{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.9936628643852978, "eval_steps": 100, "global_step": 196, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "clip_ratio": 0.0, "completion_length": 515.8303871154785, "epoch": 0.010139416983523447, "grad_norm": 0.4515054523944855, "learning_rate": 5e-08, "loss": 0.0123, "num_tokens": 610184.0, "reward": 0.26004465483129025, "reward_std": 0.3379838764667511, "rewards/accuracy_reward": 0.1986607126891613, "rewards/format_reward": 0.06138392956927419, "step": 1 }, { "clip_ratio": 0.0, "completion_length": 467.8660888671875, "epoch": 0.020278833967046894, "grad_norm": 0.6836522221565247, "learning_rate": 1e-07, "loss": 0.0058, "num_tokens": 1164448.0, "reward": 0.24776786752045155, "reward_std": 0.3431246988475323, "rewards/accuracy_reward": 0.171875, "rewards/format_reward": 0.07589285867288709, "step": 2 }, { "clip_ratio": 0.0, "completion_length": 528.7500343322754, "epoch": 0.030418250950570342, "grad_norm": 0.5222043395042419, "learning_rate": 1.5e-07, "loss": 0.005, "num_tokens": 1778328.0, "reward": 0.26450893841683865, "reward_std": 0.3584165945649147, "rewards/accuracy_reward": 0.2187499962747097, "rewards/format_reward": 0.045758930034935474, "step": 3 }, { "clip_ratio": 0.0, "completion_length": 508.4922103881836, "epoch": 0.04055766793409379, "grad_norm": 0.4442610442638397, "learning_rate": 2e-07, "loss": 0.0255, "num_tokens": 2370889.0, "reward": 0.26004465110599995, "reward_std": 0.3489377610385418, "rewards/accuracy_reward": 0.19642856996506453, "rewards/format_reward": 0.06361607229337096, "step": 4 }, { "clip_ratio": 0.0, "completion_length": 519.3482360839844, "epoch": 0.050697084917617236, "grad_norm": 0.504645824432373, "learning_rate": 2.5e-07, "loss": 0.0363, "num_tokens": 2963009.0, "reward": 0.2700893022119999, "reward_std": 0.3758334591984749, "rewards/accuracy_reward": 0.1897321417927742, "rewards/format_reward": 0.08035714202560484, "step": 5 }, { "clip_ratio": 0.0, "completion_length": 509.5201072692871, "epoch": 0.060836501901140684, "grad_norm": 0.4883345365524292, "learning_rate": 3e-07, "loss": 0.0146, "num_tokens": 3548731.0, "reward": 0.2924107275903225, "reward_std": 0.3819224089384079, "rewards/accuracy_reward": 0.21875, "rewards/format_reward": 0.07366071455180645, "step": 6 }, { "clip_ratio": 0.0, "completion_length": 506.1562728881836, "epoch": 0.07097591888466413, "grad_norm": 0.7536768317222595, "learning_rate": 3.5e-07, "loss": -0.0085, "num_tokens": 4126671.0, "reward": 0.3169642947614193, "reward_std": 0.37958647310733795, "rewards/accuracy_reward": 0.2555803544819355, "rewards/format_reward": 0.06138392980210483, "step": 7 }, { "clip_ratio": 0.0, "completion_length": 512.7143173217773, "epoch": 0.08111533586818757, "grad_norm": 0.5751662850379944, "learning_rate": 4e-07, "loss": -0.0186, "num_tokens": 4713615.0, "reward": 0.2879464440047741, "reward_std": 0.3758578822016716, "rewards/accuracy_reward": 0.2087053544819355, "rewards/format_reward": 0.07924107136204839, "step": 8 }, { "clip_ratio": 0.0, "completion_length": 527.1506843566895, "epoch": 0.09125475285171103, "grad_norm": 0.727789044380188, "learning_rate": 4.5e-07, "loss": 0.0193, "num_tokens": 5309270.0, "reward": 0.28794644214212894, "reward_std": 0.38178177177906036, "rewards/accuracy_reward": 0.19866071455180645, "rewards/format_reward": 0.08928571408614516, "step": 9 }, { "clip_ratio": 0.0, "completion_length": 462.6607437133789, "epoch": 0.10139416983523447, "grad_norm": 0.6003168821334839, "learning_rate": 5e-07, "loss": 0.0127, "num_tokens": 5850542.0, "reward": 0.3270089440047741, "reward_std": 0.42656392976641655, "rewards/accuracy_reward": 0.2176339253783226, "rewards/format_reward": 0.109375, "step": 10 }, { "clip_ratio": 0.0, "completion_length": 500.8236846923828, "epoch": 0.11153358681875793, "grad_norm": 0.6242951154708862, "learning_rate": 5.5e-07, "loss": 0.0068, "num_tokens": 6436816.0, "reward": 0.3191964440047741, "reward_std": 0.4202234297990799, "rewards/accuracy_reward": 0.171875, "rewards/format_reward": 0.14732142817229033, "step": 11 }, { "clip_ratio": 0.0, "completion_length": 517.0971221923828, "epoch": 0.12167300380228137, "grad_norm": 1.3958944082260132, "learning_rate": 6e-07, "loss": 0.0251, "num_tokens": 7031447.0, "reward": 0.3738839402794838, "reward_std": 0.4328419789671898, "rewards/accuracy_reward": 0.19531249720603228, "rewards/format_reward": 0.17857142724096775, "step": 12 }, { "clip_ratio": 0.0, "completion_length": 540.4654197692871, "epoch": 0.13181242078580482, "grad_norm": 1.9689440727233887, "learning_rate": 6.5e-07, "loss": 0.0087, "num_tokens": 7643888.0, "reward": 0.4464285969734192, "reward_std": 0.5018927529454231, "rewards/accuracy_reward": 0.2310267835855484, "rewards/format_reward": 0.2154017835855484, "step": 13 }, { "clip_ratio": 0.0, "completion_length": 489.197566986084, "epoch": 0.14195183776932827, "grad_norm": 0.798046350479126, "learning_rate": 7e-07, "loss": -0.0009, "num_tokens": 8214385.0, "reward": 0.4988839514553547, "reward_std": 0.5417631343007088, "rewards/accuracy_reward": 0.18973214458674192, "rewards/format_reward": 0.3113839253783226, "step": 14 }, { "clip_ratio": 0.0, "completion_length": 507.252254486084, "epoch": 0.1520912547528517, "grad_norm": 0.4880025386810303, "learning_rate": 7.5e-07, "loss": 0.0114, "num_tokens": 8797323.0, "reward": 0.7120536044239998, "reward_std": 0.5666098818182945, "rewards/accuracy_reward": 0.19754464272409678, "rewards/format_reward": 0.5145089402794838, "step": 15 }, { "clip_ratio": 0.0, "completion_length": 464.38841247558594, "epoch": 0.16223067173637515, "grad_norm": 0.3857059180736542, "learning_rate": 8e-07, "loss": -0.0014, "num_tokens": 9340343.0, "reward": 0.8526786044239998, "reward_std": 0.5239133685827255, "rewards/accuracy_reward": 0.21540178544819355, "rewards/format_reward": 0.6372767761349678, "step": 16 }, { "clip_ratio": 0.0, "completion_length": 508.8448944091797, "epoch": 0.17237008871989862, "grad_norm": 0.44979313015937805, "learning_rate": 8.499999999999999e-07, "loss": 0.0215, "num_tokens": 9930780.0, "reward": 0.8381696939468384, "reward_std": 0.540903877466917, "rewards/accuracy_reward": 0.2020089291036129, "rewards/format_reward": 0.6361607238650322, "step": 17 }, { "clip_ratio": 0.0, "completion_length": 480.3794860839844, "epoch": 0.18250950570342206, "grad_norm": 0.32462233304977417, "learning_rate": 9e-07, "loss": 0.03, "num_tokens": 10484544.0, "reward": 0.9810268357396126, "reward_std": 0.5262674763798714, "rewards/accuracy_reward": 0.2488839253783226, "rewards/format_reward": 0.7321428582072258, "step": 18 }, { "clip_ratio": 0.0, "completion_length": 455.48104095458984, "epoch": 0.1926489226869455, "grad_norm": 0.34758785367012024, "learning_rate": 9.499999999999999e-07, "loss": 0.023, "num_tokens": 11024639.0, "reward": 1.0424107536673546, "reward_std": 0.5019765421748161, "rewards/accuracy_reward": 0.2578124962747097, "rewards/format_reward": 0.784598208963871, "step": 19 }, { "clip_ratio": 0.0, "completion_length": 525.8482437133789, "epoch": 0.20278833967046894, "grad_norm": 0.44885051250457764, "learning_rate": 1e-06, "loss": 0.044, "num_tokens": 11616223.0, "reward": 1.1584821939468384, "reward_std": 0.47795333340764046, "rewards/accuracy_reward": 0.29352678544819355, "rewards/format_reward": 0.8649553582072258, "step": 20 }, { "clip_ratio": 0.0, "completion_length": 486.7299346923828, "epoch": 0.21292775665399238, "grad_norm": 0.5567921996116638, "learning_rate": 9.999203468625015e-07, "loss": 0.0113, "num_tokens": 12173605.0, "reward": 1.2120536416769028, "reward_std": 0.46485012769699097, "rewards/accuracy_reward": 0.31473214365541935, "rewards/format_reward": 0.8973214328289032, "step": 21 }, { "clip_ratio": 0.0, "completion_length": 542.1808204650879, "epoch": 0.22306717363751585, "grad_norm": 0.3349709212779999, "learning_rate": 9.99681412828496e-07, "loss": 0.0304, "num_tokens": 12792559.0, "reward": 1.1808036118745804, "reward_std": 0.4690566882491112, "rewards/accuracy_reward": 0.3024553544819355, "rewards/format_reward": 0.878348208963871, "step": 22 }, { "clip_ratio": 0.0, "completion_length": 527.556941986084, "epoch": 0.2332065906210393, "grad_norm": 0.32437098026275635, "learning_rate": 9.992832740253644e-07, "loss": 0.0241, "num_tokens": 13403498.0, "reward": 1.2622768431901932, "reward_std": 0.4536837413907051, "rewards/accuracy_reward": 0.3504464253783226, "rewards/format_reward": 0.9118303507566452, "step": 23 }, { "clip_ratio": 0.0, "completion_length": 549.2321624755859, "epoch": 0.24334600760456274, "grad_norm": 0.683305025100708, "learning_rate": 9.987260573051267e-07, "loss": 0.0239, "num_tokens": 14025154.0, "reward": 1.3013393580913544, "reward_std": 0.4284757189452648, "rewards/accuracy_reward": 0.3816964216530323, "rewards/format_reward": 0.9196428656578064, "step": 24 }, { "clip_ratio": 0.0, "completion_length": 525.8181076049805, "epoch": 0.2534854245880862, "grad_norm": 0.34933602809906006, "learning_rate": 9.98009940204023e-07, "loss": 0.0268, "num_tokens": 14627615.0, "reward": 1.4040179401636124, "reward_std": 0.41444528847932816, "rewards/accuracy_reward": 0.4531249962747097, "rewards/format_reward": 0.9508928656578064, "step": 25 }, { "clip_ratio": 0.0, "completion_length": 540.1384201049805, "epoch": 0.26362484157160965, "grad_norm": 0.3170299828052521, "learning_rate": 9.971351508859486e-07, "loss": 0.0192, "num_tokens": 15230411.0, "reward": 1.4040178954601288, "reward_std": 0.45906292274594307, "rewards/accuracy_reward": 0.4698660746216774, "rewards/format_reward": 0.9341517835855484, "step": 26 }, { "clip_ratio": 0.0, "completion_length": 555.2734718322754, "epoch": 0.2737642585551331, "grad_norm": 0.2808490991592407, "learning_rate": 9.961019680697591e-07, "loss": 0.0915, "num_tokens": 15856112.0, "reward": 1.460937574505806, "reward_std": 0.4059113450348377, "rewards/accuracy_reward": 0.5189732201397419, "rewards/format_reward": 0.9419642835855484, "step": 27 }, { "clip_ratio": 0.0, "completion_length": 596.7377471923828, "epoch": 0.28390367553865653, "grad_norm": 0.2290179580450058, "learning_rate": 9.949107209404663e-07, "loss": 0.0252, "num_tokens": 16524429.0, "reward": 1.5558036267757416, "reward_std": 0.35689178109169006, "rewards/accuracy_reward": 0.6015625037252903, "rewards/format_reward": 0.9542410671710968, "step": 28 }, { "clip_ratio": 0.0, "completion_length": 586.0000267028809, "epoch": 0.29404309252217997, "grad_norm": 0.2592538595199585, "learning_rate": 9.935617890443554e-07, "loss": 0.0619, "num_tokens": 17190221.0, "reward": 1.5513393580913544, "reward_std": 0.34823960438370705, "rewards/accuracy_reward": 0.5937500074505806, "rewards/format_reward": 0.9575892984867096, "step": 29 }, { "clip_ratio": 0.0, "completion_length": 659.8370819091797, "epoch": 0.3041825095057034, "grad_norm": 0.1614958941936493, "learning_rate": 9.92055602168058e-07, "loss": 0.0418, "num_tokens": 17926923.0, "reward": 1.5948661416769028, "reward_std": 0.315753024071455, "rewards/accuracy_reward": 0.6261160783469677, "rewards/format_reward": 0.9687499925494194, "step": 30 }, { "clip_ratio": 0.0, "completion_length": 628.6830673217773, "epoch": 0.31432192648922685, "grad_norm": 0.13593189418315887, "learning_rate": 9.90392640201615e-07, "loss": 0.034, "num_tokens": 18617831.0, "reward": 1.6785715073347092, "reward_std": 0.24619607254862785, "rewards/accuracy_reward": 0.6975446492433548, "rewards/format_reward": 0.9810267835855484, "step": 31 }, { "clip_ratio": 0.0, "completion_length": 635.0971374511719, "epoch": 0.3244613434727503, "grad_norm": 0.14025020599365234, "learning_rate": 9.885734329855797e-07, "loss": 0.0489, "num_tokens": 19306622.0, "reward": 1.6830357909202576, "reward_std": 0.25594867765903473, "rewards/accuracy_reward": 0.7087053582072258, "rewards/format_reward": 0.9743303656578064, "step": 32 }, { "clip_ratio": 0.0, "completion_length": 662.7310485839844, "epoch": 0.33460076045627374, "grad_norm": 0.14114561676979065, "learning_rate": 9.865985601422017e-07, "loss": 0.0403, "num_tokens": 20034485.0, "reward": 1.5926340073347092, "reward_std": 0.2743757776916027, "rewards/accuracy_reward": 0.6274896934628487, "rewards/format_reward": 0.9698660746216774, "step": 33 }, { "clip_ratio": 0.0, "completion_length": 674.9565048217773, "epoch": 0.34474017743979724, "grad_norm": 0.13052859902381897, "learning_rate": 9.844686508907537e-07, "loss": 0.036, "num_tokens": 20766734.0, "reward": 1.6595983058214188, "reward_std": 0.22791905514895916, "rewards/accuracy_reward": 0.676339291036129, "rewards/format_reward": 0.983258917927742, "step": 34 }, { "clip_ratio": 0.0, "completion_length": 661.467658996582, "epoch": 0.3548795944233207, "grad_norm": 0.14716672897338867, "learning_rate": 9.821843838470534e-07, "loss": 0.0801, "num_tokens": 21502137.0, "reward": 1.658482238650322, "reward_std": 0.2525596469640732, "rewards/accuracy_reward": 0.6774553507566452, "rewards/format_reward": 0.9810267835855484, "step": 35 }, { "clip_ratio": 0.0, "completion_length": 649.9118576049805, "epoch": 0.3650190114068441, "grad_norm": 0.1431976705789566, "learning_rate": 9.797464868072486e-07, "loss": 0.0625, "num_tokens": 22212570.0, "reward": 1.6852679550647736, "reward_std": 0.2624010592699051, "rewards/accuracy_reward": 0.7098214253783226, "rewards/format_reward": 0.9754464253783226, "step": 36 }, { "clip_ratio": 0.0, "completion_length": 653.161865234375, "epoch": 0.37515842839036756, "grad_norm": 0.10097628086805344, "learning_rate": 9.771557365159319e-07, "loss": 0.0255, "num_tokens": 22923331.0, "reward": 1.6283482909202576, "reward_std": 0.19372693821787834, "rewards/accuracy_reward": 0.6428571417927742, "rewards/format_reward": 0.9854910671710968, "step": 37 }, { "clip_ratio": 0.0, "completion_length": 655.9073944091797, "epoch": 0.385297845373891, "grad_norm": 0.1087433397769928, "learning_rate": 9.744129584186597e-07, "loss": 0.0387, "num_tokens": 23637632.0, "reward": 1.6417411416769028, "reward_std": 0.2093806378543377, "rewards/accuracy_reward": 0.6629464328289032, "rewards/format_reward": 0.9787946417927742, "step": 38 }, { "clip_ratio": 0.0, "completion_length": 703.9297103881836, "epoch": 0.39543726235741444, "grad_norm": 0.10768112540245056, "learning_rate": 9.71519026398956e-07, "loss": 0.0512, "num_tokens": 24406137.0, "reward": 1.6138393729925156, "reward_std": 0.19195392541587353, "rewards/accuracy_reward": 0.6428571455180645, "rewards/format_reward": 0.9709821417927742, "step": 39 }, { "clip_ratio": 0.0, "completion_length": 626.8917617797852, "epoch": 0.4055766793409379, "grad_norm": 0.11539029330015182, "learning_rate": 9.68474862499881e-07, "loss": 0.0277, "num_tokens": 25087096.0, "reward": 1.7276786267757416, "reward_std": 0.16980798542499542, "rewards/accuracy_reward": 0.7366071343421936, "rewards/format_reward": 0.9910714253783226, "step": 40 }, { "clip_ratio": 0.0, "completion_length": 641.186408996582, "epoch": 0.4157160963244613, "grad_norm": 0.0916220024228096, "learning_rate": 9.652814366302568e-07, "loss": 0.0297, "num_tokens": 25787575.0, "reward": 1.7444197237491608, "reward_std": 0.15002207644283772, "rewards/accuracy_reward": 0.7566964328289032, "rewards/format_reward": 0.9877232015132904, "step": 41 }, { "clip_ratio": 0.0, "completion_length": 635.1529388427734, "epoch": 0.42585551330798477, "grad_norm": 0.1067587211728096, "learning_rate": 9.619397662556433e-07, "loss": 0.0296, "num_tokens": 26485640.0, "reward": 1.6930804401636124, "reward_std": 0.1523909503594041, "rewards/accuracy_reward": 0.7064732164144516, "rewards/format_reward": 0.9866071343421936, "step": 42 }, { "clip_ratio": 0.0, "completion_length": 686.982177734375, "epoch": 0.43599493029150826, "grad_norm": 0.09716176986694336, "learning_rate": 9.5845091607416e-07, "loss": 0.0241, "num_tokens": 27219880.0, "reward": 1.7142858058214188, "reward_std": 0.19766423292458057, "rewards/accuracy_reward": 0.731026791036129, "rewards/format_reward": 0.9832589253783226, "step": 43 }, { "clip_ratio": 0.0, "completion_length": 692.6172103881836, "epoch": 0.4461343472750317, "grad_norm": 0.0998157262802124, "learning_rate": 9.548159976772592e-07, "loss": 0.0304, "num_tokens": 27970513.0, "reward": 1.6975447237491608, "reward_std": 0.2129265610128641, "rewards/accuracy_reward": 0.7120535746216774, "rewards/format_reward": 0.9854910597205162, "step": 44 }, { "clip_ratio": 0.0, "completion_length": 717.6428833007812, "epoch": 0.45627376425855515, "grad_norm": 0.110740527510643, "learning_rate": 9.510361691955606e-07, "loss": 0.0306, "num_tokens": 28735969.0, "reward": 1.6662947237491608, "reward_std": 0.25920049473643303, "rewards/accuracy_reward": 0.6796875, "rewards/format_reward": 0.9866071343421936, "step": 45 }, { "clip_ratio": 0.0, "completion_length": 651.1373062133789, "epoch": 0.4664131812420786, "grad_norm": 0.10963339358568192, "learning_rate": 9.471126349298556e-07, "loss": 0.031, "num_tokens": 29449604.0, "reward": 1.7377232760190964, "reward_std": 0.19720236957073212, "rewards/accuracy_reward": 0.75, "rewards/format_reward": 0.987723208963871, "step": 46 }, { "clip_ratio": 0.0, "completion_length": 658.0803909301758, "epoch": 0.47655259822560203, "grad_norm": 0.11225859075784683, "learning_rate": 9.430466449674013e-07, "loss": 0.0419, "num_tokens": 30187596.0, "reward": 1.656250074505806, "reward_std": 0.19186341715976596, "rewards/accuracy_reward": 0.6662946417927742, "rewards/format_reward": 0.9899553507566452, "step": 47 }, { "clip_ratio": 0.0, "completion_length": 715.3538360595703, "epoch": 0.4866920152091255, "grad_norm": 0.09579791873693466, "learning_rate": 9.388394947836278e-07, "loss": 0.0133, "num_tokens": 30955289.0, "reward": 1.6250000596046448, "reward_std": 0.1923097288236022, "rewards/accuracy_reward": 0.6372767835855484, "rewards/format_reward": 0.987723208963871, "step": 48 }, { "clip_ratio": 0.0, "completion_length": 694.5614242553711, "epoch": 0.4968314321926489, "grad_norm": 0.10001735389232635, "learning_rate": 9.344925248293835e-07, "loss": 0.0175, "num_tokens": 31698624.0, "reward": 1.6986608058214188, "reward_std": 0.1780099142342806, "rewards/accuracy_reward": 0.7087053582072258, "rewards/format_reward": 0.9899553507566452, "step": 49 }, { "clip_ratio": 0.0, "completion_length": 673.1484756469727, "epoch": 0.5069708491761724, "grad_norm": 0.10267457365989685, "learning_rate": 9.300071201038501e-07, "loss": 0.0246, "num_tokens": 32427941.0, "reward": 1.6919643878936768, "reward_std": 0.203399364836514, "rewards/accuracy_reward": 0.703125, "rewards/format_reward": 0.9888392835855484, "step": 50 }, { "clip_ratio": 0.0, "completion_length": 617.7199020385742, "epoch": 0.5171102661596958, "grad_norm": 0.10793677717447281, "learning_rate": 9.253847097132655e-07, "loss": 0.0196, "num_tokens": 33112242.0, "reward": 1.7388393580913544, "reward_std": 0.20277155563235283, "rewards/accuracy_reward": 0.7477678582072258, "rewards/format_reward": 0.9910714253783226, "step": 51 }, { "clip_ratio": 0.0, "completion_length": 659.7522659301758, "epoch": 0.5272496831432193, "grad_norm": 0.09107255190610886, "learning_rate": 9.206267664155906e-07, "loss": 0.034, "num_tokens": 33823756.0, "reward": 1.7142857909202576, "reward_std": 0.15979534853249788, "rewards/accuracy_reward": 0.7243303582072258, "rewards/format_reward": 0.9899553507566452, "step": 52 }, { "clip_ratio": 0.0, "completion_length": 681.6015930175781, "epoch": 0.5373891001267427, "grad_norm": 0.10401283204555511, "learning_rate": 9.157348061512726e-07, "loss": 0.0255, "num_tokens": 34570551.0, "reward": 1.6975447088479996, "reward_std": 0.19325901474803686, "rewards/accuracy_reward": 0.7131696417927742, "rewards/format_reward": 0.9843749925494194, "step": 53 }, { "clip_ratio": 0.0, "completion_length": 733.9632034301758, "epoch": 0.5475285171102662, "grad_norm": 0.09657347947359085, "learning_rate": 9.107103875602458e-07, "loss": 0.0269, "num_tokens": 35346734.0, "reward": 1.665178656578064, "reward_std": 0.19283783435821533, "rewards/accuracy_reward": 0.6808035671710968, "rewards/format_reward": 0.9843749925494194, "step": 54 }, { "clip_ratio": 0.0, "completion_length": 695.3471298217773, "epoch": 0.5576679340937896, "grad_norm": 0.11091943085193634, "learning_rate": 9.055551114853295e-07, "loss": 0.0383, "num_tokens": 36100157.0, "reward": 1.6729911416769028, "reward_std": 0.2295996304601431, "rewards/accuracy_reward": 0.6874999925494194, "rewards/format_reward": 0.9854910671710968, "step": 55 }, { "clip_ratio": 0.0, "completion_length": 623.5301666259766, "epoch": 0.5678073510773131, "grad_norm": 0.1108093112707138, "learning_rate": 9.002706204621802e-07, "loss": 0.0293, "num_tokens": 36782520.0, "reward": 1.7477679401636124, "reward_std": 0.20551220327615738, "rewards/accuracy_reward": 0.7544642835855484, "rewards/format_reward": 0.9933035671710968, "step": 56 }, { "clip_ratio": 0.0, "completion_length": 666.6451263427734, "epoch": 0.5779467680608364, "grad_norm": 0.10158971697092056, "learning_rate": 8.948585981959578e-07, "loss": 0.0202, "num_tokens": 37511418.0, "reward": 1.7053572535514832, "reward_std": 0.17258675582706928, "rewards/accuracy_reward": 0.7154017761349678, "rewards/format_reward": 0.9899553433060646, "step": 57 }, { "clip_ratio": 0.0, "completion_length": 651.3370819091797, "epoch": 0.5880861850443599, "grad_norm": 0.09192176908254623, "learning_rate": 8.893207690248775e-07, "loss": 0.0075, "num_tokens": 38216080.0, "reward": 1.7600447535514832, "reward_std": 0.1495125712826848, "rewards/accuracy_reward": 0.7633928582072258, "rewards/format_reward": 0.9966517835855484, "step": 58 }, { "clip_ratio": 0.0, "completion_length": 651.7176742553711, "epoch": 0.5982256020278834, "grad_norm": 0.11554893106222153, "learning_rate": 8.836588973708128e-07, "loss": 0.045, "num_tokens": 38941611.0, "reward": 1.688616156578064, "reward_std": 0.20678117126226425, "rewards/accuracy_reward": 0.7042410708963871, "rewards/format_reward": 0.9843749925494194, "step": 59 }, { "clip_ratio": 0.0, "completion_length": 673.3739242553711, "epoch": 0.6083650190114068, "grad_norm": 0.11070550233125687, "learning_rate": 8.778747871771291e-07, "loss": 0.0224, "num_tokens": 39682202.0, "reward": 1.7388393729925156, "reward_std": 0.18264728784561157, "rewards/accuracy_reward": 0.7444196492433548, "rewards/format_reward": 0.9944196417927742, "step": 60 }, { "clip_ratio": 0.0, "completion_length": 649.6138610839844, "epoch": 0.6185044359949303, "grad_norm": 0.11238072067499161, "learning_rate": 8.719702813339247e-07, "loss": 0.0075, "num_tokens": 40386600.0, "reward": 1.6919643729925156, "reward_std": 0.20359325036406517, "rewards/accuracy_reward": 0.7020089328289032, "rewards/format_reward": 0.9899553507566452, "step": 61 }, { "clip_ratio": 0.0, "completion_length": 687.9553833007812, "epoch": 0.6286438529784537, "grad_norm": 0.09478152543306351, "learning_rate": 8.659472610908627e-07, "loss": 0.0211, "num_tokens": 41132400.0, "reward": 1.6729911416769028, "reward_std": 0.17133308481425047, "rewards/accuracy_reward": 0.6785714291036129, "rewards/format_reward": 0.9944196343421936, "step": 62 }, { "clip_ratio": 0.0, "completion_length": 615.4788284301758, "epoch": 0.6387832699619772, "grad_norm": 0.12343282252550125, "learning_rate": 8.598076454577814e-07, "loss": 0.0078, "num_tokens": 41808165.0, "reward": 1.7689732909202576, "reward_std": 0.18264547362923622, "rewards/accuracy_reward": 0.7734375, "rewards/format_reward": 0.995535708963871, "step": 63 }, { "clip_ratio": 0.0, "completion_length": 636.3192367553711, "epoch": 0.6489226869455006, "grad_norm": 0.08991697430610657, "learning_rate": 8.535533905932737e-07, "loss": 0.0267, "num_tokens": 42526435.0, "reward": 1.7232143729925156, "reward_std": 0.14434537151828408, "rewards/accuracy_reward": 0.7287946492433548, "rewards/format_reward": 0.9944196417927742, "step": 64 }, { "clip_ratio": 0.0, "completion_length": 627.6741409301758, "epoch": 0.6590621039290241, "grad_norm": 0.091641865670681, "learning_rate": 8.471864891814304e-07, "loss": 0.0296, "num_tokens": 43207935.0, "reward": 1.7488840073347092, "reward_std": 0.16924036061391234, "rewards/accuracy_reward": 0.7589285671710968, "rewards/format_reward": 0.9899553507566452, "step": 65 }, { "clip_ratio": 0.0, "completion_length": 697.7109680175781, "epoch": 0.6692015209125475, "grad_norm": 0.09379775822162628, "learning_rate": 8.407089697969456e-07, "loss": 0.0402, "num_tokens": 43957948.0, "reward": 1.705357238650322, "reward_std": 0.17076639458537102, "rewards/accuracy_reward": 0.719866082072258, "rewards/format_reward": 0.9854910671710968, "step": 66 }, { "clip_ratio": 0.0, "completion_length": 719.4453353881836, "epoch": 0.679340937896071, "grad_norm": 0.09100989252328873, "learning_rate": 8.341228962587881e-07, "loss": 0.0285, "num_tokens": 44738699.0, "reward": 1.6875001192092896, "reward_std": 0.18435446824878454, "rewards/accuracy_reward": 0.7031250149011612, "rewards/format_reward": 0.9843749850988388, "step": 67 }, { "clip_ratio": 0.0, "completion_length": 604.7611961364746, "epoch": 0.6894803548795945, "grad_norm": 0.1007450744509697, "learning_rate": 8.274303669726426e-07, "loss": 0.022, "num_tokens": 45401885.0, "reward": 1.7589286416769028, "reward_std": 0.1566725349985063, "rewards/accuracy_reward": 0.7678571455180645, "rewards/format_reward": 0.9910714253783226, "step": 68 }, { "clip_ratio": 0.0, "completion_length": 717.0123062133789, "epoch": 0.6996197718631179, "grad_norm": 0.10416340827941895, "learning_rate": 8.206335142623304e-07, "loss": 0.0253, "num_tokens": 46174088.0, "reward": 1.6651786267757416, "reward_std": 0.19855336844921112, "rewards/accuracy_reward": 0.671875, "rewards/format_reward": 0.9933035597205162, "step": 69 }, { "clip_ratio": 0.0, "completion_length": 669.8147583007812, "epoch": 0.7097591888466414, "grad_norm": 0.11590211093425751, "learning_rate": 8.137345036904259e-07, "loss": 0.0433, "num_tokens": 46893618.0, "reward": 1.7075893580913544, "reward_std": 0.21949494630098343, "rewards/accuracy_reward": 0.723214291036129, "rewards/format_reward": 0.9843749925494194, "step": 70 }, { "clip_ratio": 0.0, "completion_length": 648.5803756713867, "epoch": 0.7198986058301647, "grad_norm": 0.10192694514989853, "learning_rate": 8.067355333682797e-07, "loss": 0.0238, "num_tokens": 47613514.0, "reward": 1.66183041036129, "reward_std": 0.17979866359382868, "rewards/accuracy_reward": 0.6796875074505806, "rewards/format_reward": 0.9821428582072258, "step": 71 }, { "clip_ratio": 0.0, "completion_length": 658.716552734375, "epoch": 0.7300380228136882, "grad_norm": 0.10270233452320099, "learning_rate": 7.996388332556734e-07, "loss": 0.0461, "num_tokens": 48332444.0, "reward": 1.7109375894069672, "reward_std": 0.17445513978600502, "rewards/accuracy_reward": 0.7187500149011612, "rewards/format_reward": 0.9921874925494194, "step": 72 }, { "clip_ratio": 0.0, "completion_length": 727.381721496582, "epoch": 0.7401774397972116, "grad_norm": 0.09961652010679245, "learning_rate": 7.924466644503264e-07, "loss": 0.0144, "num_tokens": 49116674.0, "reward": 1.6529018580913544, "reward_std": 0.20062358025461435, "rewards/accuracy_reward": 0.6662946529686451, "rewards/format_reward": 0.9866071417927742, "step": 73 }, { "clip_ratio": 0.0, "completion_length": 648.5078430175781, "epoch": 0.7503168567807351, "grad_norm": 0.10154278576374054, "learning_rate": 7.85161318467482e-07, "loss": 0.0184, "num_tokens": 49833121.0, "reward": 1.765625074505806, "reward_std": 0.17119968868792057, "rewards/accuracy_reward": 0.7734374925494194, "rewards/format_reward": 0.9921874925494194, "step": 74 }, { "clip_ratio": 0.0, "completion_length": 657.3549385070801, "epoch": 0.7604562737642585, "grad_norm": 0.11170431971549988, "learning_rate": 7.777851165098011e-07, "loss": 0.0291, "num_tokens": 50552791.0, "reward": 1.7522321939468384, "reward_std": 0.18582619447261095, "rewards/accuracy_reward": 0.7589285671710968, "rewards/format_reward": 0.9933035597205162, "step": 75 }, { "clip_ratio": 0.0, "completion_length": 618.2310562133789, "epoch": 0.770595690747782, "grad_norm": 0.09891778975725174, "learning_rate": 7.703204087277988e-07, "loss": 0.0087, "num_tokens": 51233966.0, "reward": 1.6707590073347092, "reward_std": 0.16447317646816373, "rewards/accuracy_reward": 0.6785714402794838, "rewards/format_reward": 0.9921874925494194, "step": 76 }, { "clip_ratio": 0.0, "completion_length": 691.1406631469727, "epoch": 0.7807351077313055, "grad_norm": 0.0960167646408081, "learning_rate": 7.627695734710564e-07, "loss": 0.0304, "num_tokens": 51975452.0, "reward": 1.7042411714792252, "reward_std": 0.1650120271369815, "rewards/accuracy_reward": 0.7220982164144516, "rewards/format_reward": 0.9821428507566452, "step": 77 }, { "clip_ratio": 0.0, "completion_length": 684.7500228881836, "epoch": 0.7908745247148289, "grad_norm": 0.09193353354930878, "learning_rate": 7.551350165304499e-07, "loss": 0.0281, "num_tokens": 52706764.0, "reward": 1.7488840222358704, "reward_std": 0.16149256937205791, "rewards/accuracy_reward": 0.7622767835855484, "rewards/format_reward": 0.9866071417927742, "step": 78 }, { "clip_ratio": 0.0, "completion_length": 661.9587326049805, "epoch": 0.8010139416983524, "grad_norm": 0.08813660591840744, "learning_rate": 7.474191703716338e-07, "loss": 0.0161, "num_tokens": 53430439.0, "reward": 1.7209822237491608, "reward_std": 0.14920138847082853, "rewards/accuracy_reward": 0.723214291036129, "rewards/format_reward": 0.9977678507566452, "step": 79 }, { "clip_ratio": 0.0, "completion_length": 657.2868576049805, "epoch": 0.8111533586818758, "grad_norm": 0.08803500235080719, "learning_rate": 7.396244933600284e-07, "loss": 0.0076, "num_tokens": 54143656.0, "reward": 1.7421875894069672, "reward_std": 0.15661878790706396, "rewards/accuracy_reward": 0.7433035746216774, "rewards/format_reward": 0.9988839253783226, "step": 80 }, { "clip_ratio": 0.0, "completion_length": 653.8337326049805, "epoch": 0.8212927756653993, "grad_norm": 0.11003640294075012, "learning_rate": 7.317534689775527e-07, "loss": 0.0296, "num_tokens": 54849195.0, "reward": 1.7332590073347092, "reward_std": 0.1878818403929472, "rewards/accuracy_reward": 0.737723208963871, "rewards/format_reward": 0.9955357015132904, "step": 81 }, { "clip_ratio": 0.0, "completion_length": 646.4810562133789, "epoch": 0.8314321926489227, "grad_norm": 0.10392062366008759, "learning_rate": 7.238086050313562e-07, "loss": 0.0264, "num_tokens": 55563698.0, "reward": 1.7165179401636124, "reward_std": 0.20238416641950607, "rewards/accuracy_reward": 0.7254464328289032, "rewards/format_reward": 0.9910714253783226, "step": 82 }, { "clip_ratio": 0.0, "completion_length": 661.6440048217773, "epoch": 0.8415716096324461, "grad_norm": 0.11124070733785629, "learning_rate": 7.157924328548002e-07, "loss": 0.0144, "num_tokens": 56290123.0, "reward": 1.7410715073347092, "reward_std": 0.16893591731786728, "rewards/accuracy_reward": 0.7477678582072258, "rewards/format_reward": 0.9933035671710968, "step": 83 }, { "clip_ratio": 0.0, "completion_length": 667.6004867553711, "epoch": 0.8517110266159695, "grad_norm": 0.1015985757112503, "learning_rate": 7.077075065009433e-07, "loss": 0.0289, "num_tokens": 57021605.0, "reward": 1.7712054401636124, "reward_std": 0.1773331630975008, "rewards/accuracy_reward": 0.7790178582072258, "rewards/format_reward": 0.9921874925494194, "step": 84 }, { "clip_ratio": 0.0, "completion_length": 654.0971298217773, "epoch": 0.861850443599493, "grad_norm": 0.10410565882921219, "learning_rate": 6.995564019287869e-07, "loss": 0.0127, "num_tokens": 57737492.0, "reward": 1.6863840073347092, "reward_std": 0.2015578467398882, "rewards/accuracy_reward": 0.6919642835855484, "rewards/format_reward": 0.9944196343421936, "step": 85 }, { "clip_ratio": 0.0, "completion_length": 705.3315048217773, "epoch": 0.8719898605830165, "grad_norm": 0.12336766719818115, "learning_rate": 6.913417161825449e-07, "loss": 0.0313, "num_tokens": 58504381.0, "reward": 1.6819197237491608, "reward_std": 0.19116687029600143, "rewards/accuracy_reward": 0.6908482164144516, "rewards/format_reward": 0.991071417927742, "step": 86 }, { "clip_ratio": 0.0, "completion_length": 663.8895492553711, "epoch": 0.8821292775665399, "grad_norm": 0.09500221908092499, "learning_rate": 6.830660665641897e-07, "loss": 0.0007, "num_tokens": 59228946.0, "reward": 1.7243304252624512, "reward_std": 0.1611752938479185, "rewards/accuracy_reward": 0.731026791036129, "rewards/format_reward": 0.9933035671710968, "step": 87 }, { "clip_ratio": 0.0, "completion_length": 676.3370895385742, "epoch": 0.8922686945500634, "grad_norm": 0.10983681678771973, "learning_rate": 6.747320897995492e-07, "loss": 0.0173, "num_tokens": 59963936.0, "reward": 1.7120536118745804, "reward_std": 0.16041494347155094, "rewards/accuracy_reward": 0.7287946343421936, "rewards/format_reward": 0.983258917927742, "step": 88 }, { "clip_ratio": 0.0, "completion_length": 627.8236999511719, "epoch": 0.9024081115335868, "grad_norm": 0.10716623067855835, "learning_rate": 6.66342441198212e-07, "loss": 0.0255, "num_tokens": 60650370.0, "reward": 1.7276786416769028, "reward_std": 0.16107469610869884, "rewards/accuracy_reward": 0.7366071417927742, "rewards/format_reward": 0.991071417927742, "step": 89 }, { "clip_ratio": 0.0, "completion_length": 641.5011367797852, "epoch": 0.9125475285171103, "grad_norm": 0.12808987498283386, "learning_rate": 6.578997938075125e-07, "loss": 0.024, "num_tokens": 61349627.0, "reward": 1.6953125596046448, "reward_std": 0.18346151895821095, "rewards/accuracy_reward": 0.7109375, "rewards/format_reward": 0.9843749925494194, "step": 90 }, { "clip_ratio": 0.0, "completion_length": 647.2109603881836, "epoch": 0.9226869455006337, "grad_norm": 0.10632172971963882, "learning_rate": 6.494068375608646e-07, "loss": 0.0176, "num_tokens": 62064152.0, "reward": 1.7388393729925156, "reward_std": 0.17309935204684734, "rewards/accuracy_reward": 0.745535708963871, "rewards/format_reward": 0.9933035597205162, "step": 91 }, { "clip_ratio": 0.0, "completion_length": 684.0725860595703, "epoch": 0.9328263624841572, "grad_norm": 0.10359478741884232, "learning_rate": 6.408662784207149e-07, "loss": 0.0327, "num_tokens": 62816457.0, "reward": 1.7265625596046448, "reward_std": 0.18876438401639462, "rewards/accuracy_reward": 0.7399553656578064, "rewards/format_reward": 0.9866071343421936, "step": 92 }, { "clip_ratio": 0.0, "completion_length": 619.8136520385742, "epoch": 0.9429657794676806, "grad_norm": 0.110443115234375, "learning_rate": 6.322808375163895e-07, "loss": 0.0174, "num_tokens": 63498690.0, "reward": 1.7377233058214188, "reward_std": 0.16478892508894205, "rewards/accuracy_reward": 0.7399553582072258, "rewards/format_reward": 0.9977678507566452, "step": 93 }, { "clip_ratio": 0.0, "completion_length": 650.7254791259766, "epoch": 0.9531051964512041, "grad_norm": 0.10725186765193939, "learning_rate": 6.236532502771077e-07, "loss": 0.0106, "num_tokens": 64215340.0, "reward": 1.7031250894069672, "reward_std": 0.18828019313514233, "rewards/accuracy_reward": 0.7109375, "rewards/format_reward": 0.9921875, "step": 94 }, { "clip_ratio": 0.0, "completion_length": 643.1205749511719, "epoch": 0.9632446134347274, "grad_norm": 0.13609780371189117, "learning_rate": 6.149862655604403e-07, "loss": 0.0263, "num_tokens": 64917264.0, "reward": 1.7589286714792252, "reward_std": 0.19211565144360065, "rewards/accuracy_reward": 0.7734375074505806, "rewards/format_reward": 0.9854910671710968, "step": 95 }, { "clip_ratio": 0.0, "completion_length": 640.569221496582, "epoch": 0.973384030418251, "grad_norm": 0.11930994689464569, "learning_rate": 6.062826447764883e-07, "loss": 0.0198, "num_tokens": 65613958.0, "reward": 1.757812574505806, "reward_std": 0.14032188896089792, "rewards/accuracy_reward": 0.7622767761349678, "rewards/format_reward": 0.9955357015132904, "step": 96 }, { "clip_ratio": 0.0, "completion_length": 654.4654388427734, "epoch": 0.9835234474017744, "grad_norm": 0.09683021903038025, "learning_rate": 5.975451610080642e-07, "loss": 0.0272, "num_tokens": 66328167.0, "reward": 1.6875000894069672, "reward_std": 0.1803465113043785, "rewards/accuracy_reward": 0.6953125074505806, "rewards/format_reward": 0.9921874925494194, "step": 97 }, { "clip_ratio": 0.0, "completion_length": 631.353515625, "epoch": 0.9936628643852978, "grad_norm": 0.1017894446849823, "learning_rate": 5.887765981271517e-07, "loss": 0.0292, "num_tokens": 67045777.0, "reward": 1.7265625894069672, "reward_std": 0.1662213634699583, "rewards/accuracy_reward": 0.7366071417927742, "rewards/format_reward": 0.9899553433060646, "step": 98 }, { "clip_ratio": 0.0, "completion_length": 662.9151916503906, "epoch": 1.0101394169835234, "grad_norm": 0.09788638353347778, "learning_rate": 5.7997974990793e-07, "loss": 0.0044, "num_tokens": 67768741.0, "reward": 1.7566965073347092, "reward_std": 0.1626696866005659, "rewards/accuracy_reward": 0.7578125, "rewards/format_reward": 0.9988839253783226, "step": 99 }, { "clip_ratio": 0.0, "completion_length": 659.5960006713867, "epoch": 1.020278833967047, "grad_norm": 0.09406567364931107, "learning_rate": 5.711574191366427e-07, "loss": 0.0125, "num_tokens": 68487115.0, "reward": 1.7321429550647736, "reward_std": 0.1440184023231268, "rewards/accuracy_reward": 0.7366071417927742, "rewards/format_reward": 0.995535708963871, "step": 100 }, { "clip_ratio": 0.0, "completion_length": 664.0189971923828, "epoch": 1.0304182509505704, "grad_norm": 0.10347539931535721, "learning_rate": 5.623124167185929e-07, "loss": 0.0152, "num_tokens": 69204604.0, "reward": 1.7410715222358704, "reward_std": 0.153701429720968, "rewards/accuracy_reward": 0.746651791036129, "rewards/format_reward": 0.9944196417927742, "step": 101 }, { "clip_ratio": 0.0, "completion_length": 623.9464569091797, "epoch": 1.0405576679340938, "grad_norm": 0.09895368665456772, "learning_rate": 5.534475607825565e-07, "loss": 0.033, "num_tokens": 69890948.0, "reward": 1.7165179401636124, "reward_std": 0.13192464038729668, "rewards/accuracy_reward": 0.7232142835855484, "rewards/format_reward": 0.9933035671710968, "step": 102 }, { "clip_ratio": 0.0, "completion_length": 683.225471496582, "epoch": 1.0506970849176172, "grad_norm": 0.11877602338790894, "learning_rate": 5.445656757828879e-07, "loss": 0.0322, "num_tokens": 70634422.0, "reward": 1.6361607909202576, "reward_std": 0.17785416916012764, "rewards/accuracy_reward": 0.645089291036129, "rewards/format_reward": 0.9910714253783226, "step": 103 }, { "clip_ratio": 0.0, "completion_length": 662.256721496582, "epoch": 1.0608365019011408, "grad_norm": 0.10182492434978485, "learning_rate": 5.356695915996161e-07, "loss": 0.0056, "num_tokens": 71362164.0, "reward": 1.7165179252624512, "reward_std": 0.15059774369001389, "rewards/accuracy_reward": 0.7187499925494194, "rewards/format_reward": 0.9977678582072258, "step": 104 }, { "clip_ratio": 0.0, "completion_length": 626.7701187133789, "epoch": 1.0709759188846641, "grad_norm": 0.1029505580663681, "learning_rate": 5.267621426368075e-07, "loss": 0.0062, "num_tokens": 72042966.0, "reward": 1.797991156578064, "reward_std": 0.14701888849958777, "rewards/accuracy_reward": 0.8013392835855484, "rewards/format_reward": 0.9966517761349678, "step": 105 }, { "clip_ratio": 0.0, "completion_length": 577.3560562133789, "epoch": 1.0811153358681875, "grad_norm": 0.10245715826749802, "learning_rate": 5.178461669194903e-07, "loss": 0.013, "num_tokens": 72687485.0, "reward": 1.7946429401636124, "reward_std": 0.14055991917848587, "rewards/accuracy_reward": 0.7991071417927742, "rewards/format_reward": 0.9955357164144516, "step": 106 }, { "clip_ratio": 0.0, "completion_length": 634.2232437133789, "epoch": 1.091254752851711, "grad_norm": 0.09775260090827942, "learning_rate": 5.08924505189423e-07, "loss": 0.0073, "num_tokens": 73381709.0, "reward": 1.77120541036129, "reward_std": 0.13471420388668776, "rewards/accuracy_reward": 0.7745535597205162, "rewards/format_reward": 0.9966517835855484, "step": 107 }, { "clip_ratio": 0.0, "completion_length": 671.7980308532715, "epoch": 1.1013941698352345, "grad_norm": 0.10289379954338074, "learning_rate": 5e-07, "loss": 0.0137, "num_tokens": 74107760.0, "reward": 1.7578125894069672, "reward_std": 0.16642380319535732, "rewards/accuracy_reward": 0.765625, "rewards/format_reward": 0.9921875, "step": 108 }, { "clip_ratio": 0.0, "completion_length": 700.1719055175781, "epoch": 1.111533586818758, "grad_norm": 0.08667897433042526, "learning_rate": 4.91075494810577e-07, "loss": 0.0214, "num_tokens": 74853106.0, "reward": 1.7477679252624512, "reward_std": 0.14406709745526314, "rewards/accuracy_reward": 0.7555803433060646, "rewards/format_reward": 0.9921874925494194, "step": 109 }, { "clip_ratio": 0.0, "completion_length": 654.7210083007812, "epoch": 1.1216730038022813, "grad_norm": 0.10183614492416382, "learning_rate": 4.821538330805098e-07, "loss": 0.0228, "num_tokens": 75566256.0, "reward": 1.7477679401636124, "reward_std": 0.16557483188807964, "rewards/accuracy_reward": 0.7600446417927742, "rewards/format_reward": 0.9877232015132904, "step": 110 }, { "clip_ratio": 0.0, "completion_length": 668.025691986084, "epoch": 1.131812420785805, "grad_norm": 0.11714053899049759, "learning_rate": 4.732378573631924e-07, "loss": 0.0246, "num_tokens": 76287535.0, "reward": 1.7421875894069672, "reward_std": 0.1860833875834942, "rewards/accuracy_reward": 0.7511160746216774, "rewards/format_reward": 0.9910714253783226, "step": 111 }, { "clip_ratio": 0.0, "completion_length": 669.5022659301758, "epoch": 1.1419518377693283, "grad_norm": 0.11465004831552505, "learning_rate": 4.643304084003838e-07, "loss": 0.0245, "num_tokens": 77014857.0, "reward": 1.7254465222358704, "reward_std": 0.19485904835164547, "rewards/accuracy_reward": 0.7321428656578064, "rewards/format_reward": 0.9933035671710968, "step": 112 }, { "clip_ratio": 0.0, "completion_length": 656.6093978881836, "epoch": 1.1520912547528517, "grad_norm": 0.160664901137352, "learning_rate": 4.55434324217112e-07, "loss": 0.0228, "num_tokens": 77732419.0, "reward": 1.6763393431901932, "reward_std": 0.1846153624355793, "rewards/accuracy_reward": 0.684151791036129, "rewards/format_reward": 0.9921874925494194, "step": 113 }, { "clip_ratio": 0.0, "completion_length": 655.0524826049805, "epoch": 1.162230671736375, "grad_norm": 0.10615188628435135, "learning_rate": 4.4655243921744367e-07, "loss": 0.0136, "num_tokens": 78444306.0, "reward": 1.695312574505806, "reward_std": 0.14989868365228176, "rewards/accuracy_reward": 0.7020089253783226, "rewards/format_reward": 0.9933035671710968, "step": 114 }, { "clip_ratio": 0.0, "completion_length": 655.928596496582, "epoch": 1.1723700887198987, "grad_norm": 0.1310528963804245, "learning_rate": 4.37687583281407e-07, "loss": 0.0203, "num_tokens": 79159234.0, "reward": 1.7176340222358704, "reward_std": 0.15918411500751972, "rewards/accuracy_reward": 0.7243303656578064, "rewards/format_reward": 0.9933035671710968, "step": 115 }, { "clip_ratio": 0.0, "completion_length": 618.7656478881836, "epoch": 1.182509505703422, "grad_norm": 0.11190109699964523, "learning_rate": 4.2884258086335745e-07, "loss": 0.0031, "num_tokens": 79852888.0, "reward": 1.774553656578064, "reward_std": 0.17635854333639145, "rewards/accuracy_reward": 0.7767857164144516, "rewards/format_reward": 0.9977678507566452, "step": 116 }, { "clip_ratio": 0.0, "completion_length": 629.7489166259766, "epoch": 1.1926489226869454, "grad_norm": 0.09713304787874222, "learning_rate": 4.2002025009206987e-07, "loss": 0.0098, "num_tokens": 80549095.0, "reward": 1.7287947237491608, "reward_std": 0.12099508009850979, "rewards/accuracy_reward": 0.7321428619325161, "rewards/format_reward": 0.9966517761349678, "step": 117 }, { "clip_ratio": 0.0, "completion_length": 615.9765968322754, "epoch": 1.202788339670469, "grad_norm": 0.11283177137374878, "learning_rate": 4.1122340187284845e-07, "loss": 0.0315, "num_tokens": 81223298.0, "reward": 1.7388393580913544, "reward_std": 0.13626329507678747, "rewards/accuracy_reward": 0.745535708963871, "rewards/format_reward": 0.9933035671710968, "step": 118 }, { "clip_ratio": 0.0, "completion_length": 641.6629791259766, "epoch": 1.2129277566539924, "grad_norm": 0.0984925627708435, "learning_rate": 4.0245483899193586e-07, "loss": 0.0086, "num_tokens": 81925604.0, "reward": 1.7310268431901932, "reward_std": 0.16442857310175896, "rewards/accuracy_reward": 0.7354910671710968, "rewards/format_reward": 0.995535708963871, "step": 119 }, { "clip_ratio": 0.0, "completion_length": 628.5569534301758, "epoch": 1.2230671736375158, "grad_norm": 0.09760965406894684, "learning_rate": 3.937173552235116e-07, "loss": 0.0041, "num_tokens": 82621599.0, "reward": 1.7053572088479996, "reward_std": 0.11313999397680163, "rewards/accuracy_reward": 0.7098214253783226, "rewards/format_reward": 0.9955357164144516, "step": 120 }, { "clip_ratio": 0.0, "completion_length": 645.1350860595703, "epoch": 1.2332065906210392, "grad_norm": 0.08554679900407791, "learning_rate": 3.850137344395598e-07, "loss": 0.0151, "num_tokens": 83322856.0, "reward": 1.7354911714792252, "reward_std": 0.11471607023850083, "rewards/accuracy_reward": 0.7399553582072258, "rewards/format_reward": 0.995535708963871, "step": 121 }, { "clip_ratio": 0.0, "completion_length": 688.2924423217773, "epoch": 1.2433460076045628, "grad_norm": 0.11858882755041122, "learning_rate": 3.763467497228922e-07, "loss": 0.0248, "num_tokens": 84077598.0, "reward": 1.678571492433548, "reward_std": 0.15971866622567177, "rewards/accuracy_reward": 0.6930803582072258, "rewards/format_reward": 0.9854910597205162, "step": 122 }, { "clip_ratio": 0.0, "completion_length": 655.8861923217773, "epoch": 1.2534854245880862, "grad_norm": 0.09887714684009552, "learning_rate": 3.677191624836106e-07, "loss": 0.0222, "num_tokens": 84786984.0, "reward": 1.7790179252624512, "reward_std": 0.15774498507380486, "rewards/accuracy_reward": 0.787946417927742, "rewards/format_reward": 0.991071417927742, "step": 123 }, { "clip_ratio": 0.0, "completion_length": 663.8951187133789, "epoch": 1.2636248415716096, "grad_norm": 0.11492931097745895, "learning_rate": 3.591337215792851e-07, "loss": 0.019, "num_tokens": 85510002.0, "reward": 1.7433036416769028, "reward_std": 0.18467565812170506, "rewards/accuracy_reward": 0.7522321417927742, "rewards/format_reward": 0.9910714253783226, "step": 124 }, { "clip_ratio": 0.0, "completion_length": 623.6551513671875, "epoch": 1.2737642585551332, "grad_norm": 0.10936412215232849, "learning_rate": 3.505931624391355e-07, "loss": 0.0138, "num_tokens": 86187061.0, "reward": 1.7689732760190964, "reward_std": 0.15780174732208252, "rewards/accuracy_reward": 0.7712053433060646, "rewards/format_reward": 0.9977678507566452, "step": 125 }, { "clip_ratio": 0.0, "completion_length": 634.5870895385742, "epoch": 1.2839036755386566, "grad_norm": 0.11182162165641785, "learning_rate": 3.421002061924876e-07, "loss": 0.0187, "num_tokens": 86882515.0, "reward": 1.7410715073347092, "reward_std": 0.17043365351855755, "rewards/accuracy_reward": 0.746651791036129, "rewards/format_reward": 0.9944196417927742, "step": 126 }, { "clip_ratio": 0.0, "completion_length": 646.7031555175781, "epoch": 1.29404309252218, "grad_norm": 0.10373136401176453, "learning_rate": 3.3365755880178807e-07, "loss": 0.0117, "num_tokens": 87583665.0, "reward": 1.781250074505806, "reward_std": 0.1267631771042943, "rewards/accuracy_reward": 0.7845982238650322, "rewards/format_reward": 0.9966517761349678, "step": 127 }, { "clip_ratio": 0.0, "completion_length": 675.5312728881836, "epoch": 1.3041825095057034, "grad_norm": 0.10996940732002258, "learning_rate": 3.2526791020045087e-07, "loss": 0.0199, "num_tokens": 88321541.0, "reward": 1.702008992433548, "reward_std": 0.18257617205381393, "rewards/accuracy_reward": 0.7120535671710968, "rewards/format_reward": 0.9899553582072258, "step": 128 }, { "clip_ratio": 0.0, "completion_length": 629.1652069091797, "epoch": 1.3143219264892267, "grad_norm": 0.1037248894572258, "learning_rate": 3.169339334358104e-07, "loss": 0.0005, "num_tokens": 89018617.0, "reward": 1.733258992433548, "reward_std": 0.15833040233701468, "rewards/accuracy_reward": 0.7366071417927742, "rewards/format_reward": 0.9966517761349678, "step": 129 }, { "clip_ratio": 0.0, "completion_length": 661.2622985839844, "epoch": 1.3244613434727504, "grad_norm": 0.1321147084236145, "learning_rate": 3.086582838174551e-07, "loss": 0.0124, "num_tokens": 89743596.0, "reward": 1.7109375596046448, "reward_std": 0.19106812588870525, "rewards/accuracy_reward": 0.7198660671710968, "rewards/format_reward": 0.9910714253783226, "step": 130 }, { "clip_ratio": 0.0, "completion_length": 659.3091735839844, "epoch": 1.3346007604562737, "grad_norm": 0.09277363866567612, "learning_rate": 3.004435980712129e-07, "loss": 0.0114, "num_tokens": 90457801.0, "reward": 1.727678656578064, "reward_std": 0.13339263666421175, "rewards/accuracy_reward": 0.7366071492433548, "rewards/format_reward": 0.9910714253783226, "step": 131 }, { "clip_ratio": 0.0, "completion_length": 630.7388687133789, "epoch": 1.3447401774397973, "grad_norm": 0.09743496030569077, "learning_rate": 2.922924934990568e-07, "loss": 0.0247, "num_tokens": 91148095.0, "reward": 1.7700893729925156, "reward_std": 0.1285113412886858, "rewards/accuracy_reward": 0.7734375, "rewards/format_reward": 0.9966517835855484, "step": 132 }, { "clip_ratio": 0.0, "completion_length": 611.0044898986816, "epoch": 1.3548795944233207, "grad_norm": 0.12098115682601929, "learning_rate": 2.8420756714519954e-07, "loss": 0.0017, "num_tokens": 91839859.0, "reward": 1.7444197237491608, "reward_std": 0.1756706228479743, "rewards/accuracy_reward": 0.745535708963871, "rewards/format_reward": 0.9988839253783226, "step": 133 }, { "clip_ratio": 0.0, "completion_length": 638.6317291259766, "epoch": 1.3650190114068441, "grad_norm": 0.10373256355524063, "learning_rate": 2.7619139496864376e-07, "loss": 0.0222, "num_tokens": 92548057.0, "reward": 1.6875000894069672, "reward_std": 0.15588384959846735, "rewards/accuracy_reward": 0.6975446492433548, "rewards/format_reward": 0.9899553507566452, "step": 134 }, { "clip_ratio": 0.0, "completion_length": 606.5056076049805, "epoch": 1.3751584283903675, "grad_norm": 0.13831515610218048, "learning_rate": 2.6824653102244727e-07, "loss": 0.0172, "num_tokens": 93225558.0, "reward": 1.751116156578064, "reward_std": 0.16407952923327684, "rewards/accuracy_reward": 0.7578125, "rewards/format_reward": 0.9933035597205162, "step": 135 }, { "clip_ratio": 0.0, "completion_length": 625.7645416259766, "epoch": 1.385297845373891, "grad_norm": 0.1232142373919487, "learning_rate": 2.603755066399718e-07, "loss": 0.01, "num_tokens": 93916323.0, "reward": 1.7801340371370316, "reward_std": 0.1467695259489119, "rewards/accuracy_reward": 0.7823660671710968, "rewards/format_reward": 0.9977678507566452, "step": 136 }, { "clip_ratio": 0.0, "completion_length": 686.4006958007812, "epoch": 1.3954372623574145, "grad_norm": 0.11197436600923538, "learning_rate": 2.5258082962836614e-07, "loss": 0.0197, "num_tokens": 94665898.0, "reward": 1.6082590073347092, "reward_std": 0.14906749688088894, "rewards/accuracy_reward": 0.613839291036129, "rewards/format_reward": 0.9944196343421936, "step": 137 }, { "clip_ratio": 0.0, "completion_length": 634.619457244873, "epoch": 1.4055766793409379, "grad_norm": 0.09477974474430084, "learning_rate": 2.4486498346955023e-07, "loss": 0.0172, "num_tokens": 95357349.0, "reward": 1.7488840222358704, "reward_std": 0.14468519762158394, "rewards/accuracy_reward": 0.7555803582072258, "rewards/format_reward": 0.9933035671710968, "step": 138 }, { "clip_ratio": 0.0, "completion_length": 669.0736923217773, "epoch": 1.4157160963244613, "grad_norm": 0.11749354004859924, "learning_rate": 2.372304265289436e-07, "loss": 0.0018, "num_tokens": 96081135.0, "reward": 1.726562574505806, "reward_std": 0.17081871442496777, "rewards/accuracy_reward": 0.7276785746216774, "rewards/format_reward": 0.9988839253783226, "step": 139 }, { "clip_ratio": 0.0, "completion_length": 634.8705596923828, "epoch": 1.4258555133079849, "grad_norm": 0.10797657072544098, "learning_rate": 2.2967959127220137e-07, "loss": 0.0227, "num_tokens": 96771939.0, "reward": 1.7611608058214188, "reward_std": 0.13173524383455515, "rewards/accuracy_reward": 0.7689732015132904, "rewards/format_reward": 0.9921874925494194, "step": 140 }, { "clip_ratio": 0.0, "completion_length": 620.8694534301758, "epoch": 1.4359949302915083, "grad_norm": 0.12707878649234772, "learning_rate": 2.2221488349019902e-07, "loss": 0.0104, "num_tokens": 97476598.0, "reward": 1.7220982909202576, "reward_std": 0.17165770754218102, "rewards/accuracy_reward": 0.7265625, "rewards/format_reward": 0.995535708963871, "step": 141 }, { "clip_ratio": 0.0, "completion_length": 622.1797180175781, "epoch": 1.4461343472750317, "grad_norm": 0.1114574521780014, "learning_rate": 2.1483868153251788e-07, "loss": 0.0302, "num_tokens": 98165615.0, "reward": 1.7801340222358704, "reward_std": 0.15362949669361115, "rewards/accuracy_reward": 0.7924107164144516, "rewards/format_reward": 0.987723208963871, "step": 142 }, { "clip_ratio": 0.0, "completion_length": 687.9710159301758, "epoch": 1.456273764258555, "grad_norm": 0.10450585186481476, "learning_rate": 2.0755333554967346e-07, "loss": 0.0098, "num_tokens": 98919237.0, "reward": 1.679687574505806, "reward_std": 0.17744049057364464, "rewards/accuracy_reward": 0.6830357126891613, "rewards/format_reward": 0.9966517835855484, "step": 143 }, { "clip_ratio": 0.0, "completion_length": 690.1897621154785, "epoch": 1.4664131812420786, "grad_norm": 0.10399246960878372, "learning_rate": 2.0036116674432652e-07, "loss": 0.0263, "num_tokens": 99674095.0, "reward": 1.6863840222358704, "reward_std": 0.15410742722451687, "rewards/accuracy_reward": 0.6941964402794838, "rewards/format_reward": 0.9921874925494194, "step": 144 }, { "clip_ratio": 0.0, "completion_length": 599.2310600280762, "epoch": 1.476552598225602, "grad_norm": 0.09947122633457184, "learning_rate": 1.9326446663172035e-07, "loss": 0.0121, "num_tokens": 100342486.0, "reward": 1.7868304401636124, "reward_std": 0.13388300314545631, "rewards/accuracy_reward": 0.7879464253783226, "rewards/format_reward": 0.9988839253783226, "step": 145 }, { "clip_ratio": 0.0, "completion_length": 595.1373062133789, "epoch": 1.4866920152091254, "grad_norm": 0.11993813514709473, "learning_rate": 1.8626549630957395e-07, "loss": 0.0045, "num_tokens": 101003873.0, "reward": 1.7544643580913544, "reward_std": 0.14560767449438572, "rewards/accuracy_reward": 0.7555803433060646, "rewards/format_reward": 0.9988839253783226, "step": 146 }, { "clip_ratio": 0.0, "completion_length": 645.3449020385742, "epoch": 1.496831432192649, "grad_norm": 0.1037568524479866, "learning_rate": 1.7936648573766954e-07, "loss": 0.0197, "num_tokens": 101718814.0, "reward": 1.727678656578064, "reward_std": 0.1583123281598091, "rewards/accuracy_reward": 0.7366071492433548, "rewards/format_reward": 0.9910714253783226, "step": 147 }, { "clip_ratio": 0.0, "completion_length": 665.5960121154785, "epoch": 1.5069708491761724, "grad_norm": 0.11654334515333176, "learning_rate": 1.725696330273575e-07, "loss": 0.0334, "num_tokens": 102445532.0, "reward": 1.6908482909202576, "reward_std": 0.15235634427517653, "rewards/accuracy_reward": 0.6997767873108387, "rewards/format_reward": 0.991071417927742, "step": 148 }, { "clip_ratio": 0.0, "completion_length": 615.9922103881836, "epoch": 1.5171102661596958, "grad_norm": 0.11088323593139648, "learning_rate": 1.65877103741212e-07, "loss": 0.0102, "num_tokens": 103122189.0, "reward": 1.7834822088479996, "reward_std": 0.15306678228080273, "rewards/accuracy_reward": 0.7901785671710968, "rewards/format_reward": 0.9933035597205162, "step": 149 }, { "clip_ratio": 0.0, "completion_length": 649.7199020385742, "epoch": 1.5272496831432192, "grad_norm": 0.1219962015748024, "learning_rate": 1.592910302030544e-07, "loss": 0.0174, "num_tokens": 103823642.0, "reward": 1.7399554401636124, "reward_std": 0.18228058330714703, "rewards/accuracy_reward": 0.7488839402794838, "rewards/format_reward": 0.9910714253783226, "step": 150 }, { "clip_ratio": 0.0, "completion_length": 668.3158836364746, "epoch": 1.5373891001267426, "grad_norm": 0.12323982268571854, "learning_rate": 1.5281351081856976e-07, "loss": 0.0218, "num_tokens": 104553709.0, "reward": 1.7120536416769028, "reward_std": 0.16315976157784462, "rewards/accuracy_reward": 0.723214291036129, "rewards/format_reward": 0.9888392761349678, "step": 151 }, { "clip_ratio": 0.0, "completion_length": 625.3861770629883, "epoch": 1.5475285171102662, "grad_norm": 0.1106390506029129, "learning_rate": 1.4644660940672627e-07, "loss": 0.0111, "num_tokens": 105239647.0, "reward": 1.781250074505806, "reward_std": 0.12235020846128464, "rewards/accuracy_reward": 0.7868303582072258, "rewards/format_reward": 0.9944196343421936, "step": 152 }, { "clip_ratio": 0.0, "completion_length": 619.7009201049805, "epoch": 1.5576679340937896, "grad_norm": 0.10685606300830841, "learning_rate": 1.4019235454221856e-07, "loss": 0.012, "num_tokens": 105928971.0, "reward": 1.728794738650322, "reward_std": 0.13385361433029175, "rewards/accuracy_reward": 0.7321428656578064, "rewards/format_reward": 0.9966517835855484, "step": 153 }, { "clip_ratio": 0.0, "completion_length": 624.036865234375, "epoch": 1.5678073510773132, "grad_norm": 0.11761969327926636, "learning_rate": 1.3405273890913737e-07, "loss": 0.0085, "num_tokens": 106610124.0, "reward": 1.7343751192092896, "reward_std": 0.17772767413407564, "rewards/accuracy_reward": 0.7354910746216774, "rewards/format_reward": 0.9988839253783226, "step": 154 }, { "clip_ratio": 0.0, "completion_length": 635.014533996582, "epoch": 1.5779467680608366, "grad_norm": 0.11695584654808044, "learning_rate": 1.280297186660752e-07, "loss": 0.0378, "num_tokens": 107297473.0, "reward": 1.7935268580913544, "reward_std": 0.17546693049371243, "rewards/accuracy_reward": 0.8035714328289032, "rewards/format_reward": 0.9899553582072258, "step": 155 }, { "clip_ratio": 0.0, "completion_length": 708.3571853637695, "epoch": 1.58808618504436, "grad_norm": 0.11352672427892685, "learning_rate": 1.2212521282287093e-07, "loss": 0.0027, "num_tokens": 108052089.0, "reward": 1.7042411267757416, "reward_std": 0.17979852855205536, "rewards/accuracy_reward": 0.7053571417927742, "rewards/format_reward": 0.9988839253783226, "step": 156 }, { "clip_ratio": 0.0, "completion_length": 633.8192291259766, "epoch": 1.5982256020278833, "grad_norm": 0.10707947611808777, "learning_rate": 1.1634110262918717e-07, "loss": 0.0118, "num_tokens": 108763999.0, "reward": 1.703125074505806, "reward_std": 0.14697580970823765, "rewards/accuracy_reward": 0.7109375, "rewards/format_reward": 0.9921874925494194, "step": 157 }, { "clip_ratio": 0.0, "completion_length": 676.0625381469727, "epoch": 1.6083650190114067, "grad_norm": 0.11339569836854935, "learning_rate": 1.1067923097512255e-07, "loss": 0.0181, "num_tokens": 109500463.0, "reward": 1.7220983058214188, "reward_std": 0.18152274563908577, "rewards/accuracy_reward": 0.7287946492433548, "rewards/format_reward": 0.9933035671710968, "step": 158 }, { "clip_ratio": 0.0, "completion_length": 639.6294937133789, "epoch": 1.6185044359949303, "grad_norm": 0.10624842345714569, "learning_rate": 1.0514140180404202e-07, "loss": -0.0033, "num_tokens": 110202627.0, "reward": 1.7232143729925156, "reward_std": 0.16777998208999634, "rewards/accuracy_reward": 0.7265624925494194, "rewards/format_reward": 0.9966517761349678, "step": 159 }, { "clip_ratio": 0.0, "completion_length": 681.1741485595703, "epoch": 1.6286438529784537, "grad_norm": 0.11243726313114166, "learning_rate": 9.972937953781984e-08, "loss": 0.0145, "num_tokens": 110959927.0, "reward": 1.678571492433548, "reward_std": 0.16837417893111706, "rewards/accuracy_reward": 0.6852678656578064, "rewards/format_reward": 0.9933035671710968, "step": 160 }, { "clip_ratio": 0.0, "completion_length": 605.4766006469727, "epoch": 1.6387832699619773, "grad_norm": 0.09180891513824463, "learning_rate": 9.444488851467041e-08, "loss": 0.019, "num_tokens": 111620106.0, "reward": 1.7968750894069672, "reward_std": 0.11058733332902193, "rewards/accuracy_reward": 0.8035714402794838, "rewards/format_reward": 0.9933035671710968, "step": 161 }, { "clip_ratio": 0.0, "completion_length": 650.7544937133789, "epoch": 1.6489226869455007, "grad_norm": 0.11670933663845062, "learning_rate": 8.928961243975436e-08, "loss": 0.0121, "num_tokens": 112336038.0, "reward": 1.7321429252624512, "reward_std": 0.1541335443034768, "rewards/accuracy_reward": 0.7402987703680992, "rewards/format_reward": 0.9988839253783226, "step": 162 }, { "clip_ratio": 0.0, "completion_length": 630.4565010070801, "epoch": 1.659062103929024, "grad_norm": 0.10085251182317734, "learning_rate": 8.426519384872732e-08, "loss": 0.0036, "num_tokens": 113023063.0, "reward": 1.7455358058214188, "reward_std": 0.15896250680088997, "rewards/accuracy_reward": 0.7500000074505806, "rewards/format_reward": 0.995535708963871, "step": 163 }, { "clip_ratio": 0.0, "completion_length": 636.2277069091797, "epoch": 1.6692015209125475, "grad_norm": 0.1013065055012703, "learning_rate": 7.937323358440934e-08, "loss": 0.0099, "num_tokens": 113724155.0, "reward": 1.7533483058214188, "reward_std": 0.13801472913473845, "rewards/accuracy_reward": 0.7555803582072258, "rewards/format_reward": 0.9977678582072258, "step": 164 }, { "clip_ratio": 0.0, "completion_length": 632.6138687133789, "epoch": 1.6793409378960709, "grad_norm": 0.09690902382135391, "learning_rate": 7.461529028673463e-08, "loss": 0.0219, "num_tokens": 114423785.0, "reward": 1.7834822237491608, "reward_std": 0.1157834529876709, "rewards/accuracy_reward": 0.7868303582072258, "rewards/format_reward": 0.9966517835855484, "step": 165 }, { "clip_ratio": 0.0, "completion_length": 642.2444458007812, "epoch": 1.6894803548795945, "grad_norm": 0.111725352704525, "learning_rate": 6.999287989614971e-08, "loss": 0.0065, "num_tokens": 115131676.0, "reward": 1.7578126043081284, "reward_std": 0.15221323259174824, "rewards/accuracy_reward": 0.7600446417927742, "rewards/format_reward": 0.9977678507566452, "step": 166 }, { "clip_ratio": 0.0, "completion_length": 625.8181190490723, "epoch": 1.6996197718631179, "grad_norm": 0.10648291558027267, "learning_rate": 6.550747517061656e-08, "loss": 0.0144, "num_tokens": 115818953.0, "reward": 1.7890625894069672, "reward_std": 0.1342666009441018, "rewards/accuracy_reward": 0.793526791036129, "rewards/format_reward": 0.9955357164144516, "step": 167 }, { "clip_ratio": 0.0, "completion_length": 649.0558166503906, "epoch": 1.7097591888466415, "grad_norm": 0.10185572504997253, "learning_rate": 6.116050521637218e-08, "loss": 0.0065, "num_tokens": 116528835.0, "reward": 1.7834822237491608, "reward_std": 0.1334951389580965, "rewards/accuracy_reward": 0.7868303507566452, "rewards/format_reward": 0.9966517761349678, "step": 168 }, { "clip_ratio": 0.0, "completion_length": 623.458740234375, "epoch": 1.7198986058301649, "grad_norm": 0.10859635472297668, "learning_rate": 5.6953355032598795e-08, "loss": 0.0139, "num_tokens": 117224526.0, "reward": 1.7455358058214188, "reward_std": 0.15123825334012508, "rewards/accuracy_reward": 0.7522321417927742, "rewards/format_reward": 0.9933035671710968, "step": 169 }, { "clip_ratio": 0.0, "completion_length": 650.3861846923828, "epoch": 1.7300380228136882, "grad_norm": 0.11233941465616226, "learning_rate": 5.288736507014435e-08, "loss": 0.0094, "num_tokens": 117931272.0, "reward": 1.7790179401636124, "reward_std": 0.16706989519298077, "rewards/accuracy_reward": 0.7857142835855484, "rewards/format_reward": 0.9933035671710968, "step": 170 }, { "clip_ratio": 0.0, "completion_length": 621.7712326049805, "epoch": 1.7401774397972116, "grad_norm": 0.11732303351163864, "learning_rate": 4.896383080443933e-08, "loss": 0.0146, "num_tokens": 118613131.0, "reward": 1.7321429252624512, "reward_std": 0.16958792228251696, "rewards/accuracy_reward": 0.738839291036129, "rewards/format_reward": 0.9933035671710968, "step": 171 }, { "clip_ratio": 0.0, "completion_length": 619.6361885070801, "epoch": 1.750316856780735, "grad_norm": 0.11354614794254303, "learning_rate": 4.518400232274078e-08, "loss": 0.0137, "num_tokens": 119298829.0, "reward": 1.7466518580913544, "reward_std": 0.16586073767393827, "rewards/accuracy_reward": 0.7564388811588287, "rewards/format_reward": 0.9966517835855484, "step": 172 }, { "clip_ratio": 0.0, "completion_length": 603.4085083007812, "epoch": 1.7604562737642584, "grad_norm": 0.11991310119628906, "learning_rate": 4.1549083925840165e-08, "loss": 0.0226, "num_tokens": 119956931.0, "reward": 1.8147322237491608, "reward_std": 0.12004590220749378, "rewards/accuracy_reward": 0.8191964328289032, "rewards/format_reward": 0.995535708963871, "step": 173 }, { "clip_ratio": 0.0, "completion_length": 602.5100708007812, "epoch": 1.770595690747782, "grad_norm": 0.13672617077827454, "learning_rate": 3.806023374435663e-08, "loss": 0.0039, "num_tokens": 120625308.0, "reward": 1.7700893729925156, "reward_std": 0.17995007149875164, "rewards/accuracy_reward": 0.7712053582072258, "rewards/format_reward": 0.9988839253783226, "step": 174 }, { "clip_ratio": 0.0, "completion_length": 647.302490234375, "epoch": 1.7807351077313056, "grad_norm": 0.10016340017318726, "learning_rate": 3.4718563369743213e-08, "loss": 0.0119, "num_tokens": 121338619.0, "reward": 1.7265626043081284, "reward_std": 0.1343973781913519, "rewards/accuracy_reward": 0.7310267984867096, "rewards/format_reward": 0.995535708963871, "step": 175 }, { "clip_ratio": 0.0, "completion_length": 626.5011367797852, "epoch": 1.790874524714829, "grad_norm": 0.10784109681844711, "learning_rate": 3.15251375001192e-08, "loss": 0.0193, "num_tokens": 122040068.0, "reward": 1.7946429550647736, "reward_std": 0.13197333365678787, "rewards/accuracy_reward": 0.8013392835855484, "rewards/format_reward": 0.9933035597205162, "step": 176 }, { "clip_ratio": 0.0, "completion_length": 627.9163284301758, "epoch": 1.8010139416983524, "grad_norm": 0.11255474388599396, "learning_rate": 2.8480973601043955e-08, "loss": 0.0134, "num_tokens": 122737489.0, "reward": 1.726562574505806, "reward_std": 0.10841276589781046, "rewards/accuracy_reward": 0.7299107126891613, "rewards/format_reward": 0.9966517835855484, "step": 177 }, { "clip_ratio": 0.0, "completion_length": 641.2310562133789, "epoch": 1.8111533586818758, "grad_norm": 0.12718521058559418, "learning_rate": 2.558704158134023e-08, "loss": 0.0123, "num_tokens": 123452056.0, "reward": 1.712053656578064, "reward_std": 0.16341756749898195, "rewards/accuracy_reward": 0.7154017984867096, "rewards/format_reward": 0.9966517761349678, "step": 178 }, { "clip_ratio": 0.0, "completion_length": 661.897346496582, "epoch": 1.8212927756653992, "grad_norm": 0.10139390826225281, "learning_rate": 2.2844263484068093e-08, "loss": 0.0179, "num_tokens": 124172116.0, "reward": 1.7165179401636124, "reward_std": 0.14442832861095667, "rewards/accuracy_reward": 0.7243303582072258, "rewards/format_reward": 0.9921874925494194, "step": 179 }, { "clip_ratio": 0.0, "completion_length": 620.7277069091797, "epoch": 1.8314321926489225, "grad_norm": 0.11224541813135147, "learning_rate": 2.025351319275137e-08, "loss": 0.0239, "num_tokens": 124859256.0, "reward": 1.751116156578064, "reward_std": 0.14368766359984875, "rewards/accuracy_reward": 0.7578124925494194, "rewards/format_reward": 0.9933035671710968, "step": 180 }, { "clip_ratio": 0.0, "completion_length": 659.0647583007812, "epoch": 1.8415716096324461, "grad_norm": 0.11130577325820923, "learning_rate": 1.781561615294652e-08, "loss": 0.0061, "num_tokens": 125584242.0, "reward": 1.7154018580913544, "reward_std": 0.1235003275796771, "rewards/accuracy_reward": 0.7198660597205162, "rewards/format_reward": 0.995535708963871, "step": 181 }, { "clip_ratio": 0.0, "completion_length": 663.6573867797852, "epoch": 1.8517110266159695, "grad_norm": 0.12259829789400101, "learning_rate": 1.553134910924636e-08, "loss": 0.0177, "num_tokens": 126314983.0, "reward": 1.671875074505806, "reward_std": 0.17521491926163435, "rewards/accuracy_reward": 0.6819196417927742, "rewards/format_reward": 0.9899553582072258, "step": 182 }, { "clip_ratio": 0.0, "completion_length": 610.7846221923828, "epoch": 1.8618504435994931, "grad_norm": 0.1277763843536377, "learning_rate": 1.340143985779829e-08, "loss": 0.0134, "num_tokens": 126989006.0, "reward": 1.7466518729925156, "reward_std": 0.13783067418262362, "rewards/accuracy_reward": 0.7511160671710968, "rewards/format_reward": 0.995535708963871, "step": 183 }, { "clip_ratio": 0.0, "completion_length": 625.2879867553711, "epoch": 1.8719898605830165, "grad_norm": 0.10394810885190964, "learning_rate": 1.1426567014420297e-08, "loss": 0.0122, "num_tokens": 127675968.0, "reward": 1.8035715073347092, "reward_std": 0.12675560358911753, "rewards/accuracy_reward": 0.8102678507566452, "rewards/format_reward": 0.9933035671710968, "step": 184 }, { "clip_ratio": 0.0, "completion_length": 661.4832916259766, "epoch": 1.88212927756654, "grad_norm": 0.10645025223493576, "learning_rate": 9.607359798384784e-09, "loss": 0.022, "num_tokens": 128384137.0, "reward": 1.7500000596046448, "reward_std": 0.1677253060042858, "rewards/accuracy_reward": 0.7555803582072258, "rewards/format_reward": 0.9944196343421936, "step": 185 }, { "clip_ratio": 0.0, "completion_length": 628.4553909301758, "epoch": 1.8922686945500633, "grad_norm": 0.12001987546682358, "learning_rate": 7.944397831941951e-09, "loss": 0.0204, "num_tokens": 129093729.0, "reward": 1.710937574505806, "reward_std": 0.1775853894650936, "rewards/accuracy_reward": 0.7176339253783226, "rewards/format_reward": 0.9933035597205162, "step": 186 }, { "clip_ratio": 0.0, "completion_length": 682.1506958007812, "epoch": 1.9024081115335867, "grad_norm": 0.11876320838928223, "learning_rate": 6.438210955644452e-09, "loss": 0.0227, "num_tokens": 129847360.0, "reward": 1.6707590073347092, "reward_std": 0.14890040084719658, "rewards/accuracy_reward": 0.6774553619325161, "rewards/format_reward": 0.9933035746216774, "step": 187 }, { "clip_ratio": 0.0, "completion_length": 632.9118576049805, "epoch": 1.9125475285171103, "grad_norm": 0.10546144843101501, "learning_rate": 5.0892790595336575e-09, "loss": 0.0116, "num_tokens": 130551465.0, "reward": 1.7399554401636124, "reward_std": 0.12469578720629215, "rewards/accuracy_reward": 0.7433035746216774, "rewards/format_reward": 0.9966517835855484, "step": 188 }, { "clip_ratio": 0.0, "completion_length": 633.2812728881836, "epoch": 1.9226869455006337, "grad_norm": 0.11886601895093918, "learning_rate": 3.898031930240797e-09, "loss": 0.011, "num_tokens": 131243885.0, "reward": 1.7555804252624512, "reward_std": 0.14942110795527697, "rewards/accuracy_reward": 0.7578125, "rewards/format_reward": 0.9977678507566452, "step": 189 }, { "clip_ratio": 0.0, "completion_length": 657.8850631713867, "epoch": 1.9328263624841573, "grad_norm": 0.14054881036281586, "learning_rate": 2.8648491140513264e-09, "loss": 0.0237, "num_tokens": 131959798.0, "reward": 1.7622768580913544, "reward_std": 0.1422654939815402, "rewards/accuracy_reward": 0.7656249925494194, "rewards/format_reward": 0.9966517761349678, "step": 190 }, { "clip_ratio": 0.0, "completion_length": 580.6707878112793, "epoch": 1.9429657794676807, "grad_norm": 0.10732496529817581, "learning_rate": 1.9900597959770505e-09, "loss": 0.0063, "num_tokens": 132605207.0, "reward": 1.8348215222358704, "reward_std": 0.1142951613292098, "rewards/accuracy_reward": 0.8370535671710968, "rewards/format_reward": 0.9977678507566452, "step": 191 }, { "clip_ratio": 0.0, "completion_length": 662.7555999755859, "epoch": 1.953105196451204, "grad_norm": 0.10711626708507538, "learning_rate": 1.2739426948732424e-09, "loss": 0.0112, "num_tokens": 133333004.0, "reward": 1.7488840073347092, "reward_std": 0.1441500186920166, "rewards/accuracy_reward": 0.7566964253783226, "rewards/format_reward": 0.9921874925494194, "step": 192 }, { "clip_ratio": 0.0, "completion_length": 649.7835159301758, "epoch": 1.9632446134347274, "grad_norm": 0.12485576421022415, "learning_rate": 7.16725974635568e-10, "loss": 0.0127, "num_tokens": 134041050.0, "reward": 1.789062574505806, "reward_std": 0.15979287773370743, "rewards/accuracy_reward": 0.7946428582072258, "rewards/format_reward": 0.9944196343421936, "step": 193 }, { "clip_ratio": 0.0, "completion_length": 666.5379867553711, "epoch": 1.9733840304182508, "grad_norm": 0.11115089803934097, "learning_rate": 3.185871715041255e-10, "loss": -0.0016, "num_tokens": 134771116.0, "reward": 1.7220982909202576, "reward_std": 0.14188571739941835, "rewards/accuracy_reward": 0.7243303507566452, "rewards/format_reward": 0.9977678507566452, "step": 194 }, { "clip_ratio": 0.0, "completion_length": 599.9765777587891, "epoch": 1.9835234474017744, "grad_norm": 0.11773505806922913, "learning_rate": 7.96531374983589e-11, "loss": 0.0201, "num_tokens": 135429175.0, "reward": 1.7845983058214188, "reward_std": 0.16611923463642597, "rewards/accuracy_reward": 0.7890625, "rewards/format_reward": 0.995535708963871, "step": 195 }, { "clip_ratio": 0.0, "completion_length": 647.62109375, "epoch": 1.9936628643852978, "grad_norm": 0.12847015261650085, "learning_rate": 0.0, "loss": 0.0274, "num_tokens": 136136408.0, "reward": 1.7332590222358704, "reward_std": 0.1759201493114233, "rewards/accuracy_reward": 0.7377232164144516, "rewards/format_reward": 0.995535708963871, "step": 196 }, { "epoch": 1.9936628643852978, "step": 196, "total_flos": 0.0, "train_loss": 0.020176043500826808, "train_runtime": 41639.173, "train_samples_per_second": 0.53, "train_steps_per_second": 0.005 } ], "logging_steps": 1, "max_steps": 196, "num_input_tokens_seen": 0, "num_train_epochs": 2, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 16, "trial_name": null, "trial_params": null }