| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.998691442030882, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010468463752944255, | |
| "grad_norm": 10.21879193399377, | |
| "learning_rate": 6.25e-08, | |
| "logits/chosen": -0.49816179275512695, | |
| "logits/rejected": -0.5135830640792847, | |
| "logps/chosen": -1.1745846271514893, | |
| "logps/rejected": -1.3595685958862305, | |
| "loss": -0.035, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.1745846271514893, | |
| "rewards/margins": 0.18498393893241882, | |
| "rewards/rejected": -1.3595685958862305, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 2.7101486132785926, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": -0.5216172933578491, | |
| "logits/rejected": -0.4984247088432312, | |
| "logps/chosen": -1.1580203771591187, | |
| "logps/rejected": -1.262268304824829, | |
| "loss": -0.0307, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -1.1580203771591187, | |
| "rewards/margins": 0.1042477935552597, | |
| "rewards/rejected": -1.262268304824829, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 5.679326865922718, | |
| "learning_rate": 1.875e-07, | |
| "logits/chosen": -0.46298742294311523, | |
| "logits/rejected": -0.45184773206710815, | |
| "logps/chosen": -1.1014564037322998, | |
| "logps/rejected": -1.357534408569336, | |
| "loss": -0.0466, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.1014564037322998, | |
| "rewards/margins": 0.2560780644416809, | |
| "rewards/rejected": -1.357534408569336, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 3.747855502948124, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": -0.4439946711063385, | |
| "logits/rejected": -0.45637279748916626, | |
| "logps/chosen": -1.143970251083374, | |
| "logps/rejected": -1.249586582183838, | |
| "loss": -0.0343, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.143970251083374, | |
| "rewards/margins": 0.10561631619930267, | |
| "rewards/rejected": -1.249586582183838, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 4.753383333287764, | |
| "learning_rate": 3.125e-07, | |
| "logits/chosen": -0.5120896697044373, | |
| "logits/rejected": -0.4878992438316345, | |
| "logps/chosen": -1.1282974481582642, | |
| "logps/rejected": -1.20094895362854, | |
| "loss": -0.0405, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.1282974481582642, | |
| "rewards/margins": 0.07265140116214752, | |
| "rewards/rejected": -1.20094895362854, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 5.921390795122043, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": -0.5101591348648071, | |
| "logits/rejected": -0.5148754715919495, | |
| "logps/chosen": -1.0519568920135498, | |
| "logps/rejected": -1.1866623163223267, | |
| "loss": -0.0316, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.0519568920135498, | |
| "rewards/margins": 0.13470537960529327, | |
| "rewards/rejected": -1.1866623163223267, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 2.925534488741588, | |
| "learning_rate": 4.3749999999999994e-07, | |
| "logits/chosen": -0.523992657661438, | |
| "logits/rejected": -0.4819146990776062, | |
| "logps/chosen": -1.0300133228302002, | |
| "logps/rejected": -1.1558997631072998, | |
| "loss": -0.0438, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.0300133228302002, | |
| "rewards/margins": 0.12588649988174438, | |
| "rewards/rejected": -1.1558997631072998, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 2.41534372661926, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": -0.5426082611083984, | |
| "logits/rejected": -0.5616889595985413, | |
| "logps/chosen": -0.9607571363449097, | |
| "logps/rejected": -1.2749502658843994, | |
| "loss": -0.0597, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.9607571363449097, | |
| "rewards/margins": 0.31419312953948975, | |
| "rewards/rejected": -1.2749502658843994, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 3.5651412454593747, | |
| "learning_rate": 5.625e-07, | |
| "logits/chosen": -0.5252939462661743, | |
| "logits/rejected": -0.5452172160148621, | |
| "logps/chosen": -0.9730573892593384, | |
| "logps/rejected": -1.1473325490951538, | |
| "loss": -0.0559, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.9730573892593384, | |
| "rewards/margins": 0.174275204539299, | |
| "rewards/rejected": -1.1473325490951538, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 4.149376436833766, | |
| "learning_rate": 5.999678242522831e-07, | |
| "logits/chosen": -0.5748304128646851, | |
| "logits/rejected": -0.6049160957336426, | |
| "logps/chosen": -0.9965342283248901, | |
| "logps/rejected": -1.2540613412857056, | |
| "loss": -0.0464, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.9965342283248901, | |
| "rewards/margins": 0.25752708315849304, | |
| "rewards/rejected": -1.2540613412857056, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 2.9691689665037373, | |
| "learning_rate": 5.996059263493219e-07, | |
| "logits/chosen": -0.5686520338058472, | |
| "logits/rejected": -0.5740174055099487, | |
| "logps/chosen": -0.9575145840644836, | |
| "logps/rejected": -1.1497360467910767, | |
| "loss": -0.051, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.9575145840644836, | |
| "rewards/margins": 0.1922214925289154, | |
| "rewards/rejected": -1.1497360467910767, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 2.6253711163069537, | |
| "learning_rate": 5.988423976115163e-07, | |
| "logits/chosen": -0.6203932762145996, | |
| "logits/rejected": -0.633256196975708, | |
| "logps/chosen": -1.0102077722549438, | |
| "logps/rejected": -1.1339367628097534, | |
| "loss": -0.062, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0102077722549438, | |
| "rewards/margins": 0.12372910976409912, | |
| "rewards/rejected": -1.1339367628097534, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 6.143760308041501, | |
| "learning_rate": 5.976782615723061e-07, | |
| "logits/chosen": -0.5896092653274536, | |
| "logits/rejected": -0.5990695357322693, | |
| "logps/chosen": -0.9534575343132019, | |
| "logps/rejected": -1.3051955699920654, | |
| "loss": -0.0542, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.9534575343132019, | |
| "rewards/margins": 0.3517380654811859, | |
| "rewards/rejected": -1.3051955699920654, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 2.425535094463044, | |
| "learning_rate": 5.961150787913738e-07, | |
| "logits/chosen": -0.5713956952095032, | |
| "logits/rejected": -0.5825516581535339, | |
| "logps/chosen": -0.9968436360359192, | |
| "logps/rejected": -1.1801879405975342, | |
| "loss": -0.0566, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.9968436360359192, | |
| "rewards/margins": 0.18334420025348663, | |
| "rewards/rejected": -1.1801879405975342, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 3.936791088850445, | |
| "learning_rate": 5.941549447626671e-07, | |
| "logits/chosen": -0.5951550006866455, | |
| "logits/rejected": -0.601603090763092, | |
| "logps/chosen": -0.9617071151733398, | |
| "logps/rejected": -1.2262237071990967, | |
| "loss": -0.0623, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9617071151733398, | |
| "rewards/margins": 0.2645165026187897, | |
| "rewards/rejected": -1.2262237071990967, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 3.1037840831495154, | |
| "learning_rate": 5.918004871053251e-07, | |
| "logits/chosen": -0.6256829500198364, | |
| "logits/rejected": -0.6605373620986938, | |
| "logps/chosen": -0.9740772247314453, | |
| "logps/rejected": -1.2311257123947144, | |
| "loss": -0.0545, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.9740772247314453, | |
| "rewards/margins": 0.25704842805862427, | |
| "rewards/rejected": -1.2311257123947144, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 3.517502431455176, | |
| "learning_rate": 5.890548620412763e-07, | |
| "logits/chosen": -0.6996882557868958, | |
| "logits/rejected": -0.6915110349655151, | |
| "logps/chosen": -0.9287127256393433, | |
| "logps/rejected": -1.1633974313735962, | |
| "loss": -0.0628, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.9287127256393433, | |
| "rewards/margins": 0.23468473553657532, | |
| "rewards/rejected": -1.1633974313735962, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 3.017999402299648, | |
| "learning_rate": 5.859217501642258e-07, | |
| "logits/chosen": -0.6648964285850525, | |
| "logits/rejected": -0.6696600914001465, | |
| "logps/chosen": -1.020616888999939, | |
| "logps/rejected": -1.240464210510254, | |
| "loss": -0.0681, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.020616888999939, | |
| "rewards/margins": 0.2198474109172821, | |
| "rewards/rejected": -1.240464210510254, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 4.972742617412184, | |
| "learning_rate": 5.824053515057091e-07, | |
| "logits/chosen": -0.6936327815055847, | |
| "logits/rejected": -0.6650416254997253, | |
| "logps/chosen": -1.010335922241211, | |
| "logps/rejected": -1.212616205215454, | |
| "loss": -0.0554, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -1.010335922241211, | |
| "rewards/margins": 0.20228025317192078, | |
| "rewards/rejected": -1.212616205215454, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 4.305605640056122, | |
| "learning_rate": 5.785103799048218e-07, | |
| "logits/chosen": -0.6884914636611938, | |
| "logits/rejected": -0.7089473009109497, | |
| "logps/chosen": -1.0421329736709595, | |
| "logps/rejected": -1.3643444776535034, | |
| "loss": -0.0577, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.0421329736709595, | |
| "rewards/margins": 0.32221153378486633, | |
| "rewards/rejected": -1.3643444776535034, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 4.401763969089556, | |
| "learning_rate": 5.742420566891749e-07, | |
| "logits/chosen": -0.6878229379653931, | |
| "logits/rejected": -0.6928119659423828, | |
| "logps/chosen": -0.9930433034896851, | |
| "logps/rejected": -1.3710923194885254, | |
| "loss": -0.0667, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9930433034896851, | |
| "rewards/margins": 0.3780490458011627, | |
| "rewards/rejected": -1.3710923194885254, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 6.081502646176958, | |
| "learning_rate": 5.696061036755478e-07, | |
| "logits/chosen": -0.7217090725898743, | |
| "logits/rejected": -0.7287099361419678, | |
| "logps/chosen": -1.03863525390625, | |
| "logps/rejected": -1.384284257888794, | |
| "loss": -0.0744, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.03863525390625, | |
| "rewards/margins": 0.3456490635871887, | |
| "rewards/rejected": -1.384284257888794, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 4.443555154250914, | |
| "learning_rate": 5.64608735499618e-07, | |
| "logits/chosen": -0.6417717337608337, | |
| "logits/rejected": -0.6641663312911987, | |
| "logps/chosen": -1.0144702196121216, | |
| "logps/rejected": -1.4784486293792725, | |
| "loss": -0.0904, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0144702196121216, | |
| "rewards/margins": 0.46397843956947327, | |
| "rewards/rejected": -1.4784486293792725, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 4.21476483098945, | |
| "learning_rate": 5.592566512850545e-07, | |
| "logits/chosen": -0.6441653966903687, | |
| "logits/rejected": -0.6482559442520142, | |
| "logps/chosen": -1.126665711402893, | |
| "logps/rejected": -1.395385980606079, | |
| "loss": -0.0685, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.126665711402893, | |
| "rewards/margins": 0.2687203884124756, | |
| "rewards/rejected": -1.395385980606079, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 5.871972559220826, | |
| "learning_rate": 5.535570256631384e-07, | |
| "logits/chosen": -0.6881023645401001, | |
| "logits/rejected": -0.6507277488708496, | |
| "logps/chosen": -1.030970811843872, | |
| "logps/rejected": -1.3154019117355347, | |
| "loss": -0.0735, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.030970811843872, | |
| "rewards/margins": 0.28443121910095215, | |
| "rewards/rejected": -1.3154019117355347, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 3.8967984948431176, | |
| "learning_rate": 5.475174991549528e-07, | |
| "logits/chosen": -0.6212294697761536, | |
| "logits/rejected": -0.6462045907974243, | |
| "logps/chosen": -1.055466651916504, | |
| "logps/rejected": -1.2383654117584229, | |
| "loss": -0.0712, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.055466651916504, | |
| "rewards/margins": 0.18289880454540253, | |
| "rewards/rejected": -1.2383654117584229, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 4.954091803088075, | |
| "learning_rate": 5.411461679290317e-07, | |
| "logits/chosen": -0.6356550455093384, | |
| "logits/rejected": -0.6022456884384155, | |
| "logps/chosen": -1.1107438802719116, | |
| "logps/rejected": -1.643451452255249, | |
| "loss": -0.0826, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.1107438802719116, | |
| "rewards/margins": 0.532707691192627, | |
| "rewards/rejected": -1.643451452255249, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 6.396288923391441, | |
| "learning_rate": 5.34451572948201e-07, | |
| "logits/chosen": -0.5937852263450623, | |
| "logits/rejected": -0.5691717863082886, | |
| "logps/chosen": -1.0965867042541504, | |
| "logps/rejected": -1.5603969097137451, | |
| "loss": -0.0915, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0965867042541504, | |
| "rewards/margins": 0.4638102948665619, | |
| "rewards/rejected": -1.5603969097137451, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 4.065147326048241, | |
| "learning_rate": 5.274426885201582e-07, | |
| "logits/chosen": -0.5843586921691895, | |
| "logits/rejected": -0.5910850763320923, | |
| "logps/chosen": -1.1310163736343384, | |
| "logps/rejected": -1.4976770877838135, | |
| "loss": -0.0849, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.1310163736343384, | |
| "rewards/margins": 0.3666609227657318, | |
| "rewards/rejected": -1.4976770877838135, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 4.014586941737209, | |
| "learning_rate": 5.201289102671411e-07, | |
| "logits/chosen": -0.546478271484375, | |
| "logits/rejected": -0.5195995569229126, | |
| "logps/chosen": -1.0407493114471436, | |
| "logps/rejected": -1.436926245689392, | |
| "loss": -0.0898, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.0407493114471436, | |
| "rewards/margins": 0.3961769640445709, | |
| "rewards/rejected": -1.436926245689392, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 6.577906404260762, | |
| "learning_rate": 5.12520042530811e-07, | |
| "logits/chosen": -0.5657233595848083, | |
| "logits/rejected": -0.5251718163490295, | |
| "logps/chosen": -1.1016614437103271, | |
| "logps/rejected": -1.4624617099761963, | |
| "loss": -0.0923, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1016614437103271, | |
| "rewards/margins": 0.36080020666122437, | |
| "rewards/rejected": -1.4624617099761963, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 5.297524478469823, | |
| "learning_rate": 5.046262852292346e-07, | |
| "logits/chosen": -0.4731084704399109, | |
| "logits/rejected": -0.4319906234741211, | |
| "logps/chosen": -1.1443315744400024, | |
| "logps/rejected": -1.5339367389678955, | |
| "loss": -0.0887, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.1443315744400024, | |
| "rewards/margins": 0.38960522413253784, | |
| "rewards/rejected": -1.5339367389678955, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 6.8937105755776065, | |
| "learning_rate": 4.964582201835856e-07, | |
| "logits/chosen": -0.4270719885826111, | |
| "logits/rejected": -0.3794442415237427, | |
| "logps/chosen": -1.1176317930221558, | |
| "logps/rejected": -1.639172911643982, | |
| "loss": -0.0965, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.1176317930221558, | |
| "rewards/margins": 0.5215411186218262, | |
| "rewards/rejected": -1.639172911643982, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 4.460892002893491, | |
| "learning_rate": 4.880267969328908e-07, | |
| "logits/chosen": -0.32646840810775757, | |
| "logits/rejected": -0.22941944003105164, | |
| "logps/chosen": -1.2465507984161377, | |
| "logps/rejected": -1.7467823028564453, | |
| "loss": -0.0851, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2465507984161377, | |
| "rewards/margins": 0.5002316236495972, | |
| "rewards/rejected": -1.7467823028564453, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 10.536194204055317, | |
| "learning_rate": 4.793433180558423e-07, | |
| "logits/chosen": -0.2930789887905121, | |
| "logits/rejected": -0.15167564153671265, | |
| "logps/chosen": -1.1869597434997559, | |
| "logps/rejected": -1.7165695428848267, | |
| "loss": -0.088, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.1869597434997559, | |
| "rewards/margins": 0.5296097993850708, | |
| "rewards/rejected": -1.7165695428848267, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 14.792822735814111, | |
| "learning_rate": 4.704194240193467e-07, | |
| "logits/chosen": -0.14493033289909363, | |
| "logits/rejected": -0.08077652007341385, | |
| "logps/chosen": -1.1866719722747803, | |
| "logps/rejected": -1.724832534790039, | |
| "loss": -0.0932, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1866719722747803, | |
| "rewards/margins": 0.5381606221199036, | |
| "rewards/rejected": -1.724832534790039, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 5.012043699075584, | |
| "learning_rate": 4.6126707757412686e-07, | |
| "logits/chosen": -0.19216454029083252, | |
| "logits/rejected": -0.06508170068264008, | |
| "logps/chosen": -1.2008949518203735, | |
| "logps/rejected": -2.0086560249328613, | |
| "loss": -0.1086, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.2008949518203735, | |
| "rewards/margins": 0.807761013507843, | |
| "rewards/rejected": -2.0086560249328613, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 6.108194509753534, | |
| "learning_rate": 4.5189854771829086e-07, | |
| "logits/chosen": -0.14359740912914276, | |
| "logits/rejected": -0.04716472700238228, | |
| "logps/chosen": -1.2239508628845215, | |
| "logps/rejected": -1.738565444946289, | |
| "loss": -0.0845, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.2239508628845215, | |
| "rewards/margins": 0.5146147012710571, | |
| "rewards/rejected": -1.738565444946289, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 12.413660235408443, | |
| "learning_rate": 4.4232639325036807e-07, | |
| "logits/chosen": -0.16371646523475647, | |
| "logits/rejected": -0.08154813945293427, | |
| "logps/chosen": -1.3641008138656616, | |
| "logps/rejected": -1.8681023120880127, | |
| "loss": -0.0883, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.3641008138656616, | |
| "rewards/margins": 0.5040014386177063, | |
| "rewards/rejected": -1.8681023120880127, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 5.664592254947892, | |
| "learning_rate": 4.32563445933859e-07, | |
| "logits/chosen": -0.3630138039588928, | |
| "logits/rejected": -0.31867215037345886, | |
| "logps/chosen": -1.2496740818023682, | |
| "logps/rejected": -1.7695592641830444, | |
| "loss": -0.0991, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.2496740818023682, | |
| "rewards/margins": 0.5198851823806763, | |
| "rewards/rejected": -1.7695592641830444, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 8.615954171040675, | |
| "learning_rate": 4.226227932958664e-07, | |
| "logits/chosen": -0.2460884302854538, | |
| "logits/rejected": -0.1599723994731903, | |
| "logps/chosen": -1.0948972702026367, | |
| "logps/rejected": -1.7847169637680054, | |
| "loss": -0.1097, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0948972702026367, | |
| "rewards/margins": 0.6898195743560791, | |
| "rewards/rejected": -1.7847169637680054, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 15.728835997323346, | |
| "learning_rate": 4.1251776108286854e-07, | |
| "logits/chosen": -0.16430199146270752, | |
| "logits/rejected": -0.09520161151885986, | |
| "logps/chosen": -1.3307850360870361, | |
| "logps/rejected": -1.7031431198120117, | |
| "loss": -0.0888, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.3307850360870361, | |
| "rewards/margins": 0.37235820293426514, | |
| "rewards/rejected": -1.7031431198120117, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 9.404924688577674, | |
| "learning_rate": 4.022618953971514e-07, | |
| "logits/chosen": -0.11717267334461212, | |
| "logits/rejected": -0.0059516532346606255, | |
| "logps/chosen": -1.2097961902618408, | |
| "logps/rejected": -1.8957335948944092, | |
| "loss": -0.1077, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -1.2097961902618408, | |
| "rewards/margins": 0.6859374046325684, | |
| "rewards/rejected": -1.8957335948944092, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 7.3279364906626485, | |
| "learning_rate": 3.918689445378477e-07, | |
| "logits/chosen": -0.09480677545070648, | |
| "logits/rejected": 0.13137385249137878, | |
| "logps/chosen": -1.2646963596343994, | |
| "logps/rejected": -1.9554264545440674, | |
| "loss": -0.0804, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.2646963596343994, | |
| "rewards/margins": 0.690730094909668, | |
| "rewards/rejected": -1.9554264545440674, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 5.249390449834869, | |
| "learning_rate": 3.813528405709251e-07, | |
| "logits/chosen": 0.017690464854240417, | |
| "logits/rejected": 0.24274654686450958, | |
| "logps/chosen": -1.2406516075134277, | |
| "logps/rejected": -2.0346226692199707, | |
| "loss": -0.0963, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.2406516075134277, | |
| "rewards/margins": 0.7939712405204773, | |
| "rewards/rejected": -2.0346226692199707, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 5.840384716420158, | |
| "learning_rate": 3.707276806528282e-07, | |
| "logits/chosen": -0.034771956503391266, | |
| "logits/rejected": 0.25075453519821167, | |
| "logps/chosen": -1.307732105255127, | |
| "logps/rejected": -2.3069005012512207, | |
| "loss": -0.1106, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.307732105255127, | |
| "rewards/margins": 0.9991682171821594, | |
| "rewards/rejected": -2.3069005012512207, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 5.708169285500334, | |
| "learning_rate": 3.6000770813281334e-07, | |
| "logits/chosen": -0.04531233012676239, | |
| "logits/rejected": 0.1216268315911293, | |
| "logps/chosen": -1.0852802991867065, | |
| "logps/rejected": -1.7644765377044678, | |
| "loss": -0.1067, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.0852802991867065, | |
| "rewards/margins": 0.6791960597038269, | |
| "rewards/rejected": -1.7644765377044678, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 5.2328738831340615, | |
| "learning_rate": 3.4920729345930654e-07, | |
| "logits/chosen": -0.15298782289028168, | |
| "logits/rejected": -0.021981507539749146, | |
| "logps/chosen": -1.181814432144165, | |
| "logps/rejected": -1.84041428565979, | |
| "loss": -0.0907, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.181814432144165, | |
| "rewards/margins": 0.6585996747016907, | |
| "rewards/rejected": -1.84041428565979, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 4.454620572722302, | |
| "learning_rate": 3.383409149158814e-07, | |
| "logits/chosen": -0.14406219124794006, | |
| "logits/rejected": 0.021218538284301758, | |
| "logps/chosen": -1.3196625709533691, | |
| "logps/rejected": -1.8837369680404663, | |
| "loss": -0.1005, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.3196625709533691, | |
| "rewards/margins": 0.5640743970870972, | |
| "rewards/rejected": -1.8837369680404663, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 16.864698084247127, | |
| "learning_rate": 3.2742313921268035e-07, | |
| "logits/chosen": -0.04993357136845589, | |
| "logits/rejected": 0.09763963520526886, | |
| "logps/chosen": -1.1896789073944092, | |
| "logps/rejected": -1.9447057247161865, | |
| "loss": -0.1026, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.1896789073944092, | |
| "rewards/margins": 0.7550268769264221, | |
| "rewards/rejected": -1.9447057247161865, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 4.006389891258429, | |
| "learning_rate": 3.1646860195929825e-07, | |
| "logits/chosen": 0.054258793592453, | |
| "logits/rejected": 0.3202013075351715, | |
| "logps/chosen": -1.37346613407135, | |
| "logps/rejected": -2.1130452156066895, | |
| "loss": -0.1166, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.37346613407135, | |
| "rewards/margins": 0.7395793199539185, | |
| "rewards/rejected": -2.1130452156066895, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 11.296019715540536, | |
| "learning_rate": 3.054919880453032e-07, | |
| "logits/chosen": 0.08209206908941269, | |
| "logits/rejected": 0.305607408285141, | |
| "logps/chosen": -1.1226253509521484, | |
| "logps/rejected": -2.1423118114471436, | |
| "loss": -0.1085, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1226253509521484, | |
| "rewards/margins": 1.0196868181228638, | |
| "rewards/rejected": -2.1423118114471436, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 12.257070737010752, | |
| "learning_rate": 2.9450801195469686e-07, | |
| "logits/chosen": 0.13827550411224365, | |
| "logits/rejected": 0.33571916818618774, | |
| "logps/chosen": -1.2839715480804443, | |
| "logps/rejected": -1.8086509704589844, | |
| "loss": -0.1095, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.2839715480804443, | |
| "rewards/margins": 0.5246793031692505, | |
| "rewards/rejected": -1.8086509704589844, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 7.650172024925882, | |
| "learning_rate": 2.835313980407017e-07, | |
| "logits/chosen": 0.2891995906829834, | |
| "logits/rejected": 0.4797445237636566, | |
| "logps/chosen": -1.3987174034118652, | |
| "logps/rejected": -1.9334684610366821, | |
| "loss": -0.0939, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.3987174034118652, | |
| "rewards/margins": 0.5347510576248169, | |
| "rewards/rejected": -1.9334684610366821, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 9.750902438039956, | |
| "learning_rate": 2.7257686078731973e-07, | |
| "logits/chosen": 0.3225773572921753, | |
| "logits/rejected": 0.6765245795249939, | |
| "logps/chosen": -1.2831330299377441, | |
| "logps/rejected": -2.208591938018799, | |
| "loss": -0.111, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.2831330299377441, | |
| "rewards/margins": 0.9254589080810547, | |
| "rewards/rejected": -2.208591938018799, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 7.363369128681095, | |
| "learning_rate": 2.6165908508411857e-07, | |
| "logits/chosen": 0.29373425245285034, | |
| "logits/rejected": 0.5836144685745239, | |
| "logps/chosen": -1.2897415161132812, | |
| "logps/rejected": -1.9501053094863892, | |
| "loss": -0.1031, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.2897415161132812, | |
| "rewards/margins": 0.6603637933731079, | |
| "rewards/rejected": -1.9501053094863892, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 13.183162679494803, | |
| "learning_rate": 2.5079270654069354e-07, | |
| "logits/chosen": 0.18902722001075745, | |
| "logits/rejected": 0.29711785912513733, | |
| "logps/chosen": -1.2677323818206787, | |
| "logps/rejected": -2.1915152072906494, | |
| "loss": -0.1109, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.2677323818206787, | |
| "rewards/margins": 0.9237826466560364, | |
| "rewards/rejected": -2.1915152072906494, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 6.687429291622309, | |
| "learning_rate": 2.399922918671867e-07, | |
| "logits/chosen": 0.022776653990149498, | |
| "logits/rejected": 0.2391856163740158, | |
| "logps/chosen": -1.3181854486465454, | |
| "logps/rejected": -2.3116376399993896, | |
| "loss": -0.116, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -1.3181854486465454, | |
| "rewards/margins": 0.993452250957489, | |
| "rewards/rejected": -2.3116376399993896, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 5.281502633271325, | |
| "learning_rate": 2.2927231934717176e-07, | |
| "logits/chosen": -0.0016543983947485685, | |
| "logits/rejected": 0.17889878153800964, | |
| "logps/chosen": -1.3387067317962646, | |
| "logps/rejected": -2.2106659412384033, | |
| "loss": -0.1028, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.3387067317962646, | |
| "rewards/margins": 0.8719590902328491, | |
| "rewards/rejected": -2.2106659412384033, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 7.373204926178573, | |
| "learning_rate": 2.1864715942907487e-07, | |
| "logits/chosen": -0.007572178728878498, | |
| "logits/rejected": 0.14625847339630127, | |
| "logps/chosen": -1.4602124691009521, | |
| "logps/rejected": -2.1073670387268066, | |
| "loss": -0.1046, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.4602124691009521, | |
| "rewards/margins": 0.6471549272537231, | |
| "rewards/rejected": -2.1073670387268066, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 4.736107509318446, | |
| "learning_rate": 2.081310554621522e-07, | |
| "logits/chosen": 0.19010517001152039, | |
| "logits/rejected": 0.3840242624282837, | |
| "logps/chosen": -1.4273982048034668, | |
| "logps/rejected": -2.3235268592834473, | |
| "loss": -0.1246, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.4273982048034668, | |
| "rewards/margins": 0.89612877368927, | |
| "rewards/rejected": -2.3235268592834473, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 23.409368660389333, | |
| "learning_rate": 1.9773810460284862e-07, | |
| "logits/chosen": 0.2859404981136322, | |
| "logits/rejected": 0.3260739743709564, | |
| "logps/chosen": -1.2994334697723389, | |
| "logps/rejected": -2.061152935028076, | |
| "loss": -0.1079, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.2994334697723389, | |
| "rewards/margins": 0.7617196440696716, | |
| "rewards/rejected": -2.061152935028076, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 22.542628144583606, | |
| "learning_rate": 1.874822389171314e-07, | |
| "logits/chosen": 0.22070816159248352, | |
| "logits/rejected": 0.5137720704078674, | |
| "logps/chosen": -1.2794992923736572, | |
| "logps/rejected": -2.3934812545776367, | |
| "loss": -0.1217, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.2794992923736572, | |
| "rewards/margins": 1.11398184299469, | |
| "rewards/rejected": -2.3934812545776367, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 9.283703258871379, | |
| "learning_rate": 1.7737720670413356e-07, | |
| "logits/chosen": 0.24648892879486084, | |
| "logits/rejected": 0.33991724252700806, | |
| "logps/chosen": -1.365140676498413, | |
| "logps/rejected": -2.166224241256714, | |
| "loss": -0.1155, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.365140676498413, | |
| "rewards/margins": 0.8010835647583008, | |
| "rewards/rejected": -2.166224241256714, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 7.959464499237016, | |
| "learning_rate": 1.6743655406614095e-07, | |
| "logits/chosen": 0.14846596121788025, | |
| "logits/rejected": 0.3873682916164398, | |
| "logps/chosen": -1.3291183710098267, | |
| "logps/rejected": -2.1744306087493896, | |
| "loss": -0.1209, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.3291183710098267, | |
| "rewards/margins": 0.845312237739563, | |
| "rewards/rejected": -2.1744306087493896, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 10.067079168168558, | |
| "learning_rate": 1.5767360674963198e-07, | |
| "logits/chosen": 0.08330237120389938, | |
| "logits/rejected": 0.29010123014450073, | |
| "logps/chosen": -1.225388765335083, | |
| "logps/rejected": -1.9910337924957275, | |
| "loss": -0.1019, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.225388765335083, | |
| "rewards/margins": 0.7656451463699341, | |
| "rewards/rejected": -1.9910337924957275, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 9.133946690623846, | |
| "learning_rate": 1.4810145228170922e-07, | |
| "logits/chosen": -0.036666952073574066, | |
| "logits/rejected": 0.15409226715564728, | |
| "logps/chosen": -1.2429730892181396, | |
| "logps/rejected": -1.8629066944122314, | |
| "loss": -0.0986, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2429730892181396, | |
| "rewards/margins": 0.6199334859848022, | |
| "rewards/rejected": -1.8629066944122314, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 10.242157594508342, | |
| "learning_rate": 1.3873292242587306e-07, | |
| "logits/chosen": 0.028749173507094383, | |
| "logits/rejected": 0.23107752203941345, | |
| "logps/chosen": -1.430831789970398, | |
| "logps/rejected": -2.205451488494873, | |
| "loss": -0.1094, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.430831789970398, | |
| "rewards/margins": 0.7746195197105408, | |
| "rewards/rejected": -2.205451488494873, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 5.706648979502026, | |
| "learning_rate": 1.295805759806533e-07, | |
| "logits/chosen": 0.007116416003555059, | |
| "logits/rejected": 0.24577870965003967, | |
| "logps/chosen": -1.297646164894104, | |
| "logps/rejected": -2.2263598442077637, | |
| "loss": -0.1067, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.297646164894104, | |
| "rewards/margins": 0.9287137985229492, | |
| "rewards/rejected": -2.2263598442077637, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 7.792589886443301, | |
| "learning_rate": 1.2065668194415777e-07, | |
| "logits/chosen": 0.1685815453529358, | |
| "logits/rejected": 0.2779741585254669, | |
| "logps/chosen": -1.3478834629058838, | |
| "logps/rejected": -2.127234935760498, | |
| "loss": -0.1044, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.3478834629058838, | |
| "rewards/margins": 0.7793514132499695, | |
| "rewards/rejected": -2.127234935760498, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 8.782679171257993, | |
| "learning_rate": 1.1197320306710923e-07, | |
| "logits/chosen": 0.1377880573272705, | |
| "logits/rejected": 0.32959434390068054, | |
| "logps/chosen": -1.1817530393600464, | |
| "logps/rejected": -2.1875174045562744, | |
| "loss": -0.1184, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.1817530393600464, | |
| "rewards/margins": 1.0057642459869385, | |
| "rewards/rejected": -2.1875174045562744, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 6.862511018163533, | |
| "learning_rate": 1.035417798164145e-07, | |
| "logits/chosen": 0.01301775686442852, | |
| "logits/rejected": 0.2229972630739212, | |
| "logps/chosen": -1.1256128549575806, | |
| "logps/rejected": -1.9575172662734985, | |
| "loss": -0.1212, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.1256128549575806, | |
| "rewards/margins": 0.8319045901298523, | |
| "rewards/rejected": -1.9575172662734985, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 9.505229420957962, | |
| "learning_rate": 9.537371477076535e-08, | |
| "logits/chosen": 0.11367920786142349, | |
| "logits/rejected": 0.356434166431427, | |
| "logps/chosen": -1.4002251625061035, | |
| "logps/rejected": -2.392739772796631, | |
| "loss": -0.1187, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.4002251625061035, | |
| "rewards/margins": 0.9925147294998169, | |
| "rewards/rejected": -2.392739772796631, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 10.965226669329015, | |
| "learning_rate": 8.747995746918898e-08, | |
| "logits/chosen": 0.20804345607757568, | |
| "logits/rejected": 0.45443806052207947, | |
| "logps/chosen": -1.363114356994629, | |
| "logps/rejected": -2.378051280975342, | |
| "loss": -0.1203, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.363114356994629, | |
| "rewards/margins": 1.0149368047714233, | |
| "rewards/rejected": -2.378051280975342, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 6.16813334899156, | |
| "learning_rate": 7.987108973285888e-08, | |
| "logits/chosen": 0.26107341051101685, | |
| "logits/rejected": 0.22848454117774963, | |
| "logps/chosen": -1.387460708618164, | |
| "logps/rejected": -2.197563886642456, | |
| "loss": -0.1101, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.387460708618164, | |
| "rewards/margins": 0.8101032972335815, | |
| "rewards/rejected": -2.197563886642456, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 12.918343537826871, | |
| "learning_rate": 7.255731147984174e-08, | |
| "logits/chosen": 0.22713570296764374, | |
| "logits/rejected": 0.3289734423160553, | |
| "logps/chosen": -1.4456373453140259, | |
| "logps/rejected": -2.084005832672119, | |
| "loss": -0.117, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.4456373453140259, | |
| "rewards/margins": 0.638368546962738, | |
| "rewards/rejected": -2.084005832672119, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 6.060272813011622, | |
| "learning_rate": 6.554842705179898e-08, | |
| "logits/chosen": 0.12520398199558258, | |
| "logits/rejected": 0.29297417402267456, | |
| "logps/chosen": -1.1945627927780151, | |
| "logps/rejected": -2.0077576637268066, | |
| "loss": -0.1195, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.1945627927780151, | |
| "rewards/margins": 0.8131949305534363, | |
| "rewards/rejected": -2.0077576637268066, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 17.713346450689524, | |
| "learning_rate": 5.885383207096832e-08, | |
| "logits/chosen": 0.13393369317054749, | |
| "logits/rejected": 0.3967200517654419, | |
| "logps/chosen": -1.2682616710662842, | |
| "logps/rejected": -2.303426504135132, | |
| "loss": -0.1115, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.2682616710662842, | |
| "rewards/margins": 1.0351645946502686, | |
| "rewards/rejected": -2.303426504135132, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 6.0463426455293385, | |
| "learning_rate": 5.2482500845047165e-08, | |
| "logits/chosen": 0.09920735657215118, | |
| "logits/rejected": 0.4265053868293762, | |
| "logps/chosen": -1.2478829622268677, | |
| "logps/rejected": -2.2664952278137207, | |
| "loss": -0.1059, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.2478829622268677, | |
| "rewards/margins": 1.018612027168274, | |
| "rewards/rejected": -2.2664952278137207, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 7.849951661296689, | |
| "learning_rate": 4.644297433686162e-08, | |
| "logits/chosen": 0.20306222140789032, | |
| "logits/rejected": 0.37799519300460815, | |
| "logps/chosen": -1.167974591255188, | |
| "logps/rejected": -2.061584949493408, | |
| "loss": -0.1211, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.167974591255188, | |
| "rewards/margins": 0.8936103582382202, | |
| "rewards/rejected": -2.061584949493408, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 9.305371166707902, | |
| "learning_rate": 4.074334871494558e-08, | |
| "logits/chosen": 0.12849920988082886, | |
| "logits/rejected": 0.27528905868530273, | |
| "logps/chosen": -1.4915854930877686, | |
| "logps/rejected": -2.3600451946258545, | |
| "loss": -0.0987, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.4915854930877686, | |
| "rewards/margins": 0.8684597015380859, | |
| "rewards/rejected": -2.3600451946258545, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 7.721976646915507, | |
| "learning_rate": 3.5391264500382e-08, | |
| "logits/chosen": 0.15187588334083557, | |
| "logits/rejected": 0.3307904601097107, | |
| "logps/chosen": -1.2111918926239014, | |
| "logps/rejected": -2.036677837371826, | |
| "loss": -0.1121, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.2111918926239014, | |
| "rewards/margins": 0.8254860043525696, | |
| "rewards/rejected": -2.036677837371826, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 7.610516456597608, | |
| "learning_rate": 3.0393896324452226e-08, | |
| "logits/chosen": 0.14874114096164703, | |
| "logits/rejected": 0.44959506392478943, | |
| "logps/chosen": -1.3082196712493896, | |
| "logps/rejected": -2.282323122024536, | |
| "loss": -0.1227, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3082196712493896, | |
| "rewards/margins": 0.9741032719612122, | |
| "rewards/rejected": -2.282323122024536, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 7.385339190408951, | |
| "learning_rate": 2.5757943310825026e-08, | |
| "logits/chosen": 0.149868905544281, | |
| "logits/rejected": 0.3381766676902771, | |
| "logps/chosen": -1.2985490560531616, | |
| "logps/rejected": -2.2531745433807373, | |
| "loss": -0.1037, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.2985490560531616, | |
| "rewards/margins": 0.9546254873275757, | |
| "rewards/rejected": -2.2531745433807373, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 5.355483863860493, | |
| "learning_rate": 2.148962009517823e-08, | |
| "logits/chosen": 0.22056706249713898, | |
| "logits/rejected": 0.38606297969818115, | |
| "logps/chosen": -1.380305528640747, | |
| "logps/rejected": -2.1738598346710205, | |
| "loss": -0.1192, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.380305528640747, | |
| "rewards/margins": 0.793554425239563, | |
| "rewards/rejected": -2.1738598346710205, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 10.975418852072519, | |
| "learning_rate": 1.759464849429082e-08, | |
| "logits/chosen": 0.17666544020175934, | |
| "logits/rejected": 0.34687134623527527, | |
| "logps/chosen": -1.3843985795974731, | |
| "logps/rejected": -2.1382744312286377, | |
| "loss": -0.1132, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3843985795974731, | |
| "rewards/margins": 0.7538760304450989, | |
| "rewards/rejected": -2.1382744312286377, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 8.94416140345287, | |
| "learning_rate": 1.4078249835774169e-08, | |
| "logits/chosen": 0.16215559840202332, | |
| "logits/rejected": 0.30989065766334534, | |
| "logps/chosen": -1.3915765285491943, | |
| "logps/rejected": -2.4042232036590576, | |
| "loss": -0.1263, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.3915765285491943, | |
| "rewards/margins": 1.0126469135284424, | |
| "rewards/rejected": -2.4042232036590576, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 9.337983704389515, | |
| "learning_rate": 1.0945137958723705e-08, | |
| "logits/chosen": 0.33620771765708923, | |
| "logits/rejected": 0.4521760940551758, | |
| "logps/chosen": -1.2900922298431396, | |
| "logps/rejected": -2.0487289428710938, | |
| "loss": -0.1176, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.2900922298431396, | |
| "rewards/margins": 0.7586369514465332, | |
| "rewards/rejected": -2.0487289428710938, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 7.908523894571034, | |
| "learning_rate": 8.19951289467482e-09, | |
| "logits/chosen": 0.16882726550102234, | |
| "logits/rejected": 0.34132689237594604, | |
| "logps/chosen": -1.2635581493377686, | |
| "logps/rejected": -2.0731050968170166, | |
| "loss": -0.1061, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.2635581493377686, | |
| "rewards/margins": 0.809546947479248, | |
| "rewards/rejected": -2.0731050968170166, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 4.689735375861538, | |
| "learning_rate": 5.84505523733293e-09, | |
| "logits/chosen": 0.29893386363983154, | |
| "logits/rejected": 0.4135954976081848, | |
| "logps/chosen": -1.2849633693695068, | |
| "logps/rejected": -2.084097385406494, | |
| "loss": -0.1116, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.2849633693695068, | |
| "rewards/margins": 0.7991340160369873, | |
| "rewards/rejected": -2.084097385406494, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 10.72581336853298, | |
| "learning_rate": 3.8849212086261466e-09, | |
| "logits/chosen": 0.24026727676391602, | |
| "logits/rejected": 0.4279538691043854, | |
| "logps/chosen": -1.4815961122512817, | |
| "logps/rejected": -2.078247547149658, | |
| "loss": -0.111, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.4815961122512817, | |
| "rewards/margins": 0.596651554107666, | |
| "rewards/rejected": -2.078247547149658, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 8.569087978528602, | |
| "learning_rate": 2.3217384276938756e-09, | |
| "logits/chosen": 0.3192306160926819, | |
| "logits/rejected": 0.4579745829105377, | |
| "logps/chosen": -1.2749390602111816, | |
| "logps/rejected": -2.1251206398010254, | |
| "loss": -0.1167, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.2749390602111816, | |
| "rewards/margins": 0.8501815795898438, | |
| "rewards/rejected": -2.1251206398010254, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 8.07831527929665, | |
| "learning_rate": 1.1576023884836472e-09, | |
| "logits/chosen": 0.14747342467308044, | |
| "logits/rejected": 0.4368831515312195, | |
| "logps/chosen": -1.3681727647781372, | |
| "logps/rejected": -2.154388427734375, | |
| "loss": -0.1145, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.3681727647781372, | |
| "rewards/margins": 0.7862156629562378, | |
| "rewards/rejected": -2.154388427734375, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 10.228166489996983, | |
| "learning_rate": 3.940736506780395e-10, | |
| "logits/chosen": 0.13486842811107635, | |
| "logits/rejected": 0.33838891983032227, | |
| "logps/chosen": -1.2538591623306274, | |
| "logps/rejected": -1.9874036312103271, | |
| "loss": -0.1067, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.2538591623306274, | |
| "rewards/margins": 0.7335445284843445, | |
| "rewards/rejected": -1.9874036312103271, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 21.73443269439748, | |
| "learning_rate": 3.2175747716822744e-11, | |
| "logits/chosen": 0.1331544667482376, | |
| "logits/rejected": 0.45593690872192383, | |
| "logps/chosen": -1.3638273477554321, | |
| "logps/rejected": -2.1485755443573, | |
| "loss": -0.0976, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.3638273477554321, | |
| "rewards/margins": 0.7847483158111572, | |
| "rewards/rejected": -2.1485755443573, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": -0.09237800735347676, | |
| "train_runtime": 17432.5225, | |
| "train_samples_per_second": 3.507, | |
| "train_steps_per_second": 0.027 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "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": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |