{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.38095238095238093,
  "eval_steps": 10,
  "global_step": 120,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "clip_ratio": 0.0,
      "completion_length": 973.6428833007812,
      "epoch": 0.0031746031746031746,
      "grad_norm": 2.777648063498741,
      "kl": 0.0,
      "learning_rate": 1e-07,
      "loss": 0.1021,
      "reward": 0.9749529957771301,
      "reward_std": 0.8150795698165894,
      "rewards/": 3.446193218231201,
      "rewards/math_compute_score": 0.3571428656578064,
      "step": 1
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 708.4285888671875,
      "epoch": 0.006349206349206349,
      "grad_norm": 0.8174320460798857,
      "kl": 0.0,
      "learning_rate": 2e-07,
      "loss": 0.0372,
      "reward": 1.2198399305343628,
      "reward_std": 0.6601618528366089,
      "rewards/": 3.9563426971435547,
      "rewards/math_compute_score": 0.535714328289032,
      "step": 2
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 891.607177734375,
      "epoch": 0.009523809523809525,
      "grad_norm": 4.681867395183935,
      "kl": 0.00020313262939453125,
      "learning_rate": 3e-07,
      "loss": -0.0069,
      "reward": 0.8394444584846497,
      "reward_std": 0.7244595885276794,
      "rewards/": 3.268650770187378,
      "rewards/math_compute_score": 0.2321428656578064,
      "step": 3
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 672.0535888671875,
      "epoch": 0.012698412698412698,
      "grad_norm": 1.9071109509944568,
      "kl": 0.00019073486328125,
      "learning_rate": 4e-07,
      "loss": 0.1393,
      "reward": 0.837346613407135,
      "reward_std": 0.7451086640357971,
      "rewards/": 2.186732769012451,
      "rewards/math_compute_score": 0.5,
      "step": 4
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 843.0178833007812,
      "epoch": 0.015873015873015872,
      "grad_norm": 3.3988529526704374,
      "kl": 0.00021457672119140625,
      "learning_rate": 5e-07,
      "loss": -0.0491,
      "reward": 1.047133207321167,
      "reward_std": 0.6228165626525879,
      "rewards/": 3.7356655597686768,
      "rewards/math_compute_score": 0.3750000298023224,
      "step": 5
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 953.6785888671875,
      "epoch": 0.01904761904761905,
      "grad_norm": 2.637160712333027,
      "kl": 0.00018310546875,
      "learning_rate": 6e-07,
      "loss": 0.1669,
      "reward": 1.0524885654449463,
      "reward_std": 0.8631003499031067,
      "rewards/": 3.9767284393310547,
      "rewards/math_compute_score": 0.3214285969734192,
      "step": 6
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 767.0357666015625,
      "epoch": 0.022222222222222223,
      "grad_norm": 4.305675207109899,
      "kl": 0.0002536773681640625,
      "learning_rate": 7e-07,
      "loss": 0.2786,
      "reward": 1.2252023220062256,
      "reward_std": 0.6430084109306335,
      "rewards/": 3.6974401473999023,
      "rewards/math_compute_score": 0.6071428656578064,
      "step": 7
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 639.1785888671875,
      "epoch": 0.025396825396825397,
      "grad_norm": 3.7743752869617477,
      "kl": 0.000499725341796875,
      "learning_rate": 8e-07,
      "loss": 0.1966,
      "reward": 1.1576731204986572,
      "reward_std": 0.6582887172698975,
      "rewards/": 3.788365125656128,
      "rewards/math_compute_score": 0.5,
      "step": 8
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 806.607177734375,
      "epoch": 0.02857142857142857,
      "grad_norm": 2.64690799480574,
      "kl": 0.000507354736328125,
      "learning_rate": 9e-07,
      "loss": 0.1066,
      "reward": 0.9516462087631226,
      "reward_std": 0.6632688045501709,
      "rewards/": 3.1153740882873535,
      "rewards/math_compute_score": 0.4107142984867096,
      "step": 9
    },
    {
      "epoch": 0.031746031746031744,
      "grad_norm": 2.1254890231202204,
      "learning_rate": 1e-06,
      "loss": 0.0976,
      "step": 10
    },
    {
      "epoch": 0.031746031746031744,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 881.2059733072916,
      "eval_kl": 0.0006243387858072916,
      "eval_loss": 0.10283354669809341,
      "eval_reward": 1.3367964426676433,
      "eval_reward_std": 0.7034762501716614,
      "eval_rewards/": 4.350648641586304,
      "eval_rewards/math_compute_score": 0.5833333730697632,
      "eval_runtime": 110.5471,
      "eval_samples_per_second": 0.181,
      "eval_steps_per_second": 0.009,
      "step": 10
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 852.982177734375,
      "epoch": 0.03492063492063492,
      "grad_norm": 0.9239839479287824,
      "kl": 0.0006341934204101562,
      "learning_rate": 9.99973476170006e-07,
      "loss": 0.0414,
      "reward": 1.064503788948059,
      "reward_std": 0.7399034202098846,
      "rewards/": 3.608232855796814,
      "rewards/math_compute_score": 0.4285714626312256,
      "step": 11
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 678.5892944335938,
      "epoch": 0.0380952380952381,
      "grad_norm": 1.877318279677736,
      "kl": 0.000621795654296875,
      "learning_rate": 9.998939074940787e-07,
      "loss": 0.1241,
      "reward": 1.0163637399673462,
      "reward_std": 0.5952743887901306,
      "rewards/": 3.224675416946411,
      "rewards/math_compute_score": 0.4642857313156128,
      "step": 12
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 605.7678833007812,
      "epoch": 0.04126984126984127,
      "grad_norm": 2.4604298210996953,
      "kl": 0.00101470947265625,
      "learning_rate": 9.997613024140818e-07,
      "loss": 0.0281,
      "reward": 1.1692917346954346,
      "reward_std": 0.7951300740242004,
      "rewards/": 3.989314556121826,
      "rewards/math_compute_score": 0.4642857313156128,
      "step": 13
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 706.5714721679688,
      "epoch": 0.044444444444444446,
      "grad_norm": 1.7791723044352024,
      "kl": 0.00089263916015625,
      "learning_rate": 9.995756749987941e-07,
      "loss": 0.1917,
      "reward": 1.2231166362762451,
      "reward_std": 0.5831072926521301,
      "rewards/": 3.829869031906128,
      "rewards/math_compute_score": 0.5714285969734192,
      "step": 14
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 752.1964721679688,
      "epoch": 0.047619047619047616,
      "grad_norm": 2.2482045589194892,
      "kl": 0.0009918212890625,
      "learning_rate": 9.993370449424152e-07,
      "loss": 0.0832,
      "reward": 1.3100411891937256,
      "reward_std": 0.7690545320510864,
      "rewards/": 4.335920333862305,
      "rewards/math_compute_score": 0.5535714626312256,
      "step": 15
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 766.5535888671875,
      "epoch": 0.050793650793650794,
      "grad_norm": 3.1628124598253606,
      "kl": 0.000843048095703125,
      "learning_rate": 9.990454375624776e-07,
      "loss": 0.38,
      "reward": 1.3946969509124756,
      "reward_std": 0.43135303258895874,
      "rewards/": 4.687770366668701,
      "rewards/math_compute_score": 0.5714285969734192,
      "step": 16
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 838.9285888671875,
      "epoch": 0.05396825396825397,
      "grad_norm": 0.8809772166643767,
      "kl": 0.0009765625,
      "learning_rate": 9.987008837971594e-07,
      "loss": 0.2734,
      "reward": 1.2452915906906128,
      "reward_std": 0.6284029483795166,
      "rewards/": 4.369315147399902,
      "rewards/math_compute_score": 0.4642857313156128,
      "step": 17
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 623.8214721679688,
      "epoch": 0.05714285714285714,
      "grad_norm": 1.037602508755578,
      "kl": 0.00156402587890625,
      "learning_rate": 9.98303420202003e-07,
      "loss": 0.1688,
      "reward": 1.3429791927337646,
      "reward_std": 0.6546451449394226,
      "rewards/": 4.357753276824951,
      "rewards/math_compute_score": 0.5892857313156128,
      "step": 18
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 785.5535888671875,
      "epoch": 0.06031746031746032,
      "grad_norm": 0.9507117942449174,
      "kl": 0.0013580322265625,
      "learning_rate": 9.978530889460349e-07,
      "loss": 0.0918,
      "reward": 1.3252233266830444,
      "reward_std": 0.42738255858421326,
      "rewards/": 3.9832592010498047,
      "rewards/math_compute_score": 0.660714328289032,
      "step": 19
    },
    {
      "epoch": 0.06349206349206349,
      "grad_norm": 1.0293500184503375,
      "learning_rate": 9.973499378072946e-07,
      "loss": 0.0123,
      "step": 20
    },
    {
      "epoch": 0.06349206349206349,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 773.2273966471354,
      "eval_kl": 0.004353841145833333,
      "eval_loss": 0.17447665333747864,
      "eval_reward": 1.47151780128479,
      "eval_reward_std": 0.5722967584927877,
      "eval_rewards/": 4.524255673090617,
      "eval_rewards/math_compute_score": 0.7083333730697632,
      "eval_runtime": 105.3036,
      "eval_samples_per_second": 0.19,
      "eval_steps_per_second": 0.009,
      "step": 20
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 580.1339721679688,
      "epoch": 0.06666666666666667,
      "grad_norm": 1.9887336347273274,
      "kl": 0.00174713134765625,
      "learning_rate": 9.967940201677625e-07,
      "loss": 0.2513,
      "reward": 1.5847508311271667,
      "reward_std": 0.5660464465618134,
      "rewards/": 4.9237542152404785,
      "rewards/math_compute_score": 0.7500000298023224,
      "step": 21
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 699.7142944335938,
      "epoch": 0.06984126984126984,
      "grad_norm": 0.8687096793533198,
      "kl": 0.000827789306640625,
      "learning_rate": 9.96185395007699e-07,
      "loss": 0.1428,
      "reward": 1.5009138584136963,
      "reward_std": 0.5467565655708313,
      "rewards/": 4.3617119789123535,
      "rewards/math_compute_score": 0.785714328289032,
      "step": 22
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 662.9464721679688,
      "epoch": 0.07301587301587302,
      "grad_norm": 1.281704967643667,
      "kl": 0.0009765625,
      "learning_rate": 9.95524126899385e-07,
      "loss": 0.2754,
      "reward": 1.4940431118011475,
      "reward_std": 0.49719491600990295,
      "rewards/": 4.684500694274902,
      "rewards/math_compute_score": 0.6964285969734192,
      "step": 23
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.1785888671875,
      "epoch": 0.0761904761904762,
      "grad_norm": 2.6268302071766283,
      "kl": 0.00179290771484375,
      "learning_rate": 9.94810286000272e-07,
      "loss": 0.1202,
      "reward": 1.4679704904556274,
      "reward_std": 0.5678864121437073,
      "rewards/": 3.9827098846435547,
      "rewards/math_compute_score": 0.8392857313156128,
      "step": 24
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 816.5178833007812,
      "epoch": 0.07936507936507936,
      "grad_norm": 1.6260567552924308,
      "kl": 0.00225830078125,
      "learning_rate": 9.940439480455385e-07,
      "loss": 0.1839,
      "reward": 1.5339845418930054,
      "reward_std": 0.5200893878936768,
      "rewards/": 5.027064800262451,
      "rewards/math_compute_score": 0.660714328289032,
      "step": 25
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 586.8035888671875,
      "epoch": 0.08253968253968254,
      "grad_norm": 3.176548889137236,
      "kl": 0.0013885498046875,
      "learning_rate": 9.932251943400553e-07,
      "loss": 0.0802,
      "reward": 1.6504465341567993,
      "reward_std": 0.38086333870887756,
      "rewards/": 4.895089626312256,
      "rewards/math_compute_score": 0.8392857313156128,
      "step": 26
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 666.607177734375,
      "epoch": 0.08571428571428572,
      "grad_norm": 2.5123015337902737,
      "kl": 0.00098419189453125,
      "learning_rate": 9.923541117497585e-07,
      "loss": 0.0904,
      "reward": 1.4960030317306519,
      "reward_std": 0.601716160774231,
      "rewards/": 4.694301128387451,
      "rewards/math_compute_score": 0.6964285969734192,
      "step": 27
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 636.4285888671875,
      "epoch": 0.08888888888888889,
      "grad_norm": 1.2874582194461706,
      "kl": 0.00299072265625,
      "learning_rate": 9.914307926924344e-07,
      "loss": 0.1484,
      "reward": 1.4589931964874268,
      "reward_std": 0.6707652807235718,
      "rewards/": 4.223536968231201,
      "rewards/math_compute_score": 0.7678571939468384,
      "step": 28
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 690.5892944335938,
      "epoch": 0.09206349206349207,
      "grad_norm": 2.693607802209877,
      "kl": 0.0023345947265625,
      "learning_rate": 9.904553351279136e-07,
      "loss": 0.1137,
      "reward": 1.4248205423355103,
      "reward_std": 0.6116536259651184,
      "rewards/": 4.409816265106201,
      "rewards/math_compute_score": 0.6785714626312256,
      "step": 29
    },
    {
      "epoch": 0.09523809523809523,
      "grad_norm": 1.5994282855822706,
      "learning_rate": 9.894278425476788e-07,
      "loss": 0.1336,
      "step": 30
    },
    {
      "epoch": 0.09523809523809523,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 845.4524129231771,
      "eval_kl": 0.009012858072916666,
      "eval_loss": 0.13642674684524536,
      "eval_reward": 1.517225941022237,
      "eval_reward_std": 0.5625964005788168,
      "eval_rewards/": 4.728986581166585,
      "eval_rewards/math_compute_score": 0.7142857511838278,
      "eval_runtime": 106.4531,
      "eval_samples_per_second": 0.188,
      "eval_steps_per_second": 0.009,
      "step": 30
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 710.9107360839844,
      "epoch": 0.09841269841269841,
      "grad_norm": 0.9974627094215855,
      "kl": 0.00218963623046875,
      "learning_rate": 9.88348423963884e-07,
      "loss": 0.2872,
      "reward": 1.5053049325942993,
      "reward_std": 0.5521042495965958,
      "rewards/": 4.9550957679748535,
      "rewards/math_compute_score": 0.6428571939468384,
      "step": 31
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 708.1250610351562,
      "epoch": 0.10158730158730159,
      "grad_norm": 2.7592195903517727,
      "kl": 0.0031280517578125,
      "learning_rate": 9.872171938977893e-07,
      "loss": 0.2145,
      "reward": 1.5618199110031128,
      "reward_std": 0.49139147996902466,
      "rewards/": 4.451956748962402,
      "rewards/math_compute_score": 0.8392857313156128,
      "step": 32
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 653.625,
      "epoch": 0.10476190476190476,
      "grad_norm": 1.4899641756718722,
      "kl": 0.0025787353515625,
      "learning_rate": 9.860342723676104e-07,
      "loss": 0.2142,
      "reward": 1.5433281660079956,
      "reward_std": 0.5415727496147156,
      "rewards/": 4.502354621887207,
      "rewards/math_compute_score": 0.8035714626312256,
      "step": 33
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 646.9464721679688,
      "epoch": 0.10793650793650794,
      "grad_norm": 0.7771150726203957,
      "kl": 0.002044677734375,
      "learning_rate": 9.847997848757854e-07,
      "loss": 0.1302,
      "reward": 1.4443954229354858,
      "reward_std": 0.5867473483085632,
      "rewards/": 4.293405055999756,
      "rewards/math_compute_score": 0.7321428656578064,
      "step": 34
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 601.8392944335938,
      "epoch": 0.1111111111111111,
      "grad_norm": 1.8258651083690958,
      "kl": 0.0035247802734375,
      "learning_rate": 9.835138623956602e-07,
      "loss": -0.0484,
      "reward": 1.6701312065124512,
      "reward_std": 0.38416627049446106,
      "rewards/": 4.493513107299805,
      "rewards/math_compute_score": 0.9642857313156128,
      "step": 35
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 633.232177734375,
      "epoch": 0.11428571428571428,
      "grad_norm": 1.9714220532856719,
      "kl": 0.0037384033203125,
      "learning_rate": 9.821766413575914e-07,
      "loss": 0.057,
      "reward": 1.6821569204330444,
      "reward_std": 0.4740881323814392,
      "rewards/": 5.553641319274902,
      "rewards/math_compute_score": 0.7142857313156128,
      "step": 36
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 620.75,
      "epoch": 0.11746031746031746,
      "grad_norm": 1.6010557178740101,
      "kl": 0.00238037109375,
      "learning_rate": 9.80788263634473e-07,
      "loss": 0.0491,
      "reward": 1.4776437282562256,
      "reward_std": 0.514848530292511,
      "rewards/": 4.459647178649902,
      "rewards/math_compute_score": 0.7321428656578064,
      "step": 37
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 663.375,
      "epoch": 0.12063492063492064,
      "grad_norm": 3.177882116146041,
      "kl": 0.0025634765625,
      "learning_rate": 9.793488765266838e-07,
      "loss": 0.2726,
      "reward": 1.4468194246292114,
      "reward_std": 0.5143493413925171,
      "rewards/": 4.519810676574707,
      "rewards/math_compute_score": 0.6785714626312256,
      "step": 38
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 662.8035888671875,
      "epoch": 0.12380952380952381,
      "grad_norm": 0.9741870004361116,
      "kl": 0.00445556640625,
      "learning_rate": 9.778586327464597e-07,
      "loss": 0.1164,
      "reward": 1.3739641904830933,
      "reward_std": 0.5605769157409668,
      "rewards/": 4.012678146362305,
      "rewards/math_compute_score": 0.7142857313156128,
      "step": 39
    },
    {
      "epoch": 0.12698412698412698,
      "grad_norm": 1.1959685615621873,
      "learning_rate": 9.763176904016913e-07,
      "loss": 0.1094,
      "step": 40
    },
    {
      "epoch": 0.12698412698412698,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 792.1571655273438,
      "eval_kl": 0.025288899739583332,
      "eval_loss": 0.11785109341144562,
      "eval_reward": 1.5712554057439168,
      "eval_reward_std": 0.5905763109525045,
      "eval_rewards/": 4.7134199142456055,
      "eval_rewards/math_compute_score": 0.785714328289032,
      "eval_runtime": 105.8496,
      "eval_samples_per_second": 0.189,
      "eval_steps_per_second": 0.009,
      "step": 40
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 713.8839721679688,
      "epoch": 0.13015873015873017,
      "grad_norm": 0.8987806956191959,
      "kl": 0.002960205078125,
      "learning_rate": 9.747262129791495e-07,
      "loss": 0.0717,
      "reward": 1.5012034177780151,
      "reward_std": 0.5221634805202484,
      "rewards/": 4.827445030212402,
      "rewards/math_compute_score": 0.6696428656578064,
      "step": 41
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 840.2857666015625,
      "epoch": 0.13333333333333333,
      "grad_norm": 1.08064899340382,
      "kl": 0.00262451171875,
      "learning_rate": 9.730843693271413e-07,
      "loss": 0.1193,
      "reward": 1.4552539587020874,
      "reward_std": 0.6179671287536621,
      "rewards/": 4.776269435882568,
      "rewards/math_compute_score": 0.625,
      "step": 42
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 641.0357666015625,
      "epoch": 0.1365079365079365,
      "grad_norm": 0.766263291574053,
      "kl": 0.0024871826171875,
      "learning_rate": 9.713923336375936e-07,
      "loss": 0.1407,
      "reward": 1.6835448741912842,
      "reward_std": 0.5881065726280212,
      "rewards/": 5.703439235687256,
      "rewards/math_compute_score": 0.6785714626312256,
      "step": 43
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 638.7857666015625,
      "epoch": 0.13968253968253969,
      "grad_norm": 1.032192482137044,
      "kl": 0.002838134765625,
      "learning_rate": 9.696502854275748e-07,
      "loss": 0.1778,
      "reward": 1.4399136304855347,
      "reward_std": 0.570347249507904,
      "rewards/": 4.27099609375,
      "rewards/math_compute_score": 0.7321428656578064,
      "step": 44
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 716.5714721679688,
      "epoch": 0.14285714285714285,
      "grad_norm": 1.0335824481684432,
      "kl": 0.0021820068359375,
      "learning_rate": 9.678584095202469e-07,
      "loss": 0.2178,
      "reward": 1.5996861457824707,
      "reward_std": 0.7114024758338928,
      "rewards/": 5.355573654174805,
      "rewards/math_compute_score": 0.660714328289032,
      "step": 45
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 502.0357360839844,
      "epoch": 0.14603174603174604,
      "grad_norm": 0.8112268442860173,
      "kl": 0.004425048828125,
      "learning_rate": 9.660168960252575e-07,
      "loss": 0.0694,
      "reward": 1.4711426496505737,
      "reward_std": 0.35930201411247253,
      "rewards/": 4.355712890625,
      "rewards/math_compute_score": 0.7500000596046448,
      "step": 46
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 811.8214721679688,
      "epoch": 0.1492063492063492,
      "grad_norm": 0.859203113128272,
      "kl": 0.00098419189453125,
      "learning_rate": 9.641259403185704e-07,
      "loss": 0.1777,
      "reward": 1.622921347618103,
      "reward_std": 0.4903363585472107,
      "rewards/": 5.900321006774902,
      "rewards/math_compute_score": 0.5535714626312256,
      "step": 47
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 765.607177734375,
      "epoch": 0.1523809523809524,
      "grad_norm": 0.8078928566280016,
      "kl": 0.0020294189453125,
      "learning_rate": 9.621857430217365e-07,
      "loss": 0.2052,
      "reward": 1.5972657203674316,
      "reward_std": 0.3894428610801697,
      "rewards/": 5.414899826049805,
      "rewards/math_compute_score": 0.6428571939468384,
      "step": 48
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 637.0178833007812,
      "epoch": 0.15555555555555556,
      "grad_norm": 4.86317816941763,
      "kl": 0.002838134765625,
      "learning_rate": 9.601965099806084e-07,
      "loss": 0.1317,
      "reward": 1.5043946504592896,
      "reward_std": 0.5993847250938416,
      "rewards/": 4.379115581512451,
      "rewards/math_compute_score": 0.785714328289032,
      "step": 49
    },
    {
      "epoch": 0.15873015873015872,
      "grad_norm": 0.6567682116832987,
      "learning_rate": 9.581584522435023e-07,
      "loss": 0.1677,
      "step": 50
    },
    {
      "epoch": 0.15873015873015872,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 841.1322021484375,
      "eval_kl": 0.004923502604166667,
      "eval_loss": 0.10817983746528625,
      "eval_reward": 1.6196069717407227,
      "eval_reward_std": 0.539500097433726,
      "eval_rewards/": 5.074225107828776,
      "eval_rewards/math_compute_score": 0.7559524178504944,
      "eval_runtime": 107.9307,
      "eval_samples_per_second": 0.185,
      "eval_steps_per_second": 0.009,
      "step": 50
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 603.4107513427734,
      "epoch": 0.1619047619047619,
      "grad_norm": 0.7984893297808615,
      "kl": 0.00357818603515625,
      "learning_rate": 9.56071786038806e-07,
      "loss": 0.1832,
      "reward": 1.4798433780670166,
      "reward_std": 0.45238083600997925,
      "rewards/": 4.363502264022827,
      "rewards/math_compute_score": 0.7589285969734192,
      "step": 51
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 742.607177734375,
      "epoch": 0.16507936507936508,
      "grad_norm": 0.5100556612157693,
      "kl": 0.00167083740234375,
      "learning_rate": 9.53936732752038e-07,
      "loss": -0.0658,
      "reward": 1.6291016340255737,
      "reward_std": 0.38449862599372864,
      "rewards/": 5.359793663024902,
      "rewards/math_compute_score": 0.6964285969734192,
      "step": 52
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 679.2857666015625,
      "epoch": 0.16825396825396827,
      "grad_norm": 0.5788655761199448,
      "kl": 0.002838134765625,
      "learning_rate": 9.517535189023601e-07,
      "loss": 0.0362,
      "reward": 1.597886562347412,
      "reward_std": 0.4612652361392975,
      "rewards/": 4.418004035949707,
      "rewards/math_compute_score": 0.8928571939468384,
      "step": 53
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 677.6785888671875,
      "epoch": 0.17142857142857143,
      "grad_norm": 1.116601202976549,
      "kl": 0.00189971923828125,
      "learning_rate": 9.495223761185441e-07,
      "loss": 0.1404,
      "reward": 1.7574794292449951,
      "reward_std": 0.4653353691101074,
      "rewards/": 5.858826160430908,
      "rewards/math_compute_score": 0.7321428656578064,
      "step": 54
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 571.9642944335938,
      "epoch": 0.1746031746031746,
      "grad_norm": 5.49249475780532,
      "kl": 0.0086669921875,
      "learning_rate": 9.472435411143977e-07,
      "loss": 0.1024,
      "reward": 1.579323172569275,
      "reward_std": 0.49340134859085083,
      "rewards/": 4.825186729431152,
      "rewards/math_compute_score": 0.7678571939468384,
      "step": 55
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 534.5714721679688,
      "epoch": 0.17777777777777778,
      "grad_norm": 0.6735986705301703,
      "kl": 0.003204345703125,
      "learning_rate": 9.449172556636497e-07,
      "loss": 0.0415,
      "reward": 1.7549248933792114,
      "reward_std": 0.41859832406044006,
      "rewards/": 5.346052169799805,
      "rewards/math_compute_score": 0.8571429252624512,
      "step": 56
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 786.3750610351562,
      "epoch": 0.18095238095238095,
      "grad_norm": 1.2145506026406598,
      "kl": 0.0016632080078125,
      "learning_rate": 9.425437665742997e-07,
      "loss": 0.1078,
      "reward": 1.463396430015564,
      "reward_std": 0.5382161736488342,
      "rewards/": 4.602696418762207,
      "rewards/math_compute_score": 0.6785714626312256,
      "step": 57
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 651.6607666015625,
      "epoch": 0.18412698412698414,
      "grad_norm": 1.6110105562648922,
      "kl": 0.0028839111328125,
      "learning_rate": 9.401233256624316e-07,
      "loss": 0.0788,
      "reward": 1.6237270832061768,
      "reward_std": 0.5656346082687378,
      "rewards/": 4.7614922523498535,
      "rewards/math_compute_score": 0.8392857313156128,
      "step": 58
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 692.2678833007812,
      "epoch": 0.1873015873015873,
      "grad_norm": 0.6039435728831022,
      "kl": 0.0020599365234375,
      "learning_rate": 9.376561897254987e-07,
      "loss": 0.0331,
      "reward": 1.7711775302886963,
      "reward_std": 0.4943082630634308,
      "rewards/": 5.570173263549805,
      "rewards/math_compute_score": 0.8214285969734192,
      "step": 59
    },
    {
      "epoch": 0.19047619047619047,
      "grad_norm": 0.6468115451400324,
      "learning_rate": 9.351426205150776e-07,
      "loss": 0.0748,
      "step": 60
    },
    {
      "epoch": 0.19047619047619047,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 789.8095296223959,
      "eval_kl": 0.023050944010416668,
      "eval_loss": 0.12776777148246765,
      "eval_reward": 1.6391727129618328,
      "eval_reward_std": 0.5289413332939148,
      "eval_rewards/": 5.076815923055013,
      "eval_rewards/math_compute_score": 0.7797619501749674,
      "eval_runtime": 107.6708,
      "eval_samples_per_second": 0.186,
      "eval_steps_per_second": 0.009,
      "step": 60
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 599.1518249511719,
      "epoch": 0.19365079365079366,
      "grad_norm": 0.62860654069817,
      "kl": 0.002593994140625,
      "learning_rate": 9.32582884709098e-07,
      "loss": -0.0003,
      "reward": 1.5030343532562256,
      "reward_std": 0.49661755561828613,
      "rewards/": 4.658028841018677,
      "rewards/math_compute_score": 0.7142857313156128,
      "step": 61
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 664.8035888671875,
      "epoch": 0.19682539682539682,
      "grad_norm": 1.3582868033212785,
      "kl": 0.0115966796875,
      "learning_rate": 9.299772538835491e-07,
      "loss": 0.1527,
      "reward": 1.7017022371292114,
      "reward_std": 0.452860027551651,
      "rewards/": 4.937081813812256,
      "rewards/math_compute_score": 0.8928571939468384,
      "step": 62
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 746.6428833007812,
      "epoch": 0.2,
      "grad_norm": 0.7888407267831482,
      "kl": 0.002288818359375,
      "learning_rate": 9.273260044836673e-07,
      "loss": 0.1191,
      "reward": 1.4725233316421509,
      "reward_std": 0.5076785087585449,
      "rewards/": 4.505473613739014,
      "rewards/math_compute_score": 0.7142857313156128,
      "step": 63
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 630.9285888671875,
      "epoch": 0.20317460317460317,
      "grad_norm": 0.5701838250380957,
      "kl": 0.003143310546875,
      "learning_rate": 9.246294177946062e-07,
      "loss": 0.175,
      "reward": 1.4169644117355347,
      "reward_std": 0.4153939485549927,
      "rewards/": 4.084821701049805,
      "rewards/math_compute_score": 0.7500000596046448,
      "step": 64
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 746.1607666015625,
      "epoch": 0.20634920634920634,
      "grad_norm": 0.43485838557829426,
      "kl": 0.003326416015625,
      "learning_rate": 9.218877799115927e-07,
      "loss": 0.0591,
      "reward": 1.405147910118103,
      "reward_std": 0.6980858445167542,
      "rewards/": 4.1685967445373535,
      "rewards/math_compute_score": 0.7142857313156128,
      "step": 65
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 754.3750610351562,
      "epoch": 0.20952380952380953,
      "grad_norm": 0.5397143184004072,
      "kl": 0.0023040771484375,
      "learning_rate": 9.191013817095761e-07,
      "loss": 0.1156,
      "reward": 1.5029298067092896,
      "reward_std": 0.5369086265563965,
      "rewards/": 4.657505989074707,
      "rewards/math_compute_score": 0.7142857313156128,
      "step": 66
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 473.4107360839844,
      "epoch": 0.2126984126984127,
      "grad_norm": 0.6430268032171578,
      "kl": 0.0027313232421875,
      "learning_rate": 9.162705188123646e-07,
      "loss": -0.0227,
      "reward": 1.8193360567092896,
      "reward_std": 0.5557112097740173,
      "rewards/": 5.1681084632873535,
      "rewards/math_compute_score": 0.9821429252624512,
      "step": 67
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 584.4642944335938,
      "epoch": 0.21587301587301588,
      "grad_norm": 0.6983774100649138,
      "kl": 0.00494384765625,
      "learning_rate": 9.133954915612634e-07,
      "loss": 0.062,
      "reward": 1.5165389776229858,
      "reward_std": 0.41740962862968445,
      "rewards/": 4.296979904174805,
      "rewards/math_compute_score": 0.8214285969734192,
      "step": 68
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 609.1607666015625,
      "epoch": 0.21904761904761905,
      "grad_norm": 0.6699339431654112,
      "kl": 0.00439453125,
      "learning_rate": 9.104766049832087e-07,
      "loss": 0.0443,
      "reward": 1.5293978452682495,
      "reward_std": 0.5475279092788696,
      "rewards/": 4.575560569763184,
      "rewards/math_compute_score": 0.7678571939468384,
      "step": 69
    },
    {
      "epoch": 0.2222222222222222,
      "grad_norm": 0.7988695058458649,
      "learning_rate": 9.075141687584056e-07,
      "loss": -0.0167,
      "step": 70
    },
    {
      "epoch": 0.2222222222222222,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 901.5059814453125,
      "eval_kl": 0.2381591796875,
      "eval_loss": 0.2867256999015808,
      "eval_reward": 1.6091902653376262,
      "eval_reward_std": 0.5438455939292908,
      "eval_rewards/": 4.926903247833252,
      "eval_rewards/math_compute_score": 0.7797619501749674,
      "eval_runtime": 111.1812,
      "eval_samples_per_second": 0.18,
      "eval_steps_per_second": 0.009,
      "step": 70
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 625.7410888671875,
      "epoch": 0.2253968253968254,
      "grad_norm": 0.5540727634053975,
      "kl": 0.003814697265625,
      "learning_rate": 9.045084971874737e-07,
      "loss": -0.0056,
      "reward": 1.7840471863746643,
      "reward_std": 0.4358938932418823,
      "rewards/": 5.313093185424805,
      "rewards/math_compute_score": 0.9017857611179352,
      "step": 71
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 791.2678833007812,
      "epoch": 0.22857142857142856,
      "grad_norm": 0.49864634439649874,
      "kl": 0.0018157958984375,
      "learning_rate": 9.014599091580998e-07,
      "loss": 0.0311,
      "reward": 1.6201382875442505,
      "reward_std": 0.5860312581062317,
      "rewards/": 5.314976692199707,
      "rewards/math_compute_score": 0.6964285969734192,
      "step": 72
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 515.8392944335938,
      "epoch": 0.23174603174603176,
      "grad_norm": 0.9973211460805157,
      "kl": 0.00482177734375,
      "learning_rate": 8.983687281112064e-07,
      "loss": -0.0697,
      "reward": 1.6238142251968384,
      "reward_std": 0.35913076996803284,
      "rewards/": 4.761928081512451,
      "rewards/math_compute_score": 0.8392857313156128,
      "step": 73
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 643.357177734375,
      "epoch": 0.23492063492063492,
      "grad_norm": 0.877503000341696,
      "kl": 0.002105712890625,
      "learning_rate": 8.952352820066358e-07,
      "loss": -0.0014,
      "reward": 1.4080777168273926,
      "reward_std": 0.5432181358337402,
      "rewards/": 4.397531032562256,
      "rewards/math_compute_score": 0.660714328289032,
      "step": 74
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 478.7857360839844,
      "epoch": 0.23809523809523808,
      "grad_norm": 0.6202183598487062,
      "kl": 0.0023345947265625,
      "learning_rate": 8.920599032883552e-07,
      "loss": 0.1503,
      "reward": 1.7339845895767212,
      "reward_std": 0.3223447799682617,
      "rewards/": 5.2413506507873535,
      "rewards/math_compute_score": 0.8571429252624512,
      "step": 75
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 727.3392944335938,
      "epoch": 0.24126984126984127,
      "grad_norm": 0.9654723599639436,
      "kl": 0.00201416015625,
      "learning_rate": 8.888429288491855e-07,
      "loss": 0.1141,
      "reward": 1.7490864992141724,
      "reward_std": 0.5959608554840088,
      "rewards/": 5.0311455726623535,
      "rewards/math_compute_score": 0.9285714626312256,
      "step": 76
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 607.982177734375,
      "epoch": 0.24444444444444444,
      "grad_norm": 0.7445671259261306,
      "kl": 0.00186920166015625,
      "learning_rate": 8.855846999950595e-07,
      "loss": 0.0564,
      "reward": 1.5656529664993286,
      "reward_std": 0.44271785020828247,
      "rewards/": 4.113978862762451,
      "rewards/math_compute_score": 0.9285714626312256,
      "step": 77
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 730.5357666015625,
      "epoch": 0.24761904761904763,
      "grad_norm": 1.1259492375136075,
      "kl": 0.00174713134765625,
      "learning_rate": 8.822855624088097e-07,
      "loss": 0.0594,
      "reward": 1.7164901494979858,
      "reward_std": 0.46802079677581787,
      "rewards/": 5.296735763549805,
      "rewards/math_compute_score": 0.8214285969734192,
      "step": 78
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 520.625,
      "epoch": 0.2507936507936508,
      "grad_norm": 0.6608772516145822,
      "kl": 0.0037689208984375,
      "learning_rate": 8.789458661134942e-07,
      "loss": 0.113,
      "reward": 1.6773438453674316,
      "reward_std": 0.41135743260383606,
      "rewards/": 4.9581475257873535,
      "rewards/math_compute_score": 0.8571429252624512,
      "step": 79
    },
    {
      "epoch": 0.25396825396825395,
      "grad_norm": 1.1086447024025683,
      "learning_rate": 8.755659654352599e-07,
      "loss": 0.0852,
      "step": 80
    },
    {
      "epoch": 0.25396825396825395,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 762.126220703125,
      "eval_kl": 0.057149251302083336,
      "eval_loss": 0.1757144182920456,
      "eval_reward": 1.6997971932093303,
      "eval_reward_std": 0.5428444643815359,
      "eval_rewards/": 5.284700234731038,
      "eval_rewards/math_compute_score": 0.8035714824994405,
      "eval_runtime": 105.3925,
      "eval_samples_per_second": 0.19,
      "eval_steps_per_second": 0.009,
      "step": 80
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 554.4107208251953,
      "epoch": 0.2571428571428571,
      "grad_norm": 0.6297992518515485,
      "kl": 0.0034942626953125,
      "learning_rate": 8.721462189657509e-07,
      "loss": 0.0229,
      "reward": 1.6316068172454834,
      "reward_std": 0.46411074697971344,
      "rewards/": 5.0508904457092285,
      "rewards/math_compute_score": 0.7767857611179352,
      "step": 81
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 612.8035888671875,
      "epoch": 0.26031746031746034,
      "grad_norm": 0.5403276172661425,
      "kl": 0.0034637451171875,
      "learning_rate": 8.686869895240631e-07,
      "loss": 0.0812,
      "reward": 1.6373475790023804,
      "reward_std": 0.49845850467681885,
      "rewards/": 4.901022911071777,
      "rewards/math_compute_score": 0.8214285969734192,
      "step": 82
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 608.0535888671875,
      "epoch": 0.2634920634920635,
      "grad_norm": 0.915800232301419,
      "kl": 0.003448486328125,
      "learning_rate": 8.651886441182508e-07,
      "loss": 0.0353,
      "reward": 1.547105312347412,
      "reward_std": 0.6321852803230286,
      "rewards/": 5.235526084899902,
      "rewards/math_compute_score": 0.625,
      "step": 83
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 619.4107666015625,
      "epoch": 0.26666666666666666,
      "grad_norm": 0.4435988410022463,
      "kl": 0.0033111572265625,
      "learning_rate": 8.616515539063894e-07,
      "loss": -0.0685,
      "reward": 1.8496233224868774,
      "reward_std": 0.49867764115333557,
      "rewards/": 5.676688194274902,
      "rewards/math_compute_score": 0.8928571939468384,
      "step": 84
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 751.5000610351562,
      "epoch": 0.2698412698412698,
      "grad_norm": 0.8497969196784788,
      "kl": 0.0025177001953125,
      "learning_rate": 8.580760941571966e-07,
      "loss": 0.2484,
      "reward": 1.459954023361206,
      "reward_std": 0.6413030624389648,
      "rewards/": 4.871198654174805,
      "rewards/math_compute_score": 0.6071428656578064,
      "step": 85
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 565.1607666015625,
      "epoch": 0.273015873015873,
      "grad_norm": 0.6258059646403542,
      "kl": 0.0048828125,
      "learning_rate": 8.544626442102187e-07,
      "loss": 0.038,
      "reward": 1.6110281944274902,
      "reward_std": 0.610092043876648,
      "rewards/": 5.197998046875,
      "rewards/math_compute_score": 0.7142857313156128,
      "step": 86
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 488.39288330078125,
      "epoch": 0.2761904761904762,
      "grad_norm": 0.8138401874703006,
      "kl": 0.004638671875,
      "learning_rate": 8.508115874355839e-07,
      "loss": 0.0085,
      "reward": 1.3848702907562256,
      "reward_std": 0.39325088262557983,
      "rewards/": 4.210065841674805,
      "rewards/math_compute_score": 0.6785714626312256,
      "step": 87
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 536.5892944335938,
      "epoch": 0.27936507936507937,
      "grad_norm": 2.4320450027460465,
      "kl": 0.00323486328125,
      "learning_rate": 8.47123311193329e-07,
      "loss": -0.0545,
      "reward": 1.82899010181427,
      "reward_std": 0.4799407124519348,
      "rewards/": 5.430664539337158,
      "rewards/math_compute_score": 0.9285714626312256,
      "step": 88
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 728.8928833007812,
      "epoch": 0.28253968253968254,
      "grad_norm": 0.5888691350508202,
      "kl": 0.0026397705078125,
      "learning_rate": 8.433982067923021e-07,
      "loss": 0.1658,
      "reward": 1.526370644569397,
      "reward_std": 0.6073156595230103,
      "rewards/": 4.560424327850342,
      "rewards/math_compute_score": 0.7678571939468384,
      "step": 89
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.672547715543188,
      "learning_rate": 8.396366694486466e-07,
      "loss": -0.0451,
      "step": 90
    },
    {
      "epoch": 0.2857142857142857,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 751.4702657063802,
      "eval_kl": 0.01361083984375,
      "eval_loss": 0.09576481580734253,
      "eval_reward": 1.6055153608322144,
      "eval_reward_std": 0.609013577302297,
      "eval_rewards/": 4.837100187937419,
      "eval_rewards/math_compute_score": 0.797619084517161,
      "eval_runtime": 104.6076,
      "eval_samples_per_second": 0.191,
      "eval_steps_per_second": 0.01,
      "step": 90
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 719.5357666015625,
      "epoch": 0.28888888888888886,
      "grad_norm": 0.5180346375049427,
      "kl": 0.002506256103515625,
      "learning_rate": 8.358390982438705e-07,
      "loss": 0.1519,
      "reward": 1.9254953861236572,
      "reward_std": 0.503593385219574,
      "rewards/": 6.09176230430603,
      "rewards/math_compute_score": 0.8839286267757416,
      "step": 91
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 452.1607360839844,
      "epoch": 0.2920634920634921,
      "grad_norm": 0.9788308883352828,
      "kl": 0.0079345703125,
      "learning_rate": 8.320058960825058e-07,
      "loss": 0.0102,
      "reward": 1.6602121591567993,
      "reward_std": 0.3053475022315979,
      "rewards/": 4.658203125,
      "rewards/math_compute_score": 0.910714328289032,
      "step": 92
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 693.5535888671875,
      "epoch": 0.29523809523809524,
      "grad_norm": 0.5847042829798909,
      "kl": 0.00244140625,
      "learning_rate": 8.281374696493626e-07,
      "loss": 0.2312,
      "reward": 1.8354074954986572,
      "reward_std": 0.37926599383354187,
      "rewards/": 6.248465538024902,
      "rewards/math_compute_score": 0.7321428656578064,
      "step": 93
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 660.232177734375,
      "epoch": 0.2984126984126984,
      "grad_norm": 0.5205561829484078,
      "kl": 0.0027008056640625,
      "learning_rate": 8.242342293663809e-07,
      "loss": 0.1094,
      "reward": 1.840959906578064,
      "reward_std": 0.3830914795398712,
      "rewards/": 5.347656726837158,
      "rewards/math_compute_score": 0.9642857313156128,
      "step": 94
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 609.8035888671875,
      "epoch": 0.30158730158730157,
      "grad_norm": 0.44482437448212625,
      "kl": 0.004425048828125,
      "learning_rate": 8.202965893490876e-07,
      "loss": 0.0873,
      "reward": 1.7079870700836182,
      "reward_std": 0.5448174476623535,
      "rewards/": 5.111363410949707,
      "rewards/math_compute_score": 0.8571429252624512,
      "step": 95
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 671.5357666015625,
      "epoch": 0.3047619047619048,
      "grad_norm": 1.1320823935790385,
      "kl": 0.002166748046875,
      "learning_rate": 8.163249673626602e-07,
      "loss": 0.0481,
      "reward": 1.4774868488311768,
      "reward_std": 0.45228156447410583,
      "rewards/": 4.9588623046875,
      "rewards/math_compute_score": 0.6071428656578064,
      "step": 96
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 544.3214721679688,
      "epoch": 0.30793650793650795,
      "grad_norm": 0.5094127089703688,
      "kl": 0.003997802734375,
      "learning_rate": 8.123197847776042e-07,
      "loss": 0.0533,
      "reward": 1.765764594078064,
      "reward_std": 0.5276929140090942,
      "rewards/": 5.757394313812256,
      "rewards/math_compute_score": 0.7678571939468384,
      "step": 97
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 593.8214721679688,
      "epoch": 0.3111111111111111,
      "grad_norm": 0.5934812502311124,
      "kl": 0.0042724609375,
      "learning_rate": 8.082814665250476e-07,
      "loss": -0.0354,
      "reward": 1.6365864276885986,
      "reward_std": 0.46695929765701294,
      "rewards/": 5.040074348449707,
      "rewards/math_compute_score": 0.785714328289032,
      "step": 98
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 511.4464416503906,
      "epoch": 0.3142857142857143,
      "grad_norm": 0.6289350290275575,
      "kl": 0.004669189453125,
      "learning_rate": 8.042104410516575e-07,
      "loss": -0.0663,
      "reward": 1.6093192100524902,
      "reward_std": 0.39656367897987366,
      "rewards/": 4.546596050262451,
      "rewards/math_compute_score": 0.8750000596046448,
      "step": 99
    },
    {
      "epoch": 0.31746031746031744,
      "grad_norm": 0.5581381958692002,
      "learning_rate": 8.001071402741842e-07,
      "loss": 0.0344,
      "step": 100
    },
    {
      "epoch": 0.31746031746031744,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 700.452392578125,
      "eval_kl": 0.01019287109375,
      "eval_loss": -0.026086583733558655,
      "eval_reward": 1.755845586458842,
      "eval_reward_std": 0.4914539158344269,
      "eval_rewards/": 5.422084490458171,
      "eval_rewards/math_compute_score": 0.8392857710520426,
      "eval_runtime": 91.4693,
      "eval_samples_per_second": 0.219,
      "eval_steps_per_second": 0.011,
      "step": 100
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 552.2410888671875,
      "epoch": 0.32063492063492066,
      "grad_norm": 0.7136178771883936,
      "kl": 0.00354766845703125,
      "learning_rate": 7.959719995336363e-07,
      "loss": 0.0693,
      "reward": 1.6664237976074219,
      "reward_std": 0.429623618721962,
      "rewards/": 4.939261436462402,
      "rewards/math_compute_score": 0.848214328289032,
      "step": 101
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 663.25,
      "epoch": 0.3238095238095238,
      "grad_norm": 0.5439969780060891,
      "kl": 0.0037994384765625,
      "learning_rate": 7.918054575490943e-07,
      "loss": 0.0424,
      "reward": 1.4987915754318237,
      "reward_std": 0.5196777582168579,
      "rewards/": 4.208243370056152,
      "rewards/math_compute_score": 0.8214285969734192,
      "step": 102
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 471.8750305175781,
      "epoch": 0.326984126984127,
      "grad_norm": 0.7220192489284563,
      "kl": 0.0033416748046875,
      "learning_rate": 7.876079563711631e-07,
      "loss": 0.1121,
      "reward": 1.7968298196792603,
      "reward_std": 0.36027005314826965,
      "rewards/": 5.127005577087402,
      "rewards/math_compute_score": 0.9642857313156128,
      "step": 103
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 640.5892944335938,
      "epoch": 0.33015873015873015,
      "grad_norm": 0.8015986414478035,
      "kl": 0.004974365234375,
      "learning_rate": 7.83379941335073e-07,
      "loss": 0.2035,
      "reward": 1.7956823110580444,
      "reward_std": 0.35563212633132935,
      "rewards/": 5.621268272399902,
      "rewards/math_compute_score": 0.8392857313156128,
      "step": 104
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 799.8750610351562,
      "epoch": 0.3333333333333333,
      "grad_norm": 0.6682617366427315,
      "kl": 0.00250244140625,
      "learning_rate": 7.791218610134322e-07,
      "loss": 0.2286,
      "reward": 1.4982212781906128,
      "reward_std": 0.5870651602745056,
      "rewards/": 5.4911065101623535,
      "rewards/math_compute_score": 0.5,
      "step": 105
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 680.9107666015625,
      "epoch": 0.33650793650793653,
      "grad_norm": 0.6146294713011484,
      "kl": 0.002471923828125,
      "learning_rate": 7.748341671686354e-07,
      "loss": 0.0038,
      "reward": 1.9384348392486572,
      "reward_std": 0.49243679642677307,
      "rewards/": 6.192173957824707,
      "rewards/math_compute_score": 0.8750000596046448,
      "step": 106
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 663.4285888671875,
      "epoch": 0.3396825396825397,
      "grad_norm": 0.7562852855642146,
      "kl": 0.0038299560546875,
      "learning_rate": 7.705173147049325e-07,
      "loss": 0.1061,
      "reward": 1.6097028255462646,
      "reward_std": 0.5770988464355469,
      "rewards/": 4.691371440887451,
      "rewards/math_compute_score": 0.8392857313156128,
      "step": 107
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 760.1250610351562,
      "epoch": 0.34285714285714286,
      "grad_norm": 0.8460361870531922,
      "kl": 0.005615234375,
      "learning_rate": 7.661717616201668e-07,
      "loss": 0.1169,
      "reward": 1.7678015232086182,
      "reward_std": 0.46296006441116333,
      "rewards/": 5.910435676574707,
      "rewards/math_compute_score": 0.7321428656578064,
      "step": 108
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 544.4107666015625,
      "epoch": 0.346031746031746,
      "grad_norm": 0.8867581981573962,
      "kl": 0.00518798828125,
      "learning_rate": 7.617979689571839e-07,
      "loss": 0.0749,
      "reward": 1.6010881662368774,
      "reward_std": 0.40945473313331604,
      "rewards/": 4.7197265625,
      "rewards/math_compute_score": 0.8214285969734192,
      "step": 109
    },
    {
      "epoch": 0.3492063492063492,
      "grad_norm": 0.613697351604427,
      "learning_rate": 7.573964007549154e-07,
      "loss": -0.0219,
      "step": 110
    },
    {
      "epoch": 0.3492063492063492,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 747.9107259114584,
      "eval_kl": 0.0235595703125,
      "eval_loss": 0.026410922408103943,
      "eval_reward": 1.7149322032928467,
      "eval_reward_std": 0.49799948930740356,
      "eval_rewards/": 5.3365654945373535,
      "eval_rewards/math_compute_score": 0.8095238407452902,
      "eval_runtime": 103.4902,
      "eval_samples_per_second": 0.193,
      "eval_steps_per_second": 0.01,
      "step": 110
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 603.4642944335938,
      "epoch": 0.3523809523809524,
      "grad_norm": 0.9237700659841631,
      "kl": 0.0062713623046875,
      "learning_rate": 7.529675239991482e-07,
      "loss": -0.0713,
      "reward": 1.6208950281143188,
      "reward_std": 0.5017957538366318,
      "rewards/": 5.03304648399353,
      "rewards/math_compute_score": 0.7678571939468384,
      "step": 111
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 537.6785888671875,
      "epoch": 0.35555555555555557,
      "grad_norm": 0.5989679853358942,
      "kl": 0.003326416015625,
      "learning_rate": 7.485118085729789e-07,
      "loss": 0.0532,
      "reward": 1.7066690921783447,
      "reward_std": 0.5952469110488892,
      "rewards/": 4.819059371948242,
      "rewards/math_compute_score": 0.9285714626312256,
      "step": 112
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 557.6607666015625,
      "epoch": 0.35873015873015873,
      "grad_norm": 0.792413828293772,
      "kl": 0.0034637451171875,
      "learning_rate": 7.440297272069614e-07,
      "loss": 0.0875,
      "reward": 1.5029088258743286,
      "reward_std": 0.5444967746734619,
      "rewards/": 4.800258159637451,
      "rewards/math_compute_score": 0.6785714626312256,
      "step": 113
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 674.9464721679688,
      "epoch": 0.3619047619047619,
      "grad_norm": 0.8399630521739422,
      "kl": 0.002349853515625,
      "learning_rate": 7.395217554289523e-07,
      "loss": 0.065,
      "reward": 1.6198941469192505,
      "reward_std": 0.5176466107368469,
      "rewards/": 4.956613063812256,
      "rewards/math_compute_score": 0.785714328289032,
      "step": 114
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 476.21429443359375,
      "epoch": 0.36507936507936506,
      "grad_norm": 1.01617746278495,
      "kl": 0.004150390625,
      "learning_rate": 7.3498837151366e-07,
      "loss": -0.0269,
      "reward": 1.5831892490386963,
      "reward_std": 0.44976264238357544,
      "rewards/": 5.058803081512451,
      "rewards/math_compute_score": 0.7142857313156128,
      "step": 115
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 715.6964721679688,
      "epoch": 0.3682539682539683,
      "grad_norm": 0.6977639971716434,
      "kl": 0.00323486328125,
      "learning_rate": 7.304300564319013e-07,
      "loss": 0.0349,
      "reward": 1.998925805091858,
      "reward_std": 0.49372294545173645,
      "rewards/": 6.423200607299805,
      "rewards/math_compute_score": 0.8928571939468384,
      "step": 116
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 738.1250610351562,
      "epoch": 0.37142857142857144,
      "grad_norm": 0.5710554895504717,
      "kl": 0.0037689208984375,
      "learning_rate": 7.258472937995735e-07,
      "loss": -0.023,
      "reward": 1.9152623414993286,
      "reward_std": 0.43735089898109436,
      "rewards/": 6.004883289337158,
      "rewards/math_compute_score": 0.8928571939468384,
      "step": 117
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 696.607177734375,
      "epoch": 0.3746031746031746,
      "grad_norm": 0.9187825212513634,
      "kl": 0.0034637451171875,
      "learning_rate": 7.212405698263446e-07,
      "loss": -0.0131,
      "reward": 1.705580472946167,
      "reward_std": 0.509798526763916,
      "rewards/": 6.027902126312256,
      "rewards/math_compute_score": 0.625,
      "step": 118
    },
    {
      "clip_ratio": 0.0,
      "completion_length": 647.75,
      "epoch": 0.37777777777777777,
      "grad_norm": 0.8884977568265239,
      "kl": 0.005096435546875,
      "learning_rate": 7.166103732640681e-07,
      "loss": 0.1126,
      "reward": 1.5224602222442627,
      "reward_std": 0.5471193194389343,
      "rewards/": 4.3265862464904785,
      "rewards/math_compute_score": 0.8214285969734192,
      "step": 119
    },
    {
      "epoch": 0.38095238095238093,
      "grad_norm": 0.5323455430197922,
      "learning_rate": 7.119571953549304e-07,
      "loss": 0.0312,
      "step": 120
    },
    {
      "epoch": 0.38095238095238093,
      "eval_clip_ratio": 0.0,
      "eval_completion_length": 752.0238240559896,
      "eval_kl": 0.0135345458984375,
      "eval_loss": -0.006800096947699785,
      "eval_reward": 1.8065431118011475,
      "eval_reward_std": 0.4533117413520813,
      "eval_rewards/": 5.604143778483073,
      "eval_rewards/math_compute_score": 0.8571429053942362,
      "eval_runtime": 102.4477,
      "eval_samples_per_second": 0.195,
      "eval_steps_per_second": 0.01,
      "step": 120
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 315,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 40,
  "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": 8,
  "trial_name": null,
  "trial_params": null
}