss": 0.0049,
      "reward": 1.8518972992897034,
      "reward_std": 0.11522487178444862,
      "rewards/accuracy_reward": 0.8675222992897034,
      "rewards/format_reward": 0.984375,
      "step": 1229
    },
    {
      "completion_length": 73.5390625,
      "epoch": 5.616438356164384,
      "grad_norm": 2.039097309112549,
      "kl": 0.150390625,
      "learning_rate": 4.383561643835616e-07,
      "loss": 0.006,
      "reward": 1.8411458730697632,
      "reward_std": 0.12557167932391167,
      "rewards/accuracy_reward": 0.8411458134651184,
      "rewards/format_reward": 1.0,
      "step": 1230
    },
    {
      "completion_length": 62.3515625,
      "epoch": 5.621004566210045,
      "grad_norm": 1.1969468593597412,
      "kl": 0.14990234375,
      "learning_rate": 4.378995433789954e-07,
      "loss": 0.006,
      "reward": 1.891406238079071,
      "reward_std": 0.055242715403437614,
      "rewards/accuracy_reward": 0.8914062678813934,
      "rewards/format_reward": 1.0,
      "step": 1231
    },
    {
      "completion_length": 65.2421875,
      "epoch": 5.6255707762557075,
      "grad_norm": 0.9585725665092468,
      "kl": 0.095703125,
      "learning_rate": 4.374429223744292e-07,
      "loss": 0.0038,
      "reward": 1.806380271911621,
      "reward_std": 0.053401291370391846,
      "rewards/accuracy_reward": 0.8063801527023315,
      "rewards/format_reward": 1.0,
      "step": 1232
    },
    {
      "completion_length": 94.8828125,
      "epoch": 5.63013698630137,
      "grad_norm": 1.548374891281128,
      "kl": 0.121826171875,
      "learning_rate": 4.36986301369863e-07,
      "loss": 0.0049,
      "reward": 1.8385416865348816,
      "reward_std": 0.07087302953004837,
      "rewards/accuracy_reward": 0.8463541269302368,
      "rewards/format_reward": 0.9921875,
      "step": 1233
    },
    {
      "completion_length": 65.015625,
      "epoch": 5.634703196347032,
      "grad_norm": 3.663956880569458,
      "kl": 0.10302734375,
      "learning_rate": 4.365296803652968e-07,
      "loss": 0.0041,
      "reward": 1.701562523841858,
      "reward_std": 0.11350465193390846,
      "rewards/accuracy_reward": 0.7015624940395355,
      "rewards/format_reward": 1.0,
      "step": 1234
    },
    {
      "completion_length": 76.859375,
      "epoch": 5.639269406392694,
      "grad_norm": 4.082939624786377,
      "kl": 0.123046875,
      "learning_rate": 4.3607305936073057e-07,
      "loss": 0.0049,
      "reward": 1.7964844107627869,
      "reward_std": 0.1577010676264763,
      "rewards/accuracy_reward": 0.7964843213558197,
      "rewards/format_reward": 1.0,
      "step": 1235
    },
    {
      "completion_length": 72.140625,
      "epoch": 5.6438356164383565,
      "grad_norm": 2.4908390045166016,
      "kl": 0.15625,
      "learning_rate": 4.3561643835616436e-07,
      "loss": 0.0062,
      "reward": 1.7734375596046448,
      "reward_std": 0.13743899390101433,
      "rewards/accuracy_reward": 0.7734374701976776,
      "rewards/format_reward": 1.0,
      "step": 1236
    },
    {
      "completion_length": 73.1953125,
      "epoch": 5.648401826484018,
      "grad_norm": 2.1884684562683105,
      "kl": 0.122314453125,
      "learning_rate": 4.351598173515982e-07,
      "loss": 0.0049,
      "reward": 1.7967329621315002,
      "reward_std": 0.22378800809383392,
      "rewards/accuracy_reward": 0.8123579323291779,
      "rewards/format_reward": 0.984375,
      "step": 1237
    },
    {
      "completion_length": 74.21875,
      "epoch": 5.65296803652968,
      "grad_norm": 2.7931785583496094,
      "kl": 0.149169921875,
      "learning_rate": 4.3470319634703193e-07,
      "loss": 0.006,
      "reward": 1.7240886092185974,
      "reward_std": 0.15288061648607254,
      "rewards/accuracy_reward": 0.7240885198116302,
      "rewards/format_reward": 1.0,
      "step": 1238
    },
    {
      "completion_length": 74.875,
      "epoch": 5.657534246575342,
      "grad_norm": 7.83119535446167,
      "kl": 0.398681640625,
      "learning_rate": 4.342465753424657e-07,
      "loss": 0.0159,
      "reward": 1.7018229365348816,
      "reward_std": 0.20128536224365234,
      "rewards/accuracy_reward": 0.7252604067325592,
      "rewards/format_reward": 0.9765625,
      "step": 1239
    },
    {
      "completion_length": 85.15625,
      "epoch": 5.662100456621005,
      "grad_norm": 2.2747838497161865,
      "kl": 0.084716796875,
      "learning_rate": 4.337899543378995e-07,
      "loss": 0.0034,
      "reward": 1.7470728158950806,
      "reward_std": 0.09693595767021179,
      "rewards/accuracy_reward": 0.7470727562904358,
      "rewards/format_reward": 1.0,
      "step": 1240
    },
    {
      "completion_length": 62.78125,
      "epoch": 5.666666666666667,
      "grad_norm": 3.267308235168457,
      "kl": 0.15283203125,
      "learning_rate": 4.3333333333333335e-07,
      "loss": 0.0061,
      "reward": 1.7454678416252136,
      "reward_std": 0.1368042230606079,
      "rewards/accuracy_reward": 0.7532803416252136,
      "rewards/format_reward": 0.9921875,
      "step": 1241
    },
    {
      "completion_length": 88.375,
      "epoch": 5.671232876712329,
      "grad_norm": 2.6403214931488037,
      "kl": 0.1396484375,
      "learning_rate": 4.328767123287671e-07,
      "loss": 0.0056,
      "reward": 1.7575656175613403,
      "reward_std": 0.20803096145391464,
      "rewards/accuracy_reward": 0.7810031175613403,
      "rewards/format_reward": 0.9765625,
      "step": 1242
    },
    {
      "completion_length": 79.96875,
      "epoch": 5.675799086757991,
      "grad_norm": 3.1201863288879395,
      "kl": 0.122314453125,
      "learning_rate": 4.3242009132420087e-07,
      "loss": 0.0049,
      "reward": 1.7832190990447998,
      "reward_std": 0.16100692749023438,
      "rewards/accuracy_reward": 0.7988439798355103,
      "rewards/format_reward": 0.984375,
      "step": 1243
    },
    {
      "completion_length": 81.8515625,
      "epoch": 5.680365296803653,
      "grad_norm": 2.047901153564453,
      "kl": 0.121337890625,
      "learning_rate": 4.319634703196347e-07,
      "loss": 0.0049,
      "reward": 1.760881781578064,
      "reward_std": 0.11731128394603729,
      "rewards/accuracy_reward": 0.7608817219734192,
      "rewards/format_reward": 1.0,
      "step": 1244
    },
    {
      "completion_length": 73.046875,
      "epoch": 5.684931506849315,
      "grad_norm": 1.7335350513458252,
      "kl": 0.147705078125,
      "learning_rate": 4.315068493150685e-07,
      "loss": 0.0059,
      "reward": 1.7863653898239136,
      "reward_std": 0.10918881744146347,
      "rewards/accuracy_reward": 0.8019903302192688,
      "rewards/format_reward": 0.984375,
      "step": 1245
    },
    {
      "completion_length": 71.8125,
      "epoch": 5.689497716894977,
      "grad_norm": 2.161808967590332,
      "kl": 0.149169921875,
      "learning_rate": 4.3105022831050223e-07,
      "loss": 0.006,
      "reward": 1.7140624523162842,
      "reward_std": 0.14955899119377136,
      "rewards/accuracy_reward": 0.7140624821186066,
      "rewards/format_reward": 1.0,
      "step": 1246
    },
    {
      "completion_length": 70.53125,
      "epoch": 5.69406392694064,
      "grad_norm": 1.7596287727355957,
      "kl": 0.13671875,
      "learning_rate": 4.3059360730593607e-07,
      "loss": 0.0055,
      "reward": 1.6859374642372131,
      "reward_std": 0.13941731676459312,
      "rewards/accuracy_reward": 0.6859374940395355,
      "rewards/format_reward": 1.0,
      "step": 1247
    },
    {
      "completion_length": 78.921875,
      "epoch": 5.698630136986301,
      "grad_norm": 1.425408959388733,
      "kl": 0.093994140625,
      "learning_rate": 4.3013698630136986e-07,
      "loss": 0.0038,
      "reward": 1.8960938453674316,
      "reward_std": 0.10968662425875664,
      "rewards/accuracy_reward": 0.9039062261581421,
      "rewards/format_reward": 0.9921875,
      "step": 1248
    },
    {
      "completion_length": 82.5625,
      "epoch": 5.703196347031963,
      "grad_norm": 3.0568716526031494,
      "kl": 0.115234375,
      "learning_rate": 4.2968036529680365e-07,
      "loss": 0.0046,
      "reward": 1.6471354365348816,
      "reward_std": 0.22037852555513382,
      "rewards/accuracy_reward": 0.6549479067325592,
      "rewards/format_reward": 0.9921875,
      "step": 1249
    },
    {
      "completion_length": 92.828125,
      "epoch": 5.707762557077626,
      "grad_norm": 27.536405563354492,
      "kl": 0.10205078125,
      "learning_rate": 4.292237442922374e-07,
      "loss": 0.0041,
      "reward": 1.75323486328125,
      "reward_std": 0.09919268637895584,
      "rewards/accuracy_reward": 0.76104736328125,
      "rewards/format_reward": 0.9921875,
      "step": 1250
    },
    {
      "completion_length": 75.4609375,
      "epoch": 5.712328767123288,
      "grad_norm": 4.183087348937988,
      "kl": 0.100830078125,
      "learning_rate": 4.287671232876712e-07,
      "loss": 0.004,
      "reward": 1.625,
      "reward_std": 0.13258251547813416,
      "rewards/accuracy_reward": 0.6249999403953552,
      "rewards/format_reward": 1.0,
      "step": 1251
    },
    {
      "completion_length": 74.15625,
      "epoch": 5.71689497716895,
      "grad_norm": 7.003372669219971,
      "kl": 0.138671875,
      "learning_rate": 4.28310502283105e-07,
      "loss": 0.0055,
      "reward": 1.631250023841858,
      "reward_std": 0.23694030940532684,
      "rewards/accuracy_reward": 0.6390624642372131,
      "rewards/format_reward": 0.9921875,
      "step": 1252
    },
    {
      "completion_length": 71.2265625,
      "epoch": 5.7214611872146115,
      "grad_norm": 4.267425060272217,
      "kl": 0.159912109375,
      "learning_rate": 4.278538812785388e-07,
      "loss": 0.0064,
      "reward": 1.7780134677886963,
      "reward_std": 0.16748925298452377,
      "rewards/accuracy_reward": 0.7858259081840515,
      "rewards/format_reward": 0.9921875,
      "step": 1253
    },
    {
      "completion_length": 76.8984375,
      "epoch": 5.726027397260274,
      "grad_norm": 2.602168083190918,
      "kl": 0.140625,
      "learning_rate": 4.273972602739726e-07,
      "loss": 0.0056,
      "reward": 1.7774969339370728,
      "reward_std": 0.1920681744813919,
      "rewards/accuracy_reward": 0.7853094637393951,
      "rewards/format_reward": 0.9921875,
      "step": 1254
    },
    {
      "completion_length": 80.0703125,
      "epoch": 5.730593607305936,
      "grad_norm": 3.621494770050049,
      "kl": 0.14501953125,
      "learning_rate": 4.2694063926940637e-07,
      "loss": 0.0058,
      "reward": 1.7632812857627869,
      "reward_std": 0.18639566004276276,
      "rewards/accuracy_reward": 0.7710936963558197,
      "rewards/format_reward": 0.9921875,
      "step": 1255
    },
    {
      "completion_length": 81.65625,
      "epoch": 5.735159817351598,
      "grad_norm": 2.0993213653564453,
      "kl": 0.143310546875,
      "learning_rate": 4.2648401826484016e-07,
      "loss": 0.0057,
      "reward": 1.765625,
      "reward_std": 0.1767766885459423,
      "rewards/accuracy_reward": 0.7812499403953552,
      "rewards/format_reward": 0.984375,
      "step": 1256
    },
    {
      "completion_length": 63.765625,
      "epoch": 5.739726027397261,
      "grad_norm": 2.3051700592041016,
      "kl": 0.19140625,
      "learning_rate": 4.26027397260274e-07,
      "loss": 0.0077,
      "reward": 1.800000011920929,
      "reward_std": 0.11946406960487366,
      "rewards/accuracy_reward": 0.8078124821186066,
      "rewards/format_reward": 0.9921875,
      "step": 1257
    },
    {
      "completion_length": 90.7265625,
      "epoch": 5.744292237442922,
      "grad_norm": 5.680175304412842,
      "kl": 0.13720703125,
      "learning_rate": 4.2557077625570773e-07,
      "loss": 0.0055,
      "reward": 1.7391927242279053,
      "reward_std": 0.1934903860092163,
      "rewards/accuracy_reward": 0.7548176646232605,
      "rewards/format_reward": 0.984375,
      "step": 1258
    },
    {
      "completion_length": 67.1171875,
      "epoch": 5.748858447488584,
      "grad_norm": 3.6025280952453613,
      "kl": 0.1533203125,
      "learning_rate": 4.251141552511415e-07,
      "loss": 0.0061,
      "reward": 1.6899954080581665,
      "reward_std": 0.177649587392807,
      "rewards/accuracy_reward": 0.6899954080581665,
      "rewards/format_reward": 1.0,
      "step": 1259
    },
    {
      "completion_length": 78.46875,
      "epoch": 5.7534246575342465,
      "grad_norm": 2.8901851177215576,
      "kl": 0.150390625,
      "learning_rate": 4.246575342465753e-07,
      "loss": 0.006,
      "reward": 1.65723317861557,
      "reward_std": 0.15903639048337936,
      "rewards/accuracy_reward": 0.6728581488132477,
      "rewards/format_reward": 0.984375,
      "step": 1260
    },
    {
      "completion_length": 90.2890625,
      "epoch": 5.757990867579909,
      "grad_norm": 30.16987419128418,
      "kl": 0.11328125,
      "learning_rate": 4.2420091324200915e-07,
      "loss": 0.0045,
      "reward": 1.80078125,
      "reward_std": 0.13798470050096512,
      "rewards/accuracy_reward": 0.8007811605930328,
      "rewards/format_reward": 1.0,
      "step": 1261
    },
    {
      "completion_length": 69.9375,
      "epoch": 5.762557077625571,
      "grad_norm": 4.615162372589111,
      "kl": 0.2236328125,
      "learning_rate": 4.237442922374429e-07,
      "loss": 0.0089,
      "reward": 1.7741782069206238,
      "reward_std": 0.1786958873271942,
      "rewards/accuracy_reward": 0.7819906771183014,
      "rewards/format_reward": 0.9921875,
      "step": 1262
    },
    {
      "completion_length": 85.2421875,
      "epoch": 5.767123287671232,
      "grad_norm": 2.68100905418396,
      "kl": 0.10888671875,
      "learning_rate": 4.2328767123287667e-07,
      "loss": 0.0044,
      "reward": 1.7717448472976685,
      "reward_std": 0.1444082222878933,
      "rewards/accuracy_reward": 0.7795572876930237,
      "rewards/format_reward": 0.9921875,
      "step": 1263
    },
    {
      "completion_length": 89.0390625,
      "epoch": 5.771689497716895,
      "grad_norm": 3.867462158203125,
      "kl": 0.141845703125,
      "learning_rate": 4.228310502283105e-07,
      "loss": 0.0057,
      "reward": 1.6650173664093018,
      "reward_std": 0.21590976417064667,
      "rewards/accuracy_reward": 0.6806423366069794,
      "rewards/format_reward": 0.984375,
      "step": 1264
    },
    {
      "completion_length": 72.515625,
      "epoch": 5.776255707762557,
      "grad_norm": 9.51895523071289,
      "kl": 0.103271484375,
      "learning_rate": 4.223744292237443e-07,
      "loss": 0.0041,
      "reward": 1.7899226546287537,
      "reward_std": 0.1994732916355133,
      "rewards/accuracy_reward": 0.8133600950241089,
      "rewards/format_reward": 0.9765625,
      "step": 1265
    },
    {
      "completion_length": 87.1328125,
      "epoch": 5.780821917808219,
      "grad_norm": 3.7063732147216797,
      "kl": 0.142578125,
      "learning_rate": 4.2191780821917803e-07,
      "loss": 0.0057,
      "reward": 1.7083333730697632,
      "reward_std": 0.19097717106342316,
      "rewards/accuracy_reward": 0.7161458134651184,
      "rewards/format_reward": 0.9921875,
      "step": 1266
    },
    {
      "completion_length": 85.3828125,
      "epoch": 5.7853881278538815,
      "grad_norm": 2.703822374343872,
      "kl": 0.13037109375,
      "learning_rate": 4.2146118721461187e-07,
      "loss": 0.0052,
      "reward": 1.5992187857627869,
      "reward_std": 0.2338416427373886,
      "rewards/accuracy_reward": 0.6148437559604645,
      "rewards/format_reward": 0.984375,
      "step": 1267
    },
    {
      "completion_length": 70.9453125,
      "epoch": 5.789954337899544,
      "grad_norm": 4.01165246963501,
      "kl": 0.17236328125,
      "learning_rate": 4.2100456621004566e-07,
      "loss": 0.0069,
      "reward": 1.794851839542389,
      "reward_std": 0.17533704824745655,
      "rewards/accuracy_reward": 0.8104767203330994,
      "rewards/format_reward": 0.984375,
      "step": 1268
    },
    {
      "completion_length": 78.515625,
      "epoch": 5.794520547945205,
      "grad_norm": 4.9043684005737305,
      "kl": 0.137451171875,
      "learning_rate": 4.2054794520547945e-07,
      "loss": 0.0055,
      "reward": 1.6109731793403625,
      "reward_std": 0.2282901182770729,
      "rewards/accuracy_reward": 0.6187856197357178,
      "rewards/format_reward": 0.9921875,
      "step": 1269
    },
    {
      "completion_length": 89.8671875,
      "epoch": 5.799086757990867,
      "grad_norm": 2.3333468437194824,
      "kl": 0.099853515625,
      "learning_rate": 4.200913242009132e-07,
      "loss": 0.004,
      "reward": 1.7177083492279053,
      "reward_std": 0.17229503020644188,
      "rewards/accuracy_reward": 0.7255208194255829,
      "rewards/format_reward": 0.9921875,
      "step": 1270
    },
    {
      "completion_length": 71.6171875,
      "epoch": 5.80365296803653,
      "grad_norm": 2.1021461486816406,
      "kl": 0.153564453125,
      "learning_rate": 4.19634703196347e-07,
      "loss": 0.0061,
      "reward": 1.6888020634651184,
      "reward_std": 0.30169400572776794,
      "rewards/accuracy_reward": 0.7200521230697632,
      "rewards/format_reward": 0.96875,
      "step": 1271
    },
    {
      "completion_length": 84.703125,
      "epoch": 5.808219178082192,
      "grad_norm": 4.3470282554626465,
      "kl": 0.148193359375,
      "learning_rate": 4.191780821917808e-07,
      "loss": 0.0059,
      "reward": 1.7307292222976685,
      "reward_std": 0.1593589335680008,
      "rewards/accuracy_reward": 0.7463541030883789,
      "rewards/format_reward": 0.984375,
      "step": 1272
    },
    {
      "completion_length": 63.3203125,
      "epoch": 5.812785388127854,
      "grad_norm": 5.333644390106201,
      "kl": 0.159912109375,
      "learning_rate": 4.187214611872146e-07,
      "loss": 0.0064,
      "reward": 1.6519964933395386,
      "reward_std": 0.18434231728315353,
      "rewards/accuracy_reward": 0.6519964933395386,
      "rewards/format_reward": 1.0,
      "step": 1273
    },
    {
      "completion_length": 67.6953125,
      "epoch": 5.817351598173516,
      "grad_norm": 2.9642715454101562,
      "kl": 0.1298828125,
      "learning_rate": 4.182648401826484e-07,
      "loss": 0.0052,
      "reward": 1.6983258724212646,
      "reward_std": 0.15258603543043137,
      "rewards/accuracy_reward": 0.698325902223587,
      "rewards/format_reward": 1.0,
      "step": 1274
    },
    {
      "completion_length": 74.75,
      "epoch": 5.821917808219178,
      "grad_norm": 2.8370203971862793,
      "kl": 0.1337890625,
      "learning_rate": 4.1780821917808217e-07,
      "loss": 0.0053,
      "reward": 1.757942795753479,
      "reward_std": 0.24083544313907623,
      "rewards/accuracy_reward": 0.7813802063465118,
      "rewards/format_reward": 0.9765625,
      "step": 1275
    },
    {
      "completion_length": 90.6640625,
      "epoch": 5.82648401826484,
      "grad_norm": 1.1865882873535156,
      "kl": 0.076171875,
      "learning_rate": 4.1735159817351596e-07,
      "loss": 0.0031,
      "reward": 1.8952009081840515,
      "reward_std": 0.09130230359733105,
      "rewards/accuracy_reward": 0.9030132591724396,
      "rewards/format_reward": 0.9921875,
      "step": 1276
    },
    {
      "completion_length": 89.4375,
      "epoch": 5.831050228310502,
      "grad_norm": 2.1936404705047607,
      "kl": 0.094482421875,
      "learning_rate": 4.168949771689498e-07,
      "loss": 0.0038,
      "reward": 1.672282099723816,
      "reward_std": 0.1674654446542263,
      "rewards/accuracy_reward": 0.6957195997238159,
      "rewards/format_reward": 0.9765625,
      "step": 1277
    },
    {
      "completion_length": 88.984375,
      "epoch": 5.835616438356165,
      "grad_norm": 2.0885391235351562,
      "kl": 0.12646484375,
      "learning_rate": 4.1643835616438353e-07,
      "loss": 0.0051,
      "reward": 1.6463323831558228,
      "reward_std": 0.18611476570367813,
      "rewards/accuracy_reward": 0.6619572937488556,
      "rewards/format_reward": 0.984375,
      "step": 1278
    },
    {
      "completion_length": 75.7109375,
      "epoch": 5.840182648401827,
      "grad_norm": 6.192373275756836,
      "kl": 0.135009765625,
      "learning_rate": 4.159817351598173e-07,
      "loss": 0.0054,
      "reward": 1.634151816368103,
      "reward_std": 0.11692540347576141,
      "rewards/accuracy_reward": 0.6419642567634583,
      "rewards/format_reward": 0.9921875,
      "step": 1279
    },
    {
      "completion_length": 80.0078125,
      "epoch": 5.844748858447488,
      "grad_norm": 2.136561155319214,
      "kl": 0.16015625,
      "learning_rate": 4.155251141552511e-07,
      "loss": 0.0064,
      "reward": 1.7792754769325256,
      "reward_std": 0.09816346131265163,
      "rewards/accuracy_reward": 0.7870878875255585,
      "rewards/format_reward": 0.9921875,
      "step": 1280
    },
    {
      "completion_length": 99.0078125,
      "epoch": 5.8493150684931505,
      "grad_norm": 1.984879732131958,
      "kl": 0.08544921875,
      "learning_rate": 4.1506849315068495e-07,
      "loss": 0.0034,
      "reward": 1.7765625715255737,
      "reward_std": 0.06629125960171223,
      "rewards/accuracy_reward": 0.7843749523162842,
      "rewards/format_reward": 0.9921875,
      "step": 1281
    },
    {
      "completion_length": 82.71875,
      "epoch": 5.853881278538813,
      "grad_norm": 2.8610336780548096,
      "kl": 0.11669921875,
      "learning_rate": 4.146118721461187e-07,
      "loss": 0.0047,
      "reward": 1.854315459728241,
      "reward_std": 0.11991548165678978,
      "rewards/accuracy_reward": 0.8543154299259186,
      "rewards/format_reward": 1.0,
      "step": 1282
    },
    {
      "completion_length": 98.296875,
      "epoch": 5.858447488584475,
      "grad_norm": 1.845438003540039,
      "kl": 0.101318359375,
      "learning_rate": 4.1415525114155247e-07,
      "loss": 0.004,
      "reward": 1.8150991797447205,
      "reward_std": 0.07342393416911364,
      "rewards/accuracy_reward": 0.8150991201400757,
      "rewards/format_reward": 1.0,
      "step": 1283
    },
    {
      "completion_length": 67.1953125,
      "epoch": 5.863013698630137,
      "grad_norm": 5.036915302276611,
      "kl": 0.145751953125,
      "learning_rate": 4.136986301369863e-07,
      "loss": 0.0058,
      "reward": 1.7412946820259094,
      "reward_std": 0.175389152020216,
      "rewards/accuracy_reward": 0.7491071224212646,
      "rewards/format_reward": 0.9921875,
      "step": 1284
    },
    {
      "completion_length": 67.8828125,
      "epoch": 5.867579908675799,
      "grad_norm": 5.918455123901367,
      "kl": 0.150390625,
      "learning_rate": 4.132420091324201e-07,
      "loss": 0.006,
      "reward": 1.7437500357627869,
      "reward_std": 0.2584189176559448,
      "rewards/accuracy_reward": 0.7749999761581421,
      "rewards/format_reward": 0.96875,
      "step": 1285
    },
    {
      "completion_length": 74.6015625,
      "epoch": 5.872146118721461,
      "grad_norm": 1.850948691368103,
      "kl": 0.132568359375,
      "learning_rate": 4.1278538812785383e-07,
      "loss": 0.0053,
      "reward": 1.9023438096046448,
      "reward_std": 0.08417459111660719,
      "rewards/accuracy_reward": 0.9023437201976776,
      "rewards/format_reward": 1.0,
      "step": 1286
    },
    {
      "completion_length": 68.390625,
      "epoch": 5.876712328767123,
      "grad_norm": 4.7030534744262695,
      "kl": 0.166015625,
      "learning_rate": 4.1232876712328767e-07,
      "loss": 0.0067,
      "reward": 1.8146693706512451,
      "reward_std": 0.1410301998257637,
      "rewards/accuracy_reward": 0.8146693110466003,
      "rewards/format_reward": 1.0,
      "step": 1287
    },
    {
      "completion_length": 83.390625,
      "epoch": 5.8812785388127855,
      "grad_norm": 2.299928665161133,
      "kl": 0.114501953125,
      "learning_rate": 4.1187214611872146e-07,
      "loss": 0.0046,
      "reward": 1.7009549140930176,
      "reward_std": 0.18842098861932755,
      "rewards/accuracy_reward": 0.7087673544883728,
      "rewards/format_reward": 0.9921875,
      "step": 1288
    },
    {
      "completion_length": 77.59375,
      "epoch": 5.885844748858448,
      "grad_norm": 2.9478793144226074,
      "kl": 0.1025390625,
      "learning_rate": 4.1141552511415525e-07,
      "loss": 0.0041,
      "reward": 1.748046875,
      "reward_std": 0.15785659104585648,
      "rewards/accuracy_reward": 0.7480468451976776,
      "rewards/format_reward": 1.0,
      "step": 1289
    },
    {
      "completion_length": 100.3359375,
      "epoch": 5.890410958904109,
      "grad_norm": 2.3754220008850098,
      "kl": 0.112060546875,
      "learning_rate": 4.10958904109589e-07,
      "loss": 0.0045,
      "reward": 1.7859375476837158,
      "reward_std": 0.17908401414752007,
      "rewards/accuracy_reward": 0.8015624284744263,
      "rewards/format_reward": 0.984375,
      "step": 1290
    },
    {
      "completion_length": 69.46875,
      "epoch": 5.894977168949771,
      "grad_norm": 3.575843334197998,
      "kl": 0.134765625,
      "learning_rate": 4.105022831050228e-07,
      "loss": 0.0054,
      "reward": 1.845781147480011,
      "reward_std": 0.08029642701148987,
      "rewards/accuracy_reward": 0.8457811176776886,
      "rewards/format_reward": 1.0,
      "step": 1291
    },
    {
      "completion_length": 94.5234375,
      "epoch": 5.899543378995434,
      "grad_norm": 7.622490406036377,
      "kl": 0.0859375,
      "learning_rate": 4.100456621004566e-07,
      "loss": 0.0034,
      "reward": 1.796093761920929,
      "reward_std": 0.1538807898759842,
      "rewards/accuracy_reward": 0.803906261920929,
      "rewards/format_reward": 0.9921875,
      "step": 1292
    },
    {
      "completion_length": 82.234375,
      "epoch": 5.904109589041096,
      "grad_norm": 7.203249931335449,
      "kl": 0.11962890625,
      "learning_rate": 4.095890410958904e-07,
      "loss": 0.0048,
      "reward": 1.737942636013031,
      "reward_std": 0.16788282990455627,
      "rewards/accuracy_reward": 0.7457550466060638,
      "rewards/format_reward": 0.9921875,
      "step": 1293
    },
    {
      "completion_length": 82.7578125,
      "epoch": 5.908675799086758,
      "grad_norm": 2.2500412464141846,
      "kl": 0.101806640625,
      "learning_rate": 4.091324200913242e-07,
      "loss": 0.0041,
      "reward": 1.7422269582748413,
      "reward_std": 0.13895701617002487,
      "rewards/accuracy_reward": 0.7500394582748413,
      "rewards/format_reward": 0.9921875,
      "step": 1294
    },
    {
      "completion_length": 93.71875,
      "epoch": 5.91324200913242,
      "grad_norm": 2.051222324371338,
      "kl": 0.122314453125,
      "learning_rate": 4.0867579908675797e-07,
      "loss": 0.0049,
      "reward": 1.8802083730697632,
      "reward_std": 0.08680902794003487,
      "rewards/accuracy_reward": 0.8880207538604736,
      "rewards/format_reward": 0.9921875,
      "step": 1295
    },
    {
      "completion_length": 98.2421875,
      "epoch": 5.917808219178082,
      "grad_norm": 1.9777852296829224,
      "kl": 0.10009765625,
      "learning_rate": 4.0821917808219176e-07,
      "loss": 0.004,
      "reward": 1.805654764175415,
      "reward_std": 0.10066970996558666,
      "rewards/accuracy_reward": 0.8134672045707703,
      "rewards/format_reward": 0.9921875,
      "step": 1296
    },
    {
      "completion_length": 73.6875,
      "epoch": 5.922374429223744,
      "grad_norm": 2.501546859741211,
      "kl": 0.1171875,
      "learning_rate": 4.077625570776256e-07,
      "loss": 0.0047,
      "reward": 1.7920072674751282,
      "reward_std": 0.11444094032049179,
      "rewards/accuracy_reward": 0.7920072078704834,
      "rewards/format_reward": 1.0,
      "step": 1297
    },
    {
      "completion_length": 113.390625,
      "epoch": 5.926940639269406,
      "grad_norm": 1.9746527671813965,
      "kl": 0.05908203125,
      "learning_rate": 4.0730593607305933e-07,
      "loss": 0.0024,
      "reward": 1.830468773841858,
      "reward_std": 0.17702769488096237,
      "rewards/accuracy_reward": 0.8617186546325684,
      "rewards/format_reward": 0.96875,
      "step": 1298
    },
    {
      "completion_length": 89.3203125,
      "epoch": 5.931506849315069,
      "grad_norm": 1.95270574092865,
      "kl": 0.118896484375,
      "learning_rate": 4.068493150684931e-07,
      "loss": 0.0048,
      "reward": 1.784375011920929,
      "reward_std": 0.11048543453216553,
      "rewards/accuracy_reward": 0.784375011920929,
      "rewards/format_reward": 1.0,
      "step": 1299
    },
    {
      "completion_length": 80.3359375,
      "epoch": 5.936073059360731,
      "grad_norm": 2.471496820449829,
      "kl": 0.116943359375,
      "learning_rate": 4.063926940639269e-07,
      "loss": 0.0047,
      "reward": 1.7593750357627869,
      "reward_std": 0.11229449138045311,
      "rewards/accuracy_reward": 0.7671874761581421,
      "rewards/format_reward": 0.9921875,
      "step": 1300
    },
    {
      "completion_length": 82.578125,
      "epoch": 5.940639269406392,
      "grad_norm": 2.451826572418213,
      "kl": 0.15087890625,
      "learning_rate": 4.0593607305936075e-07,
      "loss": 0.006,
      "reward": 1.713808834552765,
      "reward_std": 0.16193728893995285,
      "rewards/accuracy_reward": 0.7216213047504425,
      "rewards/format_reward": 0.9921875,
      "step": 1301
    },
    {
      "completion_length": 68.0234375,
      "epoch": 5.945205479452055,
      "grad_norm": 1.9107226133346558,
      "kl": 0.087890625,
      "learning_rate": 4.054794520547945e-07,
      "loss": 0.0035,
      "reward": 1.7034350037574768,
      "reward_std": 0.09215648844838142,
      "rewards/accuracy_reward": 0.7034350335597992,
      "rewards/format_reward": 1.0,
      "step": 1302
    },
    {
      "completion_length": 87.0234375,
      "epoch": 5.949771689497717,
      "grad_norm": 1.9171605110168457,
      "kl": 0.104736328125,
      "learning_rate": 4.0502283105022827e-07,
      "loss": 0.0042,
      "reward": 1.7768229246139526,
      "reward_std": 0.09278659522533417,
      "rewards/accuracy_reward": 0.7768228650093079,
      "rewards/format_reward": 1.0,
      "step": 1303
    },
    {
      "completion_length": 65.109375,
      "epoch": 5.954337899543379,
      "grad_norm": 2.9926059246063232,
      "kl": 0.1611328125,
      "learning_rate": 4.045662100456621e-07,
      "loss": 0.0064,
      "reward": 1.6825549602508545,
      "reward_std": 0.27830731868743896,
      "rewards/accuracy_reward": 0.7216174304485321,
      "rewards/format_reward": 0.9609375,
      "step": 1304
    },
    {
      "completion_length": 73.109375,
      "epoch": 5.958904109589041,
      "grad_norm": 7.7803120613098145,
      "kl": 0.1435546875,
      "learning_rate": 4.041095890410959e-07,
      "loss": 0.0057,
      "reward": 1.7353760600090027,
      "reward_std": 0.16425937414169312,
      "rewards/accuracy_reward": 0.7431885600090027,
      "rewards/format_reward": 0.9921875,
      "step": 1305
    },
    {
      "completion_length": 83.65625,
      "epoch": 5.963470319634704,
      "grad_norm": 2.513974905014038,
      "kl": 0.105712890625,
      "learning_rate": 4.0365296803652963e-07,
      "loss": 0.0042,
      "reward": 1.8269480466842651,
      "reward_std": 0.1296938955783844,
      "rewards/accuracy_reward": 0.8347605168819427,
      "rewards/format_reward": 0.9921875,
      "step": 1306
    },
    {
      "completion_length": 61.1328125,
      "epoch": 5.968036529680365,
      "grad_norm": 3.1066365242004395,
      "kl": 0.169921875,
      "learning_rate": 4.0319634703196347e-07,
      "loss": 0.0068,
      "reward": 1.8063979148864746,
      "reward_std": 0.19071058928966522,
      "rewards/accuracy_reward": 0.8220228552818298,
      "rewards/format_reward": 0.984375,
      "step": 1307
    },
    {
      "completion_length": 81.75,
      "epoch": 5.972602739726027,
      "grad_norm": 6.998978614807129,
      "kl": 0.154296875,
      "learning_rate": 4.0273972602739726e-07,
      "loss": 0.0062,
      "reward": 1.7838542461395264,
      "reward_std": 0.13565516471862793,
      "rewards/accuracy_reward": 0.7916666567325592,
      "rewards/format_reward": 0.9921875,
      "step": 1308
    },
    {
      "completion_length": 65.1953125,
      "epoch": 5.9771689497716896,
      "grad_norm": 2.565167188644409,
      "kl": 0.11279296875,
      "learning_rate": 4.0228310502283105e-07,
      "loss": 0.0045,
      "reward": 1.685937523841858,
      "reward_std": 0.15467960759997368,
      "rewards/accuracy_reward": 0.6937499940395355,
      "rewards/format_reward": 0.9921875,
      "step": 1309
    },
    {
      "completion_length": 84.0703125,
      "epoch": 5.981735159817352,
      "grad_norm": 4.1508073806762695,
      "kl": 0.099365234375,
      "learning_rate": 4.018264840182648e-07,
      "loss": 0.004,
      "reward": 1.706423580646515,
      "reward_std": 0.1783130094408989,
      "rewards/accuracy_reward": 0.7220486104488373,
      "rewards/format_reward": 0.984375,
      "step": 1310
    },
    {
      "completion_length": 74.0703125,
      "epoch": 5.986301369863014,
      "grad_norm": 4.801496505737305,
      "kl": 0.17236328125,
      "learning_rate": 4.013698630136986e-07,
      "loss": 0.0069,
      "reward": 1.8074793815612793,
      "reward_std": 0.14287326484918594,
      "rewards/accuracy_reward": 0.8152918219566345,
      "rewards/format_reward": 0.9921875,
      "step": 1311
    },
    {
      "completion_length": 101.0390625,
      "epoch": 5.9908675799086755,
      "grad_norm": 3.006030797958374,
      "kl": 0.08642578125,
      "learning_rate": 4.009132420091324e-07,
      "loss": 0.0035,
      "reward": 1.8205991387367249,
      "reward_std": 0.14634042605757713,
      "rewards/accuracy_reward": 0.8284116387367249,
      "rewards/format_reward": 0.9921875,
      "step": 1312
    },
    {
      "completion_length": 70.0078125,
      "epoch": 5.995433789954338,
      "grad_norm": 3.449769973754883,
      "kl": 0.13720703125,
      "learning_rate": 4.004566210045662e-07,
      "loss": 0.0055,
      "reward": 1.651562511920929,
      "reward_std": 0.21917617321014404,
      "rewards/accuracy_reward": 0.659375011920929,
      "rewards/format_reward": 0.9921875,
      "step": 1313
    },
    {
      "completion_length": 44.0,
      "epoch": 6.0,
      "grad_norm": 3.673351287841797,
      "kl": 0.1494140625,
      "learning_rate": 4e-07,
      "loss": 0.0045,
      "reward": 1.625,
      "reward_std": 0.408231720328331,
      "rewards/accuracy_reward": 0.625,
      "rewards/format_reward": 1.0,
      "step": 1314
    },
    {
      "completion_length": 84.4453125,
      "epoch": 6.004566210045662,
      "grad_norm": 1.5486712455749512,
      "kl": 0.10546875,
      "learning_rate": 3.9954337899543377e-07,
      "loss": 0.0042,
      "reward": 1.8225446343421936,
      "reward_std": 0.06181819178164005,
      "rewards/accuracy_reward": 0.8225446343421936,
      "rewards/format_reward": 1.0,
      "step": 1315
    },
    {
      "completion_length": 87.6484375,
      "epoch": 6.0091324200913245,
      "grad_norm": 2.9046413898468018,
      "kl": 0.106201171875,
      "learning_rate": 3.9908675799086756e-07,
      "loss": 0.0042,
      "reward": 1.8345487117767334,
      "reward_std": 0.1487816534936428,
      "rewards/accuracy_reward": 0.8501735925674438,
      "rewards/format_reward": 0.984375,
      "step": 1316
    },
    {
      "completion_length": 91.9453125,
      "epoch": 6.013698630136986,
      "grad_norm": 3.910123348236084,
      "kl": 0.083251953125,
      "learning_rate": 3.9863013698630134e-07,
      "loss": 0.0033,
      "reward": 1.6695313453674316,
      "reward_std": 0.2122296392917633,
      "rewards/accuracy_reward": 0.6851562261581421,
      "rewards/format_reward": 0.984375,
      "step": 1317
    },
    {
      "completion_length": 71.890625,
      "epoch": 6.018264840182648,
      "grad_norm": 3.065825939178467,
      "kl": 0.124755859375,
      "learning_rate": 3.9817351598173513e-07,
      "loss": 0.005,
      "reward": 1.6242188215255737,
      "reward_std": 0.14951937273144722,
      "rewards/accuracy_reward": 0.624218761920929,
      "rewards/format_reward": 1.0,
      "step": 1318
    },
    {
      "completion_length": 97.1875,
      "epoch": 6.0228310502283104,
      "grad_norm": 1.6381162405014038,
      "kl": 0.10986328125,
      "learning_rate": 3.977168949771689e-07,
      "loss": 0.0044,
      "reward": 1.8406250476837158,
      "reward_std": 0.14283224940299988,
      "rewards/accuracy_reward": 0.8562499284744263,
      "rewards/format_reward": 0.984375,
      "step": 1319
    },
    {
      "completion_length": 74.859375,
      "epoch": 6.027397260273973,
      "grad_norm": 4.058730125427246,
      "kl": 0.13037109375,
      "learning_rate": 3.972602739726027e-07,
      "loss": 0.0052,
      "reward": 1.7753472328186035,
      "reward_std": 0.11618576943874359,
      "rewards/accuracy_reward": 0.7909722030162811,
      "rewards/format_reward": 0.984375,
      "step": 1320
    },
    {
      "completion_length": 75.2109375,
      "epoch": 6.031963470319635,
      "grad_norm": 2.673145055770874,
      "kl": 0.103271484375,
      "learning_rate": 3.9680365296803655e-07,
      "loss": 0.0041,
      "reward": 1.8099148869514465,
      "reward_std": 0.0990208312869072,
      "rewards/accuracy_reward": 0.8177272975444794,
      "rewards/format_reward": 0.9921875,
      "step": 1321
    },
    {
      "completion_length": 70.84375,
      "epoch": 6.036529680365296,
      "grad_norm": 4.576772689819336,
      "kl": 0.12890625,
      "learning_rate": 3.963470319634703e-07,
      "loss": 0.0052,
      "reward": 1.7658482193946838,
      "reward_std": 0.1474735364317894,
      "rewards/accuracy_reward": 0.7814731895923615,
      "rewards/format_reward": 0.984375,
      "step": 1322
    },
    {
      "completion_length": 75.84375,
      "epoch": 6.041095890410959,
      "grad_norm": 3.1631548404693604,
      "kl": 0.1611328125,
      "learning_rate": 3.9589041095890407e-07,
      "loss": 0.0064,
      "reward": 1.6473042964935303,
      "reward_std": 0.17740929126739502,
      "rewards/accuracy_reward": 0.6473042666912079,
      "rewards/format_reward": 1.0,
      "step": 1323
    },
    {
      "completion_length": 71.1640625,
      "epoch": 6.045662100456621,
      "grad_norm": 2.176661968231201,
      "kl": 0.1416015625,
      "learning_rate": 3.954337899543379e-07,
      "loss": 0.0057,
      "reward": 1.8772332668304443,
      "reward_std": 0.07372352294623852,
      "rewards/accuracy_reward": 0.8772332966327667,
      "rewards/format_reward": 1.0,
      "step": 1324
    },
    {
      "completion_length": 80.1328125,
      "epoch": 6.050228310502283,
      "grad_norm": 2.169936180114746,
      "kl": 0.12109375,
      "learning_rate": 3.949771689497717e-07,
      "loss": 0.0048,
      "reward": 1.6571251153945923,
      "reward_std": 0.13837599009275436,
      "rewards/accuracy_reward": 0.6571250259876251,
      "rewards/format_reward": 1.0,
      "step": 1325
    },
    {
      "completion_length": 83.4375,
      "epoch": 6.054794520547945,
      "grad_norm": 2.906425714492798,
      "kl": 0.15966796875,
      "learning_rate": 3.9452054794520543e-07,
      "loss": 0.0064,
      "reward": 1.87109375,
      "reward_std": 0.17256292700767517,
      "rewards/accuracy_reward": 0.88671875,
      "rewards/format_reward": 0.984375,
      "step": 1326
    },
    {
      "completion_length": 69.375,
      "epoch": 6.059360730593608,
      "grad_norm": 2.7249135971069336,
      "kl": 0.197998046875,
      "learning_rate": 3.940639269406392e-07,
      "loss": 0.0079,
      "reward": 1.780989646911621,
      "reward_std": 0.19701316952705383,
      "rewards/accuracy_reward": 0.7888020575046539,
      "rewards/format_reward": 0.9921875,
      "step": 1327
    },
    {
      "completion_length": 85.78125,
      "epoch": 6.063926940639269,
      "grad_norm": 2.412091016769409,
      "kl": 0.125732421875,
      "learning_rate": 3.9360730593607306e-07,
      "loss": 0.005,
      "reward": 1.6776041984558105,
      "reward_std": 0.15093514323234558,
      "rewards/accuracy_reward": 0.6932291090488434,
      "rewards/format_reward": 0.984375,
      "step": 1328
    },
    {
      "completion_length": 67.96875,
      "epoch": 6.068493150684931,
      "grad_norm": 2.928151845932007,
      "kl": 0.14501953125,
      "learning_rate": 3.9315068493150684e-07,
      "loss": 0.0058,
      "reward": 1.7759548425674438,
      "reward_std": 0.15794897824525833,
      "rewards/accuracy_reward": 0.7759548723697662,
      "rewards/format_reward": 1.0,
      "step": 1329
    },
    {
      "completion_length": 94.921875,
      "epoch": 6.073059360730594,
      "grad_norm": 1.4851278066635132,
      "kl": 0.093994140625,
      "learning_rate": 3.926940639269406e-07,
      "loss": 0.0038,
      "reward": 1.8491500616073608,
      "reward_std": 0.14507145062088966,
      "rewards/accuracy_reward": 0.8725875020027161,
      "rewards/format_reward": 0.9765625,
      "step": 1330
    },
    {
      "completion_length": 78.6484375,
      "epoch": 6.077625570776256,
      "grad_norm": 2.765091896057129,
      "kl": 0.112548828125,
      "learning_rate": 3.922374429223744e-07,
      "loss": 0.0045,
      "reward": 1.8410881161689758,
      "reward_std": 0.14643773436546326,
      "rewards/accuracy_reward": 0.864525556564331,
      "rewards/format_reward": 0.9765625,
      "step": 1331
    },
    {
      "completion_length": 86.4375,
      "epoch": 6.082191780821918,
      "grad_norm": 3.7591657638549805,
      "kl": 0.1103515625,
      "learning_rate": 3.917808219178082e-07,
      "loss": 0.0044,
      "reward": 1.7199653387069702,
      "reward_std": 0.1599731780588627,
      "rewards/accuracy_reward": 0.7355902194976807,
      "rewards/format_reward": 0.984375,
      "step": 1332
    },
    {
      "completion_length": 88.1015625,
      "epoch": 6.0867579908675795,
      "grad_norm": 2.042999267578125,
      "kl": 0.102783203125,
      "learning_rate": 3.91324200913242e-07,
      "loss": 0.0041,
      "reward": 1.8294270634651184,
      "reward_std": 0.09024603478610516,
      "rewards/accuracy_reward": 0.8372394740581512,
      "rewards/format_reward": 0.9921875,
      "step": 1333
    },
    {
      "completion_length": 96.703125,
      "epoch": 6.091324200913242,
      "grad_norm": 1.6703046560287476,
      "kl": 0.14990234375,
      "learning_rate": 3.908675799086758e-07,
      "loss": 0.006,
      "reward": 1.795138955116272,
      "reward_std": 0.09230764210224152,
      "rewards/accuracy_reward": 0.8029513657093048,
      "rewards/format_reward": 0.9921875,
      "step": 1334
    },
    {
      "completion_length": 84.5078125,
      "epoch": 6.095890410958904,
      "grad_norm": 1.742505669593811,
      "kl": 0.10205078125,
      "learning_rate": 3.9041095890410957e-07,
      "loss": 0.0041,
      "reward": 1.6895833611488342,
      "reward_std": 0.2131306305527687,
      "rewards/accuracy_reward": 0.7130208313465118,
      "rewards/format_reward": 0.9765625,
      "step": 1335
    },
    {
      "completion_length": 86.1484375,
      "epoch": 6.100456621004566,
      "grad_norm": 12.497916221618652,
      "kl": 0.13671875,
      "learning_rate": 3.8995433789954336e-07,
      "loss": 0.0055,
      "reward": 1.6923341751098633,
      "reward_std": 0.17974907904863358,
      "rewards/accuracy_reward": 0.7001466453075409,
      "rewards/format_reward": 0.9921875,
      "step": 1336
    },
    {
      "completion_length": 77.3359375,
      "epoch": 6.105022831050229,
      "grad_norm": 4.783447742462158,
      "kl": 0.1591796875,
      "learning_rate": 3.8949771689497714e-07,
      "loss": 0.0064,
      "reward": 1.6346808671951294,
      "reward_std": 0.1741974987089634,
      "rewards/accuracy_reward": 0.642493337392807,
      "rewards/format_reward": 0.9921875,
      "step": 1337
    },
    {
      "completion_length": 85.03125,
      "epoch": 6.109589041095891,
      "grad_norm": 3.4984941482543945,
      "kl": 0.140625,
      "learning_rate": 3.8904109589041093e-07,
      "loss": 0.0056,
      "reward": 1.6837969422340393,
      "reward_std": 0.16627999395132065,
      "rewards/accuracy_reward": 0.6916094422340393,
      "rewards/format_reward": 0.9921875,
      "step": 1338
    },
    {
      "completion_length": 95.484375,
      "epoch": 6.114155251141552,
      "grad_norm": 3.724400520324707,
      "kl": 0.084228515625,
      "learning_rate": 3.885844748858447e-07,
      "loss": 0.0034,
      "reward": 1.7506882548332214,
      "reward_std": 0.16492830961942673,
      "rewards/accuracy_reward": 0.7741256952285767,
      "rewards/format_reward": 0.9765625,
      "step": 1339
    },
    {
      "completion_length": 84.8359375,
      "epoch": 6.1187214611872145,
      "grad_norm": 3.958569049835205,
      "kl": 0.1142578125,
      "learning_rate": 3.881278538812785e-07,
      "loss": 0.0046,
      "reward": 1.8411458730697632,
      "reward_std": 0.12668996676802635,
      "rewards/accuracy_reward": 0.8489583134651184,
      "rewards/format_reward": 0.9921875,
      "step": 1340
    },
    {
      "completion_length": 64.8828125,
      "epoch": 6.123287671232877,
      "grad_norm": 7.010223388671875,
      "kl": 0.1708984375,
      "learning_rate": 3.8767123287671235e-07,
      "loss": 0.0068,
      "reward": 1.796571135520935,
      "reward_std": 0.147329680621624,
      "rewards/accuracy_reward": 0.8043836355209351,
      "rewards/format_reward": 0.9921875,
      "step": 1341
    },
    {
      "completion_length": 51.1953125,
      "epoch": 6.127853881278539,
      "grad_norm": 1.703456163406372,
      "kl": 0.18798828125,
      "learning_rate": 3.872146118721461e-07,
      "loss": 0.0075,
      "reward": 1.7231026887893677,
      "reward_std": 0.15178490430116653,
      "rewards/accuracy_reward": 0.7231026887893677,
      "rewards/format_reward": 1.0,
      "step": 1342
    },
    {
      "completion_length": 83.125,
      "epoch": 6.132420091324201,
      "grad_norm": 2.8786849975585938,
      "kl": 0.12744140625,
      "learning_rate": 3.8675799086757987e-07,
      "loss": 0.0051,
      "reward": 1.6835938096046448,
      "reward_std": 0.15046585351228714,
      "rewards/accuracy_reward": 0.6835937201976776,
      "rewards/format_reward": 1.0,
      "step": 1343
    },
    {
      "completion_length": 84.7109375,
      "epoch": 6.136986301369863,
      "grad_norm": 3.438633680343628,
      "kl": 0.1298828125,
      "learning_rate": 3.863013698630137e-07,
      "loss": 0.0052,
      "reward": 1.8140625357627869,
      "reward_std": 0.15467960387468338,
      "rewards/accuracy_reward": 0.8140624463558197,
      "rewards/format_reward": 1.0,
      "step": 1344
    },
    {
      "completion_length": 76.953125,
      "epoch": 6.141552511415525,
      "grad_norm": 5.209409236907959,
      "kl": 0.1064453125,
      "learning_rate": 3.858447488584475e-07,
      "loss": 0.0043,
      "reward": 1.6888335943222046,
      "reward_std": 0.1942080445587635,
      "rewards/accuracy_reward": 0.696646124124527,
      "rewards/format_reward": 0.9921875,
      "step": 1345
    },
    {
      "completion_length": 75.7578125,
      "epoch": 6.146118721461187,
      "grad_norm": 1.846853494644165,
      "kl": 0.115966796875,
      "learning_rate": 3.8538812785388123e-07,
      "loss": 0.0046,
      "reward": 1.7150331735610962,
      "reward_std": 0.19362305849790573,
      "rewards/accuracy_reward": 0.7306581139564514,
      "rewards/format_reward": 0.984375,
      "step": 1346
    },
    {
      "completion_length": 91.0390625,
      "epoch": 6.1506849315068495,
      "grad_norm": 1.0356731414794922,
      "kl": 0.11962890625,
      "learning_rate": 3.84931506849315e-07,
      "loss": 0.0048,
      "reward": 1.8451389074325562,
      "reward_std": 0.0736747458577156,
      "rewards/accuracy_reward": 0.852951318025589,
      "rewards/format_reward": 0.9921875,
      "step": 1347
    },
    {
      "completion_length": 69.703125,
      "epoch": 6.155251141552512,
      "grad_norm": 1.3840466737747192,
      "kl": 0.126953125,
      "learning_rate": 3.8447488584474886e-07,
      "loss": 0.0051,
      "reward": 1.8907551765441895,
      "reward_std": 0.06240340322256088,
      "rewards/accuracy_reward": 0.8907552063465118,
      "rewards/format_reward": 1.0,
      "step": 1348
    },
    {
      "completion_length": 90.1171875,
      "epoch": 6.159817351598173,
      "grad_norm": 1.7931628227233887,
      "kl": 0.091064453125,
      "learning_rate": 3.8401826484018264e-07,
      "loss": 0.0036,
      "reward": 1.792187511920929,
      "reward_std": 0.10482378304004669,
      "rewards/accuracy_reward": 0.7999999225139618,
      "rewards/format_reward": 0.9921875,
      "step": 1349
    },
    {
      "completion_length": 93.625,
      "epoch": 6.164383561643835,
      "grad_norm": 2.3507261276245117,
      "kl": 0.12060546875,
      "learning_rate": 3.835616438356164e-07,
      "loss": 0.0048,
      "reward": 1.7338955998420715,
      "reward_std": 0.16754615679383278,
      "rewards/accuracy_reward": 0.7573330104351044,
      "rewards/format_reward": 0.9765625,
      "step": 1350
    },
    {
      "completion_length": 62.859375,
      "epoch": 6.168949771689498,
      "grad_norm": 2.7926764488220215,
      "kl": 0.1259765625,
      "learning_rate": 3.831050228310502e-07,
      "loss": 0.005,
      "reward": 1.92203950881958,
      "reward_std": 0.05836756294593215,
      "rewards/accuracy_reward": 0.9220394492149353,
      "rewards/format_reward": 1.0,
      "step": 1351
    },
    {
      "completion_length": 72.953125,
      "epoch": 6.17351598173516,
      "grad_norm": 1.8602731227874756,
      "kl": 0.120849609375,
      "learning_rate": 3.82648401826484e-07,
      "loss": 0.0048,
      "reward": 1.8530598878860474,
      "reward_std": 0.07507604733109474,
      "rewards/accuracy_reward": 0.860872358083725,
      "rewards/format_reward": 0.9921875,
      "step": 1352
    },
    {
      "completion_length": 89.671875,
      "epoch": 6.178082191780822,
      "grad_norm": 1.9812475442886353,
      "kl": 0.107421875,
      "learning_rate": 3.821917808219178e-07,
      "loss": 0.0043,
      "reward": 1.8095996975898743,
      "reward_std": 0.07891392894089222,
      "rewards/accuracy_reward": 0.8095996379852295,
      "rewards/format_reward": 1.0,
      "step": 1353
    },
    {
      "completion_length": 84.40625,
      "epoch": 6.182648401826484,
      "grad_norm": 2.6219146251678467,
      "kl": 0.110107421875,
      "learning_rate": 3.817351598173516e-07,
      "loss": 0.0044,
      "reward": 1.7649739980697632,
      "reward_std": 0.13717181608080864,
      "rewards/accuracy_reward": 0.7805989384651184,
      "rewards/format_reward": 0.984375,
      "step": 1354
    },
    {
      "completion_length": 85.2265625,
      "epoch": 6.187214611872146,
      "grad_norm": 1.6997660398483276,
      "kl": 0.078857421875,
      "learning_rate": 3.8127853881278537e-07,
      "loss": 0.0031,
      "reward": 1.7958333492279053,
      "reward_std": 0.07365695387125015,
      "rewards/accuracy_reward": 0.7958332598209381,
      "rewards/format_reward": 1.0,
      "step": 1355
    },
    {
      "completion_length": 68.3828125,
      "epoch": 6.191780821917808,
      "grad_norm": 2.4636871814727783,
      "kl": 0.140380859375,
      "learning_rate": 3.8082191780821916e-07,
      "loss": 0.0056,
      "reward": 1.638918399810791,
      "reward_std": 0.1541249379515648,
      "rewards/accuracy_reward": 0.6467308402061462,
      "rewards/format_reward": 0.9921875,
      "step": 1356
    },
    {
      "completion_length": 96.0078125,
      "epoch": 6.19634703196347,
      "grad_norm": 2.066955804824829,
      "kl": 0.08056640625,
      "learning_rate": 3.8036529680365294e-07,
      "loss": 0.0032,
      "reward": 1.7703125476837158,
      "reward_std": 0.11048543080687523,
      "rewards/accuracy_reward": 0.7781248986721039,
      "rewards/format_reward": 0.9921875,
      "step": 1357
    },
    {
      "completion_length": 87.1640625,
      "epoch": 6.200913242009133,
      "grad_norm": 1.8385370969772339,
      "kl": 0.125,
      "learning_rate": 3.7990867579908673e-07,
      "loss": 0.005,
      "reward": 1.7934895753860474,
      "reward_std": 0.16230732202529907,
      "rewards/accuracy_reward": 0.809114545583725,
      "rewards/format_reward": 0.984375,
      "step": 1358
    },
    {
      "completion_length": 79.4140625,
      "epoch": 6.205479452054795,
      "grad_norm": 1.4888020753860474,
      "kl": 0.1396484375,
      "learning_rate": 3.794520547945205e-07,
      "loss": 0.0056,
      "reward": 1.8671875596046448,
      "reward_std": 0.08838834427297115,
      "rewards/accuracy_reward": 0.875,
      "rewards/format_reward": 0.9921875,
      "step": 1359
    },
    {
      "completion_length": 81.5859375,
      "epoch": 6.210045662100456,
      "grad_norm": 3.295711040496826,
      "kl": 0.159423828125,
      "learning_rate": 3.789954337899543e-07,
      "loss": 0.0064,
      "reward": 1.7882669568061829,
      "reward_std": 0.14686734601855278,
      "rewards/accuracy_reward": 0.7960793972015381,
      "rewards/format_reward": 0.9921875,
      "step": 1360
    },
    {
      "completion_length": 70.3828125,
      "epoch": 6.2146118721461185,
      "grad_norm": 4.181000709533691,
      "kl": 0.1220703125,
      "learning_rate": 3.7853881278538814e-07,
      "loss": 0.0049,
      "reward": 1.6776198744773865,
      "reward_std": 0.20163647830486298,
      "rewards/accuracy_reward": 0.6854324042797089,
      "rewards/format_reward": 0.9921875,
      "step": 1361
    },
    {
      "completion_length": 74.453125,
      "epoch": 6.219178082191781,
      "grad_norm": 2.4070448875427246,
      "kl": 0.123779296875,
      "learning_rate": 3.780821917808219e-07,
      "loss": 0.0049,
      "reward": 1.810937523841858,
      "reward_std": 0.1541428193449974,
      "rewards/accuracy_reward": 0.8265624046325684,
      "rewards/format_reward": 0.984375,
      "step": 1362
    },
    {
      "completion_length": 73.96875,
      "epoch": 6.223744292237443,
      "grad_norm": 4.166317939758301,
      "kl": 0.16796875,
      "learning_rate": 3.7762557077625567e-07,
      "loss": 0.0067,
      "reward": 1.6138020753860474,
      "reward_std": 0.25969336926937103,
      "rewards/accuracy_reward": 0.6294271051883698,
      "rewards/format_reward": 0.984375,
      "step": 1363
    },
    {
      "completion_length": 74.0859375,
      "epoch": 6.228310502283105,
      "grad_norm": 9.133631706237793,
      "kl": 0.107421875,
      "learning_rate": 3.771689497716895e-07,
      "loss": 0.0043,
      "reward": 1.8240530490875244,
      "reward_std": 0.12234986200928688,
      "rewards/accuracy_reward": 0.8318654298782349,
      "rewards/format_reward": 0.9921875,
      "step": 1364
    },
    {
      "completion_length": 86.6328125,
      "epoch": 6.232876712328767,
      "grad_norm": 2.3225855827331543,
      "kl": 0.19580078125,
      "learning_rate": 3.767123287671233e-07,
      "loss": 0.0078,
      "reward": 1.7432359457015991,
      "reward_std": 0.20243436098098755,
      "rewards/accuracy_reward": 0.7666734158992767,
      "rewards/format_reward": 0.9765625,
      "step": 1365
    },
    {
      "completion_length": 75.84375,
      "epoch": 6.237442922374429,
      "grad_norm": 4.120067596435547,
      "kl": 0.1038818359375,
      "learning_rate": 3.7625570776255703e-07,
      "loss": 0.0042,
      "reward": 1.7382813096046448,
      "reward_std": 0.1538807973265648,
      "rewards/accuracy_reward": 0.74609375,
      "rewards/format_reward": 0.9921875,
      "step": 1366
    },
    {
      "completion_length": 71.75,
      "epoch": 6.242009132420091,
      "grad_norm": 2.138282537460327,
      "kl": 0.1630859375,
      "learning_rate": 3.757990867579908e-07,
      "loss": 0.0065,
      "reward": 1.7145547270774841,
      "reward_std": 0.18507731705904007,
      "rewards/accuracy_reward": 0.722367137670517,
      "rewards/format_reward": 0.9921875,
      "step": 1367
    },
    {
      "completion_length": 66.390625,
      "epoch": 6.2465753424657535,
      "grad_norm": 2.4669651985168457,
      "kl": 0.20654296875,
      "learning_rate": 3.7534246575342466e-07,
      "loss": 0.0083,
      "reward": 1.818272590637207,
      "reward_std": 0.11263852939009666,
      "rewards/accuracy_reward": 0.8260850310325623,
      "rewards/format_reward": 0.9921875,
      "step": 1368
    },
    {
      "completion_length": 100.6875,
      "epoch": 6.251141552511416,
      "grad_norm": 1.6667773723602295,
      "kl": 0.113037109375,
      "learning_rate": 3.7488584474885844e-07,
      "loss": 0.0045,
      "reward": 1.682812511920929,
      "reward_std": 0.14389308914542198,
      "rewards/accuracy_reward": 0.7062499523162842,
      "rewards/format_reward": 0.9765625,
      "step": 1369
    },
    {
      "completion_length": 58.890625,
      "epoch": 6.255707762557078,
      "grad_norm": 4.199071884155273,
      "kl": 0.122314453125,
      "learning_rate": 3.744292237442922e-07,
      "loss": 0.0049,
      "reward": 1.8065972328186035,
      "reward_std": 0.10062464885413647,
      "rewards/accuracy_reward": 0.8065972328186035,
      "rewards/format_reward": 1.0,
      "step": 1370
    },
    {
      "completion_length": 86.0625,
      "epoch": 6.260273972602739,
      "grad_norm": 4.296463966369629,
      "kl": 0.129150390625,
      "learning_rate": 3.73972602739726e-07,
      "loss": 0.0052,
      "reward": 1.7294270992279053,
      "reward_std": 0.1873747929930687,
      "rewards/accuracy_reward": 0.7372395694255829,
      "rewards/format_reward": 0.9921875,
      "step": 1371
    },
    {
      "completion_length": 70.609375,
      "epoch": 6.264840182648402,
      "grad_norm": 2.9196784496307373,
      "kl": 0.138671875,
      "learning_rate": 3.735159817351598e-07,
      "loss": 0.0055,
      "reward": 1.7803341746330261,
      "reward_std": 0.09776799008250237,
      "rewards/accuracy_reward": 0.7803341746330261,
      "rewards/format_reward": 1.0,
      "step": 1372
    },
    {
      "completion_length": 89.875,
      "epoch": 6.269406392694064,
      "grad_norm": 2.074902296066284,
      "kl": 0.11474609375,
      "learning_rate": 3.730593607305936e-07,
      "loss": 0.0046,
      "reward": 1.7255208492279053,
      "reward_std": 0.18392051756381989,
      "rewards/accuracy_reward": 0.7411458194255829,
      "rewards/format_reward": 0.984375,
      "step": 1373
    },
    {
      "completion_length": 75.140625,
      "epoch": 6.273972602739726,
      "grad_norm": 4.484436511993408,
      "kl": 0.156005859375,
      "learning_rate": 3.726027397260274e-07,
      "loss": 0.0062,
      "reward": 1.7850632667541504,
      "reward_std": 0.14479004591703415,
      "rewards/accuracy_reward": 0.792875736951828,
      "rewards/format_reward": 0.9921875,
      "step": 1374
    },
    {
      "completion_length": 92.6640625,
      "epoch": 6.2785388127853885,
      "grad_norm": 3.3195595741271973,
      "kl": 0.1070556640625,
      "learning_rate": 3.7214611872146117e-07,
      "loss": 0.0043,
      "reward": 1.80859375,
      "reward_std": 0.1387247210368514,
      "rewards/accuracy_reward": 0.8164062201976776,
      "rewards/format_reward": 0.9921875,
      "step": 1375
    },
    {
      "completion_length": 87.1875,
      "epoch": 6.28310502283105,
      "grad_norm": 4.265213966369629,
      "kl": 0.153564453125,
      "learning_rate": 3.7168949771689495e-07,
      "loss": 0.0061,
      "reward": 1.7935296893119812,
      "reward_std": 0.1823706291615963,
      "rewards/accuracy_reward": 0.8325920403003693,
      "rewards/format_reward": 0.9609375,
      "step": 1376
    },
    {
      "completion_length": 84.71875,
      "epoch": 6.287671232876712,
      "grad_norm": 1.7530524730682373,
      "kl": 0.115234375,
      "learning_rate": 3.7123287671232874e-07,
      "loss": 0.0046,
      "reward": 1.7466146349906921,
      "reward_std": 0.1300597358494997,
      "rewards/accuracy_reward": 0.7622395753860474,
      "rewards/format_reward": 0.984375,
      "step": 1377
    },
    {
      "completion_length": 92.078125,
      "epoch": 6.292237442922374,
      "grad_norm": 2.0412070751190186,
      "kl": 0.1259765625,
      "learning_rate": 3.7077625570776253e-07,
      "loss": 0.005,
      "reward": 1.8295753002166748,
      "reward_std": 0.17172623425722122,
      "rewards/accuracy_reward": 0.8608251810073853,
      "rewards/format_reward": 0.96875,
      "step": 1378
    },
    {
      "completion_length": 53.4375,
      "epoch": 6.296803652968037,
      "grad_norm": 2.757150888442993,
      "kl": 0.140625,
      "learning_rate": 3.703196347031963e-07,
      "loss": 0.0056,
      "reward": 1.701302170753479,
      "reward_std": 0.29356005787849426,
      "rewards/accuracy_reward": 0.7403645813465118,
      "rewards/format_reward": 0.9609375,
      "step": 1379
    },
    {
      "completion_length": 98.2109375,
      "epoch": 6.301369863013699,
      "grad_norm": 2.7139008045196533,
      "kl": 0.07958984375,
      "learning_rate": 3.698630136986301e-07,
      "loss": 0.0032,
      "reward": 1.783593773841858,
      "reward_std": 0.09553372114896774,
      "rewards/accuracy_reward": 0.7835937440395355,
      "rewards/format_reward": 1.0,
      "step": 1380
    },
    {
      "completion_length": 90.40625,
      "epoch": 6.30593607305936,
      "grad_norm": 3.6404201984405518,
      "kl": 0.105712890625,
      "learning_rate": 3.6940639269406394e-07,
      "loss": 0.0042,
      "reward": 1.746897280216217,
      "reward_std": 0.18012897670269012,
      "rewards/accuracy_reward": 0.7625222206115723,
      "rewards/format_reward": 0.984375,
      "step": 1381
    },
    {
      "completion_length": 70.71875,
      "epoch": 6.310502283105023,
      "grad_norm": 2.44050931930542,
      "kl": 0.1953125,
      "learning_rate": 3.689497716894977e-07,
      "loss": 0.0078,
      "reward": 1.7049107551574707,
      "reward_std": 0.14146586507558823,
      "rewards/accuracy_reward": 0.7127232253551483,
      "rewards/format_reward": 0.9921875,
      "step": 1382
    },
    {
      "completion_length": 84.0859375,
      "epoch": 6.315068493150685,
      "grad_norm": 2.908999443054199,
      "kl": 0.129150390625,
      "learning_rate": 3.6849315068493147e-07,
      "loss": 0.0052,
      "reward": 1.866501271724701,
      "reward_std": 0.14921535179018974,
      "rewards/accuracy_reward": 0.8743137121200562,
      "rewards/format_reward": 0.9921875,
      "step": 1383
    },
    {
      "completion_length": 66.9609375,
      "epoch": 6.319634703196347,
      "grad_norm": 7.363396644592285,
      "kl": 0.155517578125,
      "learning_rate": 3.680365296803653e-07,
      "loss": 0.0062,
      "reward": 1.73046875,
      "reward_std": 0.14824316650629044,
      "rewards/accuracy_reward": 0.7304687201976776,
      "rewards/format_reward": 1.0,
      "step": 1384
    },
    {
      "completion_length": 63.6875,
      "epoch": 6.324200913242009,
      "grad_norm": 2.700493097305298,
      "kl": 0.153564453125,
      "learning_rate": 3.675799086757991e-07,
      "loss": 0.0061,
      "reward": 1.7351128458976746,
      "reward_std": 0.1287803202867508,
      "rewards/accuracy_reward": 0.7429253458976746,
      "rewards/format_reward": 0.9921875,
      "step": 1385
    },
    {
      "completion_length": 77.515625,
      "epoch": 6.328767123287671,
      "grad_norm": 4.386509895324707,
      "kl": 0.140869140625,
      "learning_rate": 3.6712328767123283e-07,
      "loss": 0.0056,
      "reward": 1.6953125596046448,
      "reward_std": 0.20576493442058563,
      "rewards/accuracy_reward": 0.7187499701976776,
      "rewards/format_reward": 0.9765625,
      "step": 1386
    },
    {
      "completion_length": 69.765625,
      "epoch": 6.333333333333333,
      "grad_norm": 2.0167596340179443,
      "kl": 0.117431640625,
      "learning_rate": 3.666666666666666e-07,
      "loss": 0.0047,
      "reward": 1.7052951455116272,
      "reward_std": 0.12893803417682648,
      "rewards/accuracy_reward": 0.7052951157093048,
      "rewards/format_reward": 1.0,
      "step": 1387
    },
    {
      "completion_length": 90.6875,
      "epoch": 6.337899543378995,
      "grad_norm": 4.069055557250977,
      "kl": 0.095458984375,
      "learning_rate": 3.6621004566210046e-07,
      "loss": 0.0038,
      "reward": 1.753125011920929,
      "reward_std": 0.1462521031498909,
      "rewards/accuracy_reward": 0.7531249523162842,
      "rewards/format_reward": 1.0,
      "step": 1388
    },
    {
      "completion_length": 85.3671875,
      "epoch": 6.342465753424658,
      "grad_norm": 1.5775866508483887,
      "kl": 0.1103515625,
      "learning_rate": 3.6575342465753424e-07,
      "loss": 0.0044,
      "reward": 1.7893466353416443,
      "reward_std": 0.13156647235155106,
      "rewards/accuracy_reward": 0.8127840459346771,
      "rewards/format_reward": 0.9765625,
      "step": 1389
    },
    {
      "completion_length": 71.953125,
      "epoch": 6.34703196347032,
      "grad_norm": 4.919250965118408,
      "kl": 0.16259765625,
      "learning_rate": 3.65296803652968e-07,
      "loss": 0.0065,
      "reward": 1.816933274269104,
      "reward_std": 0.18086620420217514,
      "rewards/accuracy_reward": 0.816933274269104,
      "rewards/format_reward": 1.0,
      "step": 1390
    },
    {
      "completion_length": 75.359375,
      "epoch": 6.351598173515982,
      "grad_norm": 1.6152794361114502,
      "kl": 0.12841796875,
      "learning_rate": 3.648401826484018e-07,
      "loss": 0.0051,
      "reward": 1.8754695653915405,
      "reward_std": 0.05770116671919823,
      "rewards/accuracy_reward": 0.8754695355892181,
      "rewards/format_reward": 1.0,
      "step": 1391
    },
    {
      "completion_length": 97.0625,
      "epoch": 6.3561643835616435,
      "grad_norm": 3.0562081336975098,
      "kl": 0.0654296875,
      "learning_rate": 3.643835616438356e-07,
      "loss": 0.0026,
      "reward": 1.8640625476837158,
      "reward_std": 0.11048543266952038,
      "rewards/accuracy_reward": 0.8874999284744263,
      "rewards/format_reward": 0.9765625,
      "step": 1392
    },
    {
      "completion_length": 91.1015625,
      "epoch": 6.360730593607306,
      "grad_norm": 11.379284858703613,
      "kl": 0.123046875,
      "learning_rate": 3.639269406392694e-07,
      "loss": 0.0049,
      "reward": 1.7335938215255737,
      "reward_std": 0.13041038066148758,
      "rewards/accuracy_reward": 0.741406261920929,
      "rewards/format_reward": 0.9921875,
      "step": 1393
    },
    {
      "completion_length": 88.609375,
      "epoch": 6.365296803652968,
      "grad_norm": 2.2893126010894775,
      "kl": 0.120849609375,
      "learning_rate": 3.634703196347032e-07,
      "loss": 0.0048,
      "reward": 1.7554688453674316,
      "reward_std": 0.14954836666584015,
      "rewards/accuracy_reward": 0.7710936963558197,
      "rewards/format_reward": 0.984375,
      "step": 1394
    },
    {
      "completion_length": 86.0390625,
      "epoch": 6.36986301369863,
      "grad_norm": 2.2495548725128174,
      "kl": 0.10888671875,
      "learning_rate": 3.6301369863013697e-07,
      "loss": 0.0043,
      "reward": 1.8598958253860474,
      "reward_std": 0.10234630480408669,
      "rewards/accuracy_reward": 0.8677082657814026,
      "rewards/format_reward": 0.9921875,
      "step": 1395
    },
    {
      "completion_length": 75.4609375,
      "epoch": 6.3744292237442925,
      "grad_norm": 1.356154203414917,
      "kl": 0.095947265625,
      "learning_rate": 3.6255707762557075e-07,
      "loss": 0.0038,
      "reward": 1.6979167461395264,
      "reward_std": 0.11066011898219585,
      "rewards/accuracy_reward": 0.6979166567325592,
      "rewards/format_reward": 1.0,
      "step": 1396
    },
    {
      "completion_length": 76.453125,
      "epoch": 6.378995433789954,
      "grad_norm": 2.3890748023986816,
      "kl": 0.202880859375,
      "learning_rate": 3.6210045662100454e-07,
      "loss": 0.0081,
      "reward": 1.8653646111488342,
      "reward_std": 0.15285574202425778,
      "rewards/accuracy_reward": 0.8966145813465118,
      "rewards/format_reward": 0.96875,
      "step": 1397
    },
    {
      "completion_length": 66.3359375,
      "epoch": 6.383561643835616,
      "grad_norm": 1.8695800304412842,
      "kl": 0.1298828125,
      "learning_rate": 3.6164383561643833e-07,
      "loss": 0.0052,
      "reward": 1.8802322745323181,
      "reward_std": 0.046260682400316,
      "rewards/accuracy_reward": 0.8802323341369629,
      "rewards/format_reward": 1.0,
      "step": 1398
    },
    {
      "completion_length": 89.1796875,
      "epoch": 6.3881278538812785,
      "grad_norm": 1.6961324214935303,
      "kl": 0.103271484375,
      "learning_rate": 3.611872146118721e-07,
      "loss": 0.0041,
      "reward": 1.870312511920929,
      "reward_std": 0.10539799928665161,
      "rewards/accuracy_reward": 0.8781249523162842,
      "rewards/format_reward": 0.9921875,
      "step": 1399
    },
    {
      "completion_length": 90.0859375,
      "epoch": 6.392694063926941,
      "grad_norm": 8.425874710083008,
      "kl": 0.180419921875,
      "learning_rate": 3.607305936073059e-07,
      "loss": 0.0072,
      "reward": 1.6842634081840515,
      "reward_std": 0.16779197752475739,
      "rewards/accuracy_reward": 0.6842633485794067,
      "rewards/format_reward": 1.0,
      "step": 1400
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 2190,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}