| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 1875, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 2.6595744680851065e-08, | |
| "logits/chosen": 0.08248457312583923, | |
| "logits/rejected": 0.20234191417694092, | |
| "logps/chosen": -342.9681091308594, | |
| "logps/rejected": -279.68194580078125, | |
| "loss": 0.2964, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.6595744680851066e-07, | |
| "logits/chosen": 0.12211620807647705, | |
| "logits/rejected": 0.22816644608974457, | |
| "logps/chosen": -387.83782958984375, | |
| "logps/rejected": -299.0201721191406, | |
| "loss": 0.334, | |
| "rewards/accuracies": 0.3958333432674408, | |
| "rewards/chosen": 2.3156941097113304e-05, | |
| "rewards/margins": 5.933707143412903e-05, | |
| "rewards/rejected": -3.618013215600513e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5.319148936170213e-07, | |
| "logits/chosen": 0.061443157494068146, | |
| "logits/rejected": 0.2900610864162445, | |
| "logps/chosen": -387.94866943359375, | |
| "logps/rejected": -344.0950927734375, | |
| "loss": 0.3514, | |
| "rewards/accuracies": 0.4437499940395355, | |
| "rewards/chosen": -6.4797219238244e-05, | |
| "rewards/margins": -0.0001574750494910404, | |
| "rewards/rejected": 9.267783752875403e-05, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 7.97872340425532e-07, | |
| "logits/chosen": 0.07704450935125351, | |
| "logits/rejected": 0.24550946056842804, | |
| "logps/chosen": -327.02447509765625, | |
| "logps/rejected": -287.892822265625, | |
| "loss": 0.3212, | |
| "rewards/accuracies": 0.39375001192092896, | |
| "rewards/chosen": -0.00013851793482899666, | |
| "rewards/margins": -0.00014858768554404378, | |
| "rewards/rejected": 1.0069748896057718e-05, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.0638297872340427e-06, | |
| "logits/chosen": 0.15527553856372833, | |
| "logits/rejected": 0.2848062217235565, | |
| "logps/chosen": -380.29180908203125, | |
| "logps/rejected": -296.22418212890625, | |
| "loss": 0.3505, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": 0.00010127278801519424, | |
| "rewards/margins": 7.252545765368268e-05, | |
| "rewards/rejected": 2.8747313990606926e-05, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.3297872340425533e-06, | |
| "logits/chosen": 0.061357248574495316, | |
| "logits/rejected": 0.2838377356529236, | |
| "logps/chosen": -341.8178405761719, | |
| "logps/rejected": -285.6352233886719, | |
| "loss": 0.3228, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": 0.00010316881525795907, | |
| "rewards/margins": 4.464336234377697e-05, | |
| "rewards/rejected": 5.852547474205494e-05, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.595744680851064e-06, | |
| "logits/chosen": 0.10541439056396484, | |
| "logits/rejected": 0.28058919310569763, | |
| "logps/chosen": -343.18572998046875, | |
| "logps/rejected": -294.2557373046875, | |
| "loss": 0.3285, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": 0.00032235862454399467, | |
| "rewards/margins": 0.0005498815444298089, | |
| "rewards/rejected": -0.00022752289078198373, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1.8617021276595745e-06, | |
| "logits/chosen": 0.08224678784608841, | |
| "logits/rejected": 0.2537159323692322, | |
| "logps/chosen": -377.56622314453125, | |
| "logps/rejected": -302.76885986328125, | |
| "loss": 0.3673, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.0004772460088133812, | |
| "rewards/margins": 0.0005696729640476406, | |
| "rewards/rejected": -9.242700616596267e-05, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "logits/chosen": 0.08698181807994843, | |
| "logits/rejected": 0.28228509426116943, | |
| "logps/chosen": -353.9815979003906, | |
| "logps/rejected": -302.00439453125, | |
| "loss": 0.3803, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.0008322820067405701, | |
| "rewards/margins": 0.0010860636830329895, | |
| "rewards/rejected": -0.0002537816180847585, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.393617021276596e-06, | |
| "logits/chosen": 0.13658957183361053, | |
| "logits/rejected": 0.2813904583454132, | |
| "logps/chosen": -331.66705322265625, | |
| "logps/rejected": -290.46148681640625, | |
| "loss": 0.3316, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.0012246040860190988, | |
| "rewards/margins": 0.0016386559000238776, | |
| "rewards/rejected": -0.00041405190131627023, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.6595744680851065e-06, | |
| "logits/chosen": 0.14109408855438232, | |
| "logits/rejected": 0.28268519043922424, | |
| "logps/chosen": -337.4228210449219, | |
| "logps/rejected": -265.25091552734375, | |
| "loss": 0.3472, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": 0.0017904158448800445, | |
| "rewards/margins": 0.0025100118946284056, | |
| "rewards/rejected": -0.0007195962825790048, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.9255319148936174e-06, | |
| "logits/chosen": 0.0813894122838974, | |
| "logits/rejected": 0.19894129037857056, | |
| "logps/chosen": -331.2377014160156, | |
| "logps/rejected": -252.0073699951172, | |
| "loss": 0.336, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.002608851296827197, | |
| "rewards/margins": 0.0032108586747199297, | |
| "rewards/rejected": -0.0006020068540237844, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.191489361702128e-06, | |
| "logits/chosen": 0.05477667599916458, | |
| "logits/rejected": 0.2091325968503952, | |
| "logps/chosen": -345.79229736328125, | |
| "logps/rejected": -283.46649169921875, | |
| "loss": 0.3315, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.005583181511610746, | |
| "rewards/margins": 0.006550823338329792, | |
| "rewards/rejected": -0.0009676421177573502, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.457446808510639e-06, | |
| "logits/chosen": 0.10417729616165161, | |
| "logits/rejected": 0.28965848684310913, | |
| "logps/chosen": -355.0021057128906, | |
| "logps/rejected": -291.60302734375, | |
| "loss": 0.3406, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.007259941194206476, | |
| "rewards/margins": 0.008403105661273003, | |
| "rewards/rejected": -0.001143165398389101, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.723404255319149e-06, | |
| "logits/chosen": 0.1521492302417755, | |
| "logits/rejected": 0.25287285447120667, | |
| "logps/chosen": -362.653564453125, | |
| "logps/rejected": -299.1832580566406, | |
| "loss": 0.3305, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.00862643402069807, | |
| "rewards/margins": 0.016447611153125763, | |
| "rewards/rejected": -0.007821176201105118, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.98936170212766e-06, | |
| "logits/chosen": 0.1634024679660797, | |
| "logits/rejected": 0.32300662994384766, | |
| "logps/chosen": -331.88592529296875, | |
| "logps/rejected": -299.06195068359375, | |
| "loss": 0.3256, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": 0.003903627395629883, | |
| "rewards/margins": 0.019702786579728127, | |
| "rewards/rejected": -0.015799159184098244, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.255319148936171e-06, | |
| "logits/chosen": 0.17028017342090607, | |
| "logits/rejected": 0.3108305335044861, | |
| "logps/chosen": -379.0628662109375, | |
| "logps/rejected": -322.420166015625, | |
| "loss": 0.2893, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.008702902123332024, | |
| "rewards/margins": 0.03689227253198624, | |
| "rewards/rejected": -0.04559517279267311, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.521276595744681e-06, | |
| "logits/chosen": 0.17017154395580292, | |
| "logits/rejected": 0.31650176644325256, | |
| "logps/chosen": -397.65875244140625, | |
| "logps/rejected": -400.8711242675781, | |
| "loss": 0.2877, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.037718433886766434, | |
| "rewards/margins": 0.06432106345891953, | |
| "rewards/rejected": -0.10203950107097626, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.787234042553192e-06, | |
| "logits/chosen": 0.18481381237506866, | |
| "logits/rejected": 0.33424076437950134, | |
| "logps/chosen": -473.28680419921875, | |
| "logps/rejected": -518.6297607421875, | |
| "loss": 0.2869, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.12800003588199615, | |
| "rewards/margins": 0.0953989028930664, | |
| "rewards/rejected": -0.22339896857738495, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999982660399688e-06, | |
| "logits/chosen": 0.227821946144104, | |
| "logits/rejected": 0.3258199989795685, | |
| "logps/chosen": -561.3455200195312, | |
| "logps/rejected": -633.8394165039062, | |
| "loss": 0.2913, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.2103206217288971, | |
| "rewards/margins": 0.13042964041233063, | |
| "rewards/rejected": -0.3407502770423889, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.99937579964398e-06, | |
| "logits/chosen": 0.2324659377336502, | |
| "logits/rejected": 0.3176039457321167, | |
| "logps/chosen": -520.3052368164062, | |
| "logps/rejected": -595.4998168945312, | |
| "loss": 0.2482, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.16411438584327698, | |
| "rewards/margins": 0.14954686164855957, | |
| "rewards/rejected": -0.31366124749183655, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.9979021993870645e-06, | |
| "logits/chosen": 0.27860119938850403, | |
| "logits/rejected": 0.3871223032474518, | |
| "logps/chosen": -527.0191650390625, | |
| "logps/rejected": -620.4132080078125, | |
| "loss": 0.2693, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.2112080156803131, | |
| "rewards/margins": 0.1339053362607956, | |
| "rewards/rejected": -0.3451133370399475, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.995562370647553e-06, | |
| "logits/chosen": 0.2107020914554596, | |
| "logits/rejected": 0.3118240237236023, | |
| "logps/chosen": -586.7862548828125, | |
| "logps/rejected": -676.7178955078125, | |
| "loss": 0.2707, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.2537532448768616, | |
| "rewards/margins": 0.12398215383291245, | |
| "rewards/rejected": -0.37773534655570984, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.992357124836838e-06, | |
| "logits/chosen": 0.22325322031974792, | |
| "logits/rejected": 0.3859561085700989, | |
| "logps/chosen": -596.8721923828125, | |
| "logps/rejected": -648.3790893554688, | |
| "loss": 0.2911, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.24456577003002167, | |
| "rewards/margins": 0.12123928964138031, | |
| "rewards/rejected": -0.36580508947372437, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.9882875734777044e-06, | |
| "logits/chosen": 0.24928610026836395, | |
| "logits/rejected": 0.3123549818992615, | |
| "logps/chosen": -526.9947509765625, | |
| "logps/rejected": -622.6949462890625, | |
| "loss": 0.2858, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.1741228997707367, | |
| "rewards/margins": 0.15204958617687225, | |
| "rewards/rejected": -0.32617247104644775, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.983355127818882e-06, | |
| "logits/chosen": 0.2413512021303177, | |
| "logits/rejected": 0.3579440712928772, | |
| "logps/chosen": -527.4100341796875, | |
| "logps/rejected": -615.1137084960938, | |
| "loss": 0.2276, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.18759354948997498, | |
| "rewards/margins": 0.15318110585212708, | |
| "rewards/rejected": -0.34077465534210205, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.977561498345639e-06, | |
| "logits/chosen": 0.22231824696063995, | |
| "logits/rejected": 0.3377223610877991, | |
| "logps/chosen": -557.3218994140625, | |
| "logps/rejected": -656.0249633789062, | |
| "loss": 0.2409, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.21337445080280304, | |
| "rewards/margins": 0.14147409796714783, | |
| "rewards/rejected": -0.35484856367111206, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.970908694186624e-06, | |
| "logits/chosen": 0.20074252784252167, | |
| "logits/rejected": 0.31892186403274536, | |
| "logps/chosen": -619.4715576171875, | |
| "logps/rejected": -680.501708984375, | |
| "loss": 0.2752, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.27192774415016174, | |
| "rewards/margins": 0.13113710284233093, | |
| "rewards/rejected": -0.4030648171901703, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.9633990224171305e-06, | |
| "logits/chosen": 0.21738846600055695, | |
| "logits/rejected": 0.36280277371406555, | |
| "logps/chosen": -641.8638916015625, | |
| "logps/rejected": -719.9840087890625, | |
| "loss": 0.2682, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.2767099142074585, | |
| "rewards/margins": 0.1477716863155365, | |
| "rewards/rejected": -0.4244815707206726, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.955035087259046e-06, | |
| "logits/chosen": 0.25205639004707336, | |
| "logits/rejected": 0.3571772873401642, | |
| "logps/chosen": -596.9175415039062, | |
| "logps/rejected": -665.6470947265625, | |
| "loss": 0.2488, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.23527784645557404, | |
| "rewards/margins": 0.1494690328836441, | |
| "rewards/rejected": -0.38474684953689575, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.945819789177756e-06, | |
| "logits/chosen": 0.2460106611251831, | |
| "logits/rejected": 0.340692400932312, | |
| "logps/chosen": -576.7429809570312, | |
| "logps/rejected": -725.2931518554688, | |
| "loss": 0.1927, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.23355364799499512, | |
| "rewards/margins": 0.19731679558753967, | |
| "rewards/rejected": -0.4308704435825348, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.935756323876306e-06, | |
| "logits/chosen": 0.27840182185173035, | |
| "logits/rejected": 0.3699595034122467, | |
| "logps/chosen": -655.7606201171875, | |
| "logps/rejected": -755.6095581054688, | |
| "loss": 0.2432, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.2706039547920227, | |
| "rewards/margins": 0.1736321747303009, | |
| "rewards/rejected": -0.4442361295223236, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.924848181187199e-06, | |
| "logits/chosen": 0.25754767656326294, | |
| "logits/rejected": 0.38638874888420105, | |
| "logps/chosen": -657.5533447265625, | |
| "logps/rejected": -750.8352661132812, | |
| "loss": 0.2622, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.2994315028190613, | |
| "rewards/margins": 0.16641752421855927, | |
| "rewards/rejected": -0.46584898233413696, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.913099143862173e-06, | |
| "logits/chosen": 0.25059616565704346, | |
| "logits/rejected": 0.36868372559547424, | |
| "logps/chosen": -672.0949096679688, | |
| "logps/rejected": -803.8678588867188, | |
| "loss": 0.241, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.3017386496067047, | |
| "rewards/margins": 0.20527306199073792, | |
| "rewards/rejected": -0.5070117115974426, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.900513286260416e-06, | |
| "logits/chosen": 0.2503548264503479, | |
| "logits/rejected": 0.29262545704841614, | |
| "logps/chosen": -637.6034545898438, | |
| "logps/rejected": -689.4666748046875, | |
| "loss": 0.2409, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.28344589471817017, | |
| "rewards/margins": 0.14148998260498047, | |
| "rewards/rejected": -0.424935907125473, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.887094972935645e-06, | |
| "logits/chosen": 0.21581009030342102, | |
| "logits/rejected": 0.38255298137664795, | |
| "logps/chosen": -662.43212890625, | |
| "logps/rejected": -777.831787109375, | |
| "loss": 0.2152, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.28980588912963867, | |
| "rewards/margins": 0.18855467438697815, | |
| "rewards/rejected": -0.4783605635166168, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.87284885712256e-06, | |
| "logits/chosen": 0.2862156331539154, | |
| "logits/rejected": 0.3305504024028778, | |
| "logps/chosen": -632.4459228515625, | |
| "logps/rejected": -718.5078125, | |
| "loss": 0.2728, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.294480562210083, | |
| "rewards/margins": 0.14050373435020447, | |
| "rewards/rejected": -0.43498435616493225, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.857779879123181e-06, | |
| "logits/chosen": 0.2614066004753113, | |
| "logits/rejected": 0.2966289818286896, | |
| "logps/chosen": -622.0609130859375, | |
| "logps/rejected": -742.937744140625, | |
| "loss": 0.2505, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.2878873348236084, | |
| "rewards/margins": 0.15637485682964325, | |
| "rewards/rejected": -0.4442622661590576, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.841893264593643e-06, | |
| "logits/chosen": 0.1658310443162918, | |
| "logits/rejected": 0.25932207703590393, | |
| "logps/chosen": -574.1139526367188, | |
| "logps/rejected": -716.8287353515625, | |
| "loss": 0.2169, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.25041764974594116, | |
| "rewards/margins": 0.1778055727481842, | |
| "rewards/rejected": -0.428223192691803, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.825194522732023e-06, | |
| "logits/chosen": 0.2032794952392578, | |
| "logits/rejected": 0.2511264681816101, | |
| "logps/chosen": -652.1287841796875, | |
| "logps/rejected": -765.1002807617188, | |
| "loss": 0.2461, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.31416720151901245, | |
| "rewards/margins": 0.15508690476417542, | |
| "rewards/rejected": -0.4692540764808655, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.807689444367853e-06, | |
| "logits/chosen": 0.17228561639785767, | |
| "logits/rejected": 0.28850919008255005, | |
| "logps/chosen": -621.0162963867188, | |
| "logps/rejected": -751.5133666992188, | |
| "loss": 0.2474, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.28327521681785583, | |
| "rewards/margins": 0.1886013299226761, | |
| "rewards/rejected": -0.4718765616416931, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.78938409995396e-06, | |
| "logits/chosen": 0.15390515327453613, | |
| "logits/rejected": 0.23650500178337097, | |
| "logps/chosen": -579.6705322265625, | |
| "logps/rejected": -705.4219970703125, | |
| "loss": 0.2203, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.2634733319282532, | |
| "rewards/margins": 0.17486760020256042, | |
| "rewards/rejected": -0.4383409917354584, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.770284837461342e-06, | |
| "logits/chosen": 0.15774960815906525, | |
| "logits/rejected": 0.25373905897140503, | |
| "logps/chosen": -635.8734130859375, | |
| "logps/rejected": -759.529052734375, | |
| "loss": 0.2478, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.2786625921726227, | |
| "rewards/margins": 0.18710032105445862, | |
| "rewards/rejected": -0.4657629132270813, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7503982801778015e-06, | |
| "logits/chosen": 0.17501457035541534, | |
| "logits/rejected": 0.25947824120521545, | |
| "logps/chosen": -635.3245849609375, | |
| "logps/rejected": -745.48095703125, | |
| "loss": 0.2544, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.2973678708076477, | |
| "rewards/margins": 0.17151114344596863, | |
| "rewards/rejected": -0.4688790440559387, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.729731324411104e-06, | |
| "logits/chosen": 0.17305608093738556, | |
| "logits/rejected": 0.23529882729053497, | |
| "logps/chosen": -642.1393432617188, | |
| "logps/rejected": -720.1858520507812, | |
| "loss": 0.2465, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.28984335064888, | |
| "rewards/margins": 0.16753990948200226, | |
| "rewards/rejected": -0.4573832154273987, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.7082911370974645e-06, | |
| "logits/chosen": 0.20705783367156982, | |
| "logits/rejected": 0.22554746270179749, | |
| "logps/chosen": -680.9107055664062, | |
| "logps/rejected": -766.1033325195312, | |
| "loss": 0.2615, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.32083216309547424, | |
| "rewards/margins": 0.16086088120937347, | |
| "rewards/rejected": -0.4816930890083313, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.68608515331618e-06, | |
| "logits/chosen": 0.1991531401872635, | |
| "logits/rejected": 0.230634406208992, | |
| "logps/chosen": -618.709716796875, | |
| "logps/rejected": -745.50537109375, | |
| "loss": 0.2224, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.3031608760356903, | |
| "rewards/margins": 0.1703597754240036, | |
| "rewards/rejected": -0.4735206663608551, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.663121073711269e-06, | |
| "logits/chosen": 0.1910509616136551, | |
| "logits/rejected": 0.2898029685020447, | |
| "logps/chosen": -711.2774047851562, | |
| "logps/rejected": -782.086181640625, | |
| "loss": 0.2611, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.3283078670501709, | |
| "rewards/margins": 0.1573607474565506, | |
| "rewards/rejected": -0.4856685996055603, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.63940686182103e-06, | |
| "logits/chosen": 0.1314457505941391, | |
| "logits/rejected": 0.21780280768871307, | |
| "logps/chosen": -586.25439453125, | |
| "logps/rejected": -682.1627197265625, | |
| "loss": 0.2774, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.2654445469379425, | |
| "rewards/margins": 0.14997784793376923, | |
| "rewards/rejected": -0.4154224395751953, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.614950741316425e-06, | |
| "logits/chosen": 0.16227178275585175, | |
| "logits/rejected": 0.2971927225589752, | |
| "logps/chosen": -620.1968994140625, | |
| "logps/rejected": -727.1217651367188, | |
| "loss": 0.2396, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.25877463817596436, | |
| "rewards/margins": 0.17255476117134094, | |
| "rewards/rejected": -0.4313293397426605, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.589761193149254e-06, | |
| "logits/chosen": 0.17410214245319366, | |
| "logits/rejected": 0.2344098538160324, | |
| "logps/chosen": -601.7906494140625, | |
| "logps/rejected": -691.7999267578125, | |
| "loss": 0.2486, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.2862247824668884, | |
| "rewards/margins": 0.1431434452533722, | |
| "rewards/rejected": -0.429368257522583, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.563846952611112e-06, | |
| "logits/chosen": 0.21162112057209015, | |
| "logits/rejected": 0.2572205662727356, | |
| "logps/chosen": -637.9953002929688, | |
| "logps/rejected": -775.0775756835938, | |
| "loss": 0.2253, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.2933396100997925, | |
| "rewards/margins": 0.17365200817584991, | |
| "rewards/rejected": -0.4669916033744812, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.537217006304141e-06, | |
| "logits/chosen": 0.20230130851268768, | |
| "logits/rejected": 0.2933056056499481, | |
| "logps/chosen": -694.6912231445312, | |
| "logps/rejected": -818.5560913085938, | |
| "loss": 0.2893, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.35351866483688354, | |
| "rewards/margins": 0.16594737768173218, | |
| "rewards/rejected": -0.5194661021232605, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.50988058902464e-06, | |
| "logits/chosen": 0.18275514245033264, | |
| "logits/rejected": 0.280184805393219, | |
| "logps/chosen": -717.5177001953125, | |
| "logps/rejected": -888.8621215820312, | |
| "loss": 0.1868, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.3565286695957184, | |
| "rewards/margins": 0.23444192111492157, | |
| "rewards/rejected": -0.5909705758094788, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.481847180560593e-06, | |
| "logits/chosen": 0.1679656207561493, | |
| "logits/rejected": 0.26842325925827026, | |
| "logps/chosen": -630.0240478515625, | |
| "logps/rejected": -751.3941650390625, | |
| "loss": 0.2507, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.32289132475852966, | |
| "rewards/margins": 0.16883692145347595, | |
| "rewards/rejected": -0.4917283058166504, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.453126502404253e-06, | |
| "logits/chosen": 0.14577992260456085, | |
| "logits/rejected": 0.2916201949119568, | |
| "logps/chosen": -620.7379150390625, | |
| "logps/rejected": -734.3123168945312, | |
| "loss": 0.2375, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.26863834261894226, | |
| "rewards/margins": 0.1942332684993744, | |
| "rewards/rejected": -0.46287161111831665, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.423728514380892e-06, | |
| "logits/chosen": 0.20810751616954803, | |
| "logits/rejected": 0.2806350886821747, | |
| "logps/chosen": -641.7298583984375, | |
| "logps/rejected": -756.7706909179688, | |
| "loss": 0.2485, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.3129487931728363, | |
| "rewards/margins": 0.16476401686668396, | |
| "rewards/rejected": -0.47771281003952026, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.393663411194918e-06, | |
| "logits/chosen": 0.17864790558815002, | |
| "logits/rejected": 0.28502240777015686, | |
| "logps/chosen": -667.49365234375, | |
| "logps/rejected": -784.1561279296875, | |
| "loss": 0.2486, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.29371771216392517, | |
| "rewards/margins": 0.18731026351451874, | |
| "rewards/rejected": -0.4810279905796051, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.362941618894523e-06, | |
| "logits/chosen": 0.18069639801979065, | |
| "logits/rejected": 0.26032230257987976, | |
| "logps/chosen": -668.4435424804688, | |
| "logps/rejected": -794.9130859375, | |
| "loss": 0.2149, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.29407259821891785, | |
| "rewards/margins": 0.2029459923505783, | |
| "rewards/rejected": -0.49701857566833496, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.331573791256116e-06, | |
| "logits/chosen": 0.14592161774635315, | |
| "logits/rejected": 0.21640054881572723, | |
| "logps/chosen": -656.1441650390625, | |
| "logps/rejected": -773.119140625, | |
| "loss": 0.2124, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.30144789814949036, | |
| "rewards/margins": 0.19570669531822205, | |
| "rewards/rejected": -0.49715456366539, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.299570806089786e-06, | |
| "logits/chosen": 0.13483521342277527, | |
| "logits/rejected": 0.26808056235313416, | |
| "logps/chosen": -663.4806518554688, | |
| "logps/rejected": -814.9881591796875, | |
| "loss": 0.2158, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.33338984847068787, | |
| "rewards/margins": 0.21311593055725098, | |
| "rewards/rejected": -0.5465058088302612, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.266943761467057e-06, | |
| "logits/chosen": 0.17011983692646027, | |
| "logits/rejected": 0.30834323167800903, | |
| "logps/chosen": -687.5603637695312, | |
| "logps/rejected": -845.2774658203125, | |
| "loss": 0.2322, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.3399781286716461, | |
| "rewards/margins": 0.20812830328941345, | |
| "rewards/rejected": -0.5481064319610596, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.233703971872287e-06, | |
| "logits/chosen": 0.2005978524684906, | |
| "logits/rejected": 0.2786727845668793, | |
| "logps/chosen": -676.8416748046875, | |
| "logps/rejected": -822.3328247070312, | |
| "loss": 0.2327, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.32490307092666626, | |
| "rewards/margins": 0.2102217972278595, | |
| "rewards/rejected": -0.5351248979568481, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.1998629642789925e-06, | |
| "logits/chosen": 0.2299545705318451, | |
| "logits/rejected": 0.3201584815979004, | |
| "logps/chosen": -677.6417236328125, | |
| "logps/rejected": -811.4082641601562, | |
| "loss": 0.2091, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.3028990626335144, | |
| "rewards/margins": 0.2108876258134842, | |
| "rewards/rejected": -0.5137866735458374, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.165432474152505e-06, | |
| "logits/chosen": 0.18550391495227814, | |
| "logits/rejected": 0.3030850887298584, | |
| "logps/chosen": -693.2557373046875, | |
| "logps/rejected": -851.9300537109375, | |
| "loss": 0.2046, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -0.32129475474357605, | |
| "rewards/margins": 0.23609153926372528, | |
| "rewards/rejected": -0.5573862791061401, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.130424441380308e-06, | |
| "logits/chosen": 0.21198038756847382, | |
| "logits/rejected": 0.353886216878891, | |
| "logps/chosen": -680.061767578125, | |
| "logps/rejected": -838.3746337890625, | |
| "loss": 0.2152, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.3215908408164978, | |
| "rewards/margins": 0.2143673449754715, | |
| "rewards/rejected": -0.5359582304954529, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.09485100613151e-06, | |
| "logits/chosen": 0.1963845193386078, | |
| "logits/rejected": 0.32071763277053833, | |
| "logps/chosen": -664.4266357421875, | |
| "logps/rejected": -800.628662109375, | |
| "loss": 0.2326, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.323551207780838, | |
| "rewards/margins": 0.20758795738220215, | |
| "rewards/rejected": -0.5311391949653625, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.058724504646834e-06, | |
| "logits/chosen": 0.2653447985649109, | |
| "logits/rejected": 0.32281583547592163, | |
| "logps/chosen": -645.7322998046875, | |
| "logps/rejected": -818.6993408203125, | |
| "loss": 0.2088, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.30575788021087646, | |
| "rewards/margins": 0.20940056443214417, | |
| "rewards/rejected": -0.5151584148406982, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.022057464960632e-06, | |
| "logits/chosen": 0.22752830386161804, | |
| "logits/rejected": 0.29400375485420227, | |
| "logps/chosen": -668.3670654296875, | |
| "logps/rejected": -773.338623046875, | |
| "loss": 0.2668, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.3347984850406647, | |
| "rewards/margins": 0.1586008071899414, | |
| "rewards/rejected": -0.4933992922306061, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.984862602556383e-06, | |
| "logits/chosen": 0.22624416649341583, | |
| "logits/rejected": 0.3623146116733551, | |
| "logps/chosen": -635.141357421875, | |
| "logps/rejected": -833.0682373046875, | |
| "loss": 0.2039, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.32188528776168823, | |
| "rewards/margins": 0.22867591679096222, | |
| "rewards/rejected": -0.5505611896514893, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.947152815957187e-06, | |
| "logits/chosen": 0.24318163096904755, | |
| "logits/rejected": 0.31874382495880127, | |
| "logps/chosen": -668.0684814453125, | |
| "logps/rejected": -803.7855224609375, | |
| "loss": 0.2405, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.33614617586135864, | |
| "rewards/margins": 0.18720091879367828, | |
| "rewards/rejected": -0.5233470797538757, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.908941182252785e-06, | |
| "logits/chosen": 0.2103017121553421, | |
| "logits/rejected": 0.30453357100486755, | |
| "logps/chosen": -750.6541748046875, | |
| "logps/rejected": -890.9783325195312, | |
| "loss": 0.2479, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.3959191143512726, | |
| "rewards/margins": 0.2101031243801117, | |
| "rewards/rejected": -0.6060222387313843, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.8702409525646535e-06, | |
| "logits/chosen": 0.18374183773994446, | |
| "logits/rejected": 0.2802325189113617, | |
| "logps/chosen": -704.2998046875, | |
| "logps/rejected": -872.6188354492188, | |
| "loss": 0.2378, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.3979497253894806, | |
| "rewards/margins": 0.18762901425361633, | |
| "rewards/rejected": -0.5855787396430969, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.8310655474507495e-06, | |
| "logits/chosen": 0.2030530869960785, | |
| "logits/rejected": 0.2757278084754944, | |
| "logps/chosen": -781.7509155273438, | |
| "logps/rejected": -940.8536376953125, | |
| "loss": 0.1757, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.4106892943382263, | |
| "rewards/margins": 0.23548904061317444, | |
| "rewards/rejected": -0.6461783647537231, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.7914285522515002e-06, | |
| "logits/chosen": 0.1338091641664505, | |
| "logits/rejected": 0.2605449855327606, | |
| "logps/chosen": -802.19091796875, | |
| "logps/rejected": -937.0023193359375, | |
| "loss": 0.226, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.42191141843795776, | |
| "rewards/margins": 0.2278059720993042, | |
| "rewards/rejected": -0.649717390537262, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.751343712378639e-06, | |
| "logits/chosen": 0.2370164394378662, | |
| "logits/rejected": 0.2781811058521271, | |
| "logps/chosen": -754.5617065429688, | |
| "logps/rejected": -908.2735595703125, | |
| "loss": 0.2125, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.3990176320075989, | |
| "rewards/margins": 0.20696358382701874, | |
| "rewards/rejected": -0.6059812307357788, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.710824928548546e-06, | |
| "logits/chosen": 0.169587641954422, | |
| "logits/rejected": 0.3024446368217468, | |
| "logps/chosen": -662.3217163085938, | |
| "logps/rejected": -817.04638671875, | |
| "loss": 0.2308, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.35774677991867065, | |
| "rewards/margins": 0.19141057133674622, | |
| "rewards/rejected": -0.5491573214530945, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.6698862519617225e-06, | |
| "logits/chosen": 0.18283501267433167, | |
| "logits/rejected": 0.27203458547592163, | |
| "logps/chosen": -694.4537353515625, | |
| "logps/rejected": -874.41162109375, | |
| "loss": 0.1928, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.3298128545284271, | |
| "rewards/margins": 0.236416295170784, | |
| "rewards/rejected": -0.5662291646003723, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.6285418794300793e-06, | |
| "logits/chosen": 0.23415608704090118, | |
| "logits/rejected": 0.27542707324028015, | |
| "logps/chosen": -682.388671875, | |
| "logps/rejected": -793.8580932617188, | |
| "loss": 0.1776, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.29099172353744507, | |
| "rewards/margins": 0.22200138866901398, | |
| "rewards/rejected": -0.5129930973052979, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.5868061484537365e-06, | |
| "logits/chosen": 0.14799238741397858, | |
| "logits/rejected": 0.28972503542900085, | |
| "logps/chosen": -674.9385375976562, | |
| "logps/rejected": -882.7672729492188, | |
| "loss": 0.1758, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.32894080877304077, | |
| "rewards/margins": 0.24991688132286072, | |
| "rewards/rejected": -0.5788577198982239, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5446935322490285e-06, | |
| "logits/chosen": 0.1383066177368164, | |
| "logits/rejected": 0.24791190028190613, | |
| "logps/chosen": -707.7462158203125, | |
| "logps/rejected": -852.08984375, | |
| "loss": 0.223, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.36430221796035767, | |
| "rewards/margins": 0.1998947560787201, | |
| "rewards/rejected": -0.5641970038414001, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.502218634729447e-06, | |
| "logits/chosen": 0.1664973497390747, | |
| "logits/rejected": 0.2503690719604492, | |
| "logps/chosen": -671.7937622070312, | |
| "logps/rejected": -855.0314331054688, | |
| "loss": 0.2005, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.32940590381622314, | |
| "rewards/margins": 0.23283663392066956, | |
| "rewards/rejected": -0.5622425079345703, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.459396185441265e-06, | |
| "logits/chosen": 0.13743051886558533, | |
| "logits/rejected": 0.2722577452659607, | |
| "logps/chosen": -701.9246826171875, | |
| "logps/rejected": -850.3126220703125, | |
| "loss": 0.2512, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.365190327167511, | |
| "rewards/margins": 0.20733347535133362, | |
| "rewards/rejected": -0.5725237727165222, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4162410344555834e-06, | |
| "logits/chosen": 0.14434996247291565, | |
| "logits/rejected": 0.24066290259361267, | |
| "logps/chosen": -690.765380859375, | |
| "logps/rejected": -829.47802734375, | |
| "loss": 0.2331, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.3407720923423767, | |
| "rewards/margins": 0.19598500430583954, | |
| "rewards/rejected": -0.5367571115493774, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3727681472185937e-06, | |
| "logits/chosen": 0.17098698019981384, | |
| "logits/rejected": 0.24693337082862854, | |
| "logps/chosen": -667.8488159179688, | |
| "logps/rejected": -805.3338012695312, | |
| "loss": 0.2299, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.3178989291191101, | |
| "rewards/margins": 0.19422563910484314, | |
| "rewards/rejected": -0.5121246576309204, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3289925993618217e-06, | |
| "logits/chosen": 0.14288540184497833, | |
| "logits/rejected": 0.34719187021255493, | |
| "logps/chosen": -716.6546020507812, | |
| "logps/rejected": -865.6365356445312, | |
| "loss": 0.1918, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.3239936828613281, | |
| "rewards/margins": 0.23718833923339844, | |
| "rewards/rejected": -0.5611820220947266, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2849295714741643e-06, | |
| "logits/chosen": 0.16771307587623596, | |
| "logits/rejected": 0.30368170142173767, | |
| "logps/chosen": -685.092529296875, | |
| "logps/rejected": -844.0833740234375, | |
| "loss": 0.2314, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.3250967860221863, | |
| "rewards/margins": 0.22138158977031708, | |
| "rewards/rejected": -0.5464783906936646, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2405943438375287e-06, | |
| "logits/chosen": 0.184052973985672, | |
| "logits/rejected": 0.25310075283050537, | |
| "logps/chosen": -602.6265869140625, | |
| "logps/rejected": -813.6980590820312, | |
| "loss": 0.2004, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.30271750688552856, | |
| "rewards/margins": 0.22967958450317383, | |
| "rewards/rejected": -0.5323971509933472, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1960022911279036e-06, | |
| "logits/chosen": 0.16844046115875244, | |
| "logits/rejected": 0.29345783591270447, | |
| "logps/chosen": -665.3712768554688, | |
| "logps/rejected": -814.3455200195312, | |
| "loss": 0.2311, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.3122372329235077, | |
| "rewards/margins": 0.22387078404426575, | |
| "rewards/rejected": -0.5361080169677734, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1511688770836844e-06, | |
| "logits/chosen": 0.13841773569583893, | |
| "logits/rejected": 0.2724143862724304, | |
| "logps/chosen": -630.0308837890625, | |
| "logps/rejected": -790.6541748046875, | |
| "loss": 0.2211, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.3019189238548279, | |
| "rewards/margins": 0.20678682625293732, | |
| "rewards/rejected": -0.508705735206604, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1061096491431307e-06, | |
| "logits/chosen": 0.1345725655555725, | |
| "logits/rejected": 0.23907344043254852, | |
| "logps/chosen": -637.2769775390625, | |
| "logps/rejected": -765.0687255859375, | |
| "loss": 0.2417, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.3158628046512604, | |
| "rewards/margins": 0.18091067671775818, | |
| "rewards/rejected": -0.49677348136901855, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0608402330527796e-06, | |
| "logits/chosen": 0.17173132300376892, | |
| "logits/rejected": 0.27450239658355713, | |
| "logps/chosen": -654.5206909179688, | |
| "logps/rejected": -773.1864624023438, | |
| "loss": 0.2347, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.29153797030448914, | |
| "rewards/margins": 0.18850579857826233, | |
| "rewards/rejected": -0.48004379868507385, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0153763274487176e-06, | |
| "logits/chosen": 0.09793679416179657, | |
| "logits/rejected": 0.21310484409332275, | |
| "logps/chosen": -548.7819213867188, | |
| "logps/rejected": -664.791748046875, | |
| "loss": 0.2495, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.27799323201179504, | |
| "rewards/margins": 0.1372607797384262, | |
| "rewards/rejected": -0.41525402665138245, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9697336984125683e-06, | |
| "logits/chosen": 0.1282065361738205, | |
| "logits/rejected": 0.24735228717327118, | |
| "logps/chosen": -664.9998779296875, | |
| "logps/rejected": -803.4345092773438, | |
| "loss": 0.2076, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.2840580344200134, | |
| "rewards/margins": 0.20879141986370087, | |
| "rewards/rejected": -0.4928494393825531, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.923928174004094e-06, | |
| "logits/chosen": 0.16821298003196716, | |
| "logits/rejected": 0.27364301681518555, | |
| "logps/chosen": -605.760009765625, | |
| "logps/rejected": -756.7783203125, | |
| "loss": 0.1963, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.2758532166481018, | |
| "rewards/margins": 0.2066863477230072, | |
| "rewards/rejected": -0.48253950476646423, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8779756387723036e-06, | |
| "logits/chosen": 0.17568457126617432, | |
| "logits/rejected": 0.29186689853668213, | |
| "logps/chosen": -685.6893310546875, | |
| "logps/rejected": -801.7379760742188, | |
| "loss": 0.2097, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.310352623462677, | |
| "rewards/margins": 0.21010994911193848, | |
| "rewards/rejected": -0.5204626321792603, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.831892028246968e-06, | |
| "logits/chosen": 0.13978341221809387, | |
| "logits/rejected": 0.301819771528244, | |
| "logps/chosen": -684.533447265625, | |
| "logps/rejected": -834.8068237304688, | |
| "loss": 0.1966, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.32525286078453064, | |
| "rewards/margins": 0.2377183884382248, | |
| "rewards/rejected": -0.5629712343215942, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7856933234124617e-06, | |
| "logits/chosen": 0.09446505457162857, | |
| "logits/rejected": 0.23469948768615723, | |
| "logps/chosen": -680.0986938476562, | |
| "logps/rejected": -874.7296752929688, | |
| "loss": 0.2222, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.37397751212120056, | |
| "rewards/margins": 0.21345043182373047, | |
| "rewards/rejected": -0.5874279141426086, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7393955451658387e-06, | |
| "logits/chosen": 0.14745548367500305, | |
| "logits/rejected": 0.24975991249084473, | |
| "logps/chosen": -698.392333984375, | |
| "logps/rejected": -879.1796875, | |
| "loss": 0.2214, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.3602352738380432, | |
| "rewards/margins": 0.23033419251441956, | |
| "rewards/rejected": -0.5905694365501404, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6930147487610667e-06, | |
| "logits/chosen": 0.2092140018939972, | |
| "logits/rejected": 0.27887001633644104, | |
| "logps/chosen": -710.7349853515625, | |
| "logps/rejected": -857.5948486328125, | |
| "loss": 0.2073, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.3808177411556244, | |
| "rewards/margins": 0.20695774257183075, | |
| "rewards/rejected": -0.5877754092216492, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6465670182413487e-06, | |
| "logits/chosen": 0.1725001335144043, | |
| "logits/rejected": 0.29407089948654175, | |
| "logps/chosen": -729.5244750976562, | |
| "logps/rejected": -877.9232177734375, | |
| "loss": 0.2245, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.3700968027114868, | |
| "rewards/margins": 0.21660438179969788, | |
| "rewards/rejected": -0.5867011547088623, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.6000684608614594e-06, | |
| "logits/chosen": 0.2023136168718338, | |
| "logits/rejected": 0.34080618619918823, | |
| "logps/chosen": -720.8736572265625, | |
| "logps/rejected": -897.9036254882812, | |
| "loss": 0.1841, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.35647687315940857, | |
| "rewards/margins": 0.2519916296005249, | |
| "rewards/rejected": -0.6084684133529663, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5535352015020338e-06, | |
| "logits/chosen": 0.19418708980083466, | |
| "logits/rejected": 0.2382660210132599, | |
| "logps/chosen": -742.7463989257812, | |
| "logps/rejected": -869.5545654296875, | |
| "loss": 0.2384, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.38469210267066956, | |
| "rewards/margins": 0.19250501692295074, | |
| "rewards/rejected": -0.5771971344947815, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.506983377077741e-06, | |
| "logits/chosen": 0.14089372754096985, | |
| "logits/rejected": 0.31221604347229004, | |
| "logps/chosen": -706.3193359375, | |
| "logps/rejected": -869.9044799804688, | |
| "loss": 0.2098, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.37956079840660095, | |
| "rewards/margins": 0.2095469981431961, | |
| "rewards/rejected": -0.5891078114509583, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.460429130941289e-06, | |
| "logits/chosen": 0.17403286695480347, | |
| "logits/rejected": 0.26741212606430054, | |
| "logps/chosen": -721.5260620117188, | |
| "logps/rejected": -841.1495971679688, | |
| "loss": 0.2229, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.3337891697883606, | |
| "rewards/margins": 0.21049503982067108, | |
| "rewards/rejected": -0.5442842245101929, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.413888607285192e-06, | |
| "logits/chosen": 0.08936522901058197, | |
| "logits/rejected": 0.208079531788826, | |
| "logps/chosen": -630.3549194335938, | |
| "logps/rejected": -773.2559204101562, | |
| "loss": 0.2153, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.31285005807876587, | |
| "rewards/margins": 0.19155475497245789, | |
| "rewards/rejected": -0.5044047832489014, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.367377945543249e-06, | |
| "logits/chosen": 0.13943253457546234, | |
| "logits/rejected": 0.25722983479499817, | |
| "logps/chosen": -671.1966552734375, | |
| "logps/rejected": -789.2672729492188, | |
| "loss": 0.2079, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.32751041650772095, | |
| "rewards/margins": 0.2003813236951828, | |
| "rewards/rejected": -0.5278916358947754, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.320913274793676e-06, | |
| "logits/chosen": 0.1597217619419098, | |
| "logits/rejected": 0.1886836737394333, | |
| "logps/chosen": -683.5093994140625, | |
| "logps/rejected": -836.6939697265625, | |
| "loss": 0.1907, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.3347502648830414, | |
| "rewards/margins": 0.21815776824951172, | |
| "rewards/rejected": -0.5529080629348755, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.27451070816582e-06, | |
| "logits/chosen": 0.11823856830596924, | |
| "logits/rejected": 0.22079403698444366, | |
| "logps/chosen": -722.5678100585938, | |
| "logps/rejected": -888.1004028320312, | |
| "loss": 0.2454, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.3936742842197418, | |
| "rewards/margins": 0.19485920667648315, | |
| "rewards/rejected": -0.5885335206985474, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.228186337252414e-06, | |
| "logits/chosen": 0.12982912361621857, | |
| "logits/rejected": 0.2598816454410553, | |
| "logps/chosen": -726.23291015625, | |
| "logps/rejected": -867.5569458007812, | |
| "loss": 0.2153, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.3581377863883972, | |
| "rewards/margins": 0.21532902121543884, | |
| "rewards/rejected": -0.5734668970108032, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1819562265292946e-06, | |
| "logits/chosen": 0.12636715173721313, | |
| "logits/rejected": 0.23508496582508087, | |
| "logps/chosen": -683.2789306640625, | |
| "logps/rejected": -821.7802734375, | |
| "loss": 0.2059, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.32184451818466187, | |
| "rewards/margins": 0.22134391963481903, | |
| "rewards/rejected": -0.5431884527206421, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1358364077845236e-06, | |
| "logits/chosen": 0.11493659019470215, | |
| "logits/rejected": 0.25907325744628906, | |
| "logps/chosen": -688.8036499023438, | |
| "logps/rejected": -853.0017700195312, | |
| "loss": 0.1887, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.3468135893344879, | |
| "rewards/margins": 0.22318044304847717, | |
| "rewards/rejected": -0.5699939727783203, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.089842874558849e-06, | |
| "logits/chosen": 0.15250439941883087, | |
| "logits/rejected": 0.25078490376472473, | |
| "logps/chosen": -663.3401489257812, | |
| "logps/rejected": -840.7469482421875, | |
| "loss": 0.2285, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.36832159757614136, | |
| "rewards/margins": 0.2164924144744873, | |
| "rewards/rejected": -0.5848139524459839, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0439915765994242e-06, | |
| "logits/chosen": 0.15975138545036316, | |
| "logits/rejected": 0.2446996420621872, | |
| "logps/chosen": -686.4671020507812, | |
| "logps/rejected": -884.5764770507812, | |
| "loss": 0.1993, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.357178658246994, | |
| "rewards/margins": 0.2013837844133377, | |
| "rewards/rejected": -0.5585624575614929, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.9982984143287186e-06, | |
| "logits/chosen": 0.1569955199956894, | |
| "logits/rejected": 0.2909889817237854, | |
| "logps/chosen": -670.5752563476562, | |
| "logps/rejected": -830.7100830078125, | |
| "loss": 0.2121, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.3219209909439087, | |
| "rewards/margins": 0.2229768931865692, | |
| "rewards/rejected": -0.5448979139328003, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.95277923333053e-06, | |
| "logits/chosen": 0.16851790249347687, | |
| "logits/rejected": 0.27851399779319763, | |
| "logps/chosen": -673.5487060546875, | |
| "logps/rejected": -770.0308837890625, | |
| "loss": 0.2135, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.2954522967338562, | |
| "rewards/margins": 0.1889936327934265, | |
| "rewards/rejected": -0.4844458997249603, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.9074498188550156e-06, | |
| "logits/chosen": 0.12967072427272797, | |
| "logits/rejected": 0.20891770720481873, | |
| "logps/chosen": -606.8577880859375, | |
| "logps/rejected": -735.3054809570312, | |
| "loss": 0.2069, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.26535481214523315, | |
| "rewards/margins": 0.20063455402851105, | |
| "rewards/rejected": -0.465989351272583, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.862325890344643e-06, | |
| "logits/chosen": 0.1516229659318924, | |
| "logits/rejected": 0.2716851532459259, | |
| "logps/chosen": -612.8181762695312, | |
| "logps/rejected": -756.4715576171875, | |
| "loss": 0.2025, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.28186386823654175, | |
| "rewards/margins": 0.21003735065460205, | |
| "rewards/rejected": -0.4919012188911438, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.817423095982972e-06, | |
| "logits/chosen": 0.1507120281457901, | |
| "logits/rejected": 0.29278236627578735, | |
| "logps/chosen": -704.4903564453125, | |
| "logps/rejected": -869.5260009765625, | |
| "loss": 0.2119, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.3547202944755554, | |
| "rewards/margins": 0.2284606695175171, | |
| "rewards/rejected": -0.5831810235977173, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.7727570072681293e-06, | |
| "logits/chosen": 0.2021058052778244, | |
| "logits/rejected": 0.2638623118400574, | |
| "logps/chosen": -692.8404541015625, | |
| "logps/rejected": -797.3179321289062, | |
| "loss": 0.2564, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.354163259267807, | |
| "rewards/margins": 0.18512588739395142, | |
| "rewards/rejected": -0.5392891764640808, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7283431136128961e-06, | |
| "logits/chosen": 0.1298045814037323, | |
| "logits/rejected": 0.24017031490802765, | |
| "logps/chosen": -702.1468505859375, | |
| "logps/rejected": -854.5265502929688, | |
| "loss": 0.2064, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.3338576555252075, | |
| "rewards/margins": 0.22743964195251465, | |
| "rewards/rejected": -0.5612972974777222, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6841968169732478e-06, | |
| "logits/chosen": 0.20339742302894592, | |
| "logits/rejected": 0.24962684512138367, | |
| "logps/chosen": -729.5333862304688, | |
| "logps/rejected": -869.4794921875, | |
| "loss": 0.2122, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.34650754928588867, | |
| "rewards/margins": 0.223586767911911, | |
| "rewards/rejected": -0.5700943470001221, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6403334265072284e-06, | |
| "logits/chosen": 0.1773589849472046, | |
| "logits/rejected": 0.29130274057388306, | |
| "logps/chosen": -689.1021728515625, | |
| "logps/rejected": -871.9329223632812, | |
| "loss": 0.1625, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.32862764596939087, | |
| "rewards/margins": 0.2522730827331543, | |
| "rewards/rejected": -0.5809007287025452, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5967681532660066e-06, | |
| "logits/chosen": 0.08677888661623001, | |
| "logits/rejected": 0.2043531835079193, | |
| "logps/chosen": -683.0452270507812, | |
| "logps/rejected": -853.3363037109375, | |
| "loss": 0.1967, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.3488999605178833, | |
| "rewards/margins": 0.2315068542957306, | |
| "rewards/rejected": -0.5804067850112915, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5535161049189463e-06, | |
| "logits/chosen": 0.12979058921337128, | |
| "logits/rejected": 0.24603179097175598, | |
| "logps/chosen": -737.0632934570312, | |
| "logps/rejected": -886.8776245117188, | |
| "loss": 0.2403, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.39016053080558777, | |
| "rewards/margins": 0.20638911426067352, | |
| "rewards/rejected": -0.5965496301651001, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.5105922805145356e-06, | |
| "logits/chosen": 0.15358899533748627, | |
| "logits/rejected": 0.32385343313217163, | |
| "logps/chosen": -662.0921630859375, | |
| "logps/rejected": -823.0596923828125, | |
| "loss": 0.2037, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.31596821546554565, | |
| "rewards/margins": 0.2252482920885086, | |
| "rewards/rejected": -0.5412164926528931, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.4680115652789823e-06, | |
| "logits/chosen": 0.14214789867401123, | |
| "logits/rejected": 0.25008469820022583, | |
| "logps/chosen": -613.7877197265625, | |
| "logps/rejected": -783.76025390625, | |
| "loss": 0.2255, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.3245735764503479, | |
| "rewards/margins": 0.18011733889579773, | |
| "rewards/rejected": -0.504690945148468, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4257887254542767e-06, | |
| "logits/chosen": 0.14355778694152832, | |
| "logits/rejected": 0.24925288558006287, | |
| "logps/chosen": -668.3030395507812, | |
| "logps/rejected": -847.2467041015625, | |
| "loss": 0.1859, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.30989786982536316, | |
| "rewards/margins": 0.23173482716083527, | |
| "rewards/rejected": -0.5416327118873596, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.3839384031775227e-06, | |
| "logits/chosen": 0.16522178053855896, | |
| "logits/rejected": 0.24320781230926514, | |
| "logps/chosen": -641.2236938476562, | |
| "logps/rejected": -764.9053344726562, | |
| "loss": 0.2218, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.3104414641857147, | |
| "rewards/margins": 0.18151786923408508, | |
| "rewards/rejected": -0.4919593334197998, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.342475111403298e-06, | |
| "logits/chosen": 0.10045696794986725, | |
| "logits/rejected": 0.22447367012500763, | |
| "logps/chosen": -676.05419921875, | |
| "logps/rejected": -803.32763671875, | |
| "loss": 0.2208, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.3212507665157318, | |
| "rewards/margins": 0.21263284981250763, | |
| "rewards/rejected": -0.5338836312294006, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3014132288708209e-06, | |
| "logits/chosen": 0.158107191324234, | |
| "logits/rejected": 0.19983355700969696, | |
| "logps/chosen": -619.661865234375, | |
| "logps/rejected": -794.2459106445312, | |
| "loss": 0.2081, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.30042386054992676, | |
| "rewards/margins": 0.20652098953723907, | |
| "rewards/rejected": -0.5069448947906494, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2607669951176549e-06, | |
| "logits/chosen": 0.13469955325126648, | |
| "logits/rejected": 0.26009881496429443, | |
| "logps/chosen": -700.5491943359375, | |
| "logps/rejected": -826.0719604492188, | |
| "loss": 0.208, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.3267166316509247, | |
| "rewards/margins": 0.2262444794178009, | |
| "rewards/rejected": -0.5529611706733704, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2205505055416891e-06, | |
| "logits/chosen": 0.12177290767431259, | |
| "logits/rejected": 0.23323854804039001, | |
| "logps/chosen": -655.70849609375, | |
| "logps/rejected": -841.9342041015625, | |
| "loss": 0.1949, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.3228956162929535, | |
| "rewards/margins": 0.23238658905029297, | |
| "rewards/rejected": -0.5552822947502136, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1807777065131002e-06, | |
| "logits/chosen": 0.13625140488147736, | |
| "logits/rejected": 0.2116333544254303, | |
| "logps/chosen": -680.4044799804688, | |
| "logps/rejected": -844.9688720703125, | |
| "loss": 0.2145, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.3188444972038269, | |
| "rewards/margins": 0.22850990295410156, | |
| "rewards/rejected": -0.5473544001579285, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1414623905380012e-06, | |
| "logits/chosen": 0.10103823244571686, | |
| "logits/rejected": 0.19432368874549866, | |
| "logps/chosen": -624.4705810546875, | |
| "logps/rejected": -799.0113525390625, | |
| "loss": 0.2225, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.2993714213371277, | |
| "rewards/margins": 0.21414069831371307, | |
| "rewards/rejected": -0.513512134552002, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.1026181914754388e-06, | |
| "logits/chosen": 0.16599547863006592, | |
| "logits/rejected": 0.23224572837352753, | |
| "logps/chosen": -686.8082275390625, | |
| "logps/rejected": -823.6770629882812, | |
| "loss": 0.2035, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.30449575185775757, | |
| "rewards/margins": 0.21346712112426758, | |
| "rewards/rejected": -0.5179629325866699, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0642585798094136e-06, | |
| "logits/chosen": 0.11574041843414307, | |
| "logits/rejected": 0.1864078789949417, | |
| "logps/chosen": -661.083984375, | |
| "logps/rejected": -803.5848999023438, | |
| "loss": 0.2521, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.3295978903770447, | |
| "rewards/margins": 0.16920098662376404, | |
| "rewards/rejected": -0.4987989068031311, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0263968579775522e-06, | |
| "logits/chosen": 0.1372244954109192, | |
| "logits/rejected": 0.2742239534854889, | |
| "logps/chosen": -652.0556640625, | |
| "logps/rejected": -803.6484375, | |
| "loss": 0.2306, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.30462098121643066, | |
| "rewards/margins": 0.20189952850341797, | |
| "rewards/rejected": -0.5065204501152039, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.89046155758058e-07, | |
| "logits/chosen": 0.18852971494197845, | |
| "logits/rejected": 0.26183754205703735, | |
| "logps/chosen": -696.4068603515625, | |
| "logps/rejected": -802.4880981445312, | |
| "loss": 0.2703, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.32309865951538086, | |
| "rewards/margins": 0.1942833513021469, | |
| "rewards/rejected": -0.517382025718689, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.52219425716534e-07, | |
| "logits/chosen": 0.18374116718769073, | |
| "logits/rejected": 0.23104730248451233, | |
| "logps/chosen": -661.356201171875, | |
| "logps/rejected": -787.6057739257812, | |
| "loss": 0.224, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.30247530341148376, | |
| "rewards/margins": 0.20218057930469513, | |
| "rewards/rejected": -0.5046559572219849, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 9.15929438714262e-07, | |
| "logits/chosen": 0.18678037822246552, | |
| "logits/rejected": 0.3186245560646057, | |
| "logps/chosen": -632.1265869140625, | |
| "logps/rejected": -761.0514526367188, | |
| "loss": 0.2275, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.2909912168979645, | |
| "rewards/margins": 0.19340907037258148, | |
| "rewards/rejected": -0.48440033197402954, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.801887794794911e-07, | |
| "logits/chosen": 0.14555346965789795, | |
| "logits/rejected": 0.2179187834262848, | |
| "logps/chosen": -666.7107543945312, | |
| "logps/rejected": -776.0972900390625, | |
| "loss": 0.2247, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.3142651319503784, | |
| "rewards/margins": 0.18122203648090363, | |
| "rewards/rejected": -0.4954872131347656, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.450098422432787e-07, | |
| "logits/chosen": 0.10284000635147095, | |
| "logits/rejected": 0.24823495745658875, | |
| "logps/chosen": -654.6766357421875, | |
| "logps/rejected": -778.5437622070312, | |
| "loss": 0.2153, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.289842426776886, | |
| "rewards/margins": 0.20343871414661407, | |
| "rewards/rejected": -0.49328118562698364, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.104048264413858e-07, | |
| "logits/chosen": 0.16312912106513977, | |
| "logits/rejected": 0.21963107585906982, | |
| "logps/chosen": -632.3480834960938, | |
| "logps/rejected": -775.2318115234375, | |
| "loss": 0.2299, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.30911892652511597, | |
| "rewards/margins": 0.1682128757238388, | |
| "rewards/rejected": -0.4773317873477936, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.763857324837321e-07, | |
| "logits/chosen": 0.14603514969348907, | |
| "logits/rejected": 0.25196802616119385, | |
| "logps/chosen": -642.275390625, | |
| "logps/rejected": -800.051513671875, | |
| "loss": 0.234, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.3225029408931732, | |
| "rewards/margins": 0.20271165668964386, | |
| "rewards/rejected": -0.5252145528793335, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.429643575928605e-07, | |
| "logits/chosen": 0.1552133560180664, | |
| "logits/rejected": 0.2241334468126297, | |
| "logps/chosen": -628.1038818359375, | |
| "logps/rejected": -822.3765869140625, | |
| "loss": 0.1891, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.3023911118507385, | |
| "rewards/margins": 0.23453156650066376, | |
| "rewards/rejected": -0.5369226932525635, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.101522917128709e-07, | |
| "logits/chosen": 0.15086951851844788, | |
| "logits/rejected": 0.25160127878189087, | |
| "logps/chosen": -729.95947265625, | |
| "logps/rejected": -846.2911987304688, | |
| "loss": 0.1995, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.3416779637336731, | |
| "rewards/margins": 0.20796218514442444, | |
| "rewards/rejected": -0.5496400594711304, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.779609134902312e-07, | |
| "logits/chosen": 0.13389447331428528, | |
| "logits/rejected": 0.1718115508556366, | |
| "logps/chosen": -662.325927734375, | |
| "logps/rejected": -822.61181640625, | |
| "loss": 0.2358, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.3139615058898926, | |
| "rewards/margins": 0.21937116980552673, | |
| "rewards/rejected": -0.5333326458930969, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.464013863278629e-07, | |
| "logits/chosen": 0.13355755805969238, | |
| "logits/rejected": 0.24505715072155, | |
| "logps/chosen": -674.4129638671875, | |
| "logps/rejected": -807.907470703125, | |
| "loss": 0.2259, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.3266414999961853, | |
| "rewards/margins": 0.20114055275917053, | |
| "rewards/rejected": -0.5277820825576782, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.154846545138696e-07, | |
| "logits/chosen": 0.17101332545280457, | |
| "logits/rejected": 0.24388432502746582, | |
| "logps/chosen": -667.0880126953125, | |
| "logps/rejected": -843.1787109375, | |
| "loss": 0.2001, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.31553584337234497, | |
| "rewards/margins": 0.22510071098804474, | |
| "rewards/rejected": -0.5406365394592285, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.852214394262515e-07, | |
| "logits/chosen": 0.1403154879808426, | |
| "logits/rejected": 0.1944664716720581, | |
| "logps/chosen": -611.7420654296875, | |
| "logps/rejected": -802.8869018554688, | |
| "loss": 0.1929, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.3096386790275574, | |
| "rewards/margins": 0.22140173614025116, | |
| "rewards/rejected": -0.531040370464325, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.556222358149191e-07, | |
| "logits/chosen": 0.13102367520332336, | |
| "logits/rejected": 0.20809051394462585, | |
| "logps/chosen": -669.8289794921875, | |
| "logps/rejected": -821.7249145507812, | |
| "loss": 0.2066, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.2995157837867737, | |
| "rewards/margins": 0.2382737696170807, | |
| "rewards/rejected": -0.5377895832061768, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.266973081622992e-07, | |
| "logits/chosen": 0.16256758570671082, | |
| "logits/rejected": 0.2884465157985687, | |
| "logps/chosen": -657.072021484375, | |
| "logps/rejected": -821.6876220703125, | |
| "loss": 0.2189, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.30884090065956116, | |
| "rewards/margins": 0.22088280320167542, | |
| "rewards/rejected": -0.5297237634658813, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.984566871237942e-07, | |
| "logits/chosen": 0.1519722193479538, | |
| "logits/rejected": 0.2986641526222229, | |
| "logps/chosen": -695.6591796875, | |
| "logps/rejected": -865.2161865234375, | |
| "loss": 0.2069, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.33254241943359375, | |
| "rewards/margins": 0.23650801181793213, | |
| "rewards/rejected": -0.5690504312515259, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.709101660493251e-07, | |
| "logits/chosen": 0.1198592409491539, | |
| "logits/rejected": 0.24108009040355682, | |
| "logps/chosen": -657.8306274414062, | |
| "logps/rejected": -859.29833984375, | |
| "loss": 0.2173, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.31234902143478394, | |
| "rewards/margins": 0.2416820079088211, | |
| "rewards/rejected": -0.5540310740470886, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.440672975871743e-07, | |
| "logits/chosen": 0.15342529118061066, | |
| "logits/rejected": 0.21858489513397217, | |
| "logps/chosen": -640.5154418945312, | |
| "logps/rejected": -774.4490966796875, | |
| "loss": 0.2209, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.3267766833305359, | |
| "rewards/margins": 0.19230851531028748, | |
| "rewards/rejected": -0.519085168838501, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.1793739037129134e-07, | |
| "logits/chosen": 0.18993167579174042, | |
| "logits/rejected": 0.26399049162864685, | |
| "logps/chosen": -640.939453125, | |
| "logps/rejected": -825.0241088867188, | |
| "loss": 0.2353, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.3261680603027344, | |
| "rewards/margins": 0.19224853813648224, | |
| "rewards/rejected": -0.5184165835380554, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.9252950579322405e-07, | |
| "logits/chosen": 0.148400217294693, | |
| "logits/rejected": 0.22420552372932434, | |
| "logps/chosen": -638.9559936523438, | |
| "logps/rejected": -800.7708129882812, | |
| "loss": 0.1952, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.30330848693847656, | |
| "rewards/margins": 0.22106847167015076, | |
| "rewards/rejected": -0.5243769884109497, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.6785245485978864e-07, | |
| "logits/chosen": 0.18423089385032654, | |
| "logits/rejected": 0.2459380328655243, | |
| "logps/chosen": -680.0947875976562, | |
| "logps/rejected": -856.1446533203125, | |
| "loss": 0.2004, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.3225480914115906, | |
| "rewards/margins": 0.2401798963546753, | |
| "rewards/rejected": -0.5627279281616211, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.43914795137566e-07, | |
| "logits/chosen": 0.1533311903476715, | |
| "logits/rejected": 0.20948585867881775, | |
| "logps/chosen": -651.2598876953125, | |
| "logps/rejected": -795.5108032226562, | |
| "loss": 0.2182, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.3288353979587555, | |
| "rewards/margins": 0.18383339047431946, | |
| "rewards/rejected": -0.512668788433075, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.207248277852901e-07, | |
| "logits/chosen": 0.1523989737033844, | |
| "logits/rejected": 0.24100735783576965, | |
| "logps/chosen": -618.7085571289062, | |
| "logps/rejected": -805.9244995117188, | |
| "loss": 0.1826, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.29542338848114014, | |
| "rewards/margins": 0.22522738575935364, | |
| "rewards/rejected": -0.5206507444381714, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.9829059467515074e-07, | |
| "logits/chosen": 0.1278158277273178, | |
| "logits/rejected": 0.22139263153076172, | |
| "logps/chosen": -732.3717041015625, | |
| "logps/rejected": -855.8250732421875, | |
| "loss": 0.2259, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.3528819680213928, | |
| "rewards/margins": 0.20918047428131104, | |
| "rewards/rejected": -0.5620623826980591, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.766198756040153e-07, | |
| "logits/chosen": 0.14292459189891815, | |
| "logits/rejected": 0.30465906858444214, | |
| "logps/chosen": -667.1749267578125, | |
| "logps/rejected": -841.5711669921875, | |
| "loss": 0.1826, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.3154456317424774, | |
| "rewards/margins": 0.2273203581571579, | |
| "rewards/rejected": -0.5427660346031189, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.5572018559553155e-07, | |
| "logits/chosen": 0.16036184132099152, | |
| "logits/rejected": 0.2627103924751282, | |
| "logps/chosen": -683.5515747070312, | |
| "logps/rejected": -796.56787109375, | |
| "loss": 0.2053, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.32863935828208923, | |
| "rewards/margins": 0.201740100979805, | |
| "rewards/rejected": -0.5303794741630554, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.3559877229404864e-07, | |
| "logits/chosen": 0.10556507110595703, | |
| "logits/rejected": 0.19262662529945374, | |
| "logps/chosen": -708.3388671875, | |
| "logps/rejected": -892.02197265625, | |
| "loss": 0.2239, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.34648841619491577, | |
| "rewards/margins": 0.2360268384218216, | |
| "rewards/rejected": -0.582515299320221, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.1626261345126576e-07, | |
| "logits/chosen": 0.13623256981372833, | |
| "logits/rejected": 0.2520672678947449, | |
| "logps/chosen": -698.7113037109375, | |
| "logps/rejected": -843.6485595703125, | |
| "loss": 0.2044, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.350619375705719, | |
| "rewards/margins": 0.2033306360244751, | |
| "rewards/rejected": -0.5539500117301941, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.9771841450646505e-07, | |
| "logits/chosen": 0.12394211441278458, | |
| "logits/rejected": 0.255685418844223, | |
| "logps/chosen": -701.68017578125, | |
| "logps/rejected": -891.2081909179688, | |
| "loss": 0.2145, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.3482310175895691, | |
| "rewards/margins": 0.23429682850837708, | |
| "rewards/rejected": -0.5825279355049133, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.7997260626118758e-07, | |
| "logits/chosen": 0.1238374263048172, | |
| "logits/rejected": 0.23963268101215363, | |
| "logps/chosen": -683.5324096679688, | |
| "logps/rejected": -842.0015869140625, | |
| "loss": 0.2257, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.31508007645606995, | |
| "rewards/margins": 0.2424265444278717, | |
| "rewards/rejected": -0.5575065612792969, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.6303134264914365e-07, | |
| "logits/chosen": 0.1542249619960785, | |
| "logits/rejected": 0.26766401529312134, | |
| "logps/chosen": -751.26171875, | |
| "logps/rejected": -901.7872924804688, | |
| "loss": 0.1951, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.35173672437667847, | |
| "rewards/margins": 0.22757640480995178, | |
| "rewards/rejected": -0.5793130993843079, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.469004986021355e-07, | |
| "logits/chosen": 0.12692458927631378, | |
| "logits/rejected": 0.2118302285671234, | |
| "logps/chosen": -675.017578125, | |
| "logps/rejected": -829.9061279296875, | |
| "loss": 0.2303, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.33988887071609497, | |
| "rewards/margins": 0.19631198048591614, | |
| "rewards/rejected": -0.5362008810043335, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.315856680127367e-07, | |
| "logits/chosen": 0.11164776980876923, | |
| "logits/rejected": 0.21648769080638885, | |
| "logps/chosen": -657.6282958984375, | |
| "logps/rejected": -799.0164794921875, | |
| "loss": 0.192, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.32347801327705383, | |
| "rewards/margins": 0.21190226078033447, | |
| "rewards/rejected": -0.5353802442550659, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1709216179442817e-07, | |
| "logits/chosen": 0.0844898596405983, | |
| "logits/rejected": 0.2031726837158203, | |
| "logps/chosen": -725.109619140625, | |
| "logps/rejected": -869.9738159179688, | |
| "loss": 0.2053, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.35323506593704224, | |
| "rewards/margins": 0.2272004634141922, | |
| "rewards/rejected": -0.5804354548454285, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 1.0342500603986421e-07, | |
| "logits/chosen": 0.16333475708961487, | |
| "logits/rejected": 0.22125931084156036, | |
| "logps/chosen": -669.9171752929688, | |
| "logps/rejected": -802.5113525390625, | |
| "loss": 0.232, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.3276098072528839, | |
| "rewards/margins": 0.21911415457725525, | |
| "rewards/rejected": -0.5467239618301392, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.058894027791643e-08, | |
| "logits/chosen": 0.15154321491718292, | |
| "logits/rejected": 0.252513587474823, | |
| "logps/chosen": -668.7503662109375, | |
| "logps/rejected": -801.6619262695312, | |
| "loss": 0.2252, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.33079949021339417, | |
| "rewards/margins": 0.20475859940052032, | |
| "rewards/rejected": -0.5355581045150757, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.858841583008592e-08, | |
| "logits/chosen": 0.1299131065607071, | |
| "logits/rejected": 0.2569180130958557, | |
| "logps/chosen": -691.34228515625, | |
| "logps/rejected": -851.5472412109375, | |
| "loss": 0.1733, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.3316085636615753, | |
| "rewards/margins": 0.2289518564939499, | |
| "rewards/rejected": -0.5605604648590088, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 6.742759426686313e-08, | |
| "logits/chosen": 0.15428964793682098, | |
| "logits/rejected": 0.30156534910202026, | |
| "logps/chosen": -670.6358642578125, | |
| "logps/rejected": -829.8116455078125, | |
| "loss": 0.2113, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.33194589614868164, | |
| "rewards/margins": 0.22803179919719696, | |
| "rewards/rejected": -0.5599777102470398, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.7110345964571104e-08, | |
| "logits/chosen": 0.1505686342716217, | |
| "logits/rejected": 0.26538941264152527, | |
| "logps/chosen": -700.7206420898438, | |
| "logps/rejected": -845.8724365234375, | |
| "loss": 0.2174, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.33285072445869446, | |
| "rewards/margins": 0.22789886593818665, | |
| "rewards/rejected": -0.5607496500015259, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.764024876318357e-08, | |
| "logits/chosen": 0.14582636952400208, | |
| "logits/rejected": 0.210880845785141, | |
| "logps/chosen": -675.3494262695312, | |
| "logps/rejected": -822.8538818359375, | |
| "loss": 0.1898, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.3319861888885498, | |
| "rewards/margins": 0.21374984085559845, | |
| "rewards/rejected": -0.5457359552383423, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.902058672559633e-08, | |
| "logits/chosen": 0.09690573811531067, | |
| "logits/rejected": 0.21973006427288055, | |
| "logps/chosen": -667.906005859375, | |
| "logps/rejected": -875.2356567382812, | |
| "loss": 0.1945, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.3273436725139618, | |
| "rewards/margins": 0.24747705459594727, | |
| "rewards/rejected": -0.5748206377029419, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.125434899876933e-08, | |
| "logits/chosen": 0.08470339328050613, | |
| "logits/rejected": 0.23671475052833557, | |
| "logps/chosen": -656.4951171875, | |
| "logps/rejected": -821.2351684570312, | |
| "loss": 0.2145, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.32644474506378174, | |
| "rewards/margins": 0.212808296084404, | |
| "rewards/rejected": -0.5392531156539917, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.4344228777145873e-08, | |
| "logits/chosen": 0.09538368880748749, | |
| "logits/rejected": 0.21011042594909668, | |
| "logps/chosen": -703.3250732421875, | |
| "logps/rejected": -883.3131713867188, | |
| "loss": 0.1995, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.33553892374038696, | |
| "rewards/margins": 0.2493852823972702, | |
| "rewards/rejected": -0.5849242210388184, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.829262236869772e-08, | |
| "logits/chosen": 0.1573551446199417, | |
| "logits/rejected": 0.2588816285133362, | |
| "logps/chosen": -683.0828247070312, | |
| "logps/rejected": -843.8508911132812, | |
| "loss": 0.2083, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.30922335386276245, | |
| "rewards/margins": 0.24052074551582336, | |
| "rewards/rejected": -0.5497440695762634, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.3101628363929586e-08, | |
| "logits/chosen": 0.09812667965888977, | |
| "logits/rejected": 0.21138449013233185, | |
| "logps/chosen": -694.7268676757812, | |
| "logps/rejected": -886.9791870117188, | |
| "loss": 0.2094, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.34521687030792236, | |
| "rewards/margins": 0.23157024383544922, | |
| "rewards/rejected": -0.5767871141433716, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.773046908123195e-09, | |
| "logits/chosen": 0.1276710033416748, | |
| "logits/rejected": 0.21167734265327454, | |
| "logps/chosen": -673.7769775390625, | |
| "logps/rejected": -854.2696533203125, | |
| "loss": 0.1992, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.3254976272583008, | |
| "rewards/margins": 0.2494417428970337, | |
| "rewards/rejected": -0.5749393701553345, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 5.308379077080817e-09, | |
| "logits/chosen": 0.11343555152416229, | |
| "logits/rejected": 0.25598517060279846, | |
| "logps/chosen": -700.44921875, | |
| "logps/rejected": -852.5792846679688, | |
| "loss": 0.2213, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -0.3366733491420746, | |
| "rewards/margins": 0.21852286159992218, | |
| "rewards/rejected": -0.555196225643158, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2.7088263565760996e-09, | |
| "logits/chosen": 0.15132495760917664, | |
| "logits/rejected": 0.24779316782951355, | |
| "logps/chosen": -710.2337646484375, | |
| "logps/rejected": -898.6566162109375, | |
| "loss": 0.2162, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.3569861650466919, | |
| "rewards/margins": 0.23351311683654785, | |
| "rewards/rejected": -0.5904992818832397, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 9.752902257023633e-10, | |
| "logits/chosen": 0.1191195398569107, | |
| "logits/rejected": 0.2252669632434845, | |
| "logps/chosen": -693.7659912109375, | |
| "logps/rejected": -813.2882690429688, | |
| "loss": 0.239, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.34295856952667236, | |
| "rewards/margins": 0.19307160377502441, | |
| "rewards/rejected": -0.5360301733016968, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.083718442532189e-10, | |
| "logits/chosen": 0.10991505533456802, | |
| "logits/rejected": 0.2280399352312088, | |
| "logps/chosen": -677.8359985351562, | |
| "logps/rejected": -832.2960815429688, | |
| "loss": 0.234, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.34933480620384216, | |
| "rewards/margins": 0.19008032977581024, | |
| "rewards/rejected": -0.539415180683136, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1875, | |
| "total_flos": 0.0, | |
| "train_loss": 0.2330256977081299, | |
| "train_runtime": 17175.1012, | |
| "train_samples_per_second": 1.747, | |
| "train_steps_per_second": 0.109 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1875, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |