| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9994353472614342, | |
| "eval_steps": 1000, | |
| "global_step": 885, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001129305477131564, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 5.617977528089888e-08, | |
| "logits/chosen": -2.8912107944488525, | |
| "logits/rejected": -2.9116690158843994, | |
| "logps/chosen": -205.43255615234375, | |
| "logps/rejected": -277.7913818359375, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01129305477131564, | |
| "grad_norm": 11.75, | |
| "learning_rate": 5.617977528089888e-07, | |
| "logits/chosen": -2.9238009452819824, | |
| "logits/rejected": -2.8953051567077637, | |
| "logps/chosen": -286.31060791015625, | |
| "logps/rejected": -234.09100341796875, | |
| "loss": 0.6921, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.0044541251845657825, | |
| "rewards/margins": 0.0024419513065367937, | |
| "rewards/rejected": 0.0020121734123677015, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02258610954263128, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 1.1235955056179777e-06, | |
| "logits/chosen": -2.911529064178467, | |
| "logits/rejected": -2.8939220905303955, | |
| "logps/chosen": -242.33349609375, | |
| "logps/rejected": -230.67916870117188, | |
| "loss": 0.6967, | |
| "rewards/accuracies": 0.375, | |
| "rewards/chosen": -0.00516792107373476, | |
| "rewards/margins": -0.006779191549867392, | |
| "rewards/rejected": 0.0016112711746245623, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03387916431394692, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 1.6853932584269663e-06, | |
| "logits/chosen": -2.933928966522217, | |
| "logits/rejected": -2.9029603004455566, | |
| "logps/chosen": -228.00210571289062, | |
| "logps/rejected": -206.8175506591797, | |
| "loss": 0.6928, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.00036367171560414135, | |
| "rewards/margins": 0.0011087121674790978, | |
| "rewards/rejected": -0.0014723839703947306, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04517221908526256, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 2.2471910112359554e-06, | |
| "logits/chosen": -2.9396324157714844, | |
| "logits/rejected": -2.9021177291870117, | |
| "logps/chosen": -239.8461456298828, | |
| "logps/rejected": -223.19216918945312, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.01020533312112093, | |
| "rewards/margins": 0.0009328271262347698, | |
| "rewards/rejected": -0.011138159781694412, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05646527385657821, | |
| "grad_norm": 11.25, | |
| "learning_rate": 2.8089887640449444e-06, | |
| "logits/chosen": -2.9262070655822754, | |
| "logits/rejected": -2.891216993331909, | |
| "logps/chosen": -252.71072387695312, | |
| "logps/rejected": -209.1496124267578, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.02404838800430298, | |
| "rewards/margins": 0.004404561128467321, | |
| "rewards/rejected": -0.028452953323721886, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06775832862789384, | |
| "grad_norm": 9.125, | |
| "learning_rate": 3.3707865168539327e-06, | |
| "logits/chosen": -2.95279860496521, | |
| "logits/rejected": -2.9229164123535156, | |
| "logps/chosen": -231.2828826904297, | |
| "logps/rejected": -223.4114227294922, | |
| "loss": 0.6893, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.043369874358177185, | |
| "rewards/margins": 0.008675651624798775, | |
| "rewards/rejected": -0.05204553157091141, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07905138339920949, | |
| "grad_norm": 16.5, | |
| "learning_rate": 3.932584269662922e-06, | |
| "logits/chosen": -2.9178593158721924, | |
| "logits/rejected": -2.88283109664917, | |
| "logps/chosen": -249.9050750732422, | |
| "logps/rejected": -236.652099609375, | |
| "loss": 0.6814, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.06348244845867157, | |
| "rewards/margins": 0.026277879253029823, | |
| "rewards/rejected": -0.08976032584905624, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.09034443817052512, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 4.494382022471911e-06, | |
| "logits/chosen": -2.893878221511841, | |
| "logits/rejected": -2.858973979949951, | |
| "logps/chosen": -244.9037322998047, | |
| "logps/rejected": -243.09042358398438, | |
| "loss": 0.674, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.09704697877168655, | |
| "rewards/margins": 0.04427679255604744, | |
| "rewards/rejected": -0.1413237750530243, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.10163749294184077, | |
| "grad_norm": 6.84375, | |
| "learning_rate": 4.999980529233026e-06, | |
| "logits/chosen": -2.9132912158966064, | |
| "logits/rejected": -2.8923027515411377, | |
| "logps/chosen": -232.01651000976562, | |
| "logps/rejected": -232.08056640625, | |
| "loss": 0.6793, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.16643305122852325, | |
| "rewards/margins": 0.03658531233668327, | |
| "rewards/rejected": -0.20301838219165802, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.11293054771315642, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 4.997644404153021e-06, | |
| "logits/chosen": -2.958055019378662, | |
| "logits/rejected": -2.9053235054016113, | |
| "logps/chosen": -232.2222900390625, | |
| "logps/rejected": -244.8426513671875, | |
| "loss": 0.6677, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.23467275500297546, | |
| "rewards/margins": 0.08466038852930069, | |
| "rewards/rejected": -0.31933316588401794, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12422360248447205, | |
| "grad_norm": 9.3125, | |
| "learning_rate": 4.991418294828522e-06, | |
| "logits/chosen": -2.896815776824951, | |
| "logits/rejected": -2.8579037189483643, | |
| "logps/chosen": -256.8934020996094, | |
| "logps/rejected": -252.8817138671875, | |
| "loss": 0.6693, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.33838751912117004, | |
| "rewards/margins": 0.08679252117872238, | |
| "rewards/rejected": -0.425180047750473, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.13551665725578768, | |
| "grad_norm": 8.3125, | |
| "learning_rate": 4.981311898183214e-06, | |
| "logits/chosen": -2.9425575733184814, | |
| "logits/rejected": -2.927373170852661, | |
| "logps/chosen": -236.5902099609375, | |
| "logps/rejected": -218.78945922851562, | |
| "loss": 0.6899, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.39529091119766235, | |
| "rewards/margins": 0.03407430276274681, | |
| "rewards/rejected": -0.42936521768569946, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.14680971202710333, | |
| "grad_norm": 9.5, | |
| "learning_rate": 4.967340954537717e-06, | |
| "logits/chosen": -2.9237935543060303, | |
| "logits/rejected": -2.896763324737549, | |
| "logps/chosen": -249.92367553710938, | |
| "logps/rejected": -257.2637023925781, | |
| "loss": 0.6441, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.41192659735679626, | |
| "rewards/margins": 0.14285169541835785, | |
| "rewards/rejected": -0.5547782778739929, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.15810276679841898, | |
| "grad_norm": 7.09375, | |
| "learning_rate": 4.9495272230946525e-06, | |
| "logits/chosen": -2.9524216651916504, | |
| "logits/rejected": -2.915919303894043, | |
| "logps/chosen": -252.70565795898438, | |
| "logps/rejected": -242.8680419921875, | |
| "loss": 0.6634, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.48054733872413635, | |
| "rewards/margins": 0.10070822387933731, | |
| "rewards/rejected": -0.5812555551528931, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.16939582156973462, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 4.927898448049522e-06, | |
| "logits/chosen": -2.9190261363983154, | |
| "logits/rejected": -2.878838062286377, | |
| "logps/chosen": -262.5068359375, | |
| "logps/rejected": -233.3218536376953, | |
| "loss": 0.6398, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.5640592575073242, | |
| "rewards/margins": 0.16638214886188507, | |
| "rewards/rejected": -0.7304413914680481, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.18068887634105024, | |
| "grad_norm": 8.5, | |
| "learning_rate": 4.9024883153802095e-06, | |
| "logits/chosen": -2.9224305152893066, | |
| "logits/rejected": -2.8813347816467285, | |
| "logps/chosen": -225.17919921875, | |
| "logps/rejected": -219.84487915039062, | |
| "loss": 0.64, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.6268054842948914, | |
| "rewards/margins": 0.16882720589637756, | |
| "rewards/rejected": -0.7956327199935913, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1919819311123659, | |
| "grad_norm": 90.5, | |
| "learning_rate": 4.873336400382367e-06, | |
| "logits/chosen": -2.952296733856201, | |
| "logits/rejected": -2.915030002593994, | |
| "logps/chosen": -253.6019744873047, | |
| "logps/rejected": -242.11801147460938, | |
| "loss": 0.6533, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.6432451009750366, | |
| "rewards/margins": 0.15737922489643097, | |
| "rewards/rejected": -0.8006243705749512, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.20327498588368154, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 4.8404881060324375e-06, | |
| "logits/chosen": -2.9133620262145996, | |
| "logits/rejected": -2.8687403202056885, | |
| "logps/chosen": -250.570068359375, | |
| "logps/rejected": -209.63168334960938, | |
| "loss": 0.6597, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.7379518747329712, | |
| "rewards/margins": 0.15293964743614197, | |
| "rewards/rejected": -0.8908915519714355, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.21456804065499718, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 4.803994592274272e-06, | |
| "logits/chosen": -2.930187225341797, | |
| "logits/rejected": -2.9065961837768555, | |
| "logps/chosen": -255.08895874023438, | |
| "logps/rejected": -240.781982421875, | |
| "loss": 0.691, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.6726095676422119, | |
| "rewards/margins": 0.09454379975795746, | |
| "rewards/rejected": -0.7671533226966858, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.22586109542631283, | |
| "grad_norm": 8.625, | |
| "learning_rate": 4.763912696339506e-06, | |
| "logits/chosen": -2.908806800842285, | |
| "logits/rejected": -2.87739896774292, | |
| "logps/chosen": -244.31326293945312, | |
| "logps/rejected": -251.88284301757812, | |
| "loss": 0.6507, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.6921225786209106, | |
| "rewards/margins": 0.16923566162586212, | |
| "rewards/rejected": -0.8613582849502563, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.23715415019762845, | |
| "grad_norm": 11.5, | |
| "learning_rate": 4.720304844225781e-06, | |
| "logits/chosen": -2.9554803371429443, | |
| "logits/rejected": -2.9158945083618164, | |
| "logps/chosen": -259.2450866699219, | |
| "logps/rejected": -220.19070434570312, | |
| "loss": 0.6435, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.634341835975647, | |
| "rewards/margins": 0.1825425922870636, | |
| "rewards/rejected": -0.8168843388557434, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2484472049689441, | |
| "grad_norm": 13.125, | |
| "learning_rate": 4.6732389534706655e-06, | |
| "logits/chosen": -2.941028118133545, | |
| "logits/rejected": -2.9083430767059326, | |
| "logps/chosen": -279.50592041015625, | |
| "logps/rejected": -249.6971893310547, | |
| "loss": 0.6602, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.7295074462890625, | |
| "rewards/margins": 0.16120071709156036, | |
| "rewards/rejected": -0.8907082676887512, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 10.25, | |
| "learning_rate": 4.622788327372736e-06, | |
| "logits/chosen": -2.9610531330108643, | |
| "logits/rejected": -2.9274752140045166, | |
| "logps/chosen": -277.1114196777344, | |
| "logps/rejected": -249.110107421875, | |
| "loss": 0.6483, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.7016745209693909, | |
| "rewards/margins": 0.18496869504451752, | |
| "rewards/rejected": -0.8866432309150696, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.27103331451157536, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 4.569031540824526e-06, | |
| "logits/chosen": -2.933371067047119, | |
| "logits/rejected": -2.890542984008789, | |
| "logps/chosen": -237.04415893554688, | |
| "logps/rejected": -225.23086547851562, | |
| "loss": 0.6613, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.6652556657791138, | |
| "rewards/margins": 0.1700500100851059, | |
| "rewards/rejected": -0.8353056907653809, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.282326369282891, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 4.512052317935184e-06, | |
| "logits/chosen": -2.958394765853882, | |
| "logits/rejected": -2.9169325828552246, | |
| "logps/chosen": -239.5146942138672, | |
| "logps/rejected": -229.6261749267578, | |
| "loss": 0.6111, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.7014538049697876, | |
| "rewards/margins": 0.305302232503891, | |
| "rewards/rejected": -1.006756067276001, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.29361942405420666, | |
| "grad_norm": 11.375, | |
| "learning_rate": 4.45193940163342e-06, | |
| "logits/chosen": -2.9349873065948486, | |
| "logits/rejected": -2.9028127193450928, | |
| "logps/chosen": -242.2169189453125, | |
| "logps/rejected": -246.52584838867188, | |
| "loss": 0.6804, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.7285553812980652, | |
| "rewards/margins": 0.13187848031520844, | |
| "rewards/rejected": -0.8604338765144348, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3049124788255223, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 4.3887864154538426e-06, | |
| "logits/chosen": -2.948988437652588, | |
| "logits/rejected": -2.934638500213623, | |
| "logps/chosen": -241.4674072265625, | |
| "logps/rejected": -244.40072631835938, | |
| "loss": 0.6575, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.7441710233688354, | |
| "rewards/margins": 0.2095833122730255, | |
| "rewards/rejected": -0.9537544250488281, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.31620553359683795, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 4.322691717721927e-06, | |
| "logits/chosen": -2.9385104179382324, | |
| "logits/rejected": -2.9133810997009277, | |
| "logps/chosen": -264.28955078125, | |
| "logps/rejected": -279.1156311035156, | |
| "loss": 0.6609, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.7867005467414856, | |
| "rewards/margins": 0.16506877541542053, | |
| "rewards/rejected": -0.9517693519592285, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3274985883681536, | |
| "grad_norm": 14.125, | |
| "learning_rate": 4.253758248364747e-06, | |
| "logits/chosen": -2.9210855960845947, | |
| "logits/rejected": -2.913820743560791, | |
| "logps/chosen": -258.10015869140625, | |
| "logps/rejected": -252.4558563232422, | |
| "loss": 0.7288, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.8864612579345703, | |
| "rewards/margins": 0.033281516283750534, | |
| "rewards/rejected": -0.9197427034378052, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.33879164313946925, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 4.182093368586034e-06, | |
| "logits/chosen": -2.935784101486206, | |
| "logits/rejected": -2.9176745414733887, | |
| "logps/chosen": -277.13153076171875, | |
| "logps/rejected": -258.02166748046875, | |
| "loss": 0.6835, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.6932904124259949, | |
| "rewards/margins": 0.10739298164844513, | |
| "rewards/rejected": -0.8006833791732788, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3500846979107849, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 4.107808693655262e-06, | |
| "logits/chosen": -2.9573862552642822, | |
| "logits/rejected": -2.9266855716705322, | |
| "logps/chosen": -260.0476989746094, | |
| "logps/rejected": -220.8186492919922, | |
| "loss": 0.656, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.6015889644622803, | |
| "rewards/margins": 0.16243910789489746, | |
| "rewards/rejected": -0.7640281319618225, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3613777526821005, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 4.0310199190712016e-06, | |
| "logits/chosen": -2.90745210647583, | |
| "logits/rejected": -2.877584934234619, | |
| "logps/chosen": -238.2447509765625, | |
| "logps/rejected": -260.91278076171875, | |
| "loss": 0.6433, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.6627223491668701, | |
| "rewards/margins": 0.18875077366828918, | |
| "rewards/rejected": -0.8514731526374817, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.37267080745341613, | |
| "grad_norm": 9.375, | |
| "learning_rate": 3.951846640370666e-06, | |
| "logits/chosen": -2.9316892623901367, | |
| "logits/rejected": -2.8965978622436523, | |
| "logps/chosen": -249.91738891601562, | |
| "logps/rejected": -269.43255615234375, | |
| "loss": 0.6525, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.5791417360305786, | |
| "rewards/margins": 0.18505547940731049, | |
| "rewards/rejected": -0.7641971707344055, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3839638622247318, | |
| "grad_norm": 10.0, | |
| "learning_rate": 3.870412166863107e-06, | |
| "logits/chosen": -2.9794411659240723, | |
| "logits/rejected": -2.947648525238037, | |
| "logps/chosen": -259.78289794921875, | |
| "logps/rejected": -249.4364776611328, | |
| "loss": 0.6399, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.7674423456192017, | |
| "rewards/margins": 0.24263504147529602, | |
| "rewards/rejected": -1.0100773572921753, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3952569169960474, | |
| "grad_norm": 13.1875, | |
| "learning_rate": 3.786843329581147e-06, | |
| "logits/chosen": -2.9657294750213623, | |
| "logits/rejected": -2.944413900375366, | |
| "logps/chosen": -254.1212158203125, | |
| "logps/rejected": -251.4638214111328, | |
| "loss": 0.6688, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9390102624893188, | |
| "rewards/margins": 0.23788292706012726, | |
| "rewards/rejected": -1.1768931150436401, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.40654997176736307, | |
| "grad_norm": 10.25, | |
| "learning_rate": 3.701270283746168e-06, | |
| "logits/chosen": -2.997607707977295, | |
| "logits/rejected": -2.9588074684143066, | |
| "logps/chosen": -250.17568969726562, | |
| "logps/rejected": -234.97073364257812, | |
| "loss": 0.6788, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.9304245114326477, | |
| "rewards/margins": 0.18224455416202545, | |
| "rewards/rejected": -1.1126692295074463, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4178430265386787, | |
| "grad_norm": 10.875, | |
| "learning_rate": 3.613826306056607e-06, | |
| "logits/chosen": -2.8931360244750977, | |
| "logits/rejected": -2.8573451042175293, | |
| "logps/chosen": -274.30328369140625, | |
| "logps/rejected": -245.84512329101562, | |
| "loss": 0.6967, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -1.037747859954834, | |
| "rewards/margins": 0.18897393345832825, | |
| "rewards/rejected": -1.2267218828201294, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.42913608130999437, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 3.5246475871146646e-06, | |
| "logits/chosen": -2.9629368782043457, | |
| "logits/rejected": -2.920045852661133, | |
| "logps/chosen": -286.22943115234375, | |
| "logps/rejected": -249.3096160888672, | |
| "loss": 0.6428, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9461029767990112, | |
| "rewards/margins": 0.26600000262260437, | |
| "rewards/rejected": -1.2121028900146484, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.44042913608131, | |
| "grad_norm": 8.5, | |
| "learning_rate": 3.433873019314731e-06, | |
| "logits/chosen": -2.905508279800415, | |
| "logits/rejected": -2.8779773712158203, | |
| "logps/chosen": -249.64663696289062, | |
| "logps/rejected": -215.81161499023438, | |
| "loss": 0.6645, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.9739023447036743, | |
| "rewards/margins": 0.22194626927375793, | |
| "rewards/rejected": -1.1958487033843994, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.45172219085262566, | |
| "grad_norm": 8.5, | |
| "learning_rate": 3.341643980523871e-06, | |
| "logits/chosen": -2.905945062637329, | |
| "logits/rejected": -2.8710389137268066, | |
| "logps/chosen": -292.928955078125, | |
| "logps/rejected": -241.6216278076172, | |
| "loss": 0.6466, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.7866181135177612, | |
| "rewards/margins": 0.2684091627597809, | |
| "rewards/rejected": -1.0550272464752197, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.46301524562394125, | |
| "grad_norm": 12.8125, | |
| "learning_rate": 3.2481041138912784e-06, | |
| "logits/chosen": -2.8903775215148926, | |
| "logits/rejected": -2.861661672592163, | |
| "logps/chosen": -260.47174072265625, | |
| "logps/rejected": -255.26931762695312, | |
| "loss": 0.6732, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -1.0014512538909912, | |
| "rewards/margins": 0.2172473967075348, | |
| "rewards/rejected": -1.2186987400054932, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4743083003952569, | |
| "grad_norm": 9.75, | |
| "learning_rate": 3.1533991041296484e-06, | |
| "logits/chosen": -2.962191104888916, | |
| "logits/rejected": -2.932969331741333, | |
| "logps/chosen": -259.97283935546875, | |
| "logps/rejected": -248.1492919921875, | |
| "loss": 0.6506, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9169241189956665, | |
| "rewards/margins": 0.20076242089271545, | |
| "rewards/rejected": -1.1176865100860596, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.48560135516657255, | |
| "grad_norm": 11.375, | |
| "learning_rate": 3.0576764506168893e-06, | |
| "logits/chosen": -2.9573709964752197, | |
| "logits/rejected": -2.9248433113098145, | |
| "logps/chosen": -281.449462890625, | |
| "logps/rejected": -258.242431640625, | |
| "loss": 0.6653, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.8910619616508484, | |
| "rewards/margins": 0.1914868950843811, | |
| "rewards/rejected": -1.0825488567352295, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4968944099378882, | |
| "grad_norm": 8.5, | |
| "learning_rate": 2.9610852376715625e-06, | |
| "logits/chosen": -2.9228568077087402, | |
| "logits/rejected": -2.8933017253875732, | |
| "logps/chosen": -258.58245849609375, | |
| "logps/rejected": -248.20846557617188, | |
| "loss": 0.6594, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.9236515164375305, | |
| "rewards/margins": 0.2783660292625427, | |
| "rewards/rejected": -1.2020175457000732, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5081874647092038, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 2.8637759023598467e-06, | |
| "logits/chosen": -2.9429585933685303, | |
| "logits/rejected": -2.9177327156066895, | |
| "logps/chosen": -246.7409210205078, | |
| "logps/rejected": -252.89218139648438, | |
| "loss": 0.6842, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.8663239479064941, | |
| "rewards/margins": 0.17053480446338654, | |
| "rewards/rejected": -1.0368586778640747, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 9.25, | |
| "learning_rate": 2.7659000001956376e-06, | |
| "logits/chosen": -2.9368577003479004, | |
| "logits/rejected": -2.899815559387207, | |
| "logps/chosen": -258.90765380859375, | |
| "logps/rejected": -234.8472137451172, | |
| "loss": 0.6159, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.8144370317459106, | |
| "rewards/margins": 0.30742961168289185, | |
| "rewards/rejected": -1.1218667030334473, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.5307735742518351, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 2.6676099690987227e-06, | |
| "logits/chosen": -2.938596248626709, | |
| "logits/rejected": -2.9030938148498535, | |
| "logps/chosen": -251.0491485595703, | |
| "logps/rejected": -227.3994140625, | |
| "loss": 0.6612, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.7350276708602905, | |
| "rewards/margins": 0.21316580474376678, | |
| "rewards/rejected": -0.9481935501098633, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5420666290231507, | |
| "grad_norm": 8.625, | |
| "learning_rate": 2.5690588919786307e-06, | |
| "logits/chosen": -2.9533700942993164, | |
| "logits/rejected": -2.9271883964538574, | |
| "logps/chosen": -265.0854187011719, | |
| "logps/rejected": -239.2405242919922, | |
| "loss": 0.6447, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.8996723294258118, | |
| "rewards/margins": 0.29516300559043884, | |
| "rewards/rejected": -1.1948354244232178, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5533596837944664, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 2.4704002583139426e-06, | |
| "logits/chosen": -2.9021379947662354, | |
| "logits/rejected": -2.8771567344665527, | |
| "logps/chosen": -264.7264709472656, | |
| "logps/rejected": -246.2521209716797, | |
| "loss": 0.6077, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.8024009466171265, | |
| "rewards/margins": 0.2971910834312439, | |
| "rewards/rejected": -1.0995919704437256, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.564652738565782, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 2.3717877250983864e-06, | |
| "logits/chosen": -2.938613176345825, | |
| "logits/rejected": -2.904693841934204, | |
| "logps/chosen": -250.9031219482422, | |
| "logps/rejected": -249.9896240234375, | |
| "loss": 0.6574, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.845720648765564, | |
| "rewards/margins": 0.24266941845417023, | |
| "rewards/rejected": -1.0883901119232178, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5759457933370977, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 2.2733748775260256e-06, | |
| "logits/chosen": -2.8921456336975098, | |
| "logits/rejected": -2.871410846710205, | |
| "logps/chosen": -226.85903930664062, | |
| "logps/rejected": -219.52587890625, | |
| "loss": 0.6478, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.7570312023162842, | |
| "rewards/margins": 0.20835983753204346, | |
| "rewards/rejected": -0.9653909802436829, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5872388481084133, | |
| "grad_norm": 10.25, | |
| "learning_rate": 2.1753149897882847e-06, | |
| "logits/chosen": -2.9432761669158936, | |
| "logits/rejected": -2.892080545425415, | |
| "logps/chosen": -238.2821044921875, | |
| "logps/rejected": -222.26644897460938, | |
| "loss": 0.5833, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.8168408274650574, | |
| "rewards/margins": 0.41468945145606995, | |
| "rewards/rejected": -1.2315301895141602, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.598531902879729, | |
| "grad_norm": 10.0, | |
| "learning_rate": 2.0777607863553423e-06, | |
| "logits/chosen": -2.9561731815338135, | |
| "logits/rejected": -2.9202933311462402, | |
| "logps/chosen": -236.2799835205078, | |
| "logps/rejected": -216.89901733398438, | |
| "loss": 0.673, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.8298661112785339, | |
| "rewards/margins": 0.18440715968608856, | |
| "rewards/rejected": -1.0142732858657837, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6098249576510446, | |
| "grad_norm": 12.5, | |
| "learning_rate": 1.9808642041136923e-06, | |
| "logits/chosen": -2.9524600505828857, | |
| "logits/rejected": -2.9057111740112305, | |
| "logps/chosen": -239.9007110595703, | |
| "logps/rejected": -232.8452606201172, | |
| "loss": 0.6733, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.8775818943977356, | |
| "rewards/margins": 0.22822308540344238, | |
| "rewards/rejected": -1.1058050394058228, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.6211180124223602, | |
| "grad_norm": 12.0625, | |
| "learning_rate": 1.884776155730342e-06, | |
| "logits/chosen": -2.935317277908325, | |
| "logits/rejected": -2.9047186374664307, | |
| "logps/chosen": -250.3077850341797, | |
| "logps/rejected": -248.65652465820312, | |
| "loss": 0.7337, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.9318073391914368, | |
| "rewards/margins": 0.11215513944625854, | |
| "rewards/rejected": -1.0439624786376953, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.6324110671936759, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 1.7896462946121873e-06, | |
| "logits/chosen": -2.951495885848999, | |
| "logits/rejected": -2.9132513999938965, | |
| "logps/chosen": -269.6780700683594, | |
| "logps/rejected": -256.81597900390625, | |
| "loss": 0.6249, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.8714116215705872, | |
| "rewards/margins": 0.29088258743286133, | |
| "rewards/rejected": -1.1622941493988037, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6437041219649915, | |
| "grad_norm": 9.75, | |
| "learning_rate": 1.695622781826638e-06, | |
| "logits/chosen": -2.954216718673706, | |
| "logits/rejected": -2.906550884246826, | |
| "logps/chosen": -257.61883544921875, | |
| "logps/rejected": -228.7353973388672, | |
| "loss": 0.6324, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.7798657417297363, | |
| "rewards/margins": 0.3046206533908844, | |
| "rewards/rejected": -1.0844862461090088, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6549971767363072, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 1.6028520553465053e-06, | |
| "logits/chosen": -2.943326473236084, | |
| "logits/rejected": -2.907252788543701, | |
| "logps/chosen": -255.4546356201172, | |
| "logps/rejected": -229.12185668945312, | |
| "loss": 0.6212, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.125480055809021, | |
| "rewards/margins": 0.27197134494781494, | |
| "rewards/rejected": -1.3974515199661255, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6662902315076228, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 1.5114786019785416e-06, | |
| "logits/chosen": -2.9562203884124756, | |
| "logits/rejected": -2.9087963104248047, | |
| "logps/chosen": -244.3138427734375, | |
| "logps/rejected": -240.51870727539062, | |
| "loss": 0.6511, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.9466648101806641, | |
| "rewards/margins": 0.2708016335964203, | |
| "rewards/rejected": -1.2174664735794067, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6775832862789385, | |
| "grad_norm": 11.0, | |
| "learning_rate": 1.4216447323308444e-06, | |
| "logits/chosen": -2.9042696952819824, | |
| "logits/rejected": -2.883103609085083, | |
| "logps/chosen": -240.0981903076172, | |
| "logps/rejected": -246.4265899658203, | |
| "loss": 0.6582, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1731393337249756, | |
| "rewards/margins": 0.262151300907135, | |
| "rewards/rejected": -1.4352905750274658, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6888763410502541, | |
| "grad_norm": 7.78125, | |
| "learning_rate": 1.3334903591696055e-06, | |
| "logits/chosen": -2.924773931503296, | |
| "logits/rejected": -2.8905885219573975, | |
| "logps/chosen": -243.14501953125, | |
| "logps/rejected": -215.37881469726562, | |
| "loss": 0.6368, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.9755071401596069, | |
| "rewards/margins": 0.27910715341567993, | |
| "rewards/rejected": -1.2546144723892212, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.7001693958215698, | |
| "grad_norm": 13.875, | |
| "learning_rate": 1.247152779510411e-06, | |
| "logits/chosen": -2.941826820373535, | |
| "logits/rejected": -2.906125783920288, | |
| "logps/chosen": -248.43362426757812, | |
| "logps/rejected": -236.13525390625, | |
| "loss": 0.7073, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -1.108798623085022, | |
| "rewards/margins": 0.21407607197761536, | |
| "rewards/rejected": -1.3228747844696045, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.7114624505928854, | |
| "grad_norm": 11.625, | |
| "learning_rate": 1.1627664607834593e-06, | |
| "logits/chosen": -2.946981191635132, | |
| "logits/rejected": -2.9171371459960938, | |
| "logps/chosen": -251.35324096679688, | |
| "logps/rejected": -261.75396728515625, | |
| "loss": 0.6583, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.9292078018188477, | |
| "rewards/margins": 0.23983097076416016, | |
| "rewards/rejected": -1.1690387725830078, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.722755505364201, | |
| "grad_norm": 14.5625, | |
| "learning_rate": 1.0804628314057659e-06, | |
| "logits/chosen": -2.920762538909912, | |
| "logits/rejected": -2.8985981941223145, | |
| "logps/chosen": -235.6101837158203, | |
| "logps/rejected": -207.3476104736328, | |
| "loss": 0.6607, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.0142499208450317, | |
| "rewards/margins": 0.2305765599012375, | |
| "rewards/rejected": -1.2448265552520752, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.7340485601355167, | |
| "grad_norm": 16.0, | |
| "learning_rate": 1.0003700760864931e-06, | |
| "logits/chosen": -2.902686595916748, | |
| "logits/rejected": -2.8842315673828125, | |
| "logps/chosen": -252.89035034179688, | |
| "logps/rejected": -250.73623657226562, | |
| "loss": 0.6813, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -1.051636815071106, | |
| "rewards/margins": 0.21588905155658722, | |
| "rewards/rejected": -1.2675259113311768, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7453416149068323, | |
| "grad_norm": 12.3125, | |
| "learning_rate": 9.226129361842498e-07, | |
| "logits/chosen": -2.9301822185516357, | |
| "logits/rejected": -2.881704807281494, | |
| "logps/chosen": -236.1824951171875, | |
| "logps/rejected": -207.68685913085938, | |
| "loss": 0.6363, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.060085415840149, | |
| "rewards/margins": 0.3055773675441742, | |
| "rewards/rejected": -1.365662693977356, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.756634669678148, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 8.473125154272563e-07, | |
| "logits/chosen": -2.8978400230407715, | |
| "logits/rejected": -2.845043659210205, | |
| "logps/chosen": -245.31027221679688, | |
| "logps/rejected": -251.3234100341797, | |
| "loss": 0.6158, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.980219841003418, | |
| "rewards/margins": 0.3760405480861664, | |
| "rewards/rejected": -1.3562604188919067, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7679277244494636, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 7.74586091298995e-07, | |
| "logits/chosen": -2.9392664432525635, | |
| "logits/rejected": -2.8935275077819824, | |
| "logps/chosen": -263.9918518066406, | |
| "logps/rejected": -226.14920043945312, | |
| "loss": 0.6485, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.0862810611724854, | |
| "rewards/margins": 0.3031725287437439, | |
| "rewards/rejected": -1.389453649520874, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 7.045469323830731e-07, | |
| "logits/chosen": -2.908051013946533, | |
| "logits/rejected": -2.877911329269409, | |
| "logps/chosen": -247.0584259033203, | |
| "logps/rejected": -250.07669067382812, | |
| "loss": 0.677, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.146324872970581, | |
| "rewards/margins": 0.17685498297214508, | |
| "rewards/rejected": -1.3231797218322754, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7905138339920948, | |
| "grad_norm": 9.625, | |
| "learning_rate": 6.373041219518059e-07, | |
| "logits/chosen": -2.9572408199310303, | |
| "logits/rejected": -2.937887191772461, | |
| "logps/chosen": -252.33755493164062, | |
| "logps/rejected": -229.02926635742188, | |
| "loss": 0.6706, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.0239284038543701, | |
| "rewards/margins": 0.20092003047466278, | |
| "rewards/rejected": -1.2248482704162598, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.8018068887634106, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 5.729623880732455e-07, | |
| "logits/chosen": -2.9367566108703613, | |
| "logits/rejected": -2.9192299842834473, | |
| "logps/chosen": -245.69668579101562, | |
| "logps/rejected": -225.72848510742188, | |
| "loss": 0.6588, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.0508708953857422, | |
| "rewards/margins": 0.26678308844566345, | |
| "rewards/rejected": -1.3176538944244385, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.8130999435347261, | |
| "grad_norm": 8.5, | |
| "learning_rate": 5.116219405012865e-07, | |
| "logits/chosen": -2.9278311729431152, | |
| "logits/rejected": -2.8987011909484863, | |
| "logps/chosen": -246.24691772460938, | |
| "logps/rejected": -206.0987548828125, | |
| "loss": 0.6105, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.0415557622909546, | |
| "rewards/margins": 0.3364638686180115, | |
| "rewards/rejected": -1.3780196905136108, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.8243929983060417, | |
| "grad_norm": 7.75, | |
| "learning_rate": 4.533783146028542e-07, | |
| "logits/chosen": -2.9237890243530273, | |
| "logits/rejected": -2.8972859382629395, | |
| "logps/chosen": -234.02517700195312, | |
| "logps/rejected": -260.5685119628906, | |
| "loss": 0.6314, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0762670040130615, | |
| "rewards/margins": 0.31192946434020996, | |
| "rewards/rejected": -1.388196587562561, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.8356860530773574, | |
| "grad_norm": 7.71875, | |
| "learning_rate": 3.9832222256528633e-07, | |
| "logits/chosen": -2.9241185188293457, | |
| "logits/rejected": -2.8836278915405273, | |
| "logps/chosen": -238.7316131591797, | |
| "logps/rejected": -258.7846984863281, | |
| "loss": 0.625, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.1147111654281616, | |
| "rewards/margins": 0.31473425030708313, | |
| "rewards/rejected": -1.4294453859329224, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.846979107848673, | |
| "grad_norm": 7.8125, | |
| "learning_rate": 3.4653941211561265e-07, | |
| "logits/chosen": -2.918570041656494, | |
| "logits/rejected": -2.8599822521209717, | |
| "logps/chosen": -228.1241912841797, | |
| "logps/rejected": -200.6188507080078, | |
| "loss": 0.591, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.1377301216125488, | |
| "rewards/margins": 0.45671454071998596, | |
| "rewards/rejected": -1.5944445133209229, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8582721626199887, | |
| "grad_norm": 11.9375, | |
| "learning_rate": 2.9811053297180454e-07, | |
| "logits/chosen": -2.9596495628356934, | |
| "logits/rejected": -2.938786268234253, | |
| "logps/chosen": -231.0944061279297, | |
| "logps/rejected": -241.2841339111328, | |
| "loss": 0.6626, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.0585596561431885, | |
| "rewards/margins": 0.2590670883655548, | |
| "rewards/rejected": -1.317626714706421, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 2.531110112339638e-07, | |
| "logits/chosen": -2.933401584625244, | |
| "logits/rejected": -2.920301914215088, | |
| "logps/chosen": -237.66879272460938, | |
| "logps/rejected": -238.02297973632812, | |
| "loss": 0.6987, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -1.0770819187164307, | |
| "rewards/margins": 0.14950065314769745, | |
| "rewards/rejected": -1.2265825271606445, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.88085827216262, | |
| "grad_norm": 12.5, | |
| "learning_rate": 2.116109319111015e-07, | |
| "logits/chosen": -2.912548065185547, | |
| "logits/rejected": -2.8936960697174072, | |
| "logps/chosen": -235.7461395263672, | |
| "logps/rejected": -246.876708984375, | |
| "loss": 0.7254, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -1.1536673307418823, | |
| "rewards/margins": 0.14096426963806152, | |
| "rewards/rejected": -1.2946317195892334, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8921513269339356, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 1.736749297664539e-07, | |
| "logits/chosen": -2.9254114627838135, | |
| "logits/rejected": -2.89033842086792, | |
| "logps/chosen": -263.29071044921875, | |
| "logps/rejected": -274.1188659667969, | |
| "loss": 0.6559, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9891961216926575, | |
| "rewards/margins": 0.2855607271194458, | |
| "rewards/rejected": -1.2747570276260376, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.9034443817052513, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 1.3936208865134666e-07, | |
| "logits/chosen": -2.930879831314087, | |
| "logits/rejected": -2.8965582847595215, | |
| "logps/chosen": -254.17044067382812, | |
| "logps/rejected": -273.36724853515625, | |
| "loss": 0.6567, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.0911872386932373, | |
| "rewards/margins": 0.27449628710746765, | |
| "rewards/rejected": -1.3656837940216064, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9147374364765669, | |
| "grad_norm": 8.125, | |
| "learning_rate": 1.0872584948439208e-07, | |
| "logits/chosen": -2.9673569202423096, | |
| "logits/rejected": -2.938894748687744, | |
| "logps/chosen": -264.1458435058594, | |
| "logps/rejected": -253.3695068359375, | |
| "loss": 0.6348, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.9376649856567383, | |
| "rewards/margins": 0.28761720657348633, | |
| "rewards/rejected": -1.225282073020935, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.9260304912478825, | |
| "grad_norm": 10.75, | |
| "learning_rate": 8.181392701932745e-08, | |
| "logits/chosen": -2.9556472301483154, | |
| "logits/rejected": -2.930018663406372, | |
| "logps/chosen": -244.6985321044922, | |
| "logps/rejected": -241.3518524169922, | |
| "loss": 0.6962, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.143431305885315, | |
| "rewards/margins": 0.20309944450855255, | |
| "rewards/rejected": -1.3465306758880615, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.9373235460191982, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 5.866823553114187e-08, | |
| "logits/chosen": -2.9636921882629395, | |
| "logits/rejected": -2.934422254562378, | |
| "logps/chosen": -249.5702667236328, | |
| "logps/rejected": -229.45053100585938, | |
| "loss": 0.6749, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.1629018783569336, | |
| "rewards/margins": 0.19255365431308746, | |
| "rewards/rejected": -1.3554556369781494, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.9486166007905138, | |
| "grad_norm": 12.6875, | |
| "learning_rate": 3.932482353621536e-08, | |
| "logits/chosen": -2.9121766090393066, | |
| "logits/rejected": -2.8778061866760254, | |
| "logps/chosen": -252.2318115234375, | |
| "logps/rejected": -242.0477294921875, | |
| "loss": 0.7142, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -1.1486766338348389, | |
| "rewards/margins": 0.10505317151546478, | |
| "rewards/rejected": -1.253730058670044, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.9599096555618295, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 2.381381764815671e-08, | |
| "logits/chosen": -2.9521381855010986, | |
| "logits/rejected": -2.9332849979400635, | |
| "logps/chosen": -248.31655883789062, | |
| "logps/rejected": -244.4917449951172, | |
| "loss": 0.7399, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.1000101566314697, | |
| "rewards/margins": 0.10928668826818466, | |
| "rewards/rejected": -1.2092968225479126, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9712027103331451, | |
| "grad_norm": 11.125, | |
| "learning_rate": 1.2159375656770056e-08, | |
| "logits/chosen": -2.964635133743286, | |
| "logits/rejected": -2.9422874450683594, | |
| "logps/chosen": -266.625732421875, | |
| "logps/rejected": -265.8684997558594, | |
| "loss": 0.693, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -1.1866648197174072, | |
| "rewards/margins": 0.21592363715171814, | |
| "rewards/rejected": -1.4025886058807373, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9824957651044608, | |
| "grad_norm": 11.5, | |
| "learning_rate": 4.379648903234668e-09, | |
| "logits/chosen": -2.930752754211426, | |
| "logits/rejected": -2.8959178924560547, | |
| "logps/chosen": -268.1014404296875, | |
| "logps/rejected": -252.4002227783203, | |
| "loss": 0.6735, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.995762825012207, | |
| "rewards/margins": 0.236598402261734, | |
| "rewards/rejected": -1.2323614358901978, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9937888198757764, | |
| "grad_norm": 9.75, | |
| "learning_rate": 4.867540100961244e-10, | |
| "logits/chosen": -2.928917646408081, | |
| "logits/rejected": -2.881701946258545, | |
| "logps/chosen": -294.4600524902344, | |
| "logps/rejected": -246.2449493408203, | |
| "loss": 0.6221, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.1883199214935303, | |
| "rewards/margins": 0.35472264885902405, | |
| "rewards/rejected": -1.543042540550232, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9994353472614342, | |
| "step": 885, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6616038349388683, | |
| "train_runtime": 9964.5835, | |
| "train_samples_per_second": 1.422, | |
| "train_steps_per_second": 0.089 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 885, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |