| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 1875, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 2.6595744680851065e-08, | |
| "logits/chosen": 0.09963524341583252, | |
| "logits/rejected": 0.17411324381828308, | |
| "logps/chosen": -378.2606506347656, | |
| "logps/rejected": -386.1599426269531, | |
| "loss": 0.21, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 2.6595744680851066e-07, | |
| "logits/chosen": 0.1553082913160324, | |
| "logits/rejected": 0.24615353345870972, | |
| "logps/chosen": -397.8062438964844, | |
| "logps/rejected": -394.25341796875, | |
| "loss": 0.2045, | |
| "rewards/accuracies": 0.2777777910232544, | |
| "rewards/chosen": -0.0005843756953254342, | |
| "rewards/margins": -3.1117440812522545e-05, | |
| "rewards/rejected": -0.0005532581708393991, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5.319148936170213e-07, | |
| "logits/chosen": 0.3212467432022095, | |
| "logits/rejected": 0.24689868092536926, | |
| "logps/chosen": -409.90167236328125, | |
| "logps/rejected": -409.5082092285156, | |
| "loss": 0.2087, | |
| "rewards/accuracies": 0.3687500059604645, | |
| "rewards/chosen": -0.0009676044574007392, | |
| "rewards/margins": 5.4393574828282e-05, | |
| "rewards/rejected": -0.0010219980031251907, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 7.97872340425532e-07, | |
| "logits/chosen": 0.22952334582805634, | |
| "logits/rejected": 0.2672143876552582, | |
| "logps/chosen": -385.3333740234375, | |
| "logps/rejected": -392.8493957519531, | |
| "loss": 0.2137, | |
| "rewards/accuracies": 0.28125, | |
| "rewards/chosen": -0.0009394015069119632, | |
| "rewards/margins": -8.656540012452751e-05, | |
| "rewards/rejected": -0.0008528361213393509, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.0638297872340427e-06, | |
| "logits/chosen": 0.24455244839191437, | |
| "logits/rejected": 0.14430980384349823, | |
| "logps/chosen": -414.5653381347656, | |
| "logps/rejected": -410.3203125, | |
| "loss": 0.2186, | |
| "rewards/accuracies": 0.35624998807907104, | |
| "rewards/chosen": -0.0006429333006963134, | |
| "rewards/margins": 7.077780173858628e-05, | |
| "rewards/rejected": -0.0007137111388146877, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.3297872340425533e-06, | |
| "logits/chosen": 0.16719885170459747, | |
| "logits/rejected": 0.19559958577156067, | |
| "logps/chosen": -388.90740966796875, | |
| "logps/rejected": -379.8285827636719, | |
| "loss": 0.2026, | |
| "rewards/accuracies": 0.35624998807907104, | |
| "rewards/chosen": -3.721771645359695e-05, | |
| "rewards/margins": -9.795947335078381e-06, | |
| "rewards/rejected": -2.742178003245499e-05, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.595744680851064e-06, | |
| "logits/chosen": 0.2627559304237366, | |
| "logits/rejected": 0.2265670746564865, | |
| "logps/chosen": -355.0706787109375, | |
| "logps/rejected": -372.27508544921875, | |
| "loss": 0.2185, | |
| "rewards/accuracies": 0.39375001192092896, | |
| "rewards/chosen": 0.00125337322242558, | |
| "rewards/margins": 0.00012534942652564496, | |
| "rewards/rejected": 0.0011280238395556808, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1.8617021276595745e-06, | |
| "logits/chosen": 0.15168386697769165, | |
| "logits/rejected": 0.3290417194366455, | |
| "logps/chosen": -432.84197998046875, | |
| "logps/rejected": -418.5240783691406, | |
| "loss": 0.2123, | |
| "rewards/accuracies": 0.40625, | |
| "rewards/chosen": 0.0038177906535565853, | |
| "rewards/margins": 0.00015494157560169697, | |
| "rewards/rejected": 0.003662849310785532, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "logits/chosen": 0.150368332862854, | |
| "logits/rejected": 0.2564127445220947, | |
| "logps/chosen": -402.29449462890625, | |
| "logps/rejected": -372.3401184082031, | |
| "loss": 0.2282, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": 0.006463122554123402, | |
| "rewards/margins": 0.00028264717548154294, | |
| "rewards/rejected": 0.00618047546595335, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.393617021276596e-06, | |
| "logits/chosen": 0.19279547035694122, | |
| "logits/rejected": 0.3653843104839325, | |
| "logps/chosen": -433.8819885253906, | |
| "logps/rejected": -433.3250427246094, | |
| "loss": 0.2121, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": 0.008460921235382557, | |
| "rewards/margins": 0.0003720732347574085, | |
| "rewards/rejected": 0.008088847622275352, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.6595744680851065e-06, | |
| "logits/chosen": 0.187942236661911, | |
| "logits/rejected": 0.30413854122161865, | |
| "logps/chosen": -415.242431640625, | |
| "logps/rejected": -411.3409118652344, | |
| "loss": 0.2096, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": 0.006011855788528919, | |
| "rewards/margins": 0.0004307249910198152, | |
| "rewards/rejected": 0.005581130273640156, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.9255319148936174e-06, | |
| "logits/chosen": 0.18410541117191315, | |
| "logits/rejected": 0.22288016974925995, | |
| "logps/chosen": -402.7239685058594, | |
| "logps/rejected": -401.08685302734375, | |
| "loss": 0.2072, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": 0.002768724225461483, | |
| "rewards/margins": 0.00032299422309733927, | |
| "rewards/rejected": 0.0024457299150526524, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.191489361702128e-06, | |
| "logits/chosen": 0.23630130290985107, | |
| "logits/rejected": 0.2844993472099304, | |
| "logps/chosen": -436.46624755859375, | |
| "logps/rejected": -418.0640563964844, | |
| "loss": 0.2094, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": 0.0015008236514404416, | |
| "rewards/margins": 0.000703630386851728, | |
| "rewards/rejected": 0.0007971933227963746, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.457446808510639e-06, | |
| "logits/chosen": 0.1779193878173828, | |
| "logits/rejected": 0.2234630137681961, | |
| "logps/chosen": -445.2749938964844, | |
| "logps/rejected": -395.0121154785156, | |
| "loss": 0.218, | |
| "rewards/accuracies": 0.45625001192092896, | |
| "rewards/chosen": 0.0037746974267065525, | |
| "rewards/margins": 0.000721289194189012, | |
| "rewards/rejected": 0.0030534083489328623, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.723404255319149e-06, | |
| "logits/chosen": 0.21873219311237335, | |
| "logits/rejected": 0.28560131788253784, | |
| "logps/chosen": -386.23992919921875, | |
| "logps/rejected": -392.177978515625, | |
| "loss": 0.2085, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.005661582108587027, | |
| "rewards/margins": 0.001634161570109427, | |
| "rewards/rejected": 0.004027421120554209, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.98936170212766e-06, | |
| "logits/chosen": 0.14144310355186462, | |
| "logits/rejected": 0.21909043192863464, | |
| "logps/chosen": -410.2815856933594, | |
| "logps/rejected": -409.62420654296875, | |
| "loss": 0.2042, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.002802160568535328, | |
| "rewards/margins": 0.002337489277124405, | |
| "rewards/rejected": 0.0004646713496185839, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.255319148936171e-06, | |
| "logits/chosen": 0.15275654196739197, | |
| "logits/rejected": 0.18240560591220856, | |
| "logps/chosen": -368.1477966308594, | |
| "logps/rejected": -379.97052001953125, | |
| "loss": 0.2084, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": -0.0024238377809524536, | |
| "rewards/margins": 0.0018925836775451899, | |
| "rewards/rejected": -0.0043164207600057125, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.521276595744681e-06, | |
| "logits/chosen": 0.16948816180229187, | |
| "logits/rejected": 0.27625542879104614, | |
| "logps/chosen": -439.19580078125, | |
| "logps/rejected": -423.7747497558594, | |
| "loss": 0.1964, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.010004468262195587, | |
| "rewards/margins": 0.0029372868593782187, | |
| "rewards/rejected": -0.01294175535440445, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.787234042553192e-06, | |
| "logits/chosen": 0.14220403134822845, | |
| "logits/rejected": 0.17016319930553436, | |
| "logps/chosen": -402.444091796875, | |
| "logps/rejected": -399.7784729003906, | |
| "loss": 0.2215, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.017413243651390076, | |
| "rewards/margins": 0.005264014936983585, | |
| "rewards/rejected": -0.022677261382341385, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999982660399688e-06, | |
| "logits/chosen": 0.12951701879501343, | |
| "logits/rejected": 0.24623069167137146, | |
| "logps/chosen": -468.02325439453125, | |
| "logps/rejected": -468.9815979003906, | |
| "loss": 0.1973, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.03261370584368706, | |
| "rewards/margins": 0.006805351935327053, | |
| "rewards/rejected": -0.039419058710336685, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.99937579964398e-06, | |
| "logits/chosen": 0.052798859775066376, | |
| "logits/rejected": 0.23226705193519592, | |
| "logps/chosen": -469.90753173828125, | |
| "logps/rejected": -448.4300842285156, | |
| "loss": 0.1876, | |
| "rewards/accuracies": 0.4437499940395355, | |
| "rewards/chosen": -0.05150008201599121, | |
| "rewards/margins": 0.006041003856807947, | |
| "rewards/rejected": -0.05754108354449272, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.9979021993870645e-06, | |
| "logits/chosen": 0.08435344696044922, | |
| "logits/rejected": 0.16326232254505157, | |
| "logps/chosen": -446.05584716796875, | |
| "logps/rejected": -458.54217529296875, | |
| "loss": 0.2059, | |
| "rewards/accuracies": 0.41874998807907104, | |
| "rewards/chosen": -0.06337223947048187, | |
| "rewards/margins": 0.008098254911601543, | |
| "rewards/rejected": -0.07147050648927689, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.995562370647553e-06, | |
| "logits/chosen": 0.14515891671180725, | |
| "logits/rejected": 0.1300167292356491, | |
| "logps/chosen": -490.72271728515625, | |
| "logps/rejected": -501.8885803222656, | |
| "loss": 0.2165, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.06726966798305511, | |
| "rewards/margins": 0.010882614180445671, | |
| "rewards/rejected": -0.07815226912498474, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.992357124836838e-06, | |
| "logits/chosen": 0.14549218118190765, | |
| "logits/rejected": 0.11253325641155243, | |
| "logps/chosen": -450.588623046875, | |
| "logps/rejected": -453.873291015625, | |
| "loss": 0.2008, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.0649234801530838, | |
| "rewards/margins": 0.012732243165373802, | |
| "rewards/rejected": -0.07765571773052216, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.9882875734777044e-06, | |
| "logits/chosen": 0.17049750685691833, | |
| "logits/rejected": 0.08599407970905304, | |
| "logps/chosen": -452.00152587890625, | |
| "logps/rejected": -478.6534118652344, | |
| "loss": 0.2, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.07021967321634293, | |
| "rewards/margins": 0.01245578657835722, | |
| "rewards/rejected": -0.08267545700073242, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.983355127818882e-06, | |
| "logits/chosen": 0.08762475103139877, | |
| "logits/rejected": 0.15931396186351776, | |
| "logps/chosen": -472.1898498535156, | |
| "logps/rejected": -486.1856994628906, | |
| "loss": 0.1904, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.08154710382223129, | |
| "rewards/margins": 0.02140663005411625, | |
| "rewards/rejected": -0.10295373201370239, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.977561498345639e-06, | |
| "logits/chosen": 0.052127450704574585, | |
| "logits/rejected": 0.011202364228665829, | |
| "logps/chosen": -450.86712646484375, | |
| "logps/rejected": -529.0736083984375, | |
| "loss": 0.1943, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.0922573059797287, | |
| "rewards/margins": 0.03915147855877876, | |
| "rewards/rejected": -0.13140878081321716, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.970908694186624e-06, | |
| "logits/chosen": 0.02654929831624031, | |
| "logits/rejected": 0.14042030274868011, | |
| "logps/chosen": -490.01922607421875, | |
| "logps/rejected": -497.9236755371094, | |
| "loss": 0.2082, | |
| "rewards/accuracies": 0.40625, | |
| "rewards/chosen": -0.10826190561056137, | |
| "rewards/margins": 0.021642591804265976, | |
| "rewards/rejected": -0.12990449368953705, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.9633990224171305e-06, | |
| "logits/chosen": 0.09107808768749237, | |
| "logits/rejected": 0.08074366301298141, | |
| "logps/chosen": -493.1717834472656, | |
| "logps/rejected": -526.7948608398438, | |
| "loss": 0.1922, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.09491624683141708, | |
| "rewards/margins": 0.02360902726650238, | |
| "rewards/rejected": -0.11852526664733887, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.955035087259046e-06, | |
| "logits/chosen": -0.01798112317919731, | |
| "logits/rejected": 0.14514566957950592, | |
| "logps/chosen": -473.822021484375, | |
| "logps/rejected": -492.75079345703125, | |
| "loss": 0.1917, | |
| "rewards/accuracies": 0.4000000059604645, | |
| "rewards/chosen": -0.0761452168226242, | |
| "rewards/margins": 0.02504543401300907, | |
| "rewards/rejected": -0.10119066387414932, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.945819789177756e-06, | |
| "logits/chosen": 0.012573105283081532, | |
| "logits/rejected": 0.08077313005924225, | |
| "logps/chosen": -461.69134521484375, | |
| "logps/rejected": -452.34393310546875, | |
| "loss": 0.2008, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": -0.07385671138763428, | |
| "rewards/margins": 0.022219108417630196, | |
| "rewards/rejected": -0.09607584029436111, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.935756323876306e-06, | |
| "logits/chosen": 0.0107968058437109, | |
| "logits/rejected": 0.017728794366121292, | |
| "logps/chosen": -512.6489868164062, | |
| "logps/rejected": -566.363525390625, | |
| "loss": 0.1829, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.09047379344701767, | |
| "rewards/margins": 0.04145172983407974, | |
| "rewards/rejected": -0.1319255381822586, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.924848181187199e-06, | |
| "logits/chosen": -0.0678163543343544, | |
| "logits/rejected": 0.10408651828765869, | |
| "logps/chosen": -535.5982055664062, | |
| "logps/rejected": -556.8341064453125, | |
| "loss": 0.1803, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.1051037460565567, | |
| "rewards/margins": 0.029761528596282005, | |
| "rewards/rejected": -0.13486526906490326, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.913099143862173e-06, | |
| "logits/chosen": -0.04091095179319382, | |
| "logits/rejected": -0.0030039071571081877, | |
| "logps/chosen": -553.9273681640625, | |
| "logps/rejected": -610.8055419921875, | |
| "loss": 0.1908, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.12315535545349121, | |
| "rewards/margins": 0.04520713910460472, | |
| "rewards/rejected": -0.16836249828338623, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.900513286260416e-06, | |
| "logits/chosen": -0.1246240958571434, | |
| "logits/rejected": -0.03937938064336777, | |
| "logps/chosen": -490.50030517578125, | |
| "logps/rejected": -532.2816162109375, | |
| "loss": 0.1887, | |
| "rewards/accuracies": 0.4124999940395355, | |
| "rewards/chosen": -0.11805293709039688, | |
| "rewards/margins": 0.02471679449081421, | |
| "rewards/rejected": -0.1427697390317917, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.887094972935645e-06, | |
| "logits/chosen": -0.08421911299228668, | |
| "logits/rejected": 0.008182203397154808, | |
| "logps/chosen": -535.670654296875, | |
| "logps/rejected": -536.1793212890625, | |
| "loss": 0.1949, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.10835033655166626, | |
| "rewards/margins": 0.03422056883573532, | |
| "rewards/rejected": -0.1425708830356598, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.87284885712256e-06, | |
| "logits/chosen": -0.07014905661344528, | |
| "logits/rejected": -0.006586342118680477, | |
| "logps/chosen": -494.0985412597656, | |
| "logps/rejected": -525.9044189453125, | |
| "loss": 0.2065, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.10610511153936386, | |
| "rewards/margins": 0.0382779985666275, | |
| "rewards/rejected": -0.14438311755657196, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.857779879123181e-06, | |
| "logits/chosen": -0.0900832936167717, | |
| "logits/rejected": -0.10245800018310547, | |
| "logps/chosen": -572.3137817382812, | |
| "logps/rejected": -617.9874267578125, | |
| "loss": 0.2156, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.12338846921920776, | |
| "rewards/margins": 0.036996591836214066, | |
| "rewards/rejected": -0.16038507223129272, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.841893264593643e-06, | |
| "logits/chosen": -0.051192570477724075, | |
| "logits/rejected": -0.02001297101378441, | |
| "logps/chosen": -547.9758911132812, | |
| "logps/rejected": -576.4624633789062, | |
| "loss": 0.1939, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.10065815597772598, | |
| "rewards/margins": 0.02802097424864769, | |
| "rewards/rejected": -0.12867912650108337, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.825194522732023e-06, | |
| "logits/chosen": -0.14838102459907532, | |
| "logits/rejected": 0.1484626680612564, | |
| "logps/chosen": -494.1825256347656, | |
| "logps/rejected": -524.7642822265625, | |
| "loss": 0.1967, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.08151207864284515, | |
| "rewards/margins": 0.032111816108226776, | |
| "rewards/rejected": -0.11362389475107193, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.807689444367853e-06, | |
| "logits/chosen": -0.046164363622665405, | |
| "logits/rejected": -0.03798365592956543, | |
| "logps/chosen": -472.62677001953125, | |
| "logps/rejected": -500.4376525878906, | |
| "loss": 0.1883, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.07655288279056549, | |
| "rewards/margins": 0.039138298481702805, | |
| "rewards/rejected": -0.1156911849975586, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.78938409995396e-06, | |
| "logits/chosen": -0.1458251178264618, | |
| "logits/rejected": -0.0393032506108284, | |
| "logps/chosen": -436.66033935546875, | |
| "logps/rejected": -447.3544006347656, | |
| "loss": 0.2138, | |
| "rewards/accuracies": 0.375, | |
| "rewards/chosen": -0.08541758358478546, | |
| "rewards/margins": 0.011176596395671368, | |
| "rewards/rejected": -0.0965941846370697, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.770284837461342e-06, | |
| "logits/chosen": -0.15639075636863708, | |
| "logits/rejected": -0.010832506231963634, | |
| "logps/chosen": -489.99151611328125, | |
| "logps/rejected": -543.11767578125, | |
| "loss": 0.1843, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.09636662900447845, | |
| "rewards/margins": 0.04293360933661461, | |
| "rewards/rejected": -0.13930022716522217, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7503982801778015e-06, | |
| "logits/chosen": -0.22076280415058136, | |
| "logits/rejected": -0.018700700253248215, | |
| "logps/chosen": -517.91455078125, | |
| "logps/rejected": -551.0287475585938, | |
| "loss": 0.1901, | |
| "rewards/accuracies": 0.45625001192092896, | |
| "rewards/chosen": -0.11141178756952286, | |
| "rewards/margins": 0.033164333552122116, | |
| "rewards/rejected": -0.14457611739635468, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.729731324411104e-06, | |
| "logits/chosen": -0.15695324540138245, | |
| "logits/rejected": -0.08840291202068329, | |
| "logps/chosen": -491.2373046875, | |
| "logps/rejected": -511.56781005859375, | |
| "loss": 0.1891, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.10444734245538712, | |
| "rewards/margins": 0.04566046595573425, | |
| "rewards/rejected": -0.15010783076286316, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.7082911370974645e-06, | |
| "logits/chosen": -0.11595731973648071, | |
| "logits/rejected": -0.128843754529953, | |
| "logps/chosen": -557.0236206054688, | |
| "logps/rejected": -633.2085571289062, | |
| "loss": 0.2007, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.12254992872476578, | |
| "rewards/margins": 0.05843643471598625, | |
| "rewards/rejected": -0.18098635971546173, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.68608515331618e-06, | |
| "logits/chosen": -0.08469126373529434, | |
| "logits/rejected": -0.03279350325465202, | |
| "logps/chosen": -486.62078857421875, | |
| "logps/rejected": -515.3419799804688, | |
| "loss": 0.1855, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.0999387800693512, | |
| "rewards/margins": 0.04227910935878754, | |
| "rewards/rejected": -0.14221788942813873, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.663121073711269e-06, | |
| "logits/chosen": -0.1471957415342331, | |
| "logits/rejected": -0.05932598561048508, | |
| "logps/chosen": -474.94915771484375, | |
| "logps/rejected": -536.3035888671875, | |
| "loss": 0.1819, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.0976475328207016, | |
| "rewards/margins": 0.04680894687771797, | |
| "rewards/rejected": -0.14445647597312927, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.63940686182103e-06, | |
| "logits/chosen": -0.17284898459911346, | |
| "logits/rejected": -0.1464369297027588, | |
| "logps/chosen": -502.36578369140625, | |
| "logps/rejected": -583.7872314453125, | |
| "loss": 0.1724, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.10467234998941422, | |
| "rewards/margins": 0.052542489022016525, | |
| "rewards/rejected": -0.15721485018730164, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.614950741316425e-06, | |
| "logits/chosen": -0.21725276112556458, | |
| "logits/rejected": -0.14681020379066467, | |
| "logps/chosen": -523.1455078125, | |
| "logps/rejected": -577.9979248046875, | |
| "loss": 0.184, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.09476028382778168, | |
| "rewards/margins": 0.056237511336803436, | |
| "rewards/rejected": -0.15099778771400452, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.589761193149254e-06, | |
| "logits/chosen": -0.2188795506954193, | |
| "logits/rejected": -0.24383120238780975, | |
| "logps/chosen": -541.2039794921875, | |
| "logps/rejected": -610.45166015625, | |
| "loss": 0.1864, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.11351511627435684, | |
| "rewards/margins": 0.044019803404808044, | |
| "rewards/rejected": -0.1575349122285843, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.563846952611112e-06, | |
| "logits/chosen": -0.2326229363679886, | |
| "logits/rejected": -0.15911969542503357, | |
| "logps/chosen": -494.03302001953125, | |
| "logps/rejected": -544.4476318359375, | |
| "loss": 0.1756, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.09819264709949493, | |
| "rewards/margins": 0.056823331862688065, | |
| "rewards/rejected": -0.1550159752368927, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.537217006304141e-06, | |
| "logits/chosen": -0.16317172348499298, | |
| "logits/rejected": -0.1998225897550583, | |
| "logps/chosen": -530.8056640625, | |
| "logps/rejected": -605.7437133789062, | |
| "loss": 0.1588, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.09498073905706406, | |
| "rewards/margins": 0.07860372215509415, | |
| "rewards/rejected": -0.1735844761133194, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.50988058902464e-06, | |
| "logits/chosen": -0.17353829741477966, | |
| "logits/rejected": -0.20073547959327698, | |
| "logps/chosen": -539.7728271484375, | |
| "logps/rejected": -576.4498901367188, | |
| "loss": 0.1832, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.10605832189321518, | |
| "rewards/margins": 0.05749889463186264, | |
| "rewards/rejected": -0.16355721652507782, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.481847180560593e-06, | |
| "logits/chosen": -0.17195938527584076, | |
| "logits/rejected": -0.15436215698719025, | |
| "logps/chosen": -502.6675720214844, | |
| "logps/rejected": -537.3475341796875, | |
| "loss": 0.2008, | |
| "rewards/accuracies": 0.45625001192092896, | |
| "rewards/chosen": -0.10046552121639252, | |
| "rewards/margins": 0.028822993859648705, | |
| "rewards/rejected": -0.12928852438926697, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.453126502404253e-06, | |
| "logits/chosen": -0.21576860547065735, | |
| "logits/rejected": -0.22100479900836945, | |
| "logps/chosen": -528.4920654296875, | |
| "logps/rejected": -600.3697509765625, | |
| "loss": 0.1657, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.11002079397439957, | |
| "rewards/margins": 0.06445158272981644, | |
| "rewards/rejected": -0.1744723618030548, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.423728514380892e-06, | |
| "logits/chosen": -0.2724868655204773, | |
| "logits/rejected": -0.18154220283031464, | |
| "logps/chosen": -534.0877075195312, | |
| "logps/rejected": -585.4654541015625, | |
| "loss": 0.1745, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.11277395486831665, | |
| "rewards/margins": 0.05171126872301102, | |
| "rewards/rejected": -0.16448521614074707, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.393663411194918e-06, | |
| "logits/chosen": -0.29137545824050903, | |
| "logits/rejected": -0.1960664540529251, | |
| "logps/chosen": -543.6236572265625, | |
| "logps/rejected": -613.0497436523438, | |
| "loss": 0.1896, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.11756434291601181, | |
| "rewards/margins": 0.049765445291996, | |
| "rewards/rejected": -0.16732977330684662, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.362941618894523e-06, | |
| "logits/chosen": -0.24844829738140106, | |
| "logits/rejected": -0.17800265550613403, | |
| "logps/chosen": -483.45697021484375, | |
| "logps/rejected": -549.9342041015625, | |
| "loss": 0.1642, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.10112373530864716, | |
| "rewards/margins": 0.06612999737262726, | |
| "rewards/rejected": -0.16725373268127441, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.331573791256116e-06, | |
| "logits/chosen": -0.35919591784477234, | |
| "logits/rejected": -0.2761411964893341, | |
| "logps/chosen": -489.17852783203125, | |
| "logps/rejected": -502.04022216796875, | |
| "loss": 0.1868, | |
| "rewards/accuracies": 0.45625001192092896, | |
| "rewards/chosen": -0.09544913470745087, | |
| "rewards/margins": 0.05491846054792404, | |
| "rewards/rejected": -0.1503675878047943, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.299570806089786e-06, | |
| "logits/chosen": -0.254047155380249, | |
| "logits/rejected": -0.24359670281410217, | |
| "logps/chosen": -517.5626220703125, | |
| "logps/rejected": -570.8585205078125, | |
| "loss": 0.1936, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.10336007177829742, | |
| "rewards/margins": 0.06461663544178009, | |
| "rewards/rejected": -0.16797669231891632, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.266943761467057e-06, | |
| "logits/chosen": -0.3495100736618042, | |
| "logits/rejected": -0.22493663430213928, | |
| "logps/chosen": -526.2481079101562, | |
| "logps/rejected": -611.931396484375, | |
| "loss": 0.1684, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.10717520862817764, | |
| "rewards/margins": 0.07210040837526321, | |
| "rewards/rejected": -0.17927563190460205, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.233703971872287e-06, | |
| "logits/chosen": -0.28394874930381775, | |
| "logits/rejected": -0.2587239146232605, | |
| "logps/chosen": -567.471435546875, | |
| "logps/rejected": -639.43310546875, | |
| "loss": 0.1944, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.11128894239664078, | |
| "rewards/margins": 0.06409231573343277, | |
| "rewards/rejected": -0.17538125813007355, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.1998629642789925e-06, | |
| "logits/chosen": -0.28350967168807983, | |
| "logits/rejected": -0.277797669172287, | |
| "logps/chosen": -476.7294921875, | |
| "logps/rejected": -528.3859252929688, | |
| "loss": 0.1746, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.07711435854434967, | |
| "rewards/margins": 0.0645236223936081, | |
| "rewards/rejected": -0.14163798093795776, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.165432474152505e-06, | |
| "logits/chosen": -0.3298090100288391, | |
| "logits/rejected": -0.27090325951576233, | |
| "logps/chosen": -543.4595947265625, | |
| "logps/rejected": -586.7649536132812, | |
| "loss": 0.1706, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.08748660981655121, | |
| "rewards/margins": 0.06710480898618698, | |
| "rewards/rejected": -0.1545914113521576, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.130424441380308e-06, | |
| "logits/chosen": -0.3356090188026428, | |
| "logits/rejected": -0.14076504111289978, | |
| "logps/chosen": -532.180908203125, | |
| "logps/rejected": -585.5838623046875, | |
| "loss": 0.1865, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.09366744011640549, | |
| "rewards/margins": 0.046517182141542435, | |
| "rewards/rejected": -0.14018462598323822, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.09485100613151e-06, | |
| "logits/chosen": -0.2658724784851074, | |
| "logits/rejected": -0.2869165539741516, | |
| "logps/chosen": -479.1338806152344, | |
| "logps/rejected": -522.8956298828125, | |
| "loss": 0.1859, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.07211341708898544, | |
| "rewards/margins": 0.050303440541028976, | |
| "rewards/rejected": -0.12241685390472412, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.058724504646834e-06, | |
| "logits/chosen": -0.3326638340950012, | |
| "logits/rejected": -0.3534831404685974, | |
| "logps/chosen": -465.98046875, | |
| "logps/rejected": -551.1593017578125, | |
| "loss": 0.1817, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.06464243680238724, | |
| "rewards/margins": 0.06546784937381744, | |
| "rewards/rejected": -0.13011027872562408, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.022057464960632e-06, | |
| "logits/chosen": -0.2847468852996826, | |
| "logits/rejected": -0.3084865212440491, | |
| "logps/chosen": -466.17431640625, | |
| "logps/rejected": -591.4989624023438, | |
| "loss": 0.169, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.06827010214328766, | |
| "rewards/margins": 0.08749465644359589, | |
| "rewards/rejected": -0.15576477348804474, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.984862602556383e-06, | |
| "logits/chosen": -0.29794925451278687, | |
| "logits/rejected": -0.2838256359100342, | |
| "logps/chosen": -505.8924255371094, | |
| "logps/rejected": -569.5626831054688, | |
| "loss": 0.173, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.08569585531949997, | |
| "rewards/margins": 0.058602042496204376, | |
| "rewards/rejected": -0.14429791271686554, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.947152815957187e-06, | |
| "logits/chosen": -0.4137333333492279, | |
| "logits/rejected": -0.201189324259758, | |
| "logps/chosen": -520.5211181640625, | |
| "logps/rejected": -567.9163208007812, | |
| "loss": 0.1771, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.07455627620220184, | |
| "rewards/margins": 0.06876102834939957, | |
| "rewards/rejected": -0.1433172971010208, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.908941182252785e-06, | |
| "logits/chosen": -0.3251053988933563, | |
| "logits/rejected": -0.21724073588848114, | |
| "logps/chosen": -457.90692138671875, | |
| "logps/rejected": -512.7808837890625, | |
| "loss": 0.1777, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.0669245719909668, | |
| "rewards/margins": 0.05744165927171707, | |
| "rewards/rejected": -0.12436623871326447, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.8702409525646535e-06, | |
| "logits/chosen": -0.20694856345653534, | |
| "logits/rejected": -0.2958626449108124, | |
| "logps/chosen": -455.8350524902344, | |
| "logps/rejected": -536.2254638671875, | |
| "loss": 0.1935, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.0645083338022232, | |
| "rewards/margins": 0.06103675439953804, | |
| "rewards/rejected": -0.12554509937763214, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.8310655474507495e-06, | |
| "logits/chosen": -0.3800198435783386, | |
| "logits/rejected": -0.2641103267669678, | |
| "logps/chosen": -480.8030700683594, | |
| "logps/rejected": -521.5682373046875, | |
| "loss": 0.2025, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.07129710167646408, | |
| "rewards/margins": 0.04021110758185387, | |
| "rewards/rejected": -0.11150821298360825, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.7914285522515002e-06, | |
| "logits/chosen": -0.35266560316085815, | |
| "logits/rejected": -0.21721109747886658, | |
| "logps/chosen": -503.35357666015625, | |
| "logps/rejected": -518.2500610351562, | |
| "loss": 0.1928, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.07125665247440338, | |
| "rewards/margins": 0.034123968333005905, | |
| "rewards/rejected": -0.10538061708211899, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.751343712378639e-06, | |
| "logits/chosen": -0.303008496761322, | |
| "logits/rejected": -0.3047345280647278, | |
| "logps/chosen": -447.089599609375, | |
| "logps/rejected": -469.61590576171875, | |
| "loss": 0.1992, | |
| "rewards/accuracies": 0.40625, | |
| "rewards/chosen": -0.06738699972629547, | |
| "rewards/margins": 0.027302343398332596, | |
| "rewards/rejected": -0.09468933939933777, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.710824928548546e-06, | |
| "logits/chosen": -0.29965728521347046, | |
| "logits/rejected": -0.2571870982646942, | |
| "logps/chosen": -450.4271545410156, | |
| "logps/rejected": -482.21502685546875, | |
| "loss": 0.1935, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.055190593004226685, | |
| "rewards/margins": 0.04364421218633652, | |
| "rewards/rejected": -0.0988348051905632, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.6698862519617225e-06, | |
| "logits/chosen": -0.290757954120636, | |
| "logits/rejected": -0.3262009024620056, | |
| "logps/chosen": -449.7730407714844, | |
| "logps/rejected": -520.2115478515625, | |
| "loss": 0.1599, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.05544726923108101, | |
| "rewards/margins": 0.06180447340011597, | |
| "rewards/rejected": -0.11725173890590668, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.6285418794300793e-06, | |
| "logits/chosen": -0.3281220495700836, | |
| "logits/rejected": -0.2547357678413391, | |
| "logps/chosen": -441.2623596191406, | |
| "logps/rejected": -495.96368408203125, | |
| "loss": 0.1915, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.0586274154484272, | |
| "rewards/margins": 0.03878774121403694, | |
| "rewards/rejected": -0.09741516411304474, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.5868061484537365e-06, | |
| "logits/chosen": -0.41046142578125, | |
| "logits/rejected": -0.29820680618286133, | |
| "logps/chosen": -447.24151611328125, | |
| "logps/rejected": -473.2769470214844, | |
| "loss": 0.1898, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.07239285856485367, | |
| "rewards/margins": 0.03644294664263725, | |
| "rewards/rejected": -0.10883580148220062, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5446935322490285e-06, | |
| "logits/chosen": -0.35609519481658936, | |
| "logits/rejected": -0.2902910113334656, | |
| "logps/chosen": -469.17706298828125, | |
| "logps/rejected": -546.7158203125, | |
| "loss": 0.1788, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.05905740335583687, | |
| "rewards/margins": 0.06316393613815308, | |
| "rewards/rejected": -0.12222135066986084, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.502218634729447e-06, | |
| "logits/chosen": -0.2686077356338501, | |
| "logits/rejected": -0.2707546055316925, | |
| "logps/chosen": -522.1178588867188, | |
| "logps/rejected": -570.613525390625, | |
| "loss": 0.1825, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.07459238916635513, | |
| "rewards/margins": 0.061357058584690094, | |
| "rewards/rejected": -0.13594944775104523, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.459396185441265e-06, | |
| "logits/chosen": -0.38130688667297363, | |
| "logits/rejected": -0.14519771933555603, | |
| "logps/chosen": -474.2879943847656, | |
| "logps/rejected": -492.1015625, | |
| "loss": 0.1804, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.049582771956920624, | |
| "rewards/margins": 0.04935313016176224, | |
| "rewards/rejected": -0.09893589466810226, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4162410344555834e-06, | |
| "logits/chosen": -0.2872348427772522, | |
| "logits/rejected": -0.2791721224784851, | |
| "logps/chosen": -485.4851989746094, | |
| "logps/rejected": -571.0114135742188, | |
| "loss": 0.1743, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.060776274651288986, | |
| "rewards/margins": 0.06890632957220078, | |
| "rewards/rejected": -0.12968261539936066, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3727681472185937e-06, | |
| "logits/chosen": -0.3343026041984558, | |
| "logits/rejected": -0.2967159152030945, | |
| "logps/chosen": -441.74713134765625, | |
| "logps/rejected": -455.7685546875, | |
| "loss": 0.1898, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.05669083446264267, | |
| "rewards/margins": 0.04281865060329437, | |
| "rewards/rejected": -0.09950948506593704, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3289925993618217e-06, | |
| "logits/chosen": -0.31694555282592773, | |
| "logits/rejected": -0.23448070883750916, | |
| "logps/chosen": -505.189208984375, | |
| "logps/rejected": -574.4694213867188, | |
| "loss": 0.1641, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.06701242923736572, | |
| "rewards/margins": 0.07507143914699554, | |
| "rewards/rejected": -0.14208386838436127, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2849295714741643e-06, | |
| "logits/chosen": -0.36678022146224976, | |
| "logits/rejected": -0.22024087607860565, | |
| "logps/chosen": -481.8533630371094, | |
| "logps/rejected": -517.580078125, | |
| "loss": 0.1795, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.08131915330886841, | |
| "rewards/margins": 0.054040372371673584, | |
| "rewards/rejected": -0.135359525680542, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2405943438375287e-06, | |
| "logits/chosen": -0.31432998180389404, | |
| "logits/rejected": -0.3275941014289856, | |
| "logps/chosen": -486.9906311035156, | |
| "logps/rejected": -556.7380981445312, | |
| "loss": 0.1659, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.0830111876130104, | |
| "rewards/margins": 0.08041650056838989, | |
| "rewards/rejected": -0.1634276807308197, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1960022911279036e-06, | |
| "logits/chosen": -0.3552636504173279, | |
| "logits/rejected": -0.12550072371959686, | |
| "logps/chosen": -442.874267578125, | |
| "logps/rejected": -486.37457275390625, | |
| "loss": 0.1946, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.06523702293634415, | |
| "rewards/margins": 0.0638941302895546, | |
| "rewards/rejected": -0.12913115322589874, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1511688770836844e-06, | |
| "logits/chosen": -0.44293370842933655, | |
| "logits/rejected": -0.206680566072464, | |
| "logps/chosen": -426.76861572265625, | |
| "logps/rejected": -496.85565185546875, | |
| "loss": 0.1725, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.06692390143871307, | |
| "rewards/margins": 0.07236063480377197, | |
| "rewards/rejected": -0.13928455114364624, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1061096491431307e-06, | |
| "logits/chosen": -0.35239410400390625, | |
| "logits/rejected": -0.22438523173332214, | |
| "logps/chosen": -493.1455078125, | |
| "logps/rejected": -538.7332153320312, | |
| "loss": 0.189, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.08153922855854034, | |
| "rewards/margins": 0.05397678166627884, | |
| "rewards/rejected": -0.13551601767539978, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0608402330527796e-06, | |
| "logits/chosen": -0.28043216466903687, | |
| "logits/rejected": -0.32117584347724915, | |
| "logps/chosen": -489.95855712890625, | |
| "logps/rejected": -528.5255737304688, | |
| "loss": 0.1694, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.07012003660202026, | |
| "rewards/margins": 0.05748220160603523, | |
| "rewards/rejected": -0.1276022493839264, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0153763274487176e-06, | |
| "logits/chosen": -0.30975884199142456, | |
| "logits/rejected": -0.28235483169555664, | |
| "logps/chosen": -523.9342041015625, | |
| "logps/rejected": -557.6265869140625, | |
| "loss": 0.1987, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.08475781232118607, | |
| "rewards/margins": 0.04346821457147598, | |
| "rewards/rejected": -0.12822601199150085, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9697336984125683e-06, | |
| "logits/chosen": -0.3483017385005951, | |
| "logits/rejected": -0.30833274126052856, | |
| "logps/chosen": -446.2549743652344, | |
| "logps/rejected": -472.79736328125, | |
| "loss": 0.1947, | |
| "rewards/accuracies": 0.45625001192092896, | |
| "rewards/chosen": -0.07903579622507095, | |
| "rewards/margins": 0.04975394904613495, | |
| "rewards/rejected": -0.1287897378206253, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.923928174004094e-06, | |
| "logits/chosen": -0.34173694252967834, | |
| "logits/rejected": -0.2428359091281891, | |
| "logps/chosen": -514.5177612304688, | |
| "logps/rejected": -543.485595703125, | |
| "loss": 0.174, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.06802085787057877, | |
| "rewards/margins": 0.06197448447346687, | |
| "rewards/rejected": -0.12999534606933594, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.8779756387723036e-06, | |
| "logits/chosen": -0.2911175787448883, | |
| "logits/rejected": -0.26822999119758606, | |
| "logps/chosen": -492.79412841796875, | |
| "logps/rejected": -546.5845947265625, | |
| "loss": 0.1747, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.068579763174057, | |
| "rewards/margins": 0.06541167199611664, | |
| "rewards/rejected": -0.13399145007133484, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.831892028246968e-06, | |
| "logits/chosen": -0.3150990307331085, | |
| "logits/rejected": -0.25644177198410034, | |
| "logps/chosen": -497.47216796875, | |
| "logps/rejected": -579.6931762695312, | |
| "loss": 0.1752, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.059874940663576126, | |
| "rewards/margins": 0.06721685081720352, | |
| "rewards/rejected": -0.12709179520606995, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7856933234124617e-06, | |
| "logits/chosen": -0.31079426407814026, | |
| "logits/rejected": -0.3151572346687317, | |
| "logps/chosen": -463.4168395996094, | |
| "logps/rejected": -555.4398193359375, | |
| "loss": 0.1846, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.0642518401145935, | |
| "rewards/margins": 0.05405072122812271, | |
| "rewards/rejected": -0.11830256134271622, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7393955451658387e-06, | |
| "logits/chosen": -0.22878186404705048, | |
| "logits/rejected": -0.21153822541236877, | |
| "logps/chosen": -510.89837646484375, | |
| "logps/rejected": -539.2826538085938, | |
| "loss": 0.175, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.05998819321393967, | |
| "rewards/margins": 0.06107606738805771, | |
| "rewards/rejected": -0.12106426060199738, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6930147487610667e-06, | |
| "logits/chosen": -0.34622281789779663, | |
| "logits/rejected": -0.25016504526138306, | |
| "logps/chosen": -429.02178955078125, | |
| "logps/rejected": -502.2627868652344, | |
| "loss": 0.1741, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.05254564806818962, | |
| "rewards/margins": 0.06378427147865295, | |
| "rewards/rejected": -0.11632990837097168, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6465670182413487e-06, | |
| "logits/chosen": -0.37311479449272156, | |
| "logits/rejected": -0.20994921028614044, | |
| "logps/chosen": -451.9130859375, | |
| "logps/rejected": -485.57061767578125, | |
| "loss": 0.1822, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.05379350110888481, | |
| "rewards/margins": 0.04739636927843094, | |
| "rewards/rejected": -0.10118986666202545, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.6000684608614594e-06, | |
| "logits/chosen": -0.34606754779815674, | |
| "logits/rejected": -0.33781155943870544, | |
| "logps/chosen": -458.77392578125, | |
| "logps/rejected": -538.665283203125, | |
| "loss": 0.1762, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.04491495341062546, | |
| "rewards/margins": 0.07754696160554886, | |
| "rewards/rejected": -0.12246191501617432, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5535352015020338e-06, | |
| "logits/chosen": -0.3478602468967438, | |
| "logits/rejected": -0.3592594563961029, | |
| "logps/chosen": -426.89324951171875, | |
| "logps/rejected": -497.1231994628906, | |
| "loss": 0.1811, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.05246494337916374, | |
| "rewards/margins": 0.06296879798173904, | |
| "rewards/rejected": -0.11543373763561249, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.506983377077741e-06, | |
| "logits/chosen": -0.30938345193862915, | |
| "logits/rejected": -0.36706987023353577, | |
| "logps/chosen": -438.70477294921875, | |
| "logps/rejected": -479.84930419921875, | |
| "loss": 0.185, | |
| "rewards/accuracies": 0.38749998807907104, | |
| "rewards/chosen": -0.06825192272663116, | |
| "rewards/margins": 0.051073990762233734, | |
| "rewards/rejected": -0.1193259134888649, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.460429130941289e-06, | |
| "logits/chosen": -0.30941054224967957, | |
| "logits/rejected": -0.3541705310344696, | |
| "logps/chosen": -500.3165588378906, | |
| "logps/rejected": -547.2388916015625, | |
| "loss": 0.1648, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.0632004514336586, | |
| "rewards/margins": 0.0647406131029129, | |
| "rewards/rejected": -0.1279410421848297, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.413888607285192e-06, | |
| "logits/chosen": -0.369188129901886, | |
| "logits/rejected": -0.2474605143070221, | |
| "logps/chosen": -482.4578552246094, | |
| "logps/rejected": -470.77178955078125, | |
| "loss": 0.2005, | |
| "rewards/accuracies": 0.4124999940395355, | |
| "rewards/chosen": -0.06637545675039291, | |
| "rewards/margins": 0.03153007850050926, | |
| "rewards/rejected": -0.09790553897619247, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.367377945543249e-06, | |
| "logits/chosen": -0.323347270488739, | |
| "logits/rejected": -0.30056172609329224, | |
| "logps/chosen": -450.601318359375, | |
| "logps/rejected": -498.0372619628906, | |
| "loss": 0.1822, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.06100805476307869, | |
| "rewards/margins": 0.048852551728487015, | |
| "rewards/rejected": -0.1098606139421463, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.320913274793676e-06, | |
| "logits/chosen": -0.37067005038261414, | |
| "logits/rejected": -0.28159070014953613, | |
| "logps/chosen": -450.1123046875, | |
| "logps/rejected": -492.94903564453125, | |
| "loss": 0.177, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.054400503635406494, | |
| "rewards/margins": 0.05063963681459427, | |
| "rewards/rejected": -0.10504014790058136, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.27451070816582e-06, | |
| "logits/chosen": -0.3574444651603699, | |
| "logits/rejected": -0.26874804496765137, | |
| "logps/chosen": -471.46002197265625, | |
| "logps/rejected": -542.9146728515625, | |
| "loss": 0.1728, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.06453348696231842, | |
| "rewards/margins": 0.07580892741680145, | |
| "rewards/rejected": -0.14034239947795868, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.228186337252414e-06, | |
| "logits/chosen": -0.2806636393070221, | |
| "logits/rejected": -0.29087376594543457, | |
| "logps/chosen": -486.2430725097656, | |
| "logps/rejected": -514.122314453125, | |
| "loss": 0.1749, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.07113119959831238, | |
| "rewards/margins": 0.05178775265812874, | |
| "rewards/rejected": -0.12291894853115082, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1819562265292946e-06, | |
| "logits/chosen": -0.4109742045402527, | |
| "logits/rejected": -0.32353344559669495, | |
| "logps/chosen": -539.9480590820312, | |
| "logps/rejected": -556.8568725585938, | |
| "loss": 0.1739, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.07678203284740448, | |
| "rewards/margins": 0.04964344576001167, | |
| "rewards/rejected": -0.12642547488212585, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1358364077845236e-06, | |
| "logits/chosen": -0.33106154203414917, | |
| "logits/rejected": -0.22853222489356995, | |
| "logps/chosen": -463.4998474121094, | |
| "logps/rejected": -510.4764709472656, | |
| "loss": 0.1836, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.0651940405368805, | |
| "rewards/margins": 0.05113544315099716, | |
| "rewards/rejected": -0.11632947623729706, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.089842874558849e-06, | |
| "logits/chosen": -0.4000775218009949, | |
| "logits/rejected": -0.16853252053260803, | |
| "logps/chosen": -520.6731567382812, | |
| "logps/rejected": -535.2789306640625, | |
| "loss": 0.1777, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.06823070347309113, | |
| "rewards/margins": 0.046103235334157944, | |
| "rewards/rejected": -0.11433394253253937, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0439915765994242e-06, | |
| "logits/chosen": -0.3381574749946594, | |
| "logits/rejected": -0.22323736548423767, | |
| "logps/chosen": -483.09814453125, | |
| "logps/rejected": -513.6597900390625, | |
| "loss": 0.2054, | |
| "rewards/accuracies": 0.4437499940395355, | |
| "rewards/chosen": -0.06595513969659805, | |
| "rewards/margins": 0.043463800102472305, | |
| "rewards/rejected": -0.10941894352436066, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.9982984143287186e-06, | |
| "logits/chosen": -0.34239357709884644, | |
| "logits/rejected": -0.3451228439807892, | |
| "logps/chosen": -495.82403564453125, | |
| "logps/rejected": -536.5828857421875, | |
| "loss": 0.1641, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.05730292201042175, | |
| "rewards/margins": 0.05639542266726494, | |
| "rewards/rejected": -0.11369834840297699, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.95277923333053e-06, | |
| "logits/chosen": -0.23302578926086426, | |
| "logits/rejected": -0.3353923261165619, | |
| "logps/chosen": -467.14642333984375, | |
| "logps/rejected": -511.3815002441406, | |
| "loss": 0.1702, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.05319993942975998, | |
| "rewards/margins": 0.055631499737501144, | |
| "rewards/rejected": -0.10883144289255142, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.9074498188550156e-06, | |
| "logits/chosen": -0.3332996964454651, | |
| "logits/rejected": -0.34316807985305786, | |
| "logps/chosen": -486.32373046875, | |
| "logps/rejected": -549.6057739257812, | |
| "loss": 0.1644, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.06252233684062958, | |
| "rewards/margins": 0.06708262860774994, | |
| "rewards/rejected": -0.1296049803495407, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.862325890344643e-06, | |
| "logits/chosen": -0.34824302792549133, | |
| "logits/rejected": -0.19805127382278442, | |
| "logps/chosen": -435.487060546875, | |
| "logps/rejected": -514.4375610351562, | |
| "loss": 0.1957, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.06748778373003006, | |
| "rewards/margins": 0.052836887538433075, | |
| "rewards/rejected": -0.12032465636730194, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.817423095982972e-06, | |
| "logits/chosen": -0.4363935589790344, | |
| "logits/rejected": -0.2699471712112427, | |
| "logps/chosen": -517.750732421875, | |
| "logps/rejected": -548.2755126953125, | |
| "loss": 0.1871, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.08836901187896729, | |
| "rewards/margins": 0.04405100643634796, | |
| "rewards/rejected": -0.13242003321647644, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.7727570072681293e-06, | |
| "logits/chosen": -0.3290655016899109, | |
| "logits/rejected": -0.3122135400772095, | |
| "logps/chosen": -466.1502990722656, | |
| "logps/rejected": -535.735107421875, | |
| "loss": 0.1699, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.07871656864881516, | |
| "rewards/margins": 0.06162920594215393, | |
| "rewards/rejected": -0.14034578204154968, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7283431136128961e-06, | |
| "logits/chosen": -0.35683125257492065, | |
| "logits/rejected": -0.28000420331954956, | |
| "logps/chosen": -494.61016845703125, | |
| "logps/rejected": -513.5484619140625, | |
| "loss": 0.1875, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.07970693707466125, | |
| "rewards/margins": 0.04146048426628113, | |
| "rewards/rejected": -0.12116742134094238, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6841968169732478e-06, | |
| "logits/chosen": -0.3199581801891327, | |
| "logits/rejected": -0.30323344469070435, | |
| "logps/chosen": -484.8779296875, | |
| "logps/rejected": -546.8536987304688, | |
| "loss": 0.1711, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.07513175904750824, | |
| "rewards/margins": 0.05645768716931343, | |
| "rewards/rejected": -0.13158944249153137, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6403334265072284e-06, | |
| "logits/chosen": -0.2838456928730011, | |
| "logits/rejected": -0.31488656997680664, | |
| "logps/chosen": -470.95025634765625, | |
| "logps/rejected": -502.6404724121094, | |
| "loss": 0.1864, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.07776617258787155, | |
| "rewards/margins": 0.044505417346954346, | |
| "rewards/rejected": -0.1222715824842453, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5967681532660066e-06, | |
| "logits/chosen": -0.2862758934497833, | |
| "logits/rejected": -0.38872987031936646, | |
| "logps/chosen": -515.7428588867188, | |
| "logps/rejected": -601.1578369140625, | |
| "loss": 0.1767, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.08725148439407349, | |
| "rewards/margins": 0.07090319693088531, | |
| "rewards/rejected": -0.15815469622612, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5535161049189463e-06, | |
| "logits/chosen": -0.35917767882347107, | |
| "logits/rejected": -0.39498478174209595, | |
| "logps/chosen": -512.1259765625, | |
| "logps/rejected": -557.6096801757812, | |
| "loss": 0.1919, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.08681392669677734, | |
| "rewards/margins": 0.055614955723285675, | |
| "rewards/rejected": -0.14242888987064362, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.5105922805145356e-06, | |
| "logits/chosen": -0.3822699189186096, | |
| "logits/rejected": -0.2847387194633484, | |
| "logps/chosen": -531.32861328125, | |
| "logps/rejected": -538.8329467773438, | |
| "loss": 0.1865, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.09042377769947052, | |
| "rewards/margins": 0.047544240951538086, | |
| "rewards/rejected": -0.1379680335521698, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.4680115652789823e-06, | |
| "logits/chosen": -0.2868042588233948, | |
| "logits/rejected": -0.39659756422042847, | |
| "logps/chosen": -440.13201904296875, | |
| "logps/rejected": -505.8741760253906, | |
| "loss": 0.168, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.06821772456169128, | |
| "rewards/margins": 0.07103266566991806, | |
| "rewards/rejected": -0.13925038278102875, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4257887254542767e-06, | |
| "logits/chosen": -0.267185777425766, | |
| "logits/rejected": -0.3715210556983948, | |
| "logps/chosen": -508.6048278808594, | |
| "logps/rejected": -551.8297119140625, | |
| "loss": 0.1861, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.08317936956882477, | |
| "rewards/margins": 0.04879546910524368, | |
| "rewards/rejected": -0.13197484612464905, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.3839384031775227e-06, | |
| "logits/chosen": -0.40115490555763245, | |
| "logits/rejected": -0.3220266103744507, | |
| "logps/chosen": -511.22564697265625, | |
| "logps/rejected": -581.5658569335938, | |
| "loss": 0.1907, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.07654689252376556, | |
| "rewards/margins": 0.05669323727488518, | |
| "rewards/rejected": -0.13324013352394104, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.342475111403298e-06, | |
| "logits/chosen": -0.4519384801387787, | |
| "logits/rejected": -0.3560634255409241, | |
| "logps/chosen": -482.3561096191406, | |
| "logps/rejected": -541.8218383789062, | |
| "loss": 0.1587, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.07603658735752106, | |
| "rewards/margins": 0.062384773045778275, | |
| "rewards/rejected": -0.13842137157917023, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3014132288708209e-06, | |
| "logits/chosen": -0.45893383026123047, | |
| "logits/rejected": -0.34163400530815125, | |
| "logps/chosen": -470.9329528808594, | |
| "logps/rejected": -551.7818603515625, | |
| "loss": 0.1698, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.07042945921421051, | |
| "rewards/margins": 0.07146573066711426, | |
| "rewards/rejected": -0.14189518988132477, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2607669951176549e-06, | |
| "logits/chosen": -0.30523496866226196, | |
| "logits/rejected": -0.35816168785095215, | |
| "logps/chosen": -522.6578979492188, | |
| "logps/rejected": -601.38818359375, | |
| "loss": 0.1701, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.08582588285207748, | |
| "rewards/margins": 0.06938043236732483, | |
| "rewards/rejected": -0.15520630776882172, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2205505055416891e-06, | |
| "logits/chosen": -0.4220711290836334, | |
| "logits/rejected": -0.37211546301841736, | |
| "logps/chosen": -488.38916015625, | |
| "logps/rejected": -604.0801391601562, | |
| "loss": 0.1738, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.07974353432655334, | |
| "rewards/margins": 0.086810402572155, | |
| "rewards/rejected": -0.16655394434928894, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1807777065131002e-06, | |
| "logits/chosen": -0.4147316515445709, | |
| "logits/rejected": -0.3836427628993988, | |
| "logps/chosen": -502.61083984375, | |
| "logps/rejected": -578.9619140625, | |
| "loss": 0.1864, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.08877801150083542, | |
| "rewards/margins": 0.07549368590116501, | |
| "rewards/rejected": -0.16427169740200043, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1414623905380012e-06, | |
| "logits/chosen": -0.38667914271354675, | |
| "logits/rejected": -0.432734876871109, | |
| "logps/chosen": -504.5563049316406, | |
| "logps/rejected": -590.4224853515625, | |
| "loss": 0.1893, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.08736555278301239, | |
| "rewards/margins": 0.06142817810177803, | |
| "rewards/rejected": -0.14879372715950012, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.1026181914754388e-06, | |
| "logits/chosen": -0.44079136848449707, | |
| "logits/rejected": -0.32765600085258484, | |
| "logps/chosen": -426.31103515625, | |
| "logps/rejected": -458.25, | |
| "loss": 0.1883, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.06336107105016708, | |
| "rewards/margins": 0.05088551715016365, | |
| "rewards/rejected": -0.11424658447504044, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0642585798094136e-06, | |
| "logits/chosen": -0.41225147247314453, | |
| "logits/rejected": -0.35755234956741333, | |
| "logps/chosen": -410.5465393066406, | |
| "logps/rejected": -444.13201904296875, | |
| "loss": 0.1893, | |
| "rewards/accuracies": 0.38749998807907104, | |
| "rewards/chosen": -0.07661779224872589, | |
| "rewards/margins": 0.043170757591724396, | |
| "rewards/rejected": -0.11978854984045029, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0263968579775522e-06, | |
| "logits/chosen": -0.43255501985549927, | |
| "logits/rejected": -0.3273727297782898, | |
| "logps/chosen": -493.48895263671875, | |
| "logps/rejected": -553.6163330078125, | |
| "loss": 0.1807, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.0860995277762413, | |
| "rewards/margins": 0.05855479836463928, | |
| "rewards/rejected": -0.1446543186903, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.89046155758058e-07, | |
| "logits/chosen": -0.49861329793930054, | |
| "logits/rejected": -0.3107123374938965, | |
| "logps/chosen": -458.6407775878906, | |
| "logps/rejected": -507.7791442871094, | |
| "loss": 0.1735, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.06812812387943268, | |
| "rewards/margins": 0.05930028110742569, | |
| "rewards/rejected": -0.12742839753627777, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.52219425716534e-07, | |
| "logits/chosen": -0.43581169843673706, | |
| "logits/rejected": -0.3831300139427185, | |
| "logps/chosen": -505.4359436035156, | |
| "logps/rejected": -514.2116088867188, | |
| "loss": 0.191, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.08728592097759247, | |
| "rewards/margins": 0.043969061225652695, | |
| "rewards/rejected": -0.13125498592853546, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 9.15929438714262e-07, | |
| "logits/chosen": -0.4287847876548767, | |
| "logits/rejected": -0.3385041356086731, | |
| "logps/chosen": -453.57763671875, | |
| "logps/rejected": -497.10394287109375, | |
| "loss": 0.1644, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.06731972098350525, | |
| "rewards/margins": 0.06951483339071274, | |
| "rewards/rejected": -0.13683456182479858, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.801887794794911e-07, | |
| "logits/chosen": -0.45681437849998474, | |
| "logits/rejected": -0.33320799469947815, | |
| "logps/chosen": -503.00701904296875, | |
| "logps/rejected": -537.3406982421875, | |
| "loss": 0.186, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.08495952188968658, | |
| "rewards/margins": 0.056953202933073044, | |
| "rewards/rejected": -0.14191272854804993, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.450098422432787e-07, | |
| "logits/chosen": -0.3651468753814697, | |
| "logits/rejected": -0.4669302999973297, | |
| "logps/chosen": -472.07421875, | |
| "logps/rejected": -524.6736450195312, | |
| "loss": 0.1901, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.0767470970749855, | |
| "rewards/margins": 0.049060117453336716, | |
| "rewards/rejected": -0.12580721080303192, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.104048264413858e-07, | |
| "logits/chosen": -0.42758917808532715, | |
| "logits/rejected": -0.3809109330177307, | |
| "logps/chosen": -458.9986877441406, | |
| "logps/rejected": -511.10076904296875, | |
| "loss": 0.1656, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.061626553535461426, | |
| "rewards/margins": 0.05785534530878067, | |
| "rewards/rejected": -0.11948190629482269, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.763857324837321e-07, | |
| "logits/chosen": -0.45241889357566833, | |
| "logits/rejected": -0.48053082823753357, | |
| "logps/chosen": -479.00628662109375, | |
| "logps/rejected": -570.8079223632812, | |
| "loss": 0.1688, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.07606343924999237, | |
| "rewards/margins": 0.077184297144413, | |
| "rewards/rejected": -0.15324774384498596, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.429643575928605e-07, | |
| "logits/chosen": -0.4044586718082428, | |
| "logits/rejected": -0.4395798146724701, | |
| "logps/chosen": -458.71307373046875, | |
| "logps/rejected": -535.9743041992188, | |
| "loss": 0.1802, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.07682407647371292, | |
| "rewards/margins": 0.062133777886629105, | |
| "rewards/rejected": -0.13895784318447113, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.101522917128709e-07, | |
| "logits/chosen": -0.5003517866134644, | |
| "logits/rejected": -0.33631640672683716, | |
| "logps/chosen": -469.9927673339844, | |
| "logps/rejected": -499.16119384765625, | |
| "loss": 0.1919, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.08875006437301636, | |
| "rewards/margins": 0.04381968826055527, | |
| "rewards/rejected": -0.13256976008415222, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.779609134902312e-07, | |
| "logits/chosen": -0.4493805766105652, | |
| "logits/rejected": -0.3728955388069153, | |
| "logps/chosen": -463.2948303222656, | |
| "logps/rejected": -491.2294006347656, | |
| "loss": 0.1858, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.07361346483230591, | |
| "rewards/margins": 0.06247217208147049, | |
| "rewards/rejected": -0.1360856294631958, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.464013863278629e-07, | |
| "logits/chosen": -0.4490404725074768, | |
| "logits/rejected": -0.4013081192970276, | |
| "logps/chosen": -473.97869873046875, | |
| "logps/rejected": -497.76239013671875, | |
| "loss": 0.1927, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.0837077870965004, | |
| "rewards/margins": 0.04747476428747177, | |
| "rewards/rejected": -0.13118253648281097, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.154846545138696e-07, | |
| "logits/chosen": -0.4343733787536621, | |
| "logits/rejected": -0.32414108514785767, | |
| "logps/chosen": -459.42059326171875, | |
| "logps/rejected": -485.1878967285156, | |
| "loss": 0.1714, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.08275729417800903, | |
| "rewards/margins": 0.050310492515563965, | |
| "rewards/rejected": -0.133067786693573, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.852214394262515e-07, | |
| "logits/chosen": -0.38409170508384705, | |
| "logits/rejected": -0.36326664686203003, | |
| "logps/chosen": -479.1509704589844, | |
| "logps/rejected": -543.0540161132812, | |
| "loss": 0.1888, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.08453266322612762, | |
| "rewards/margins": 0.05342210456728935, | |
| "rewards/rejected": -0.13795478641986847, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.556222358149191e-07, | |
| "logits/chosen": -0.3777012526988983, | |
| "logits/rejected": -0.4088537096977234, | |
| "logps/chosen": -484.7447204589844, | |
| "logps/rejected": -536.444091796875, | |
| "loss": 0.1752, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.08025570213794708, | |
| "rewards/margins": 0.061404865235090256, | |
| "rewards/rejected": -0.14166057109832764, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.266973081622992e-07, | |
| "logits/chosen": -0.4726890027523041, | |
| "logits/rejected": -0.31746143102645874, | |
| "logps/chosen": -519.90869140625, | |
| "logps/rejected": -538.6578369140625, | |
| "loss": 0.1785, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.07591833174228668, | |
| "rewards/margins": 0.06365291029214859, | |
| "rewards/rejected": -0.13957123458385468, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.984566871237942e-07, | |
| "logits/chosen": -0.5020835399627686, | |
| "logits/rejected": -0.37848636507987976, | |
| "logps/chosen": -493.8661193847656, | |
| "logps/rejected": -555.9498901367188, | |
| "loss": 0.1738, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.08122778683900833, | |
| "rewards/margins": 0.06830001622438431, | |
| "rewards/rejected": -0.14952778816223145, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.709101660493251e-07, | |
| "logits/chosen": -0.44509878754615784, | |
| "logits/rejected": -0.3348652720451355, | |
| "logps/chosen": -458.56353759765625, | |
| "logps/rejected": -519.80029296875, | |
| "loss": 0.1837, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.07414525747299194, | |
| "rewards/margins": 0.07978249341249466, | |
| "rewards/rejected": -0.153927743434906, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.440672975871743e-07, | |
| "logits/chosen": -0.4507269859313965, | |
| "logits/rejected": -0.28932178020477295, | |
| "logps/chosen": -517.053955078125, | |
| "logps/rejected": -568.24658203125, | |
| "loss": 0.1837, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.08871571719646454, | |
| "rewards/margins": 0.0724785327911377, | |
| "rewards/rejected": -0.16119424998760223, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.1793739037129134e-07, | |
| "logits/chosen": -0.44717317819595337, | |
| "logits/rejected": -0.3539753556251526, | |
| "logps/chosen": -513.8165283203125, | |
| "logps/rejected": -546.3040771484375, | |
| "loss": 0.183, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.08804842829704285, | |
| "rewards/margins": 0.05240340158343315, | |
| "rewards/rejected": -0.1404518187046051, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.9252950579322405e-07, | |
| "logits/chosen": -0.4437643885612488, | |
| "logits/rejected": -0.43425798416137695, | |
| "logps/chosen": -429.33642578125, | |
| "logps/rejected": -503.93695068359375, | |
| "loss": 0.1803, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.06738166511058807, | |
| "rewards/margins": 0.07205136865377426, | |
| "rewards/rejected": -0.13943304121494293, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.6785245485978864e-07, | |
| "logits/chosen": -0.42230138182640076, | |
| "logits/rejected": -0.33158788084983826, | |
| "logps/chosen": -474.5384826660156, | |
| "logps/rejected": -532.5901489257812, | |
| "loss": 0.1876, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.08723129332065582, | |
| "rewards/margins": 0.050881870090961456, | |
| "rewards/rejected": -0.13811317086219788, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.43914795137566e-07, | |
| "logits/chosen": -0.4339612126350403, | |
| "logits/rejected": -0.4076845645904541, | |
| "logps/chosen": -514.6712036132812, | |
| "logps/rejected": -599.03955078125, | |
| "loss": 0.1866, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.09284873306751251, | |
| "rewards/margins": 0.05939553305506706, | |
| "rewards/rejected": -0.15224425494670868, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.207248277852901e-07, | |
| "logits/chosen": -0.35802894830703735, | |
| "logits/rejected": -0.42679715156555176, | |
| "logps/chosen": -499.3904724121094, | |
| "logps/rejected": -601.59619140625, | |
| "loss": 0.1633, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.08157524466514587, | |
| "rewards/margins": 0.08413099497556686, | |
| "rewards/rejected": -0.16570623219013214, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.9829059467515074e-07, | |
| "logits/chosen": -0.47917836904525757, | |
| "logits/rejected": -0.3484811782836914, | |
| "logps/chosen": -485.848388671875, | |
| "logps/rejected": -538.3430786132812, | |
| "loss": 0.1783, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.08013377338647842, | |
| "rewards/margins": 0.0631137266755104, | |
| "rewards/rejected": -0.14324750006198883, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.766198756040153e-07, | |
| "logits/chosen": -0.4456513524055481, | |
| "logits/rejected": -0.32365337014198303, | |
| "logps/chosen": -411.17755126953125, | |
| "logps/rejected": -458.21923828125, | |
| "loss": 0.1884, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.06910572201013565, | |
| "rewards/margins": 0.057366736233234406, | |
| "rewards/rejected": -0.12647245824337006, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.5572018559553155e-07, | |
| "logits/chosen": -0.39375001192092896, | |
| "logits/rejected": -0.37620794773101807, | |
| "logps/chosen": -486.57952880859375, | |
| "logps/rejected": -546.0811157226562, | |
| "loss": 0.1909, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.07672485709190369, | |
| "rewards/margins": 0.0658622533082962, | |
| "rewards/rejected": -0.1425871104001999, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.3559877229404864e-07, | |
| "logits/chosen": -0.4211618900299072, | |
| "logits/rejected": -0.3744751513004303, | |
| "logps/chosen": -487.8899841308594, | |
| "logps/rejected": -532.0867309570312, | |
| "loss": 0.1696, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.08219683170318604, | |
| "rewards/margins": 0.05911056324839592, | |
| "rewards/rejected": -0.14130739867687225, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.1626261345126576e-07, | |
| "logits/chosen": -0.44197311997413635, | |
| "logits/rejected": -0.36461585760116577, | |
| "logps/chosen": -481.8468322753906, | |
| "logps/rejected": -576.697021484375, | |
| "loss": 0.173, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.06508458405733109, | |
| "rewards/margins": 0.08507412672042847, | |
| "rewards/rejected": -0.15015868842601776, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.9771841450646505e-07, | |
| "logits/chosen": -0.5129082798957825, | |
| "logits/rejected": -0.3829159140586853, | |
| "logps/chosen": -495.10162353515625, | |
| "logps/rejected": -520.402587890625, | |
| "loss": 0.184, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.08100192248821259, | |
| "rewards/margins": 0.04889564588665962, | |
| "rewards/rejected": -0.1298975646495819, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.7997260626118758e-07, | |
| "logits/chosen": -0.308317631483078, | |
| "logits/rejected": -0.4445367455482483, | |
| "logps/chosen": -456.50860595703125, | |
| "logps/rejected": -507.1785583496094, | |
| "loss": 0.1696, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.06736288219690323, | |
| "rewards/margins": 0.06645634770393372, | |
| "rewards/rejected": -0.13381922245025635, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.6303134264914365e-07, | |
| "logits/chosen": -0.4731658399105072, | |
| "logits/rejected": -0.40087780356407166, | |
| "logps/chosen": -472.3287048339844, | |
| "logps/rejected": -530.8348388671875, | |
| "loss": 0.1823, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.07341121137142181, | |
| "rewards/margins": 0.05679051950573921, | |
| "rewards/rejected": -0.13020172715187073, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.469004986021355e-07, | |
| "logits/chosen": -0.43905097246170044, | |
| "logits/rejected": -0.4049570560455322, | |
| "logps/chosen": -467.10504150390625, | |
| "logps/rejected": -543.2198486328125, | |
| "loss": 0.1844, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.09198932349681854, | |
| "rewards/margins": 0.05702856183052063, | |
| "rewards/rejected": -0.14901788532733917, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.315856680127367e-07, | |
| "logits/chosen": -0.4321065843105316, | |
| "logits/rejected": -0.3883914649486542, | |
| "logps/chosen": -507.26641845703125, | |
| "logps/rejected": -551.1337280273438, | |
| "loss": 0.1953, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.08602117002010345, | |
| "rewards/margins": 0.05450272560119629, | |
| "rewards/rejected": -0.14052391052246094, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1709216179442817e-07, | |
| "logits/chosen": -0.42082422971725464, | |
| "logits/rejected": -0.41699376702308655, | |
| "logps/chosen": -464.18939208984375, | |
| "logps/rejected": -534.9678344726562, | |
| "loss": 0.1756, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.07923159748315811, | |
| "rewards/margins": 0.06845531612634659, | |
| "rewards/rejected": -0.1476869136095047, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 1.0342500603986421e-07, | |
| "logits/chosen": -0.45704323053359985, | |
| "logits/rejected": -0.3829982876777649, | |
| "logps/chosen": -516.6412353515625, | |
| "logps/rejected": -563.4931640625, | |
| "loss": 0.1887, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.08942435681819916, | |
| "rewards/margins": 0.07117379456758499, | |
| "rewards/rejected": -0.16059814393520355, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.058894027791643e-08, | |
| "logits/chosen": -0.4678065776824951, | |
| "logits/rejected": -0.39768505096435547, | |
| "logps/chosen": -384.6977233886719, | |
| "logps/rejected": -438.79754638671875, | |
| "loss": 0.1838, | |
| "rewards/accuracies": 0.42500001192092896, | |
| "rewards/chosen": -0.05616065859794617, | |
| "rewards/margins": 0.054783742874860764, | |
| "rewards/rejected": -0.11094440519809723, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.858841583008592e-08, | |
| "logits/chosen": -0.45986437797546387, | |
| "logits/rejected": -0.3251883089542389, | |
| "logps/chosen": -523.9485473632812, | |
| "logps/rejected": -555.4884033203125, | |
| "loss": 0.1823, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.0797450989484787, | |
| "rewards/margins": 0.05564265325665474, | |
| "rewards/rejected": -0.13538774847984314, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 6.742759426686313e-08, | |
| "logits/chosen": -0.4108913540840149, | |
| "logits/rejected": -0.2947922945022583, | |
| "logps/chosen": -538.8464965820312, | |
| "logps/rejected": -580.0211791992188, | |
| "loss": 0.192, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.09397675096988678, | |
| "rewards/margins": 0.05662484094500542, | |
| "rewards/rejected": -0.1506015956401825, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.7110345964571104e-08, | |
| "logits/chosen": -0.31575465202331543, | |
| "logits/rejected": -0.376641184091568, | |
| "logps/chosen": -487.80633544921875, | |
| "logps/rejected": -575.765625, | |
| "loss": 0.1692, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.0703391507267952, | |
| "rewards/margins": 0.08664790540933609, | |
| "rewards/rejected": -0.1569870412349701, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.764024876318357e-08, | |
| "logits/chosen": -0.414639949798584, | |
| "logits/rejected": -0.43213406205177307, | |
| "logps/chosen": -588.9749755859375, | |
| "logps/rejected": -631.2474365234375, | |
| "loss": 0.1871, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.09701113402843475, | |
| "rewards/margins": 0.057556282728910446, | |
| "rewards/rejected": -0.1545674204826355, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.902058672559633e-08, | |
| "logits/chosen": -0.42236852645874023, | |
| "logits/rejected": -0.3948751389980316, | |
| "logps/chosen": -486.19598388671875, | |
| "logps/rejected": -501.712646484375, | |
| "loss": 0.1907, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.08502433449029922, | |
| "rewards/margins": 0.051293183118104935, | |
| "rewards/rejected": -0.13631752133369446, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.125434899876933e-08, | |
| "logits/chosen": -0.4508518576622009, | |
| "logits/rejected": -0.2803448438644409, | |
| "logps/chosen": -528.0969848632812, | |
| "logps/rejected": -578.2789306640625, | |
| "loss": 0.1681, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.0905873253941536, | |
| "rewards/margins": 0.07002563774585724, | |
| "rewards/rejected": -0.16061295568943024, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.4344228777145873e-08, | |
| "logits/chosen": -0.4513494372367859, | |
| "logits/rejected": -0.4356984496116638, | |
| "logps/chosen": -499.56884765625, | |
| "logps/rejected": -582.083740234375, | |
| "loss": 0.1743, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.08331683278083801, | |
| "rewards/margins": 0.07355960458517075, | |
| "rewards/rejected": -0.15687642991542816, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.829262236869772e-08, | |
| "logits/chosen": -0.46016424894332886, | |
| "logits/rejected": -0.29402580857276917, | |
| "logps/chosen": -451.9331970214844, | |
| "logps/rejected": -494.04815673828125, | |
| "loss": 0.1806, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.0696534812450409, | |
| "rewards/margins": 0.06958705186843872, | |
| "rewards/rejected": -0.13924053311347961, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.3101628363929586e-08, | |
| "logits/chosen": -0.501002848148346, | |
| "logits/rejected": -0.35800036787986755, | |
| "logps/chosen": -489.6319274902344, | |
| "logps/rejected": -548.3452758789062, | |
| "loss": 0.1739, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.08254210650920868, | |
| "rewards/margins": 0.06711296737194061, | |
| "rewards/rejected": -0.1496550738811493, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.773046908123195e-09, | |
| "logits/chosen": -0.42636099457740784, | |
| "logits/rejected": -0.3364613652229309, | |
| "logps/chosen": -483.1874084472656, | |
| "logps/rejected": -515.3287353515625, | |
| "loss": 0.172, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.07863263040781021, | |
| "rewards/margins": 0.057226501405239105, | |
| "rewards/rejected": -0.1358591467142105, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 5.308379077080817e-09, | |
| "logits/chosen": -0.45698413252830505, | |
| "logits/rejected": -0.3893563151359558, | |
| "logps/chosen": -492.31610107421875, | |
| "logps/rejected": -549.0144653320312, | |
| "loss": 0.1795, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.08794066309928894, | |
| "rewards/margins": 0.055044692009687424, | |
| "rewards/rejected": -0.14298535883426666, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2.7088263565760996e-09, | |
| "logits/chosen": -0.4179384112358093, | |
| "logits/rejected": -0.3843657672405243, | |
| "logps/chosen": -512.0191650390625, | |
| "logps/rejected": -557.4730224609375, | |
| "loss": 0.1648, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.07361841946840286, | |
| "rewards/margins": 0.06739214807748795, | |
| "rewards/rejected": -0.141010582447052, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 9.752902257023633e-10, | |
| "logits/chosen": -0.4060386121273041, | |
| "logits/rejected": -0.3960750699043274, | |
| "logps/chosen": -479.4234313964844, | |
| "logps/rejected": -551.9636840820312, | |
| "loss": 0.1975, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.08916465938091278, | |
| "rewards/margins": 0.05007479339838028, | |
| "rewards/rejected": -0.13923944532871246, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 1.083718442532189e-10, | |
| "logits/chosen": -0.42799386382102966, | |
| "logits/rejected": -0.3839925229549408, | |
| "logps/chosen": -454.05499267578125, | |
| "logps/rejected": -501.215087890625, | |
| "loss": 0.1847, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.07664839923381805, | |
| "rewards/margins": 0.05011936277151108, | |
| "rewards/rejected": -0.12676775455474854, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1875, | |
| "total_flos": 0.0, | |
| "train_loss": 0.1858275575319926, | |
| "train_runtime": 11460.3939, | |
| "train_samples_per_second": 2.618, | |
| "train_steps_per_second": 0.164 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1875, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |