{ "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": 544.7745819091797, "epoch": 0.010139416983523447, "grad_norm": 0.1392974704504013, "learning_rate": 5e-08, "loss": -0.0292, "num_tokens": 636118.0, "reward": 0.12946429196745157, "reward_std": 0.22425302118062973, "rewards/accuracy_reward": 0.1261160708963871, "rewards/format_reward": 0.003348214435391128, "step": 1 }, { "clip_ratio": 0.0, "completion_length": 549.2812843322754, "epoch": 0.020278833967046894, "grad_norm": 0.1566777527332306, "learning_rate": 1e-07, "loss": -0.0074, "num_tokens": 1263330.0, "reward": 0.152901791036129, "reward_std": 0.25637512281537056, "rewards/accuracy_reward": 0.14508928637951612, "rewards/format_reward": 0.007812500349245965, "step": 2 }, { "clip_ratio": 0.0, "completion_length": 585.7969131469727, "epoch": 0.030418250950570342, "grad_norm": 0.15250442922115326, "learning_rate": 1.5e-07, "loss": -0.0299, "num_tokens": 1928324.0, "reward": 0.17745536379516125, "reward_std": 0.26186119951307774, "rewards/accuracy_reward": 0.16964285913854837, "rewards/format_reward": 0.007812500349245965, "step": 3 }, { "clip_ratio": 0.0, "completion_length": 550.0346145629883, "epoch": 0.04055766793409379, "grad_norm": 0.15349256992340088, "learning_rate": 2e-07, "loss": -0.0336, "num_tokens": 2558107.0, "reward": 0.17075893748551607, "reward_std": 0.2673841342329979, "rewards/accuracy_reward": 0.16071428824216127, "rewards/format_reward": 0.010044643306173384, "step": 4 }, { "clip_ratio": 0.0, "completion_length": 578.6484642028809, "epoch": 0.050697084917617236, "grad_norm": 0.3040037155151367, "learning_rate": 2.5e-07, "loss": -0.0205, "num_tokens": 3203360.0, "reward": 0.15848214831203222, "reward_std": 0.2771828528493643, "rewards/accuracy_reward": 0.1517857126891613, "rewards/format_reward": 0.006696428754366934, "step": 5 }, { "clip_ratio": 0.0, "completion_length": 583.7399826049805, "epoch": 0.060836501901140684, "grad_norm": 0.13601042330265045, "learning_rate": 3e-07, "loss": -0.031, "num_tokens": 3855583.0, "reward": 0.15625000465661287, "reward_std": 0.25207988917827606, "rewards/accuracy_reward": 0.15066964365541935, "rewards/format_reward": 0.005580357392318547, "step": 6 }, { "clip_ratio": 0.0, "completion_length": 537.3594093322754, "epoch": 0.07097591888466413, "grad_norm": 0.16291461884975433, "learning_rate": 3.5e-07, "loss": -0.0137, "num_tokens": 4461481.0, "reward": 0.17522322200238705, "reward_std": 0.2866264618933201, "rewards/accuracy_reward": 0.16183035727590322, "rewards/format_reward": 0.013392857508733869, "step": 7 }, { "clip_ratio": 0.0, "completion_length": 537.2634048461914, "epoch": 0.08111533586818757, "grad_norm": 0.19215919077396393, "learning_rate": 4e-07, "loss": -0.0325, "num_tokens": 5070421.0, "reward": 0.1785714365541935, "reward_std": 0.2945236265659332, "rewards/accuracy_reward": 0.1696428582072258, "rewards/format_reward": 0.008928571827709675, "step": 8 }, { "clip_ratio": 0.0, "completion_length": 575.7366256713867, "epoch": 0.09125475285171103, "grad_norm": 0.14393264055252075, "learning_rate": 4.5e-07, "loss": -0.014, "num_tokens": 5709609.0, "reward": 0.14843750465661287, "reward_std": 0.22998391091823578, "rewards/accuracy_reward": 0.14285714272409678, "rewards/format_reward": 0.005580357392318547, "step": 9 }, { "clip_ratio": 0.0, "completion_length": 519.7522468566895, "epoch": 0.10139416983523447, "grad_norm": 0.16316372156143188, "learning_rate": 5e-07, "loss": -0.0266, "num_tokens": 6302035.0, "reward": 0.1930803656578064, "reward_std": 0.2917625065892935, "rewards/accuracy_reward": 0.1863839291036129, "rewards/format_reward": 0.006696428870782256, "step": 10 }, { "clip_ratio": 0.0, "completion_length": 595.2377395629883, "epoch": 0.11153358681875793, "grad_norm": 0.14569994807243347, "learning_rate": 5.5e-07, "loss": -0.0097, "num_tokens": 6972904.0, "reward": 0.17299108020961285, "reward_std": 0.27831926569342613, "rewards/accuracy_reward": 0.16183035913854837, "rewards/format_reward": 0.01116071455180645, "step": 11 }, { "clip_ratio": 0.0, "completion_length": 575.3605155944824, "epoch": 0.12167300380228137, "grad_norm": 0.15014010667800903, "learning_rate": 6e-07, "loss": -0.0216, "num_tokens": 7619739.0, "reward": 0.17187501303851604, "reward_std": 0.2696072347462177, "rewards/accuracy_reward": 0.16183035634458065, "rewards/format_reward": 0.010044643306173384, "step": 12 }, { "clip_ratio": 0.0, "completion_length": 579.7835006713867, "epoch": 0.13181242078580482, "grad_norm": 0.14231476187705994, "learning_rate": 6.5e-07, "loss": 0.0066, "num_tokens": 8267409.0, "reward": 0.18191965110599995, "reward_std": 0.27643212117254734, "rewards/accuracy_reward": 0.17299107369035482, "rewards/format_reward": 0.008928571827709675, "step": 13 }, { "clip_ratio": 0.0, "completion_length": 560.2623062133789, "epoch": 0.14195183776932827, "grad_norm": 0.1491479128599167, "learning_rate": 7e-07, "loss": -0.007, "num_tokens": 8901580.0, "reward": 0.17745536752045155, "reward_std": 0.28200731612741947, "rewards/accuracy_reward": 0.1704155197367072, "rewards/format_reward": 0.007812500349245965, "step": 14 }, { "clip_ratio": 0.0, "completion_length": 563.2969017028809, "epoch": 0.1520912547528517, "grad_norm": 0.16368582844734192, "learning_rate": 7.5e-07, "loss": -0.0445, "num_tokens": 9534734.0, "reward": 0.22544643841683865, "reward_std": 0.3410548157989979, "rewards/accuracy_reward": 0.20870535541325808, "rewards/format_reward": 0.016741071944124997, "step": 15 }, { "clip_ratio": 0.0, "completion_length": 575.6607360839844, "epoch": 0.16223067173637515, "grad_norm": 0.155720517039299, "learning_rate": 8e-07, "loss": -0.0125, "num_tokens": 10177454.0, "reward": 0.2332589402794838, "reward_std": 0.33312445878982544, "rewards/accuracy_reward": 0.21986606996506453, "rewards/format_reward": 0.01339285762514919, "step": 16 }, { "clip_ratio": 0.0, "completion_length": 605.8471221923828, "epoch": 0.17237008871989862, "grad_norm": 0.18089956045150757, "learning_rate": 8.499999999999999e-07, "loss": -0.0189, "num_tokens": 10854805.0, "reward": 0.22433037124574184, "reward_std": 0.3147235903888941, "rewards/accuracy_reward": 0.20647321455180645, "rewards/format_reward": 0.017857143306173384, "step": 17 }, { "clip_ratio": 0.0, "completion_length": 620.2031478881836, "epoch": 0.18250950570342206, "grad_norm": 0.14688049256801605, "learning_rate": 9e-07, "loss": -0.0215, "num_tokens": 11533851.0, "reward": 0.26227680034935474, "reward_std": 0.32751886174082756, "rewards/accuracy_reward": 0.2522321417927742, "rewards/format_reward": 0.010044643189758062, "step": 18 }, { "clip_ratio": 0.0, "completion_length": 567.4386444091797, "epoch": 0.1926489226869455, "grad_norm": 0.18814197182655334, "learning_rate": 9.499999999999999e-07, "loss": -0.0224, "num_tokens": 12174260.0, "reward": 0.33370537497103214, "reward_std": 0.3732391260564327, "rewards/accuracy_reward": 0.3169642798602581, "rewards/format_reward": 0.016741071944124997, "step": 19 }, { "clip_ratio": 0.0, "completion_length": 620.7399749755859, "epoch": 0.20278833967046894, "grad_norm": 0.1669970452785492, "learning_rate": 1e-06, "loss": -0.0292, "num_tokens": 12850867.0, "reward": 0.3582589402794838, "reward_std": 0.3860209658741951, "rewards/accuracy_reward": 0.3359375037252903, "rewards/format_reward": 0.02232142922002822, "step": 20 }, { "clip_ratio": 0.0, "completion_length": 607.200927734375, "epoch": 0.21292775665399238, "grad_norm": 1.3311952352523804, "learning_rate": 9.999203468625015e-07, "loss": -0.0164, "num_tokens": 13516191.0, "reward": 0.3962053768336773, "reward_std": 0.3798532336950302, "rewards/accuracy_reward": 0.3694196417927742, "rewards/format_reward": 0.026785714901052415, "step": 21 }, { "clip_ratio": 0.0, "completion_length": 682.4051742553711, "epoch": 0.22306717363751585, "grad_norm": 0.14331556856632233, "learning_rate": 9.99681412828496e-07, "loss": 0.0035, "num_tokens": 14260786.0, "reward": 0.4140625186264515, "reward_std": 0.3847779668867588, "rewards/accuracy_reward": 0.3750000037252903, "rewards/format_reward": 0.03906250128056854, "step": 22 }, { "clip_ratio": 0.0, "completion_length": 633.8303909301758, "epoch": 0.2332065906210393, "grad_norm": 0.1424940973520279, "learning_rate": 9.992832740253644e-07, "loss": 0.01, "num_tokens": 14966946.0, "reward": 0.427455373108387, "reward_std": 0.3340139761567116, "rewards/accuracy_reward": 0.4073660746216774, "rewards/format_reward": 0.02008928614668548, "step": 23 }, { "clip_ratio": 0.0, "completion_length": 693.4665451049805, "epoch": 0.24334600760456274, "grad_norm": 0.1358347088098526, "learning_rate": 9.987260573051267e-07, "loss": 0.021, "num_tokens": 15717836.0, "reward": 0.4095982313156128, "reward_std": 0.3687121346592903, "rewards/accuracy_reward": 0.3805803619325161, "rewards/format_reward": 0.029017857974395156, "step": 24 }, { "clip_ratio": 0.0, "completion_length": 679.7701263427734, "epoch": 0.2534854245880862, "grad_norm": 0.12858213484287262, "learning_rate": 9.98009940204023e-07, "loss": 0.0077, "num_tokens": 16458238.0, "reward": 0.4821428805589676, "reward_std": 0.3479448612779379, "rewards/accuracy_reward": 0.4531250074505806, "rewards/format_reward": 0.029017857857979834, "step": 25 }, { "clip_ratio": 0.0, "completion_length": 643.7042846679688, "epoch": 0.26362484157160965, "grad_norm": 0.13453267514705658, "learning_rate": 9.971351508859486e-07, "loss": 0.0002, "num_tokens": 17153829.0, "reward": 0.4787946566939354, "reward_std": 0.33469830453395844, "rewards/accuracy_reward": 0.457589291036129, "rewards/format_reward": 0.02120535762514919, "step": 26 }, { "clip_ratio": 0.0, "completion_length": 647.9174499511719, "epoch": 0.2737642585551331, "grad_norm": 0.13937269151210785, "learning_rate": 9.961019680697591e-07, "loss": 0.031, "num_tokens": 17862539.0, "reward": 0.5033482313156128, "reward_std": 0.34470784291625023, "rewards/accuracy_reward": 0.4676339291036129, "rewards/format_reward": 0.0357142862631008, "step": 27 }, { "clip_ratio": 0.0, "completion_length": 729.0391006469727, "epoch": 0.28390367553865653, "grad_norm": 0.12249790877103806, "learning_rate": 9.949107209404663e-07, "loss": 0.039, "num_tokens": 18649398.0, "reward": 0.5368303805589676, "reward_std": 0.3495071418583393, "rewards/accuracy_reward": 0.5122767835855484, "rewards/format_reward": 0.02455357217695564, "step": 28 }, { "clip_ratio": 0.0, "completion_length": 696.9040451049805, "epoch": 0.29404309252217997, "grad_norm": 0.12198278307914734, "learning_rate": 9.935617890443554e-07, "loss": 0.0252, "num_tokens": 19414560.0, "reward": 0.541294664144516, "reward_std": 0.30416675843298435, "rewards/accuracy_reward": 0.5145089328289032, "rewards/format_reward": 0.026785714901052415, "step": 29 }, { "clip_ratio": 0.0, "completion_length": 760.1384353637695, "epoch": 0.3041825095057034, "grad_norm": 0.12740154564380646, "learning_rate": 9.92055602168058e-07, "loss": 0.0327, "num_tokens": 20241132.0, "reward": 0.5569196678698063, "reward_std": 0.30499308556318283, "rewards/accuracy_reward": 0.5145089253783226, "rewards/format_reward": 0.04241071501746774, "step": 30 }, { "clip_ratio": 0.0, "completion_length": 727.1250381469727, "epoch": 0.31432192648922685, "grad_norm": 0.13364288210868835, "learning_rate": 9.90392640201615e-07, "loss": 0.0145, "num_tokens": 21020244.0, "reward": 0.6316964700818062, "reward_std": 0.35084592178463936, "rewards/accuracy_reward": 0.5758928619325161, "rewards/format_reward": 0.05580357206054032, "step": 31 }, { "clip_ratio": 0.0, "completion_length": 743.0826187133789, "epoch": 0.3244613434727503, "grad_norm": 0.1573316901922226, "learning_rate": 9.885734329855797e-07, "loss": 0.047, "num_tokens": 21805790.0, "reward": 0.6562500223517418, "reward_std": 0.3445834815502167, "rewards/accuracy_reward": 0.5792410708963871, "rewards/format_reward": 0.0770089291036129, "step": 32 }, { "clip_ratio": 0.0, "completion_length": 750.4531631469727, "epoch": 0.33460076045627374, "grad_norm": 0.15323247015476227, "learning_rate": 9.865985601422017e-07, "loss": 0.0446, "num_tokens": 22612252.0, "reward": 0.6238839514553547, "reward_std": 0.41414720192551613, "rewards/accuracy_reward": 0.49690933898091316, "rewards/format_reward": 0.13058035727590322, "step": 33 }, { "clip_ratio": 0.0, "completion_length": 765.1116485595703, "epoch": 0.34474017743979724, "grad_norm": 0.22268784046173096, "learning_rate": 9.844686508907537e-07, "loss": 0.0329, "num_tokens": 23425280.0, "reward": 0.7979911044239998, "reward_std": 0.4762897975742817, "rewards/accuracy_reward": 0.5691964253783226, "rewards/format_reward": 0.22879464365541935, "step": 34 }, { "clip_ratio": 0.0, "completion_length": 733.958740234375, "epoch": 0.3548795944233207, "grad_norm": 0.21707987785339355, "learning_rate": 9.821843838470534e-07, "loss": 0.034, "num_tokens": 24225635.0, "reward": 0.9587054029107094, "reward_std": 0.53031075745821, "rewards/accuracy_reward": 0.5680803619325161, "rewards/format_reward": 0.390625, "step": 35 }, { "clip_ratio": 0.0, "completion_length": 737.1797103881836, "epoch": 0.3650190114068441, "grad_norm": 0.1912337839603424, "learning_rate": 9.797464868072486e-07, "loss": 0.033, "num_tokens": 25014260.0, "reward": 1.112723246216774, "reward_std": 0.574417520314455, "rewards/accuracy_reward": 0.5881696492433548, "rewards/format_reward": 0.5245535708963871, "step": 36 }, { "clip_ratio": 0.0, "completion_length": 740.9598541259766, "epoch": 0.37515842839036756, "grad_norm": 0.22646164894104004, "learning_rate": 9.771557365159319e-07, "loss": 0.0325, "num_tokens": 25803688.0, "reward": 1.1941964775323868, "reward_std": 0.5038099698722363, "rewards/accuracy_reward": 0.5524553544819355, "rewards/format_reward": 0.6417410746216774, "step": 37 }, { "clip_ratio": 0.0, "completion_length": 741.780158996582, "epoch": 0.385297845373891, "grad_norm": 0.2826821208000183, "learning_rate": 9.744129584186597e-07, "loss": 0.0653, "num_tokens": 26594931.0, "reward": 1.340401828289032, "reward_std": 0.47122547030448914, "rewards/accuracy_reward": 0.5691964253783226, "rewards/format_reward": 0.7712053582072258, "step": 38 }, { "clip_ratio": 0.0, "completion_length": 783.4699020385742, "epoch": 0.39543726235741444, "grad_norm": 0.1316017210483551, "learning_rate": 9.71519026398956e-07, "loss": 0.0629, "num_tokens": 27434704.0, "reward": 1.3526786416769028, "reward_std": 0.40762051939964294, "rewards/accuracy_reward": 0.5200892835855484, "rewards/format_reward": 0.832589291036129, "step": 39 }, { "clip_ratio": 0.0, "completion_length": 709.8928909301758, "epoch": 0.4055766793409379, "grad_norm": 0.12696978449821472, "learning_rate": 9.68474862499881e-07, "loss": 0.0423, "num_tokens": 28190032.0, "reward": 1.491071492433548, "reward_std": 0.35216479003429413, "rewards/accuracy_reward": 0.5848214253783226, "rewards/format_reward": 0.9062499925494194, "step": 40 }, { "clip_ratio": 0.0, "completion_length": 700.1216735839844, "epoch": 0.4157160963244613, "grad_norm": 0.1216612383723259, "learning_rate": 9.652814366302568e-07, "loss": 0.0672, "num_tokens": 28943317.0, "reward": 1.5658482760190964, "reward_std": 0.34097408317029476, "rewards/accuracy_reward": 0.6484375, "rewards/format_reward": 0.917410708963871, "step": 41 }, { "clip_ratio": 0.0, "completion_length": 703.3582916259766, "epoch": 0.42585551330798477, "grad_norm": 0.11864102631807327, "learning_rate": 9.619397662556433e-07, "loss": 0.0279, "num_tokens": 29702494.0, "reward": 1.537946492433548, "reward_std": 0.2801562137901783, "rewards/accuracy_reward": 0.5892857164144516, "rewards/format_reward": 0.9486607164144516, "step": 42 }, { "clip_ratio": 0.0, "completion_length": 775.4074020385742, "epoch": 0.43599493029150826, "grad_norm": 0.10538308322429657, "learning_rate": 9.5845091607416e-07, "loss": 0.0566, "num_tokens": 30515963.0, "reward": 1.5178572088479996, "reward_std": 0.33493360690772533, "rewards/accuracy_reward": 0.5758928544819355, "rewards/format_reward": 0.9419642984867096, "step": 43 }, { "clip_ratio": 0.0, "completion_length": 792.2288360595703, "epoch": 0.4461343472750317, "grad_norm": 0.10541502386331558, "learning_rate": 9.548159976772592e-07, "loss": 0.0749, "num_tokens": 31355848.0, "reward": 1.53683041036129, "reward_std": 0.3079599365592003, "rewards/accuracy_reward": 0.5747767873108387, "rewards/format_reward": 0.9620535746216774, "step": 44 }, { "clip_ratio": 0.0, "completion_length": 785.3750228881836, "epoch": 0.45627376425855515, "grad_norm": 0.11054891347885132, "learning_rate": 9.510361691955606e-07, "loss": 0.0749, "num_tokens": 32181992.0, "reward": 1.4888393431901932, "reward_std": 0.33042189106345177, "rewards/accuracy_reward": 0.5234375037252903, "rewards/format_reward": 0.9654017835855484, "step": 45 }, { "clip_ratio": 0.0, "completion_length": 743.7846298217773, "epoch": 0.4664131812420786, "grad_norm": 0.10712749511003494, "learning_rate": 9.471126349298556e-07, "loss": 0.0636, "num_tokens": 32978639.0, "reward": 1.5468750447034836, "reward_std": 0.34593483060598373, "rewards/accuracy_reward": 0.5770089402794838, "rewards/format_reward": 0.9698660746216774, "step": 46 }, { "clip_ratio": 0.0, "completion_length": 732.7355194091797, "epoch": 0.47655259822560203, "grad_norm": 0.1037302166223526, "learning_rate": 9.430466449674013e-07, "loss": 0.0546, "num_tokens": 33783522.0, "reward": 1.537946492433548, "reward_std": 0.2701996862888336, "rewards/accuracy_reward": 0.5558035708963871, "rewards/format_reward": 0.9821428582072258, "step": 47 }, { "clip_ratio": 0.0, "completion_length": 794.5357513427734, "epoch": 0.4866920152091255, "grad_norm": 0.09865893423557281, "learning_rate": 9.388394947836278e-07, "loss": 0.0496, "num_tokens": 34622162.0, "reward": 1.4854911416769028, "reward_std": 0.27795105054974556, "rewards/accuracy_reward": 0.5133928544819355, "rewards/format_reward": 0.972098208963871, "step": 48 }, { "clip_ratio": 0.0, "completion_length": 784.8024978637695, "epoch": 0.4968314321926489, "grad_norm": 0.09806529432535172, "learning_rate": 9.344925248293835e-07, "loss": 0.0469, "num_tokens": 35446353.0, "reward": 1.5401786416769028, "reward_std": 0.2609728015959263, "rewards/accuracy_reward": 0.5658482201397419, "rewards/format_reward": 0.9743303582072258, "step": 49 }, { "clip_ratio": 0.0, "completion_length": 755.7600708007812, "epoch": 0.5069708491761724, "grad_norm": 0.09463509172201157, "learning_rate": 9.300071201038501e-07, "loss": 0.0447, "num_tokens": 36249690.0, "reward": 1.5591518729925156, "reward_std": 0.27370808832347393, "rewards/accuracy_reward": 0.5837053693830967, "rewards/format_reward": 0.9754464253783226, "step": 50 }, { "clip_ratio": 0.0, "completion_length": 683.0134201049805, "epoch": 0.5171102661596958, "grad_norm": 0.1009419783949852, "learning_rate": 9.253847097132655e-07, "loss": 0.0259, "num_tokens": 36992494.0, "reward": 1.6160715073347092, "reward_std": 0.2591102607548237, "rewards/accuracy_reward": 0.6316964328289032, "rewards/format_reward": 0.984375, "step": 51 }, { "clip_ratio": 0.0, "completion_length": 744.6250228881836, "epoch": 0.5272496831432193, "grad_norm": 0.10415738075971603, "learning_rate": 9.206267664155906e-07, "loss": 0.0411, "num_tokens": 37780054.0, "reward": 1.561383992433548, "reward_std": 0.2772454805672169, "rewards/accuracy_reward": 0.5825892947614193, "rewards/format_reward": 0.9787946417927742, "step": 52 }, { "clip_ratio": 0.0, "completion_length": 768.4509201049805, "epoch": 0.5373891001267427, "grad_norm": 0.09972584992647171, "learning_rate": 9.157348061512726e-07, "loss": 0.0551, "num_tokens": 38604666.0, "reward": 1.5491072237491608, "reward_std": 0.27522944658994675, "rewards/accuracy_reward": 0.5792410783469677, "rewards/format_reward": 0.9698660746216774, "step": 53 }, { "clip_ratio": 0.0, "completion_length": 828.3806228637695, "epoch": 0.5475285171102662, "grad_norm": 0.08794871717691422, "learning_rate": 9.107103875602458e-07, "loss": 0.0472, "num_tokens": 39465447.0, "reward": 1.5133929252624512, "reward_std": 0.29366694390773773, "rewards/accuracy_reward": 0.5345982164144516, "rewards/format_reward": 0.9787946417927742, "step": 54 }, { "clip_ratio": 0.0, "completion_length": 784.3370895385742, "epoch": 0.5576679340937896, "grad_norm": 0.10152813047170639, "learning_rate": 9.055551114853295e-07, "loss": 0.049, "num_tokens": 40298605.0, "reward": 1.507812574505806, "reward_std": 0.2921369504183531, "rewards/accuracy_reward": 0.5312499925494194, "rewards/format_reward": 0.9765624925494194, "step": 55 }, { "clip_ratio": 0.0, "completion_length": 721.1373062133789, "epoch": 0.5678073510773131, "grad_norm": 0.09925797581672668, "learning_rate": 9.002706204621802e-07, "loss": 0.0244, "num_tokens": 41068424.0, "reward": 1.585937574505806, "reward_std": 0.26984195969998837, "rewards/accuracy_reward": 0.5959821343421936, "rewards/format_reward": 0.9899553582072258, "step": 56 }, { "clip_ratio": 0.0, "completion_length": 754.9319458007812, "epoch": 0.5779467680608364, "grad_norm": 0.0985010415315628, "learning_rate": 8.948585981959578e-07, "loss": 0.055, "num_tokens": 41876427.0, "reward": 1.5926340222358704, "reward_std": 0.2599882632493973, "rewards/accuracy_reward": 0.618303582072258, "rewards/format_reward": 0.9743303507566452, "step": 57 }, { "clip_ratio": 0.0, "completion_length": 735.5156631469727, "epoch": 0.5880861850443599, "grad_norm": 0.09651587158441544, "learning_rate": 8.893207690248775e-07, "loss": 0.0342, "num_tokens": 42656513.0, "reward": 1.5904018431901932, "reward_std": 0.2695194110274315, "rewards/accuracy_reward": 0.6071428507566452, "rewards/format_reward": 0.9832589328289032, "step": 58 }, { "clip_ratio": 0.0, "completion_length": 701.5614166259766, "epoch": 0.5982256020278834, "grad_norm": 0.10193536430597305, "learning_rate": 8.836588973708128e-07, "loss": 0.0322, "num_tokens": 43426704.0, "reward": 1.537946492433548, "reward_std": 0.2486334890127182, "rewards/accuracy_reward": 0.5479910708963871, "rewards/format_reward": 0.9899553507566452, "step": 59 }, { "clip_ratio": 0.0, "completion_length": 767.4029312133789, "epoch": 0.6083650190114068, "grad_norm": 0.09067587554454803, "learning_rate": 8.778747871771291e-07, "loss": 0.0462, "num_tokens": 44251545.0, "reward": 1.5535715073347092, "reward_std": 0.2701433580368757, "rewards/accuracy_reward": 0.5714285783469677, "rewards/format_reward": 0.9821428582072258, "step": 60 }, { "clip_ratio": 0.0, "completion_length": 728.3716888427734, "epoch": 0.6185044359949303, "grad_norm": 0.10155454277992249, "learning_rate": 8.719702813339247e-07, "loss": 0.0326, "num_tokens": 45026510.0, "reward": 1.577008992433548, "reward_std": 0.2642171438783407, "rewards/accuracy_reward": 0.5904017873108387, "rewards/format_reward": 0.9866071343421936, "step": 61 }, { "clip_ratio": 0.0, "completion_length": 777.1127548217773, "epoch": 0.6286438529784537, "grad_norm": 0.09913618117570877, "learning_rate": 8.659472610908627e-07, "loss": 0.0358, "num_tokens": 45852195.0, "reward": 1.5312500894069672, "reward_std": 0.27686445228755474, "rewards/accuracy_reward": 0.5446428544819355, "rewards/format_reward": 0.986607126891613, "step": 62 }, { "clip_ratio": 0.0, "completion_length": 697.9274826049805, "epoch": 0.6387832699619772, "grad_norm": 0.10096921026706696, "learning_rate": 8.598076454577814e-07, "loss": 0.0279, "num_tokens": 46601834.0, "reward": 1.6305804252624512, "reward_std": 0.2701211776584387, "rewards/accuracy_reward": 0.6428571492433548, "rewards/format_reward": 0.9877232015132904, "step": 63 }, { "clip_ratio": 0.0, "completion_length": 724.1774826049805, "epoch": 0.6489226869455006, "grad_norm": 0.09761769324541092, "learning_rate": 8.535533905932737e-07, "loss": 0.0398, "num_tokens": 47398825.0, "reward": 1.601562574505806, "reward_std": 0.23534459993243217, "rewards/accuracy_reward": 0.6093749962747097, "rewards/format_reward": 0.9921875, "step": 64 }, { "clip_ratio": 0.0, "completion_length": 717.3169937133789, "epoch": 0.6590621039290241, "grad_norm": 0.11670506000518799, "learning_rate": 8.471864891814304e-07, "loss": 0.0516, "num_tokens": 48160645.0, "reward": 1.6060268580913544, "reward_std": 0.26124117337167263, "rewards/accuracy_reward": 0.6216517798602581, "rewards/format_reward": 0.984375, "step": 65 }, { "clip_ratio": 0.0, "completion_length": 766.2712478637695, "epoch": 0.6692015209125475, "grad_norm": 0.09610045701265335, "learning_rate": 8.407089697969456e-07, "loss": 0.0429, "num_tokens": 48972088.0, "reward": 1.5725447088479996, "reward_std": 0.25589474849402905, "rewards/accuracy_reward": 0.5937499850988388, "rewards/format_reward": 0.9787946417927742, "step": 66 }, { "clip_ratio": 0.0, "completion_length": 794.2879867553711, "epoch": 0.679340937896071, "grad_norm": 0.08830226212739944, "learning_rate": 8.341228962587881e-07, "loss": 0.0485, "num_tokens": 49819898.0, "reward": 1.545758992433548, "reward_std": 0.2661575675010681, "rewards/accuracy_reward": 0.5669642947614193, "rewards/format_reward": 0.9787946343421936, "step": 67 }, { "clip_ratio": 0.0, "completion_length": 702.197582244873, "epoch": 0.6894803548795945, "grad_norm": 0.08808374404907227, "learning_rate": 8.274303669726426e-07, "loss": 0.0427, "num_tokens": 50570387.0, "reward": 1.641741156578064, "reward_std": 0.24876852426677942, "rewards/accuracy_reward": 0.6595982052385807, "rewards/format_reward": 0.9821428582072258, "step": 68 }, { "clip_ratio": 0.0, "completion_length": 780.8024978637695, "epoch": 0.6996197718631179, "grad_norm": 0.0812569409608841, "learning_rate": 8.206335142623304e-07, "loss": 0.0349, "num_tokens": 51399746.0, "reward": 1.5446429252624512, "reward_std": 0.21720573492348194, "rewards/accuracy_reward": 0.5680803582072258, "rewards/format_reward": 0.9765624925494194, "step": 69 }, { "clip_ratio": 0.0, "completion_length": 757.3337326049805, "epoch": 0.7097591888466414, "grad_norm": 0.08954399079084396, "learning_rate": 8.137345036904259e-07, "loss": 0.0439, "num_tokens": 52197693.0, "reward": 1.5390625596046448, "reward_std": 0.25285689160227776, "rewards/accuracy_reward": 0.5624999962747097, "rewards/format_reward": 0.9765624925494194, "step": 70 }, { "clip_ratio": 0.0, "completion_length": 731.8973541259766, "epoch": 0.7198986058301647, "grad_norm": 0.09004837274551392, "learning_rate": 8.067355333682797e-07, "loss": 0.0354, "num_tokens": 52992241.0, "reward": 1.562500074505806, "reward_std": 0.22563916258513927, "rewards/accuracy_reward": 0.5814732126891613, "rewards/format_reward": 0.9810267761349678, "step": 71 }, { "clip_ratio": 0.0, "completion_length": 734.0893249511719, "epoch": 0.7300380228136882, "grad_norm": 0.0921608954668045, "learning_rate": 7.996388332556734e-07, "loss": 0.0553, "num_tokens": 53778705.0, "reward": 1.561383992433548, "reward_std": 0.25903933122754097, "rewards/accuracy_reward": 0.5792410857975483, "rewards/format_reward": 0.9821428582072258, "step": 72 }, { "clip_ratio": 0.0, "completion_length": 783.1897735595703, "epoch": 0.7401774397972116, "grad_norm": 0.09094219654798508, "learning_rate": 7.924466644503264e-07, "loss": 0.0356, "num_tokens": 54612939.0, "reward": 1.5558036416769028, "reward_std": 0.25537824630737305, "rewards/accuracy_reward": 0.5691964365541935, "rewards/format_reward": 0.9866071343421936, "step": 73 }, { "clip_ratio": 0.0, "completion_length": 725.2801666259766, "epoch": 0.7503168567807351, "grad_norm": 0.09524378925561905, "learning_rate": 7.85161318467482e-07, "loss": 0.0229, "num_tokens": 55398174.0, "reward": 1.6205357909202576, "reward_std": 0.2613362595438957, "rewards/accuracy_reward": 0.6305803582072258, "rewards/format_reward": 0.9899553507566452, "step": 74 }, { "clip_ratio": 0.0, "completion_length": 710.1853103637695, "epoch": 0.7604562737642585, "grad_norm": 0.0883307084441185, "learning_rate": 7.777851165098011e-07, "loss": 0.0235, "num_tokens": 56165180.0, "reward": 1.6183036416769028, "reward_std": 0.21434970386326313, "rewards/accuracy_reward": 0.6305803619325161, "rewards/format_reward": 0.9877232015132904, "step": 75 }, { "clip_ratio": 0.0, "completion_length": 690.4687805175781, "epoch": 0.770595690747782, "grad_norm": 0.09709777683019638, "learning_rate": 7.703204087277988e-07, "loss": 0.0316, "num_tokens": 56911080.0, "reward": 1.5267857760190964, "reward_std": 0.23126200772821903, "rewards/accuracy_reward": 0.5401785671710968, "rewards/format_reward": 0.9866071492433548, "step": 76 }, { "clip_ratio": 0.0, "completion_length": 753.3359756469727, "epoch": 0.7807351077313055, "grad_norm": 0.08559317141771317, "learning_rate": 7.627695734710564e-07, "loss": 0.0362, "num_tokens": 57708293.0, "reward": 1.601562574505806, "reward_std": 0.22745812311768532, "rewards/accuracy_reward": 0.6216517761349678, "rewards/format_reward": 0.979910708963871, "step": 77 }, { "clip_ratio": 0.0, "completion_length": 771.0870895385742, "epoch": 0.7908745247148289, "grad_norm": 0.08436188846826553, "learning_rate": 7.551350165304499e-07, "loss": 0.0295, "num_tokens": 58516963.0, "reward": 1.6361608058214188, "reward_std": 0.22785574942827225, "rewards/accuracy_reward": 0.6540178693830967, "rewards/format_reward": 0.9821428582072258, "step": 78 }, { "clip_ratio": 0.0, "completion_length": 738.5379791259766, "epoch": 0.8010139416983524, "grad_norm": 0.08563719689846039, "learning_rate": 7.474191703716338e-07, "loss": 0.0205, "num_tokens": 59309253.0, "reward": 1.6015625894069672, "reward_std": 0.211506649851799, "rewards/accuracy_reward": 0.6071428544819355, "rewards/format_reward": 0.9944196417927742, "step": 79 }, { "clip_ratio": 0.0, "completion_length": 729.372802734375, "epoch": 0.8111533586818758, "grad_norm": 0.09525574743747711, "learning_rate": 7.396244933600284e-07, "loss": 0.0426, "num_tokens": 60087059.0, "reward": 1.6004464775323868, "reward_std": 0.24874755926430225, "rewards/accuracy_reward": 0.608258917927742, "rewards/format_reward": 0.9921874925494194, "step": 80 }, { "clip_ratio": 0.0, "completion_length": 742.4721298217773, "epoch": 0.8212927756653993, "grad_norm": 0.0932924821972847, "learning_rate": 7.317534689775527e-07, "loss": 0.035, "num_tokens": 60872018.0, "reward": 1.5714286267757416, "reward_std": 0.25846732780337334, "rewards/accuracy_reward": 0.5881696417927742, "rewards/format_reward": 0.9832589253783226, "step": 81 }, { "clip_ratio": 0.0, "completion_length": 709.4643173217773, "epoch": 0.8314321926489227, "grad_norm": 0.09472807496786118, "learning_rate": 7.238086050313562e-07, "loss": 0.0341, "num_tokens": 61642954.0, "reward": 1.5870536416769028, "reward_std": 0.23454621247947216, "rewards/accuracy_reward": 0.5993303582072258, "rewards/format_reward": 0.9877232015132904, "step": 82 }, { "clip_ratio": 0.0, "completion_length": 716.0413208007812, "epoch": 0.8415716096324461, "grad_norm": 0.10518212616443634, "learning_rate": 7.157924328548002e-07, "loss": 0.045, "num_tokens": 62418119.0, "reward": 1.6026786416769028, "reward_std": 0.2639338057488203, "rewards/accuracy_reward": 0.6160714328289032, "rewards/format_reward": 0.9866071343421936, "step": 83 }, { "clip_ratio": 0.0, "completion_length": 724.6919937133789, "epoch": 0.8517110266159695, "grad_norm": 0.0993557870388031, "learning_rate": 7.077075065009433e-07, "loss": 0.0374, "num_tokens": 63200755.0, "reward": 1.649553656578064, "reward_std": 0.2632820960134268, "rewards/accuracy_reward": 0.659598208963871, "rewards/format_reward": 0.9899553507566452, "step": 84 }, { "clip_ratio": 0.0, "completion_length": 694.9643096923828, "epoch": 0.861850443599493, "grad_norm": 0.0879211500287056, "learning_rate": 6.995564019287869e-07, "loss": 0.0205, "num_tokens": 63953259.0, "reward": 1.5792411267757416, "reward_std": 0.19979050569236279, "rewards/accuracy_reward": 0.5859375037252903, "rewards/format_reward": 0.9933035671710968, "step": 85 }, { "clip_ratio": 0.0, "completion_length": 764.1395416259766, "epoch": 0.8719898605830165, "grad_norm": 0.3273586332798004, "learning_rate": 6.913417161825449e-07, "loss": 0.027, "num_tokens": 64772840.0, "reward": 1.5647322237491608, "reward_std": 0.23416299000382423, "rewards/accuracy_reward": 0.5758928582072258, "rewards/format_reward": 0.9888392686843872, "step": 86 }, { "clip_ratio": 0.0, "completion_length": 723.0658798217773, "epoch": 0.8821292775665399, "grad_norm": 0.231331005692482, "learning_rate": 6.830660665641897e-07, "loss": 0.0197, "num_tokens": 65550427.0, "reward": 1.6183036714792252, "reward_std": 0.2430772576481104, "rewards/accuracy_reward": 0.6272321380674839, "rewards/format_reward": 0.9910714253783226, "step": 87 }, { "clip_ratio": 0.0, "completion_length": 756.286865234375, "epoch": 0.8922686945500634, "grad_norm": 0.09298628568649292, "learning_rate": 6.747320897995492e-07, "loss": 0.0426, "num_tokens": 66357052.0, "reward": 1.5915179252624512, "reward_std": 0.2366574928164482, "rewards/accuracy_reward": 0.6149553656578064, "rewards/format_reward": 0.9765625, "step": 88 }, { "clip_ratio": 0.0, "completion_length": 686.9029312133789, "epoch": 0.9024081115335868, "grad_norm": 0.0874035432934761, "learning_rate": 6.66342441198212e-07, "loss": 0.015, "num_tokens": 67096421.0, "reward": 1.6127232909202576, "reward_std": 0.1957838051021099, "rewards/accuracy_reward": 0.6205357126891613, "rewards/format_reward": 0.9921874925494194, "step": 89 }, { "clip_ratio": 0.0, "completion_length": 701.6573944091797, "epoch": 0.9125475285171103, "grad_norm": 0.09878430515527725, "learning_rate": 6.578997938075125e-07, "loss": 0.0477, "num_tokens": 67849578.0, "reward": 1.5691965073347092, "reward_std": 0.2424814011901617, "rewards/accuracy_reward": 0.5881696492433548, "rewards/format_reward": 0.9810267835855484, "step": 90 }, { "clip_ratio": 0.0, "completion_length": 717.0368499755859, "epoch": 0.9226869455006337, "grad_norm": 0.0928407534956932, "learning_rate": 6.494068375608646e-07, "loss": 0.0515, "num_tokens": 68626667.0, "reward": 1.5803571939468384, "reward_std": 0.2204410433769226, "rewards/accuracy_reward": 0.5993303582072258, "rewards/format_reward": 0.9810267835855484, "step": 91 }, { "clip_ratio": 0.0, "completion_length": 750.7065048217773, "epoch": 0.9328263624841572, "grad_norm": 0.09953666478395462, "learning_rate": 6.408662784207149e-07, "loss": 0.0434, "num_tokens": 69438676.0, "reward": 1.6037947237491608, "reward_std": 0.263325035572052, "rewards/accuracy_reward": 0.621651791036129, "rewards/format_reward": 0.9821428507566452, "step": 92 }, { "clip_ratio": 0.0, "completion_length": 673.391773223877, "epoch": 0.9429657794676806, "grad_norm": 0.09401652216911316, "learning_rate": 6.322808375163895e-07, "loss": 0.0215, "num_tokens": 70168915.0, "reward": 1.6227679252624512, "reward_std": 0.1998270694166422, "rewards/accuracy_reward": 0.629464291036129, "rewards/format_reward": 0.9933035746216774, "step": 93 }, { "clip_ratio": 0.0, "completion_length": 706.0636520385742, "epoch": 0.9531051964512041, "grad_norm": 0.10031445324420929, "learning_rate": 6.236532502771077e-07, "loss": 0.029, "num_tokens": 70935148.0, "reward": 1.600446492433548, "reward_std": 0.2434455920010805, "rewards/accuracy_reward": 0.6104910634458065, "rewards/format_reward": 0.9899553582072258, "step": 94 }, { "clip_ratio": 0.0, "completion_length": 702.7511444091797, "epoch": 0.9632446134347274, "grad_norm": 0.09827374666929245, "learning_rate": 6.149862655604403e-07, "loss": 0.0358, "num_tokens": 71690501.0, "reward": 1.5982143431901932, "reward_std": 0.2399199977517128, "rewards/accuracy_reward": 0.6149553507566452, "rewards/format_reward": 0.9832589253783226, "step": 95 }, { "clip_ratio": 0.0, "completion_length": 713.8114166259766, "epoch": 0.973384030418251, "grad_norm": 0.08831395208835602, "learning_rate": 6.062826447764883e-07, "loss": 0.031, "num_tokens": 72452820.0, "reward": 1.6127232760190964, "reward_std": 0.2217018250375986, "rewards/accuracy_reward": 0.6238839253783226, "rewards/format_reward": 0.9888392761349678, "step": 96 }, { "clip_ratio": 0.0, "completion_length": 701.4710083007812, "epoch": 0.9835234474017744, "grad_norm": 0.12909561395645142, "learning_rate": 5.975451610080642e-07, "loss": 0.0299, "num_tokens": 73209146.0, "reward": 1.5602679252624512, "reward_std": 0.21769515611231327, "rewards/accuracy_reward": 0.5680803507566452, "rewards/format_reward": 0.9921874925494194, "step": 97 }, { "clip_ratio": 0.0, "completion_length": 698.564453125, "epoch": 0.9936628643852978, "grad_norm": 0.09522398561239243, "learning_rate": 5.887765981271517e-07, "loss": 0.0337, "num_tokens": 73991697.0, "reward": 1.56808041036129, "reward_std": 0.24109553173184395, "rewards/accuracy_reward": 0.5792410746216774, "rewards/format_reward": 0.9888392761349678, "step": 98 }, { "clip_ratio": 0.0, "completion_length": 719.4676666259766, "epoch": 1.0101394169835234, "grad_norm": 0.09128111600875854, "learning_rate": 5.7997974990793e-07, "loss": 0.0236, "num_tokens": 74765332.0, "reward": 1.648437574505806, "reward_std": 0.22949284128844738, "rewards/accuracy_reward": 0.65625, "rewards/format_reward": 0.9921874925494194, "step": 99 }, { "clip_ratio": 0.0, "completion_length": 720.794677734375, "epoch": 1.020278833967047, "grad_norm": 0.09661993384361267, "learning_rate": 5.711574191366427e-07, "loss": 0.0317, "num_tokens": 75538540.0, "reward": 1.617187574505806, "reward_std": 0.21921591460704803, "rewards/accuracy_reward": 0.6316964328289032, "rewards/format_reward": 0.9854910671710968, "step": 100 }, { "clip_ratio": 0.0, "completion_length": 715.5815048217773, "epoch": 1.0304182509505704, "grad_norm": 0.09119318425655365, "learning_rate": 5.623124167185929e-07, "loss": 0.0238, "num_tokens": 76302229.0, "reward": 1.6149554401636124, "reward_std": 0.23784785345196724, "rewards/accuracy_reward": 0.6328125, "rewards/format_reward": 0.9821428507566452, "step": 101 }, { "clip_ratio": 0.0, "completion_length": 671.747802734375, "epoch": 1.0405576679340938, "grad_norm": 0.0963052362203598, "learning_rate": 5.534475607825565e-07, "loss": 0.0087, "num_tokens": 77031403.0, "reward": 1.6227679252624512, "reward_std": 0.22612315602600574, "rewards/accuracy_reward": 0.6272321417927742, "rewards/format_reward": 0.9955357015132904, "step": 102 }, { "clip_ratio": 0.0, "completion_length": 736.8772659301758, "epoch": 1.0506970849176172, "grad_norm": 0.09166496247053146, "learning_rate": 5.445656757828879e-07, "loss": 0.0313, "num_tokens": 77822949.0, "reward": 1.5044643431901932, "reward_std": 0.22669786028563976, "rewards/accuracy_reward": 0.5189732238650322, "rewards/format_reward": 0.9854910746216774, "step": 103 }, { "clip_ratio": 0.0, "completion_length": 719.7199020385742, "epoch": 1.0608365019011408, "grad_norm": 0.09398168325424194, "learning_rate": 5.356695915996161e-07, "loss": 0.0265, "num_tokens": 78602178.0, "reward": 1.5602679252624512, "reward_std": 0.25012633949518204, "rewards/accuracy_reward": 0.5691964328289032, "rewards/format_reward": 0.9910714253783226, "step": 104 }, { "clip_ratio": 0.0, "completion_length": 688.8962249755859, "epoch": 1.0709759188846641, "grad_norm": 0.11429024487733841, "learning_rate": 5.267621426368075e-07, "loss": 0.0246, "num_tokens": 79338645.0, "reward": 1.6383929550647736, "reward_std": 0.266002238728106, "rewards/accuracy_reward": 0.6495535746216774, "rewards/format_reward": 0.9888392761349678, "step": 105 }, { "clip_ratio": 0.0, "completion_length": 645.3884124755859, "epoch": 1.0811153358681875, "grad_norm": 0.0960325375199318, "learning_rate": 5.178461669194903e-07, "loss": 0.0291, "num_tokens": 80044121.0, "reward": 1.670758992433548, "reward_std": 0.21290560998022556, "rewards/accuracy_reward": 0.6808035671710968, "rewards/format_reward": 0.9899553582072258, "step": 106 }, { "clip_ratio": 0.0, "completion_length": 682.0167694091797, "epoch": 1.091254752851711, "grad_norm": 0.08784578740596771, "learning_rate": 5.08924505189423e-07, "loss": 0.0275, "num_tokens": 80781168.0, "reward": 1.649553656578064, "reward_std": 0.1967218518257141, "rewards/accuracy_reward": 0.65625, "rewards/format_reward": 0.9933035597205162, "step": 107 }, { "clip_ratio": 0.0, "completion_length": 718.1272659301758, "epoch": 1.1013941698352345, "grad_norm": 0.08991685509681702, "learning_rate": 5e-07, "loss": 0.0394, "num_tokens": 81548730.0, "reward": 1.626116156578064, "reward_std": 0.24366096779704094, "rewards/accuracy_reward": 0.6450892835855484, "rewards/format_reward": 0.981026791036129, "step": 108 }, { "clip_ratio": 0.0, "completion_length": 746.9297256469727, "epoch": 1.111533586818758, "grad_norm": 0.10698525607585907, "learning_rate": 4.91075494810577e-07, "loss": 0.0338, "num_tokens": 82335971.0, "reward": 1.6015625596046448, "reward_std": 0.2378039713948965, "rewards/accuracy_reward": 0.6160714253783226, "rewards/format_reward": 0.9854910671710968, "step": 109 }, { "clip_ratio": 0.0, "completion_length": 744.1752624511719, "epoch": 1.1216730038022813, "grad_norm": 0.0870419293642044, "learning_rate": 4.821538330805098e-07, "loss": 0.0489, "num_tokens": 83129272.0, "reward": 1.5982143580913544, "reward_std": 0.22117138653993607, "rewards/accuracy_reward": 0.6183035746216774, "rewards/format_reward": 0.979910708963871, "step": 110 }, { "clip_ratio": 0.0, "completion_length": 729.8170013427734, "epoch": 1.131812420785805, "grad_norm": 0.09355349838733673, "learning_rate": 4.732378573631924e-07, "loss": 0.0372, "num_tokens": 83905916.0, "reward": 1.6004465073347092, "reward_std": 0.24993346445262432, "rewards/accuracy_reward": 0.609375, "rewards/format_reward": 0.9910714253783226, "step": 111 }, { "clip_ratio": 0.0, "completion_length": 730.1663208007812, "epoch": 1.1419518377693283, "grad_norm": 0.2215433269739151, "learning_rate": 4.643304084003838e-07, "loss": 0.025, "num_tokens": 84687593.0, "reward": 1.5837054252624512, "reward_std": 0.23716635070741177, "rewards/accuracy_reward": 0.594866082072258, "rewards/format_reward": 0.9888392686843872, "step": 112 }, { "clip_ratio": 0.0, "completion_length": 704.9989242553711, "epoch": 1.1520912547528517, "grad_norm": 0.10191125422716141, "learning_rate": 4.55434324217112e-07, "loss": 0.0477, "num_tokens": 85448512.0, "reward": 1.5513393580913544, "reward_std": 0.24850638210773468, "rewards/accuracy_reward": 0.5691964291036129, "rewards/format_reward": 0.9821428433060646, "step": 113 }, { "clip_ratio": 0.0, "completion_length": 696.9486846923828, "epoch": 1.162230671736375, "grad_norm": 0.08355645090341568, "learning_rate": 4.4655243921744367e-07, "loss": 0.0175, "num_tokens": 86197938.0, "reward": 1.5982143729925156, "reward_std": 0.1789538525044918, "rewards/accuracy_reward": 0.6049107164144516, "rewards/format_reward": 0.9933035671710968, "step": 114 }, { "clip_ratio": 0.0, "completion_length": 697.7143096923828, "epoch": 1.1723700887198987, "grad_norm": 0.08945267647504807, "learning_rate": 4.37687583281407e-07, "loss": 0.0261, "num_tokens": 86950306.0, "reward": 1.5982143580913544, "reward_std": 0.19760245829820633, "rewards/accuracy_reward": 0.6071428619325161, "rewards/format_reward": 0.9910714253783226, "step": 115 }, { "clip_ratio": 0.0, "completion_length": 681.9843978881836, "epoch": 1.182509505703422, "grad_norm": 0.10237658023834229, "learning_rate": 4.2884258086335745e-07, "loss": 0.017, "num_tokens": 87700604.0, "reward": 1.625000074505806, "reward_std": 0.23482281155884266, "rewards/accuracy_reward": 0.6328124962747097, "rewards/format_reward": 0.9921875, "step": 116 }, { "clip_ratio": 0.0, "completion_length": 679.974365234375, "epoch": 1.1926489226869454, "grad_norm": 0.09299944341182709, "learning_rate": 4.2002025009206987e-07, "loss": 0.0178, "num_tokens": 88441813.0, "reward": 1.6104911416769028, "reward_std": 0.19839700311422348, "rewards/accuracy_reward": 0.6171875037252903, "rewards/format_reward": 0.9933035597205162, "step": 117 }, { "clip_ratio": 0.0, "completion_length": 686.677490234375, "epoch": 1.202788339670469, "grad_norm": 0.16542084515094757, "learning_rate": 4.1122340187284845e-07, "loss": 0.0392, "num_tokens": 89179364.0, "reward": 1.609375074505806, "reward_std": 0.2622632719576359, "rewards/accuracy_reward": 0.6183035597205162, "rewards/format_reward": 0.991071417927742, "step": 118 }, { "clip_ratio": 0.0, "completion_length": 694.8125228881836, "epoch": 1.2129277566539924, "grad_norm": 0.11430428177118301, "learning_rate": 4.0245483899193586e-07, "loss": 0.0139, "num_tokens": 89929292.0, "reward": 1.6049107760190964, "reward_std": 0.22252802923321724, "rewards/accuracy_reward": 0.610491082072258, "rewards/format_reward": 0.9944196417927742, "step": 119 }, { "clip_ratio": 0.0, "completion_length": 684.7466812133789, "epoch": 1.2230671736375158, "grad_norm": 0.08012707531452179, "learning_rate": 3.937173552235116e-07, "loss": 0.0213, "num_tokens": 90675633.0, "reward": 1.5915179401636124, "reward_std": 0.16910855192691088, "rewards/accuracy_reward": 0.6037946492433548, "rewards/format_reward": 0.9877232164144516, "step": 120 }, { "clip_ratio": 0.0, "completion_length": 690.8917694091797, "epoch": 1.2332065906210392, "grad_norm": 0.10144107788801193, "learning_rate": 3.850137344395598e-07, "loss": 0.0521, "num_tokens": 91417888.0, "reward": 1.6227679550647736, "reward_std": 0.22748247906565666, "rewards/accuracy_reward": 0.6350446417927742, "rewards/format_reward": 0.9877232164144516, "step": 121 }, { "clip_ratio": 0.0, "completion_length": 754.5803909301758, "epoch": 1.2433460076045628, "grad_norm": 0.09234697371721268, "learning_rate": 3.763467497228922e-07, "loss": 0.0348, "num_tokens": 92232024.0, "reward": 1.5513393580913544, "reward_std": 0.22778411209583282, "rewards/accuracy_reward": 0.5725446529686451, "rewards/format_reward": 0.9787946343421936, "step": 122 }, { "clip_ratio": 0.0, "completion_length": 718.1328506469727, "epoch": 1.2534854245880862, "grad_norm": 0.09252262860536575, "learning_rate": 3.677191624836106e-07, "loss": 0.0375, "num_tokens": 92997183.0, "reward": 1.6417411267757416, "reward_std": 0.2288501374423504, "rewards/accuracy_reward": 0.651785708963871, "rewards/format_reward": 0.9899553433060646, "step": 123 }, { "clip_ratio": 0.0, "completion_length": 738.9018173217773, "epoch": 1.2636248415716096, "grad_norm": 0.09162338823080063, "learning_rate": 3.591337215792851e-07, "loss": 0.0228, "num_tokens": 93787407.0, "reward": 1.6037947237491608, "reward_std": 0.22520777396857738, "rewards/accuracy_reward": 0.6149553656578064, "rewards/format_reward": 0.9888392686843872, "step": 124 }, { "clip_ratio": 0.0, "completion_length": 709.0446853637695, "epoch": 1.2737642585551332, "grad_norm": 0.09890223294496536, "learning_rate": 3.505931624391355e-07, "loss": 0.0222, "num_tokens": 94540975.0, "reward": 1.6272322088479996, "reward_std": 0.25062085315585136, "rewards/accuracy_reward": 0.6316964328289032, "rewards/format_reward": 0.995535708963871, "step": 125 }, { "clip_ratio": 0.0, "completion_length": 695.0937805175781, "epoch": 1.2839036755386566, "grad_norm": 0.09244479238986969, "learning_rate": 3.421002061924876e-07, "loss": 0.0234, "num_tokens": 95290643.0, "reward": 1.6171875894069672, "reward_std": 0.23021192848682404, "rewards/accuracy_reward": 0.6350446417927742, "rewards/format_reward": 0.9821428507566452, "step": 126 }, { "clip_ratio": 0.0, "completion_length": 705.2265853881836, "epoch": 1.29404309252218, "grad_norm": 0.08838233351707458, "learning_rate": 3.3365755880178807e-07, "loss": 0.0274, "num_tokens": 96044230.0, "reward": 1.6517858058214188, "reward_std": 0.19274440594017506, "rewards/accuracy_reward": 0.6629464291036129, "rewards/format_reward": 0.9888392761349678, "step": 127 }, { "clip_ratio": 0.0, "completion_length": 730.522346496582, "epoch": 1.3041825095057034, "grad_norm": 0.10294985771179199, "learning_rate": 3.2526791020045087e-07, "loss": 0.0338, "num_tokens": 96831378.0, "reward": 1.5580357909202576, "reward_std": 0.2531749904155731, "rewards/accuracy_reward": 0.5758928544819355, "rewards/format_reward": 0.9821428507566452, "step": 128 }, { "clip_ratio": 0.0, "completion_length": 683.2656555175781, "epoch": 1.3143219264892267, "grad_norm": 0.09631213545799255, "learning_rate": 3.169339334358104e-07, "loss": 0.019, "num_tokens": 97576928.0, "reward": 1.6283482909202576, "reward_std": 0.22065912000834942, "rewards/accuracy_reward": 0.6372767835855484, "rewards/format_reward": 0.9910714328289032, "step": 129 }, { "clip_ratio": 0.0, "completion_length": 700.1573944091797, "epoch": 1.3244613434727504, "grad_norm": 0.09407830238342285, "learning_rate": 3.086582838174551e-07, "loss": 0.0196, "num_tokens": 98336757.0, "reward": 1.6116072237491608, "reward_std": 0.20910983718931675, "rewards/accuracy_reward": 0.6183035671710968, "rewards/format_reward": 0.9933035597205162, "step": 130 }, { "clip_ratio": 0.0, "completion_length": 694.3426666259766, "epoch": 1.3346007604562737, "grad_norm": 0.10380922257900238, "learning_rate": 3.004435980712129e-07, "loss": 0.0372, "num_tokens": 99082352.0, "reward": 1.5892857611179352, "reward_std": 0.23904000967741013, "rewards/accuracy_reward": 0.6026785783469677, "rewards/format_reward": 0.9866071343421936, "step": 131 }, { "clip_ratio": 0.0, "completion_length": 689.7589569091797, "epoch": 1.3447401774397973, "grad_norm": 0.08147790282964706, "learning_rate": 2.922924934990568e-07, "loss": 0.0384, "num_tokens": 99825528.0, "reward": 1.6372768580913544, "reward_std": 0.19569051638245583, "rewards/accuracy_reward": 0.645089291036129, "rewards/format_reward": 0.9921874925494194, "step": 132 }, { "clip_ratio": 0.0, "completion_length": 652.4018096923828, "epoch": 1.3548795944233207, "grad_norm": 0.10540962219238281, "learning_rate": 2.8420756714519954e-07, "loss": 0.0049, "num_tokens": 100554384.0, "reward": 1.6138393431901932, "reward_std": 0.2248824257403612, "rewards/accuracy_reward": 0.6183035708963871, "rewards/format_reward": 0.995535708963871, "step": 133 }, { "clip_ratio": 0.0, "completion_length": 691.8192138671875, "epoch": 1.3650190114068441, "grad_norm": 0.10231253504753113, "learning_rate": 2.7619139496864376e-07, "loss": 0.0299, "num_tokens": 101310238.0, "reward": 1.5345982760190964, "reward_std": 0.22884078323841095, "rewards/accuracy_reward": 0.543526791036129, "rewards/format_reward": 0.991071417927742, "step": 134 }, { "clip_ratio": 0.0, "completion_length": 654.686408996582, "epoch": 1.3751584283903675, "grad_norm": 0.10083113610744476, "learning_rate": 2.6824653102244727e-07, "loss": 0.0201, "num_tokens": 102030909.0, "reward": 1.6171875596046448, "reward_std": 0.22183530405163765, "rewards/accuracy_reward": 0.6227678582072258, "rewards/format_reward": 0.9944196343421936, "step": 135 }, { "clip_ratio": 0.0, "completion_length": 673.8951187133789, "epoch": 1.385297845373891, "grad_norm": 0.10060420632362366, "learning_rate": 2.603755066399718e-07, "loss": 0.0253, "num_tokens": 102764799.0, "reward": 1.6138393431901932, "reward_std": 0.2160534653812647, "rewards/accuracy_reward": 0.620535708963871, "rewards/format_reward": 0.9933035671710968, "step": 136 }, { "clip_ratio": 0.0, "completion_length": 719.1886596679688, "epoch": 1.3954372623574145, "grad_norm": 0.08712632954120636, "learning_rate": 2.5258082962836614e-07, "loss": 0.0291, "num_tokens": 103543752.0, "reward": 1.502232238650322, "reward_std": 0.17017750442028046, "rewards/accuracy_reward": 0.5100446492433548, "rewards/format_reward": 0.9921874850988388, "step": 137 }, { "clip_ratio": 0.0, "completion_length": 696.5636520385742, "epoch": 1.4055766793409379, "grad_norm": 0.20818105340003967, "learning_rate": 2.4486498346955023e-07, "loss": 0.0298, "num_tokens": 104290705.0, "reward": 1.6026786416769028, "reward_std": 0.24658331461250782, "rewards/accuracy_reward": 0.6104910746216774, "rewards/format_reward": 0.9921874925494194, "step": 138 }, { "clip_ratio": 0.0, "completion_length": 707.4475784301758, "epoch": 1.4157160963244613, "grad_norm": 0.09881385415792465, "learning_rate": 2.372304265289436e-07, "loss": 0.0172, "num_tokens": 105048874.0, "reward": 1.56026791036129, "reward_std": 0.24689678102731705, "rewards/accuracy_reward": 0.5703125037252903, "rewards/format_reward": 0.9899553582072258, "step": 139 }, { "clip_ratio": 0.0, "completion_length": 685.3381958007812, "epoch": 1.4258555133079849, "grad_norm": 0.0895194411277771, "learning_rate": 2.2967959127220137e-07, "loss": 0.0223, "num_tokens": 105784897.0, "reward": 1.6450893580913544, "reward_std": 0.20549658872187138, "rewards/accuracy_reward": 0.6551339253783226, "rewards/format_reward": 0.9899553582072258, "step": 140 }, { "clip_ratio": 0.0, "completion_length": 710.3538131713867, "epoch": 1.4359949302915083, "grad_norm": 0.10867197811603546, "learning_rate": 2.2221488349019902e-07, "loss": 0.036, "num_tokens": 106569734.0, "reward": 1.5904018580913544, "reward_std": 0.25699869357049465, "rewards/accuracy_reward": 0.5993303544819355, "rewards/format_reward": 0.9910714253783226, "step": 141 }, { "clip_ratio": 0.0, "completion_length": 672.2600936889648, "epoch": 1.4461343472750317, "grad_norm": 0.09679386019706726, "learning_rate": 2.1483868153251788e-07, "loss": 0.0437, "num_tokens": 107303623.0, "reward": 1.6272322237491608, "reward_std": 0.22507693991065025, "rewards/accuracy_reward": 0.6484374925494194, "rewards/format_reward": 0.9787946343421936, "step": 142 }, { "clip_ratio": 0.0, "completion_length": 721.3147583007812, "epoch": 1.456273764258555, "grad_norm": 0.09544187784194946, "learning_rate": 2.0755333554967346e-07, "loss": 0.0345, "num_tokens": 108087121.0, "reward": 1.5524554252624512, "reward_std": 0.2506922725588083, "rewards/accuracy_reward": 0.5658482275903225, "rewards/format_reward": 0.9866071417927742, "step": 143 }, { "clip_ratio": 0.0, "completion_length": 728.9341812133789, "epoch": 1.4664131812420786, "grad_norm": 0.10036015510559082, "learning_rate": 2.0036116674432652e-07, "loss": 0.0161, "num_tokens": 108876694.0, "reward": 1.5703125447034836, "reward_std": 0.23400406539440155, "rewards/accuracy_reward": 0.5803571455180645, "rewards/format_reward": 0.9899553507566452, "step": 144 }, { "clip_ratio": 0.0, "completion_length": 633.6451225280762, "epoch": 1.476552598225602, "grad_norm": 0.10225114226341248, "learning_rate": 1.9326446663172035e-07, "loss": 0.0215, "num_tokens": 109575920.0, "reward": 1.6830357909202576, "reward_std": 0.23387996666133404, "rewards/accuracy_reward": 0.6863839328289032, "rewards/format_reward": 0.9966517761349678, "step": 145 }, { "clip_ratio": 0.0, "completion_length": 653.1841735839844, "epoch": 1.4866920152091254, "grad_norm": 1.6936026811599731, "learning_rate": 1.8626549630957395e-07, "loss": 0.028, "num_tokens": 110289317.0, "reward": 1.6517858058214188, "reward_std": 0.24972818791866302, "rewards/accuracy_reward": 0.6607142947614193, "rewards/format_reward": 0.9910714253783226, "step": 146 }, { "clip_ratio": 0.0, "completion_length": 688.7846298217773, "epoch": 1.496831432192649, "grad_norm": 0.10226480662822723, "learning_rate": 1.7936648573766954e-07, "loss": 0.0531, "num_tokens": 111043180.0, "reward": 1.5881697088479996, "reward_std": 0.2411833181977272, "rewards/accuracy_reward": 0.5993303693830967, "rewards/format_reward": 0.9888392835855484, "step": 147 }, { "clip_ratio": 0.0, "completion_length": 730.7690048217773, "epoch": 1.5069708491761724, "grad_norm": 0.09249524772167206, "learning_rate": 1.725696330273575e-07, "loss": 0.0272, "num_tokens": 111828293.0, "reward": 1.5714286267757416, "reward_std": 0.2146799899637699, "rewards/accuracy_reward": 0.5803571343421936, "rewards/format_reward": 0.9910714328289032, "step": 148 }, { "clip_ratio": 0.0, "completion_length": 666.1719055175781, "epoch": 1.5171102661596958, "grad_norm": 0.09008808434009552, "learning_rate": 1.65877103741212e-07, "loss": 0.018, "num_tokens": 112549911.0, "reward": 1.7042411416769028, "reward_std": 0.18795600719749928, "rewards/accuracy_reward": 0.706473208963871, "rewards/format_reward": 0.9977678507566452, "step": 149 }, { "clip_ratio": 0.0, "completion_length": 704.3828430175781, "epoch": 1.5272496831432192, "grad_norm": 0.09881480038166046, "learning_rate": 1.592910302030544e-07, "loss": 0.0242, "num_tokens": 113300342.0, "reward": 1.5926340222358704, "reward_std": 0.2511101048439741, "rewards/accuracy_reward": 0.6015624888241291, "rewards/format_reward": 0.9910714253783226, "step": 150 }, { "clip_ratio": 0.0, "completion_length": 697.3582992553711, "epoch": 1.5373891001267426, "grad_norm": 0.11632478982210159, "learning_rate": 1.5281351081856976e-07, "loss": 0.0248, "num_tokens": 114056431.0, "reward": 1.6015625596046448, "reward_std": 0.20664218626916409, "rewards/accuracy_reward": 0.6127232201397419, "rewards/format_reward": 0.9888392835855484, "step": 151 }, { "clip_ratio": 0.0, "completion_length": 680.4286041259766, "epoch": 1.5475285171102662, "grad_norm": 0.09516316652297974, "learning_rate": 1.4644660940672627e-07, "loss": 0.0338, "num_tokens": 114791687.0, "reward": 1.648437574505806, "reward_std": 0.19783248752355576, "rewards/accuracy_reward": 0.6584821343421936, "rewards/format_reward": 0.9899553507566452, "step": 152 }, { "clip_ratio": 0.0, "completion_length": 669.5949096679688, "epoch": 1.5576679340937896, "grad_norm": 0.09097380936145782, "learning_rate": 1.4019235454221856e-07, "loss": 0.0149, "num_tokens": 115525716.0, "reward": 1.6104911714792252, "reward_std": 0.20410908479243517, "rewards/accuracy_reward": 0.6171875, "rewards/format_reward": 0.9933035746216774, "step": 153 }, { "clip_ratio": 0.0, "completion_length": 677.7366333007812, "epoch": 1.5678073510773132, "grad_norm": 0.09991227835416794, "learning_rate": 1.3405273890913737e-07, "loss": 0.0334, "num_tokens": 116254984.0, "reward": 1.5892857760190964, "reward_std": 0.2337410505861044, "rewards/accuracy_reward": 0.6004464365541935, "rewards/format_reward": 0.9888392761349678, "step": 154 }, { "clip_ratio": 0.0, "completion_length": 691.3571853637695, "epoch": 1.5779467680608366, "grad_norm": 0.18712472915649414, "learning_rate": 1.280297186660752e-07, "loss": 0.0335, "num_tokens": 116992816.0, "reward": 1.6595982760190964, "reward_std": 0.2452637990936637, "rewards/accuracy_reward": 0.6707589328289032, "rewards/format_reward": 0.9888392835855484, "step": 155 }, { "clip_ratio": 0.0, "completion_length": 758.325927734375, "epoch": 1.58808618504436, "grad_norm": 0.09767128527164459, "learning_rate": 1.2212521282287093e-07, "loss": 0.0328, "num_tokens": 117792204.0, "reward": 1.5558036416769028, "reward_std": 0.23506544344127178, "rewards/accuracy_reward": 0.5647321492433548, "rewards/format_reward": 0.9910714104771614, "step": 156 }, { "clip_ratio": 0.0, "completion_length": 704.4464721679688, "epoch": 1.5982256020278833, "grad_norm": 0.08936415612697601, "learning_rate": 1.1634110262918717e-07, "loss": 0.0365, "num_tokens": 118567396.0, "reward": 1.5758929252624512, "reward_std": 0.21123607829213142, "rewards/accuracy_reward": 0.5970982126891613, "rewards/format_reward": 0.9787946417927742, "step": 157 }, { "clip_ratio": 0.0, "completion_length": 721.3761444091797, "epoch": 1.6083650190114067, "grad_norm": 0.9475556015968323, "learning_rate": 1.1067923097512255e-07, "loss": 0.051, "num_tokens": 119344461.0, "reward": 1.5680804252624512, "reward_std": 0.24346480891108513, "rewards/accuracy_reward": 0.5848214253783226, "rewards/format_reward": 0.9832589253783226, "step": 158 }, { "clip_ratio": 0.0, "completion_length": 720.2455749511719, "epoch": 1.6185044359949303, "grad_norm": 0.10232140868902206, "learning_rate": 1.0514140180404202e-07, "loss": 0.0421, "num_tokens": 120118857.0, "reward": 1.5736607909202576, "reward_std": 0.2383702825754881, "rewards/accuracy_reward": 0.5859375, "rewards/format_reward": 0.987723208963871, "step": 159 }, { "clip_ratio": 0.0, "completion_length": 724.2288284301758, "epoch": 1.6286438529784537, "grad_norm": 0.17790253460407257, "learning_rate": 9.972937953781984e-08, "loss": 0.0345, "num_tokens": 120914734.0, "reward": 1.5167411416769028, "reward_std": 0.22692781873047352, "rewards/accuracy_reward": 0.527901791036129, "rewards/format_reward": 0.9888392835855484, "step": 160 }, { "clip_ratio": 0.0, "completion_length": 671.4531478881836, "epoch": 1.6387832699619773, "grad_norm": 0.08403746038675308, "learning_rate": 9.444488851467041e-08, "loss": 0.0196, "num_tokens": 121634028.0, "reward": 1.6439733058214188, "reward_std": 0.18759915605187416, "rewards/accuracy_reward": 0.6517857015132904, "rewards/format_reward": 0.9921874925494194, "step": 161 }, { "clip_ratio": 0.0, "completion_length": 686.6283798217773, "epoch": 1.6489226869455007, "grad_norm": 0.11062318086624146, "learning_rate": 8.928961243975436e-08, "loss": 0.0421, "num_tokens": 122382103.0, "reward": 1.5970982909202576, "reward_std": 0.22882544435560703, "rewards/accuracy_reward": 0.6119505539536476, "rewards/format_reward": 0.991071417927742, "step": 162 }, { "clip_ratio": 0.0, "completion_length": 666.947566986084, "epoch": 1.659062103929024, "grad_norm": 0.09584940969944, "learning_rate": 8.426519384872732e-08, "loss": 0.0095, "num_tokens": 123101824.0, "reward": 1.6573661416769028, "reward_std": 0.2075268942862749, "rewards/accuracy_reward": 0.6607142835855484, "rewards/format_reward": 0.9966517835855484, "step": 163 }, { "clip_ratio": 0.0, "completion_length": 697.6964645385742, "epoch": 1.6692015209125475, "grad_norm": 0.10291486978530884, "learning_rate": 7.937323358440934e-08, "loss": 0.0226, "num_tokens": 123857992.0, "reward": 1.6450893431901932, "reward_std": 0.25325831212103367, "rewards/accuracy_reward": 0.6506696417927742, "rewards/format_reward": 0.9944196343421936, "step": 164 }, { "clip_ratio": 0.0, "completion_length": 691.475471496582, "epoch": 1.6793409378960709, "grad_norm": 0.09860701858997345, "learning_rate": 7.461529028673463e-08, "loss": 0.0197, "num_tokens": 124610362.0, "reward": 1.6473215073347092, "reward_std": 0.2233567014336586, "rewards/accuracy_reward": 0.652901791036129, "rewards/format_reward": 0.9944196343421936, "step": 165 }, { "clip_ratio": 0.0, "completion_length": 712.0446701049805, "epoch": 1.6894803548795945, "grad_norm": 0.10766546428203583, "learning_rate": 6.999287989614971e-08, "loss": 0.0403, "num_tokens": 125380794.0, "reward": 1.5881696939468384, "reward_std": 0.2603952419012785, "rewards/accuracy_reward": 0.5970982126891613, "rewards/format_reward": 0.9910714253783226, "step": 166 }, { "clip_ratio": 0.0, "completion_length": 671.4163131713867, "epoch": 1.6996197718631179, "grad_norm": 0.15764014422893524, "learning_rate": 6.550747517061656e-08, "loss": 0.0282, "num_tokens": 126108927.0, "reward": 1.6595982909202576, "reward_std": 0.20449624955654144, "rewards/accuracy_reward": 0.6662946343421936, "rewards/format_reward": 0.9933035671710968, "step": 167 }, { "clip_ratio": 0.0, "completion_length": 703.9542617797852, "epoch": 1.7097591888466415, "grad_norm": 0.09230643510818481, "learning_rate": 6.116050521637218e-08, "loss": 0.0282, "num_tokens": 126867998.0, "reward": 1.6395090073347092, "reward_std": 0.2097290549427271, "rewards/accuracy_reward": 0.6450892835855484, "rewards/format_reward": 0.9944196343421936, "step": 168 }, { "clip_ratio": 0.0, "completion_length": 672.6339569091797, "epoch": 1.7198986058301649, "grad_norm": 0.11055152863264084, "learning_rate": 5.6953355032598795e-08, "loss": 0.0277, "num_tokens": 127607750.0, "reward": 1.5736607760190964, "reward_std": 0.2487525064498186, "rewards/accuracy_reward": 0.5881696380674839, "rewards/format_reward": 0.9854910746216774, "step": 169 }, { "clip_ratio": 0.0, "completion_length": 696.6395416259766, "epoch": 1.7300380228136882, "grad_norm": 0.09632628411054611, "learning_rate": 5.288736507014435e-08, "loss": 0.0203, "num_tokens": 128355939.0, "reward": 1.640625074505806, "reward_std": 0.22130538150668144, "rewards/accuracy_reward": 0.6517857238650322, "rewards/format_reward": 0.988839291036129, "step": 170 }, { "clip_ratio": 0.0, "completion_length": 665.4888763427734, "epoch": 1.7401774397972116, "grad_norm": 0.185902938246727, "learning_rate": 4.896383080443933e-08, "loss": 0.014, "num_tokens": 129076969.0, "reward": 1.59933041036129, "reward_std": 0.22374757565557957, "rewards/accuracy_reward": 0.6037946306169033, "rewards/format_reward": 0.995535708963871, "step": 171 }, { "clip_ratio": 0.0, "completion_length": 655.0580596923828, "epoch": 1.750316856780735, "grad_norm": 0.0977133959531784, "learning_rate": 4.518400232274078e-08, "loss": 0.0151, "num_tokens": 129794405.0, "reward": 1.5870536416769028, "reward_std": 0.21878547966480255, "rewards/accuracy_reward": 0.599759615957737, "rewards/format_reward": 0.9933035597205162, "step": 172 }, { "clip_ratio": 0.0, "completion_length": 660.8594131469727, "epoch": 1.7604562737642584, "grad_norm": 0.10144926607608795, "learning_rate": 4.1549083925840165e-08, "loss": 0.0387, "num_tokens": 130503983.0, "reward": 1.6529018580913544, "reward_std": 0.22397751361131668, "rewards/accuracy_reward": 0.6662946417927742, "rewards/format_reward": 0.9866071417927742, "step": 173 }, { "clip_ratio": 0.0, "completion_length": 672.0837326049805, "epoch": 1.770595690747782, "grad_norm": 0.09892982989549637, "learning_rate": 3.806023374435663e-08, "loss": 0.0305, "num_tokens": 131234698.0, "reward": 1.6082590222358704, "reward_std": 0.21078570373356342, "rewards/accuracy_reward": 0.6149553582072258, "rewards/format_reward": 0.9933035597205162, "step": 174 }, { "clip_ratio": 0.0, "completion_length": 700.0781631469727, "epoch": 1.7807351077313056, "grad_norm": 0.08607691526412964, "learning_rate": 3.4718563369743213e-08, "loss": 0.0126, "num_tokens": 131995296.0, "reward": 1.5825893580913544, "reward_std": 0.19891513139009476, "rewards/accuracy_reward": 0.5870535746216774, "rewards/format_reward": 0.9955357164144516, "step": 175 }, { "clip_ratio": 0.0, "completion_length": 683.560302734375, "epoch": 1.790874524714829, "grad_norm": 0.0931280255317688, "learning_rate": 3.15251375001192e-08, "loss": 0.0266, "num_tokens": 132747870.0, "reward": 1.6540179401636124, "reward_std": 0.2136262021958828, "rewards/accuracy_reward": 0.6662946417927742, "rewards/format_reward": 0.987723208963871, "step": 176 }, { "clip_ratio": 0.0, "completion_length": 673.4799385070801, "epoch": 1.8010139416983524, "grad_norm": 0.09533898532390594, "learning_rate": 2.8480973601043955e-08, "loss": 0.0372, "num_tokens": 133486116.0, "reward": 1.6417411416769028, "reward_std": 0.18038041703402996, "rewards/accuracy_reward": 0.6529017798602581, "rewards/format_reward": 0.9888392761349678, "step": 177 }, { "clip_ratio": 0.0, "completion_length": 709.9821853637695, "epoch": 1.8111533586818758, "grad_norm": 0.1904437392950058, "learning_rate": 2.558704158134023e-08, "loss": 0.0261, "num_tokens": 134262284.0, "reward": 1.5825893431901932, "reward_std": 0.22126466780900955, "rewards/accuracy_reward": 0.5937500111758709, "rewards/format_reward": 0.9888392761349678, "step": 178 }, { "clip_ratio": 0.0, "completion_length": 719.8861846923828, "epoch": 1.8212927756653992, "grad_norm": 0.0858464315533638, "learning_rate": 2.2844263484068093e-08, "loss": 0.0161, "num_tokens": 135034302.0, "reward": 1.5937500447034836, "reward_std": 0.18957781419157982, "rewards/accuracy_reward": 0.6004464216530323, "rewards/format_reward": 0.9933035671710968, "step": 179 }, { "clip_ratio": 0.0, "completion_length": 670.2042694091797, "epoch": 1.8314321926489225, "grad_norm": 0.09248825907707214, "learning_rate": 2.025351319275137e-08, "loss": 0.0237, "num_tokens": 135765773.0, "reward": 1.6517857909202576, "reward_std": 0.20931300334632397, "rewards/accuracy_reward": 0.6607142835855484, "rewards/format_reward": 0.991071417927742, "step": 180 }, { "clip_ratio": 0.0, "completion_length": 697.1328506469727, "epoch": 1.8415716096324461, "grad_norm": 0.09593623131513596, "learning_rate": 1.781561615294652e-08, "loss": 0.0136, "num_tokens": 136524868.0, "reward": 1.5881697237491608, "reward_std": 0.23646764643490314, "rewards/accuracy_reward": 0.597098208963871, "rewards/format_reward": 0.9910714253783226, "step": 181 }, { "clip_ratio": 0.0, "completion_length": 709.1886520385742, "epoch": 1.8517110266159695, "grad_norm": 0.10065972059965134, "learning_rate": 1.553134910924636e-08, "loss": 0.0373, "num_tokens": 137296405.0, "reward": 1.539062574505806, "reward_std": 0.22722396813333035, "rewards/accuracy_reward": 0.5479910708963871, "rewards/format_reward": 0.991071417927742, "step": 182 }, { "clip_ratio": 0.0, "completion_length": 661.6339569091797, "epoch": 1.8618504435994931, "grad_norm": 0.09599631279706955, "learning_rate": 1.340143985779829e-08, "loss": 0.0283, "num_tokens": 138015989.0, "reward": 1.6540179550647736, "reward_std": 0.18772431463003159, "rewards/accuracy_reward": 0.661830373108387, "rewards/format_reward": 0.9921874925494194, "step": 183 }, { "clip_ratio": 0.0, "completion_length": 697.3761444091797, "epoch": 1.8719898605830165, "grad_norm": 0.09907021373510361, "learning_rate": 1.1426567014420297e-08, "loss": 0.047, "num_tokens": 138767542.0, "reward": 1.6383929550647736, "reward_std": 0.23151767998933792, "rewards/accuracy_reward": 0.6551339402794838, "rewards/format_reward": 0.9832589104771614, "step": 184 }, { "clip_ratio": 0.0, "completion_length": 706.0480117797852, "epoch": 1.88212927756654, "grad_norm": 0.09566432982683182, "learning_rate": 9.607359798384784e-09, "loss": 0.0372, "num_tokens": 139515641.0, "reward": 1.592633992433548, "reward_std": 0.2507201302796602, "rewards/accuracy_reward": 0.6015625074505806, "rewards/format_reward": 0.991071417927742, "step": 185 }, { "clip_ratio": 0.0, "completion_length": 675.1484756469727, "epoch": 1.8922686945500633, "grad_norm": 0.09894827008247375, "learning_rate": 7.944397831941951e-09, "loss": 0.031, "num_tokens": 140267070.0, "reward": 1.5837054252624512, "reward_std": 0.21183509565889835, "rewards/accuracy_reward": 0.5937499962747097, "rewards/format_reward": 0.9899553507566452, "step": 186 }, { "clip_ratio": 0.0, "completion_length": 740.9620971679688, "epoch": 1.9024081115335867, "grad_norm": 0.1215052530169487, "learning_rate": 6.438210955644452e-09, "loss": 0.0443, "num_tokens": 141073396.0, "reward": 1.5044643580913544, "reward_std": 0.2504011485725641, "rewards/accuracy_reward": 0.5256696455180645, "rewards/format_reward": 0.9787946343421936, "step": 187 }, { "clip_ratio": 0.0, "completion_length": 691.3471298217773, "epoch": 1.9125475285171103, "grad_norm": 0.08761589974164963, "learning_rate": 5.0892790595336575e-09, "loss": 0.0307, "num_tokens": 141829859.0, "reward": 1.6305804550647736, "reward_std": 0.220117699354887, "rewards/accuracy_reward": 0.6406250037252903, "rewards/format_reward": 0.9899553433060646, "step": 188 }, { "clip_ratio": 0.0, "completion_length": 687.7377548217773, "epoch": 1.9226869455006337, "grad_norm": 0.10144967585802078, "learning_rate": 3.898031930240797e-09, "loss": 0.03, "num_tokens": 142571072.0, "reward": 1.6037946790456772, "reward_std": 0.22627083584666252, "rewards/accuracy_reward": 0.6138392947614193, "rewards/format_reward": 0.9899553582072258, "step": 189 }, { "clip_ratio": 0.0, "completion_length": 743.5949020385742, "epoch": 1.9328263624841573, "grad_norm": 0.08726272732019424, "learning_rate": 2.8648491140513264e-09, "loss": 0.0307, "num_tokens": 143363781.0, "reward": 1.5948661267757416, "reward_std": 0.2067482192069292, "rewards/accuracy_reward": 0.6004464328289032, "rewards/format_reward": 0.9944196417927742, "step": 190 }, { "clip_ratio": 0.0, "completion_length": 631.5368728637695, "epoch": 1.9429657794676807, "grad_norm": 0.09442251920700073, "learning_rate": 1.9900597959770505e-09, "loss": 0.0309, "num_tokens": 144054766.0, "reward": 1.7064732909202576, "reward_std": 0.1611343901604414, "rewards/accuracy_reward": 0.7142857164144516, "rewards/format_reward": 0.9921874925494194, "step": 191 }, { "clip_ratio": 0.0, "completion_length": 702.8839721679688, "epoch": 1.953105196451204, "grad_norm": 0.09530234336853027, "learning_rate": 1.2739426948732424e-09, "loss": 0.0236, "num_tokens": 144818518.0, "reward": 1.6216518580913544, "reward_std": 0.22148274257779121, "rewards/accuracy_reward": 0.6283482164144516, "rewards/format_reward": 0.9933035597205162, "step": 192 }, { "clip_ratio": 0.0, "completion_length": 705.2678833007812, "epoch": 1.9632446134347274, "grad_norm": 0.11064296960830688, "learning_rate": 7.16725974635568e-10, "loss": 0.0467, "num_tokens": 145576278.0, "reward": 1.6149554401636124, "reward_std": 0.2686098553240299, "rewards/accuracy_reward": 0.6305803507566452, "rewards/format_reward": 0.9843749850988388, "step": 193 }, { "clip_ratio": 0.0, "completion_length": 709.090446472168, "epoch": 1.9733840304182508, "grad_norm": 0.1301695555448532, "learning_rate": 3.185871715041255e-10, "loss": 0.0256, "num_tokens": 146344471.0, "reward": 1.5904018580913544, "reward_std": 0.20222523249685764, "rewards/accuracy_reward": 0.5959821455180645, "rewards/format_reward": 0.9944196343421936, "step": 194 }, { "clip_ratio": 0.0, "completion_length": 649.944221496582, "epoch": 1.9835234474017744, "grad_norm": 0.10688284784555435, "learning_rate": 7.96531374983589e-11, "loss": 0.0264, "num_tokens": 147047301.0, "reward": 1.6361607760190964, "reward_std": 0.23337006010115147, "rewards/accuracy_reward": 0.6417410708963871, "rewards/format_reward": 0.9944196343421936, "step": 195 }, { "clip_ratio": 0.0, "completion_length": 693.314453125, "epoch": 1.9936628643852978, "grad_norm": 0.09134732186794281, "learning_rate": 0.0, "loss": 0.0173, "num_tokens": 147801661.0, "reward": 1.56026791036129, "reward_std": 0.21255235373973846, "rewards/accuracy_reward": 0.565848208963871, "rewards/format_reward": 0.9944196343421936, "step": 196 }, { "epoch": 1.9936628643852978, "step": 196, "total_flos": 0.0, "train_loss": 0.02676454978019156, "train_runtime": 31339.2496, "train_samples_per_second": 0.705, "train_steps_per_second": 0.006 } ], "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 }