| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9982631930527722, | |
| "eval_steps": 400, | |
| "global_step": 467, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01068804275217101, | |
| "grad_norm": 84.02989783534706, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -1.0397655963897705, | |
| "logits/rejected": -1.0092562437057495, | |
| "logps/chosen": -0.7628876566886902, | |
| "logps/rejected": -0.7414335012435913, | |
| "loss": 4.8657, | |
| "rewards/accuracies": 0.3687500059604645, | |
| "rewards/chosen": -7.628875732421875, | |
| "rewards/margins": -0.21454186737537384, | |
| "rewards/rejected": -7.414334774017334, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02137608550434202, | |
| "grad_norm": 65.98135057247741, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -1.0671762228012085, | |
| "logits/rejected": -0.9972389340400696, | |
| "logps/chosen": -0.7587485909461975, | |
| "logps/rejected": -0.7000675201416016, | |
| "loss": 4.7692, | |
| "rewards/accuracies": 0.41874998807907104, | |
| "rewards/chosen": -7.587485313415527, | |
| "rewards/margins": -0.5868101119995117, | |
| "rewards/rejected": -7.000675201416016, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03206412825651302, | |
| "grad_norm": 81.81057773129896, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -0.9940463900566101, | |
| "logits/rejected": -1.014291524887085, | |
| "logps/chosen": -0.7531558871269226, | |
| "logps/rejected": -0.7980529069900513, | |
| "loss": 5.0125, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -7.531558990478516, | |
| "rewards/margins": 0.44897016882896423, | |
| "rewards/rejected": -7.980528831481934, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04275217100868404, | |
| "grad_norm": 131.3611224894929, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -0.9941411018371582, | |
| "logits/rejected": -0.9676485061645508, | |
| "logps/chosen": -0.7748786807060242, | |
| "logps/rejected": -0.7950050234794617, | |
| "loss": 4.3887, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.748786926269531, | |
| "rewards/margins": 0.20126314461231232, | |
| "rewards/rejected": -7.950050354003906, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053440213760855046, | |
| "grad_norm": 41.98227986443809, | |
| "learning_rate": 5.319148936170212e-07, | |
| "logits/chosen": -1.06044602394104, | |
| "logits/rejected": -1.0300180912017822, | |
| "logps/chosen": -0.5242100358009338, | |
| "logps/rejected": -0.5184083580970764, | |
| "loss": 4.4531, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -5.242100715637207, | |
| "rewards/margins": -0.05801659822463989, | |
| "rewards/rejected": -5.184083461761475, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06412825651302605, | |
| "grad_norm": 68.84250778408698, | |
| "learning_rate": 6.382978723404255e-07, | |
| "logits/chosen": -1.0604535341262817, | |
| "logits/rejected": -1.0147985219955444, | |
| "logps/chosen": -0.5427820086479187, | |
| "logps/rejected": -0.538798451423645, | |
| "loss": 4.0481, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": -5.427820205688477, | |
| "rewards/margins": -0.03983556479215622, | |
| "rewards/rejected": -5.3879852294921875, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07481629926519706, | |
| "grad_norm": 55.83166407300918, | |
| "learning_rate": 7.446808510638297e-07, | |
| "logits/chosen": -1.1716325283050537, | |
| "logits/rejected": -1.0896761417388916, | |
| "logps/chosen": -0.45515695214271545, | |
| "logps/rejected": -0.46327242255210876, | |
| "loss": 3.8398, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -4.55156946182251, | |
| "rewards/margins": 0.08115490525960922, | |
| "rewards/rejected": -4.632723808288574, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08550434201736808, | |
| "grad_norm": 47.314667796032154, | |
| "learning_rate": 8.51063829787234e-07, | |
| "logits/chosen": -1.1367859840393066, | |
| "logits/rejected": -1.0874762535095215, | |
| "logps/chosen": -0.3599122166633606, | |
| "logps/rejected": -0.4211342930793762, | |
| "loss": 3.8501, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -3.5991222858428955, | |
| "rewards/margins": 0.6122205853462219, | |
| "rewards/rejected": -4.211342811584473, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09619238476953908, | |
| "grad_norm": 43.42431109758399, | |
| "learning_rate": 9.574468085106384e-07, | |
| "logits/chosen": -1.17836332321167, | |
| "logits/rejected": -1.1281588077545166, | |
| "logps/chosen": -0.377611368894577, | |
| "logps/rejected": -0.42245084047317505, | |
| "loss": 3.7936, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -3.7761130332946777, | |
| "rewards/margins": 0.448394775390625, | |
| "rewards/rejected": -4.224508762359619, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10688042752171009, | |
| "grad_norm": 50.23949059041095, | |
| "learning_rate": 9.998741174712533e-07, | |
| "logits/chosen": -1.1873140335083008, | |
| "logits/rejected": -1.1317476034164429, | |
| "logps/chosen": -0.34768182039260864, | |
| "logps/rejected": -0.38563376665115356, | |
| "loss": 3.7644, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -3.476818561553955, | |
| "rewards/margins": 0.3795194625854492, | |
| "rewards/rejected": -3.856337785720825, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11756847027388109, | |
| "grad_norm": 77.86452279542212, | |
| "learning_rate": 9.991050648838675e-07, | |
| "logits/chosen": -1.2336595058441162, | |
| "logits/rejected": -1.1919711828231812, | |
| "logps/chosen": -0.33353298902511597, | |
| "logps/rejected": -0.3969067931175232, | |
| "loss": 3.6758, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -3.3353302478790283, | |
| "rewards/margins": 0.6337377429008484, | |
| "rewards/rejected": -3.9690680503845215, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1282565130260521, | |
| "grad_norm": 48.54307450778913, | |
| "learning_rate": 9.97637968732563e-07, | |
| "logits/chosen": -1.2853938341140747, | |
| "logits/rejected": -1.2468154430389404, | |
| "logps/chosen": -0.3357910215854645, | |
| "logps/rejected": -0.36223435401916504, | |
| "loss": 3.6487, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -3.35791015625, | |
| "rewards/margins": 0.26443368196487427, | |
| "rewards/rejected": -3.6223437786102295, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13894455577822312, | |
| "grad_norm": 67.81573981196492, | |
| "learning_rate": 9.954748808839674e-07, | |
| "logits/chosen": -1.203680396080017, | |
| "logits/rejected": -1.1713194847106934, | |
| "logps/chosen": -0.3974061906337738, | |
| "logps/rejected": -0.4787389636039734, | |
| "loss": 3.5226, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -3.9740614891052246, | |
| "rewards/margins": 0.813327431678772, | |
| "rewards/rejected": -4.787389278411865, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14963259853039412, | |
| "grad_norm": 43.70954674356485, | |
| "learning_rate": 9.926188266120295e-07, | |
| "logits/chosen": -1.2420989274978638, | |
| "logits/rejected": -1.211631178855896, | |
| "logps/chosen": -0.43263110518455505, | |
| "logps/rejected": -0.5376982688903809, | |
| "loss": 3.6077, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -4.3263115882873535, | |
| "rewards/margins": 1.0506718158721924, | |
| "rewards/rejected": -5.376982688903809, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16032064128256512, | |
| "grad_norm": 76.0561732944734, | |
| "learning_rate": 9.890738003669027e-07, | |
| "logits/chosen": -1.196466088294983, | |
| "logits/rejected": -1.1188781261444092, | |
| "logps/chosen": -0.4406164586544037, | |
| "logps/rejected": -0.5367096662521362, | |
| "loss": 3.5757, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -4.40616512298584, | |
| "rewards/margins": 0.9609323740005493, | |
| "rewards/rejected": -5.3670973777771, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17100868403473615, | |
| "grad_norm": 51.31544624480531, | |
| "learning_rate": 9.848447601883433e-07, | |
| "logits/chosen": -1.189353346824646, | |
| "logits/rejected": -1.1714979410171509, | |
| "logps/chosen": -0.38350600004196167, | |
| "logps/rejected": -0.5140829682350159, | |
| "loss": 3.4182, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -3.835059642791748, | |
| "rewards/margins": 1.3057689666748047, | |
| "rewards/rejected": -5.140829563140869, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18169672678690715, | |
| "grad_norm": 72.27870189202869, | |
| "learning_rate": 9.799376207714444e-07, | |
| "logits/chosen": -1.2069041728973389, | |
| "logits/rejected": -1.1819300651550293, | |
| "logps/chosen": -0.4050617218017578, | |
| "logps/rejected": -0.46999722719192505, | |
| "loss": 3.4683, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -4.05061674118042, | |
| "rewards/margins": 0.6493551731109619, | |
| "rewards/rejected": -4.699972152709961, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19238476953907815, | |
| "grad_norm": 65.15839777923942, | |
| "learning_rate": 9.743592451943998e-07, | |
| "logits/chosen": -1.2592518329620361, | |
| "logits/rejected": -1.2168278694152832, | |
| "logps/chosen": -0.49755024909973145, | |
| "logps/rejected": -0.6345678567886353, | |
| "loss": 3.5477, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -4.975502014160156, | |
| "rewards/margins": 1.3701757192611694, | |
| "rewards/rejected": -6.345677852630615, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20307281229124916, | |
| "grad_norm": 86.65912439673568, | |
| "learning_rate": 9.681174353198686e-07, | |
| "logits/chosen": -1.324328064918518, | |
| "logits/rejected": -1.2326462268829346, | |
| "logps/chosen": -0.5199041962623596, | |
| "logps/rejected": -0.5949305295944214, | |
| "loss": 3.3966, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -5.199041843414307, | |
| "rewards/margins": 0.7502638697624207, | |
| "rewards/rejected": -5.949306011199951, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21376085504342018, | |
| "grad_norm": 84.91017386381003, | |
| "learning_rate": 9.612209208833646e-07, | |
| "logits/chosen": -1.223847508430481, | |
| "logits/rejected": -1.193149209022522, | |
| "logps/chosen": -0.5218724012374878, | |
| "logps/rejected": -0.6197377443313599, | |
| "loss": 3.4071, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -5.218724250793457, | |
| "rewards/margins": 0.978653609752655, | |
| "rewards/rejected": -6.197378158569336, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22444889779559118, | |
| "grad_norm": 122.37976199382105, | |
| "learning_rate": 9.536793472839324e-07, | |
| "logits/chosen": -1.2392634153366089, | |
| "logits/rejected": -1.178882360458374, | |
| "logps/chosen": -0.47783392667770386, | |
| "logps/rejected": -0.6210560202598572, | |
| "loss": 3.3437, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -4.7783403396606445, | |
| "rewards/margins": 1.432220220565796, | |
| "rewards/rejected": -6.210559844970703, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23513694054776219, | |
| "grad_norm": 117.75268692874619, | |
| "learning_rate": 9.455032620941839e-07, | |
| "logits/chosen": -1.2091577053070068, | |
| "logits/rejected": -1.1453241109848022, | |
| "logps/chosen": -0.5471813678741455, | |
| "logps/rejected": -0.6967746615409851, | |
| "loss": 3.3036, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -5.471813678741455, | |
| "rewards/margins": 1.495932936668396, | |
| "rewards/rejected": -6.967746734619141, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2458249832999332, | |
| "grad_norm": 103.23266991280943, | |
| "learning_rate": 9.367041003085648e-07, | |
| "logits/chosen": -1.2898657321929932, | |
| "logits/rejected": -1.224526047706604, | |
| "logps/chosen": -0.5657516121864319, | |
| "logps/rejected": -0.684479296207428, | |
| "loss": 3.058, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -5.657515525817871, | |
| "rewards/margins": 1.18727707862854, | |
| "rewards/rejected": -6.84479284286499, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2565130260521042, | |
| "grad_norm": 87.5488547010946, | |
| "learning_rate": 9.272941683504808e-07, | |
| "logits/chosen": -1.2546017169952393, | |
| "logits/rejected": -1.157871961593628, | |
| "logps/chosen": -0.6318168640136719, | |
| "logps/rejected": -0.8914499282836914, | |
| "loss": 2.963, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -6.318168640136719, | |
| "rewards/margins": 2.596330165863037, | |
| "rewards/rejected": -8.914499282836914, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26720106880427524, | |
| "grad_norm": 136.85902662680772, | |
| "learning_rate": 9.172866268606513e-07, | |
| "logits/chosen": -1.336275339126587, | |
| "logits/rejected": -1.2872272729873657, | |
| "logps/chosen": -0.6883447766304016, | |
| "logps/rejected": -0.8685702085494995, | |
| "loss": 2.782, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -6.883447170257568, | |
| "rewards/margins": 1.8022544384002686, | |
| "rewards/rejected": -8.685701370239258, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27788911155644624, | |
| "grad_norm": 219.78249066687138, | |
| "learning_rate": 9.066954722907638e-07, | |
| "logits/chosen": -1.365232229232788, | |
| "logits/rejected": -1.354945182800293, | |
| "logps/chosen": -0.795397162437439, | |
| "logps/rejected": -1.2445639371871948, | |
| "loss": 2.7364, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -7.953972816467285, | |
| "rewards/margins": 4.491666793823242, | |
| "rewards/rejected": -12.445638656616211, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.28857715430861725, | |
| "grad_norm": 108.29484064425847, | |
| "learning_rate": 8.955355173281707e-07, | |
| "logits/chosen": -1.3360868692398071, | |
| "logits/rejected": -1.2848550081253052, | |
| "logps/chosen": -0.8393806219100952, | |
| "logps/rejected": -1.0930553674697876, | |
| "loss": 2.7236, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -8.393805503845215, | |
| "rewards/margins": 2.536748170852661, | |
| "rewards/rejected": -10.930554389953613, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29926519706078825, | |
| "grad_norm": 111.40981489221316, | |
| "learning_rate": 8.838223701790055e-07, | |
| "logits/chosen": -1.4011452198028564, | |
| "logits/rejected": -1.3746473789215088, | |
| "logps/chosen": -0.8652567863464355, | |
| "logps/rejected": -1.1113642454147339, | |
| "loss": 2.6512, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -8.652568817138672, | |
| "rewards/margins": 2.4610743522644043, | |
| "rewards/rejected": -11.113641738891602, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30995323981295925, | |
| "grad_norm": 138.17602287907837, | |
| "learning_rate": 8.71572412738697e-07, | |
| "logits/chosen": -1.3082659244537354, | |
| "logits/rejected": -1.2775059938430786, | |
| "logps/chosen": -0.8726702928543091, | |
| "logps/rejected": -1.1962165832519531, | |
| "loss": 2.5784, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -8.726702690124512, | |
| "rewards/margins": 3.2354626655578613, | |
| "rewards/rejected": -11.962165832519531, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32064128256513025, | |
| "grad_norm": 115.35023284841587, | |
| "learning_rate": 8.588027776804058e-07, | |
| "logits/chosen": -1.3314917087554932, | |
| "logits/rejected": -1.306792974472046, | |
| "logps/chosen": -1.0279521942138672, | |
| "logps/rejected": -1.3382768630981445, | |
| "loss": 2.47, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -10.279521942138672, | |
| "rewards/margins": 3.1032464504241943, | |
| "rewards/rejected": -13.382769584655762, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33132932531730125, | |
| "grad_norm": 144.67956167588085, | |
| "learning_rate": 8.455313244934324e-07, | |
| "logits/chosen": -1.3292877674102783, | |
| "logits/rejected": -1.3038713932037354, | |
| "logps/chosen": -1.035979986190796, | |
| "logps/rejected": -1.4305903911590576, | |
| "loss": 2.4435, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -10.359800338745117, | |
| "rewards/margins": 3.946104049682617, | |
| "rewards/rejected": -14.305903434753418, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3420173680694723, | |
| "grad_norm": 106.03488572932915, | |
| "learning_rate": 8.317766145051057e-07, | |
| "logits/chosen": -1.3299791812896729, | |
| "logits/rejected": -1.3083471059799194, | |
| "logps/chosen": -1.1912587881088257, | |
| "logps/rejected": -1.648496389389038, | |
| "loss": 2.3299, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -11.912586212158203, | |
| "rewards/margins": 4.5723772048950195, | |
| "rewards/rejected": -16.484966278076172, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3527054108216433, | |
| "grad_norm": 124.30842086260031, | |
| "learning_rate": 8.175578849210894e-07, | |
| "logits/chosen": -1.3360395431518555, | |
| "logits/rejected": -1.3087151050567627, | |
| "logps/chosen": -1.2806798219680786, | |
| "logps/rejected": -1.7463128566741943, | |
| "loss": 2.1935, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -12.806798934936523, | |
| "rewards/margins": 4.656330108642578, | |
| "rewards/rejected": -17.463130950927734, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3633934535738143, | |
| "grad_norm": 159.71509664144037, | |
| "learning_rate": 8.028950219204099e-07, | |
| "logits/chosen": -1.3390699625015259, | |
| "logits/rejected": -1.3161685466766357, | |
| "logps/chosen": -1.27903151512146, | |
| "logps/rejected": -1.8541208505630493, | |
| "loss": 2.1363, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -12.790315628051758, | |
| "rewards/margins": 5.750893592834473, | |
| "rewards/rejected": -18.541208267211914, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3740814963259853, | |
| "grad_norm": 163.13967408983922, | |
| "learning_rate": 7.878085328428368e-07, | |
| "logits/chosen": -1.359133243560791, | |
| "logits/rejected": -1.3104041814804077, | |
| "logps/chosen": -1.4678348302841187, | |
| "logps/rejected": -1.8786561489105225, | |
| "loss": 2.0099, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -14.678349494934082, | |
| "rewards/margins": 4.108211517333984, | |
| "rewards/rejected": -18.786561965942383, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3847695390781563, | |
| "grad_norm": 213.96398552844718, | |
| "learning_rate": 7.723195175075135e-07, | |
| "logits/chosen": -1.3180896043777466, | |
| "logits/rejected": -1.2993403673171997, | |
| "logps/chosen": -1.395212173461914, | |
| "logps/rejected": -1.9256620407104492, | |
| "loss": 1.8566, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -13.952122688293457, | |
| "rewards/margins": 5.304497718811035, | |
| "rewards/rejected": -19.25661849975586, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3954575818303273, | |
| "grad_norm": 151.287610930137, | |
| "learning_rate": 7.564496387029531e-07, | |
| "logits/chosen": -1.354866623878479, | |
| "logits/rejected": -1.2994146347045898, | |
| "logps/chosen": -1.5145620107650757, | |
| "logps/rejected": -2.0945541858673096, | |
| "loss": 1.9317, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -15.14561939239502, | |
| "rewards/margins": 5.799921989440918, | |
| "rewards/rejected": -20.945541381835938, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4061456245824983, | |
| "grad_norm": 147.29173146280465, | |
| "learning_rate": 7.402210918896689e-07, | |
| "logits/chosen": -1.348362684249878, | |
| "logits/rejected": -1.3590227365493774, | |
| "logps/chosen": -1.7428133487701416, | |
| "logps/rejected": -2.511070966720581, | |
| "loss": 1.7298, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -17.428131103515625, | |
| "rewards/margins": 7.682579040527344, | |
| "rewards/rejected": -25.1107120513916, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4168336673346693, | |
| "grad_norm": 107.1032490301315, | |
| "learning_rate": 7.236565741578162e-07, | |
| "logits/chosen": -1.3077692985534668, | |
| "logits/rejected": -1.2952277660369873, | |
| "logps/chosen": -1.8426529169082642, | |
| "logps/rejected": -2.392831325531006, | |
| "loss": 1.9538, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -18.426528930664062, | |
| "rewards/margins": 5.5017852783203125, | |
| "rewards/rejected": -23.928314208984375, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42752171008684037, | |
| "grad_norm": 190.9422038213508, | |
| "learning_rate": 7.067792524832603e-07, | |
| "logits/chosen": -1.3160083293914795, | |
| "logits/rejected": -1.3034440279006958, | |
| "logps/chosen": -1.7327241897583008, | |
| "logps/rejected": -2.3605778217315674, | |
| "loss": 1.6663, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -17.327241897583008, | |
| "rewards/margins": 6.278534889221191, | |
| "rewards/rejected": -23.605777740478516, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43820975283901137, | |
| "grad_norm": 161.95141010920645, | |
| "learning_rate": 6.896127313264642e-07, | |
| "logits/chosen": -1.3677047491073608, | |
| "logits/rejected": -1.3167794942855835, | |
| "logps/chosen": -1.8963912725448608, | |
| "logps/rejected": -2.526191234588623, | |
| "loss": 1.8748, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -18.96391487121582, | |
| "rewards/margins": 6.297998905181885, | |
| "rewards/rejected": -25.261911392211914, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44889779559118237, | |
| "grad_norm": 150.6462623725548, | |
| "learning_rate": 6.721810196195174e-07, | |
| "logits/chosen": -1.400187373161316, | |
| "logits/rejected": -1.3892953395843506, | |
| "logps/chosen": -2.041691541671753, | |
| "logps/rejected": -2.6361751556396484, | |
| "loss": 1.827, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -20.416913986206055, | |
| "rewards/margins": 5.9448370933532715, | |
| "rewards/rejected": -26.361751556396484, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45958583834335337, | |
| "grad_norm": 126.80315819769801, | |
| "learning_rate": 6.545084971874736e-07, | |
| "logits/chosen": -1.3723266124725342, | |
| "logits/rejected": -1.356567144393921, | |
| "logps/chosen": -2.0259296894073486, | |
| "logps/rejected": -2.754607677459717, | |
| "loss": 1.6874, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -20.259296417236328, | |
| "rewards/margins": 7.286777496337891, | |
| "rewards/rejected": -27.546072006225586, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.47027388109552437, | |
| "grad_norm": 161.97036694522683, | |
| "learning_rate": 6.3661988065096e-07, | |
| "logits/chosen": -1.4270765781402588, | |
| "logits/rejected": -1.4134238958358765, | |
| "logps/chosen": -2.0706095695495605, | |
| "logps/rejected": -2.8078479766845703, | |
| "loss": 1.5613, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -20.706096649169922, | |
| "rewards/margins": 7.372382164001465, | |
| "rewards/rejected": -28.078479766845703, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48096192384769537, | |
| "grad_norm": 119.50283654492473, | |
| "learning_rate": 6.185401888577487e-07, | |
| "logits/chosen": -1.4146029949188232, | |
| "logits/rejected": -1.3837422132492065, | |
| "logps/chosen": -2.119856595993042, | |
| "logps/rejected": -2.7663817405700684, | |
| "loss": 1.7069, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -21.198566436767578, | |
| "rewards/margins": 6.465248107910156, | |
| "rewards/rejected": -27.663818359375, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4916499665998664, | |
| "grad_norm": 109.77939760201329, | |
| "learning_rate": 6.002947078916364e-07, | |
| "logits/chosen": -1.4831273555755615, | |
| "logits/rejected": -1.435591459274292, | |
| "logps/chosen": -2.0505659580230713, | |
| "logps/rejected": -2.685992956161499, | |
| "loss": 1.5078, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": -20.505659103393555, | |
| "rewards/margins": 6.354269027709961, | |
| "rewards/rejected": -26.859928131103516, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5023380093520374, | |
| "grad_norm": 147.37811459958783, | |
| "learning_rate": 5.819089557075688e-07, | |
| "logits/chosen": -1.4949449300765991, | |
| "logits/rejected": -1.469238042831421, | |
| "logps/chosen": -2.045539617538452, | |
| "logps/rejected": -2.816204071044922, | |
| "loss": 1.4911, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -20.455394744873047, | |
| "rewards/margins": 7.706644535064697, | |
| "rewards/rejected": -28.162038803100586, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5130260521042084, | |
| "grad_norm": 132.440512722152, | |
| "learning_rate": 5.634086464424742e-07, | |
| "logits/chosen": -1.461572289466858, | |
| "logits/rejected": -1.4638675451278687, | |
| "logps/chosen": -1.9458844661712646, | |
| "logps/rejected": -2.7108004093170166, | |
| "loss": 1.5773, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -19.458845138549805, | |
| "rewards/margins": 7.6491570472717285, | |
| "rewards/rejected": -27.10800552368164, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5237140948563794, | |
| "grad_norm": 236.09806301372015, | |
| "learning_rate": 5.448196544517167e-07, | |
| "logits/chosen": -1.5521364212036133, | |
| "logits/rejected": -1.4946634769439697, | |
| "logps/chosen": -2.0292551517486572, | |
| "logps/rejected": -2.9242515563964844, | |
| "loss": 1.395, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -20.292551040649414, | |
| "rewards/margins": 8.949962615966797, | |
| "rewards/rejected": -29.242517471313477, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5344021376085505, | |
| "grad_norm": 142.79859396103615, | |
| "learning_rate": 5.26167978121472e-07, | |
| "logits/chosen": -1.4808663129806519, | |
| "logits/rejected": -1.4642443656921387, | |
| "logps/chosen": -2.046252489089966, | |
| "logps/rejected": -2.9118173122406006, | |
| "loss": 1.3738, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -20.462526321411133, | |
| "rewards/margins": 8.655647277832031, | |
| "rewards/rejected": -29.1181697845459, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5450901803607214, | |
| "grad_norm": 155.0378761351227, | |
| "learning_rate": 5.074797035076318e-07, | |
| "logits/chosen": -1.5233964920043945, | |
| "logits/rejected": -1.5014541149139404, | |
| "logps/chosen": -2.2131590843200684, | |
| "logps/rejected": -2.9324965476989746, | |
| "loss": 1.5353, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -22.131589889526367, | |
| "rewards/margins": 7.193373680114746, | |
| "rewards/rejected": -29.324962615966797, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5557782231128925, | |
| "grad_norm": 130.04315061980418, | |
| "learning_rate": 4.887809678520975e-07, | |
| "logits/chosen": -1.5009315013885498, | |
| "logits/rejected": -1.474130630493164, | |
| "logps/chosen": -2.1018893718719482, | |
| "logps/rejected": -2.8382973670959473, | |
| "loss": 1.5292, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -21.018896102905273, | |
| "rewards/margins": 7.364079475402832, | |
| "rewards/rejected": -28.382976531982422, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5664662658650634, | |
| "grad_norm": 119.27303382130957, | |
| "learning_rate": 4.700979230274829e-07, | |
| "logits/chosen": -1.4742642641067505, | |
| "logits/rejected": -1.4596500396728516, | |
| "logps/chosen": -2.2155094146728516, | |
| "logps/rejected": -3.0179553031921387, | |
| "loss": 1.1693, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -22.15509605407715, | |
| "rewards/margins": 8.024459838867188, | |
| "rewards/rejected": -30.179553985595703, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5771543086172345, | |
| "grad_norm": 131.97372035922703, | |
| "learning_rate": 4.514566989613559e-07, | |
| "logits/chosen": -1.4763275384902954, | |
| "logits/rejected": -1.4486229419708252, | |
| "logps/chosen": -1.990666151046753, | |
| "logps/rejected": -2.838904857635498, | |
| "loss": 1.3013, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -19.906661987304688, | |
| "rewards/margins": 8.482388496398926, | |
| "rewards/rejected": -28.389049530029297, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5878423513694054, | |
| "grad_norm": 104.56387348722761, | |
| "learning_rate": 4.328833670911724e-07, | |
| "logits/chosen": -1.4566619396209717, | |
| "logits/rejected": -1.4216809272766113, | |
| "logps/chosen": -1.9770195484161377, | |
| "logps/rejected": -2.719089984893799, | |
| "loss": 1.4258, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -19.77019500732422, | |
| "rewards/margins": 7.4207048416137695, | |
| "rewards/rejected": -27.190898895263672, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5985303941215765, | |
| "grad_norm": 133.9300472658157, | |
| "learning_rate": 4.144039039010124e-07, | |
| "logits/chosen": -1.542373776435852, | |
| "logits/rejected": -1.5213029384613037, | |
| "logps/chosen": -2.0355706214904785, | |
| "logps/rejected": -2.883401393890381, | |
| "loss": 1.3022, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -20.3557071685791, | |
| "rewards/margins": 8.478304862976074, | |
| "rewards/rejected": -28.834014892578125, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6092184368737475, | |
| "grad_norm": 118.2550806898935, | |
| "learning_rate": 3.960441545911204e-07, | |
| "logits/chosen": -1.501997947692871, | |
| "logits/rejected": -1.4718389511108398, | |
| "logps/chosen": -2.043879270553589, | |
| "logps/rejected": -2.9162051677703857, | |
| "loss": 1.0025, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -20.438793182373047, | |
| "rewards/margins": 8.723258972167969, | |
| "rewards/rejected": -29.16205406188965, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6199064796259185, | |
| "grad_norm": 112.08802157558269, | |
| "learning_rate": 3.778297969310529e-07, | |
| "logits/chosen": -1.5190720558166504, | |
| "logits/rejected": -1.4788782596588135, | |
| "logps/chosen": -2.014256000518799, | |
| "logps/rejected": -2.7785563468933105, | |
| "loss": 1.3157, | |
| "rewards/accuracies": 0.831250011920929, | |
| "rewards/chosen": -20.142559051513672, | |
| "rewards/margins": 7.64300537109375, | |
| "rewards/rejected": -27.785564422607422, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6305945223780896, | |
| "grad_norm": 122.16012718306597, | |
| "learning_rate": 3.5978630534699865e-07, | |
| "logits/chosen": -1.4519203901290894, | |
| "logits/rejected": -1.4377477169036865, | |
| "logps/chosen": -2.052429437637329, | |
| "logps/rejected": -2.7946228981018066, | |
| "loss": 1.2035, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -20.524295806884766, | |
| "rewards/margins": 7.421932220458984, | |
| "rewards/rejected": -27.94622802734375, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6412825651302605, | |
| "grad_norm": 127.59960775053135, | |
| "learning_rate": 3.4193891529348795e-07, | |
| "logits/chosen": -1.3882105350494385, | |
| "logits/rejected": -1.365216612815857, | |
| "logps/chosen": -2.0868842601776123, | |
| "logps/rejected": -2.7921385765075684, | |
| "loss": 1.5608, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -20.86884307861328, | |
| "rewards/margins": 7.0525383949279785, | |
| "rewards/rejected": -27.9213809967041, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6519706078824316, | |
| "grad_norm": 104.5591334100815, | |
| "learning_rate": 3.243125879593286e-07, | |
| "logits/chosen": -1.5032074451446533, | |
| "logits/rejected": -1.458888053894043, | |
| "logps/chosen": -2.047602415084839, | |
| "logps/rejected": -2.731414794921875, | |
| "loss": 1.3552, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": -20.476024627685547, | |
| "rewards/margins": 6.8381242752075195, | |
| "rewards/rejected": -27.31414794921875, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6626586506346025, | |
| "grad_norm": 128.9435075925882, | |
| "learning_rate": 3.069319753571269e-07, | |
| "logits/chosen": -1.5344185829162598, | |
| "logits/rejected": -1.520206093788147, | |
| "logps/chosen": -2.182490587234497, | |
| "logps/rejected": -2.994981288909912, | |
| "loss": 1.3424, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -21.824905395507812, | |
| "rewards/margins": 8.124906539916992, | |
| "rewards/rejected": -29.949810028076172, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6733466933867736, | |
| "grad_norm": 117.862174484777, | |
| "learning_rate": 2.898213858452173e-07, | |
| "logits/chosen": -1.5316812992095947, | |
| "logits/rejected": -1.4777030944824219, | |
| "logps/chosen": -2.1463074684143066, | |
| "logps/rejected": -2.8947274684906006, | |
| "loss": 1.2773, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -21.46307373046875, | |
| "rewards/margins": 7.484200954437256, | |
| "rewards/rejected": -28.947277069091797, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6840347361389446, | |
| "grad_norm": 154.98050630680447, | |
| "learning_rate": 2.730047501302266e-07, | |
| "logits/chosen": -1.5196678638458252, | |
| "logits/rejected": -1.5156781673431396, | |
| "logps/chosen": -2.249702215194702, | |
| "logps/rejected": -3.2510273456573486, | |
| "loss": 1.0665, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -22.497024536132812, | |
| "rewards/margins": 10.013248443603516, | |
| "rewards/rejected": -32.51027297973633, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6947227788911156, | |
| "grad_norm": 180.45567365921923, | |
| "learning_rate": 2.5650558779781635e-07, | |
| "logits/chosen": -1.549286127090454, | |
| "logits/rejected": -1.5002086162567139, | |
| "logps/chosen": -2.399672031402588, | |
| "logps/rejected": -3.504227876663208, | |
| "loss": 1.2384, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -23.996719360351562, | |
| "rewards/margins": 11.045561790466309, | |
| "rewards/rejected": -35.04228210449219, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7054108216432866, | |
| "grad_norm": 105.90784807533655, | |
| "learning_rate": 2.403469744184154e-07, | |
| "logits/chosen": -1.4591710567474365, | |
| "logits/rejected": -1.4205106496810913, | |
| "logps/chosen": -2.2293872833251953, | |
| "logps/rejected": -3.00614857673645, | |
| "loss": 1.2359, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -22.293874740600586, | |
| "rewards/margins": 7.767613410949707, | |
| "rewards/rejected": -30.061487197875977, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7160988643954576, | |
| "grad_norm": 124.3214763138239, | |
| "learning_rate": 2.2455150927394878e-07, | |
| "logits/chosen": -1.495527744293213, | |
| "logits/rejected": -1.4803019762039185, | |
| "logps/chosen": -2.202894449234009, | |
| "logps/rejected": -3.0441994667053223, | |
| "loss": 1.1359, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -22.02894401550293, | |
| "rewards/margins": 8.413049697875977, | |
| "rewards/rejected": -30.441997528076172, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7267869071476286, | |
| "grad_norm": 156.03049853633283, | |
| "learning_rate": 2.0914128375069722e-07, | |
| "logits/chosen": -1.5111547708511353, | |
| "logits/rejected": -1.4789291620254517, | |
| "logps/chosen": -2.1601741313934326, | |
| "logps/rejected": -3.021275281906128, | |
| "loss": 1.2419, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -21.601741790771484, | |
| "rewards/margins": 8.611013412475586, | |
| "rewards/rejected": -30.212757110595703, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7374749498997996, | |
| "grad_norm": 130.76746865900543, | |
| "learning_rate": 1.9413785044249676e-07, | |
| "logits/chosen": -1.5371043682098389, | |
| "logits/rejected": -1.5177440643310547, | |
| "logps/chosen": -2.2243714332580566, | |
| "logps/rejected": -3.2498092651367188, | |
| "loss": 1.2299, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -22.243711471557617, | |
| "rewards/margins": 10.25438117980957, | |
| "rewards/rejected": -32.49809265136719, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7481629926519706, | |
| "grad_norm": 219.49500818808932, | |
| "learning_rate": 1.7956219300748792e-07, | |
| "logits/chosen": -1.5225059986114502, | |
| "logits/rejected": -1.5286823511123657, | |
| "logps/chosen": -2.12577486038208, | |
| "logps/rejected": -3.068633794784546, | |
| "loss": 1.164, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -21.257749557495117, | |
| "rewards/margins": 9.428587913513184, | |
| "rewards/rejected": -30.686336517333984, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7588510354041417, | |
| "grad_norm": 148.30498812771748, | |
| "learning_rate": 1.6543469682057104e-07, | |
| "logits/chosen": -1.4495574235916138, | |
| "logits/rejected": -1.4652235507965088, | |
| "logps/chosen": -2.066810131072998, | |
| "logps/rejected": -2.938152313232422, | |
| "loss": 1.1351, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -20.668102264404297, | |
| "rewards/margins": 8.713422775268555, | |
| "rewards/rejected": -29.38152503967285, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7695390781563126, | |
| "grad_norm": 116.32905868645479, | |
| "learning_rate": 1.5177512046261666e-07, | |
| "logits/chosen": -1.5009536743164062, | |
| "logits/rejected": -1.5022971630096436, | |
| "logps/chosen": -2.0494394302368164, | |
| "logps/rejected": -3.0328898429870605, | |
| "loss": 1.1972, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -20.494396209716797, | |
| "rewards/margins": 9.834505081176758, | |
| "rewards/rejected": -30.328899383544922, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7802271209084837, | |
| "grad_norm": 108.7134031593527, | |
| "learning_rate": 1.3860256808630427e-07, | |
| "logits/chosen": -1.5464454889297485, | |
| "logits/rejected": -1.4812748432159424, | |
| "logps/chosen": -2.1187500953674316, | |
| "logps/rejected": -3.0532031059265137, | |
| "loss": 1.253, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -21.187496185302734, | |
| "rewards/margins": 9.344534873962402, | |
| "rewards/rejected": -30.532033920288086, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7909151636606546, | |
| "grad_norm": 145.2134549922322, | |
| "learning_rate": 1.2593546269723647e-07, | |
| "logits/chosen": -1.474498987197876, | |
| "logits/rejected": -1.4640263319015503, | |
| "logps/chosen": -2.100693941116333, | |
| "logps/rejected": -2.9064152240753174, | |
| "loss": 1.1015, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -21.00693702697754, | |
| "rewards/margins": 8.05721378326416, | |
| "rewards/rejected": -29.064151763916016, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8016032064128257, | |
| "grad_norm": 130.4190340462438, | |
| "learning_rate": 1.1379152038770029e-07, | |
| "logits/chosen": -1.4943348169326782, | |
| "logits/rejected": -1.4995317459106445, | |
| "logps/chosen": -2.2358219623565674, | |
| "logps/rejected": -3.1511778831481934, | |
| "loss": 1.1776, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -22.35822105407715, | |
| "rewards/margins": 9.153559684753418, | |
| "rewards/rejected": -31.51177978515625, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8122912491649966, | |
| "grad_norm": 171.4769450829484, | |
| "learning_rate": 1.0218772555910954e-07, | |
| "logits/chosen": -1.5206212997436523, | |
| "logits/rejected": -1.5025417804718018, | |
| "logps/chosen": -2.1182684898376465, | |
| "logps/rejected": -2.9665935039520264, | |
| "loss": 1.2678, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -21.18268394470215, | |
| "rewards/margins": 8.48325252532959, | |
| "rewards/rejected": -29.665935516357422, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8229792919171677, | |
| "grad_norm": 116.74061042296164, | |
| "learning_rate": 9.114030716778432e-08, | |
| "logits/chosen": -1.5152442455291748, | |
| "logits/rejected": -1.4944274425506592, | |
| "logps/chosen": -2.0695395469665527, | |
| "logps/rejected": -3.1040778160095215, | |
| "loss": 1.0078, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -20.69539451599121, | |
| "rewards/margins": 10.345380783081055, | |
| "rewards/rejected": -31.0407772064209, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8336673346693386, | |
| "grad_norm": 112.33137236518094, | |
| "learning_rate": 8.066471602728803e-08, | |
| "logits/chosen": -1.5178992748260498, | |
| "logits/rejected": -1.5032987594604492, | |
| "logps/chosen": -2.170666456222534, | |
| "logps/rejected": -3.162553310394287, | |
| "loss": 1.2224, | |
| "rewards/accuracies": 0.8687499761581421, | |
| "rewards/chosen": -21.7066650390625, | |
| "rewards/margins": 9.918868064880371, | |
| "rewards/rejected": -31.62552833557129, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8443553774215097, | |
| "grad_norm": 186.4853149645633, | |
| "learning_rate": 7.077560319906694e-08, | |
| "logits/chosen": -1.5229237079620361, | |
| "logits/rejected": -1.505030632019043, | |
| "logps/chosen": -2.0735864639282227, | |
| "logps/rejected": -2.9833900928497314, | |
| "loss": 1.1526, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -20.73586654663086, | |
| "rewards/margins": 9.098031997680664, | |
| "rewards/rejected": -29.833898544311523, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "grad_norm": 96.6762011696182, | |
| "learning_rate": 6.148679950161672e-08, | |
| "logits/chosen": -1.5230014324188232, | |
| "logits/rejected": -1.5108994245529175, | |
| "logps/chosen": -2.109778642654419, | |
| "logps/rejected": -2.941293716430664, | |
| "loss": 1.1544, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -21.09778594970703, | |
| "rewards/margins": 8.315154075622559, | |
| "rewards/rejected": -29.412939071655273, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "eval_logits/chosen": -1.710108757019043, | |
| "eval_logits/rejected": -1.7197374105453491, | |
| "eval_logps/chosen": -2.08453369140625, | |
| "eval_logps/rejected": -2.94063138961792, | |
| "eval_loss": 1.1388919353485107, | |
| "eval_rewards/accuracies": 0.8678861856460571, | |
| "eval_rewards/chosen": -20.845340728759766, | |
| "eval_rewards/margins": 8.560976028442383, | |
| "eval_rewards/rejected": -29.406312942504883, | |
| "eval_runtime": 123.1178, | |
| "eval_samples_per_second": 15.928, | |
| "eval_steps_per_second": 0.999, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8657314629258517, | |
| "grad_norm": 143.6008976778444, | |
| "learning_rate": 5.2811296166831666e-08, | |
| "logits/chosen": -1.4884498119354248, | |
| "logits/rejected": -1.508371114730835, | |
| "logps/chosen": -2.2530012130737305, | |
| "logps/rejected": -3.071047782897949, | |
| "loss": 1.1786, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": -22.530010223388672, | |
| "rewards/margins": 8.18046760559082, | |
| "rewards/rejected": -30.710479736328125, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8764195056780227, | |
| "grad_norm": 145.0618422989251, | |
| "learning_rate": 4.4761226670592066e-08, | |
| "logits/chosen": -1.4991635084152222, | |
| "logits/rejected": -1.4899944067001343, | |
| "logps/chosen": -2.1319069862365723, | |
| "logps/rejected": -2.977902889251709, | |
| "loss": 1.073, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -21.31907081604004, | |
| "rewards/margins": 8.4599609375, | |
| "rewards/rejected": -29.77903175354004, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8871075484301937, | |
| "grad_norm": 121.04879854574509, | |
| "learning_rate": 3.734784976300165e-08, | |
| "logits/chosen": -1.5042963027954102, | |
| "logits/rejected": -1.4524286985397339, | |
| "logps/chosen": -2.0573108196258545, | |
| "logps/rejected": -3.0475449562072754, | |
| "loss": 1.2613, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -20.573108673095703, | |
| "rewards/margins": 9.902341842651367, | |
| "rewards/rejected": -30.475452423095703, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8977955911823647, | |
| "grad_norm": 139.53176945804844, | |
| "learning_rate": 3.058153372200695e-08, | |
| "logits/chosen": -1.5310368537902832, | |
| "logits/rejected": -1.4835598468780518, | |
| "logps/chosen": -2.005772352218628, | |
| "logps/rejected": -3.0218088626861572, | |
| "loss": 1.1364, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": -20.057723999023438, | |
| "rewards/margins": 10.160362243652344, | |
| "rewards/rejected": -30.218088150024414, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9084836339345357, | |
| "grad_norm": 136.53214771188803, | |
| "learning_rate": 2.4471741852423233e-08, | |
| "logits/chosen": -1.525810718536377, | |
| "logits/rejected": -1.51809823513031, | |
| "logps/chosen": -2.186115026473999, | |
| "logps/rejected": -3.016331911087036, | |
| "loss": 1.4659, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -21.86115074157715, | |
| "rewards/margins": 8.302164077758789, | |
| "rewards/rejected": -30.163314819335938, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9191716766867067, | |
| "grad_norm": 159.75999561219473, | |
| "learning_rate": 1.9027019250647036e-08, | |
| "logits/chosen": -1.5137279033660889, | |
| "logits/rejected": -1.5020883083343506, | |
| "logps/chosen": -2.2320022583007812, | |
| "logps/rejected": -3.141892671585083, | |
| "loss": 1.1123, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -22.320022583007812, | |
| "rewards/margins": 9.098905563354492, | |
| "rewards/rejected": -31.418926239013672, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9298597194388778, | |
| "grad_norm": 103.73846522782931, | |
| "learning_rate": 1.4254980853566246e-08, | |
| "logits/chosen": -1.469313383102417, | |
| "logits/rejected": -1.4278925657272339, | |
| "logps/chosen": -1.988031029701233, | |
| "logps/rejected": -2.880598545074463, | |
| "loss": 1.1778, | |
| "rewards/accuracies": 0.8812500238418579, | |
| "rewards/chosen": -19.880308151245117, | |
| "rewards/margins": 8.925679206848145, | |
| "rewards/rejected": -28.805988311767578, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9405477621910487, | |
| "grad_norm": 103.8822744213963, | |
| "learning_rate": 1.016230078838226e-08, | |
| "logits/chosen": -1.4933596849441528, | |
| "logits/rejected": -1.4389324188232422, | |
| "logps/chosen": -2.1481704711914062, | |
| "logps/rejected": -3.0063669681549072, | |
| "loss": 1.0981, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": -21.481706619262695, | |
| "rewards/margins": 8.581964492797852, | |
| "rewards/rejected": -30.063669204711914, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9512358049432198, | |
| "grad_norm": 105.17668873683024, | |
| "learning_rate": 6.754703038239329e-09, | |
| "logits/chosen": -1.4460530281066895, | |
| "logits/rejected": -1.431138277053833, | |
| "logps/chosen": -2.159626007080078, | |
| "logps/rejected": -3.2153289318084717, | |
| "loss": 1.1378, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": -21.59626007080078, | |
| "rewards/margins": 10.557029724121094, | |
| "rewards/rejected": -32.153289794921875, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9619238476953907, | |
| "grad_norm": 149.56429045938268, | |
| "learning_rate": 4.036953436716895e-09, | |
| "logits/chosen": -1.5528205633163452, | |
| "logits/rejected": -1.5331635475158691, | |
| "logps/chosen": -2.0413358211517334, | |
| "logps/rejected": -2.8829147815704346, | |
| "loss": 1.1398, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": -20.413360595703125, | |
| "rewards/margins": 8.415786743164062, | |
| "rewards/rejected": -28.829147338867188, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9726118904475618, | |
| "grad_norm": 144.02520089734548, | |
| "learning_rate": 2.0128530023804656e-09, | |
| "logits/chosen": -1.5203301906585693, | |
| "logits/rejected": -1.4858264923095703, | |
| "logps/chosen": -2.0674800872802734, | |
| "logps/rejected": -3.0988433361053467, | |
| "loss": 1.0231, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": -20.674800872802734, | |
| "rewards/margins": 10.313634872436523, | |
| "rewards/rejected": -30.988433837890625, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9832999331997327, | |
| "grad_norm": 123.56718367330731, | |
| "learning_rate": 6.852326227130833e-10, | |
| "logits/chosen": -1.5208638906478882, | |
| "logits/rejected": -1.5143053531646729, | |
| "logps/chosen": -2.192969799041748, | |
| "logps/rejected": -3.1692872047424316, | |
| "loss": 1.1094, | |
| "rewards/accuracies": 0.862500011920929, | |
| "rewards/chosen": -21.929697036743164, | |
| "rewards/margins": 9.763177871704102, | |
| "rewards/rejected": -31.692874908447266, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9939879759519038, | |
| "grad_norm": 128.9716796420016, | |
| "learning_rate": 5.594909486328348e-11, | |
| "logits/chosen": -1.4912371635437012, | |
| "logits/rejected": -1.496371865272522, | |
| "logps/chosen": -2.177973508834839, | |
| "logps/rejected": -3.1520798206329346, | |
| "loss": 1.2437, | |
| "rewards/accuracies": 0.856249988079071, | |
| "rewards/chosen": -21.779735565185547, | |
| "rewards/margins": 9.741061210632324, | |
| "rewards/rejected": -31.520797729492188, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9982631930527722, | |
| "step": 467, | |
| "total_flos": 0.0, | |
| "train_loss": 2.0977548006004643, | |
| "train_runtime": 12789.999, | |
| "train_samples_per_second": 4.681, | |
| "train_steps_per_second": 0.037 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 467, | |
| "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 | |
| } | |