| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.998451213216314, | |
| "eval_steps": 100, | |
| "global_step": 2904, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.7182130584192438e-09, | |
| "logits/chosen": -2.9648265838623047, | |
| "logits/rejected": -2.9711227416992188, | |
| "logps/chosen": -256.0919494628906, | |
| "logps/rejected": -234.60708618164062, | |
| "loss": 1.0, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 1.718213058419244e-08, | |
| "logits/chosen": -3.049875497817993, | |
| "logits/rejected": -3.0188238620758057, | |
| "logps/chosen": -276.6912536621094, | |
| "logps/rejected": -202.39605712890625, | |
| "loss": 1.0001, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.0005764114903286099, | |
| "rewards/margins": -0.006484686397016048, | |
| "rewards/rejected": 0.007061097305268049, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 3.436426116838488e-08, | |
| "logits/chosen": -2.988577127456665, | |
| "logits/rejected": -2.9995627403259277, | |
| "logps/chosen": -312.2018127441406, | |
| "logps/rejected": -246.76266479492188, | |
| "loss": 1.0026, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": 0.003506724489852786, | |
| "rewards/margins": -0.0012849611230194569, | |
| "rewards/rejected": 0.004791685380041599, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 5.154639175257731e-08, | |
| "logits/chosen": -3.063732624053955, | |
| "logits/rejected": -3.0357906818389893, | |
| "logps/chosen": -260.15679931640625, | |
| "logps/rejected": -224.3686065673828, | |
| "loss": 0.9974, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.0021156296133995056, | |
| "rewards/margins": 0.0043937130831182, | |
| "rewards/rejected": -0.0022780844010412693, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 6.872852233676976e-08, | |
| "logits/chosen": -3.070286273956299, | |
| "logits/rejected": -3.0322961807250977, | |
| "logps/chosen": -299.5580139160156, | |
| "logps/rejected": -250.05123901367188, | |
| "loss": 0.9971, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.00013011172995902598, | |
| "rewards/margins": -0.0006899217842146754, | |
| "rewards/rejected": 0.0008200337179005146, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 8.59106529209622e-08, | |
| "logits/chosen": -3.019392490386963, | |
| "logits/rejected": -3.024167060852051, | |
| "logps/chosen": -289.3621520996094, | |
| "logps/rejected": -224.00979614257812, | |
| "loss": 1.003, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.0025630726013332605, | |
| "rewards/margins": -0.0055747563019394875, | |
| "rewards/rejected": 0.003011685097590089, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 1.0309278350515462e-07, | |
| "logits/chosen": -3.0416665077209473, | |
| "logits/rejected": -3.020573616027832, | |
| "logps/chosen": -247.55380249023438, | |
| "logps/rejected": -226.4866943359375, | |
| "loss": 0.9977, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": 0.0035637759137898684, | |
| "rewards/margins": 0.0046446239575743675, | |
| "rewards/rejected": -0.0010808479273691773, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 1.202749140893471e-07, | |
| "logits/chosen": -3.0761704444885254, | |
| "logits/rejected": -3.058954954147339, | |
| "logps/chosen": -305.7156677246094, | |
| "logps/rejected": -251.75009155273438, | |
| "loss": 0.9938, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.004508626647293568, | |
| "rewards/margins": 0.007502266205847263, | |
| "rewards/rejected": -0.0029936402570456266, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.3745704467353952e-07, | |
| "logits/chosen": -3.049072742462158, | |
| "logits/rejected": -3.0234384536743164, | |
| "logps/chosen": -293.57989501953125, | |
| "logps/rejected": -240.2385711669922, | |
| "loss": 0.9985, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.002101506572216749, | |
| "rewards/margins": 0.002054845681414008, | |
| "rewards/rejected": 4.6660610678372905e-05, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 1.5463917525773197e-07, | |
| "logits/chosen": -3.079655885696411, | |
| "logits/rejected": -3.0430655479431152, | |
| "logps/chosen": -259.3849792480469, | |
| "logps/rejected": -216.38330078125, | |
| "loss": 0.9993, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.001981315901502967, | |
| "rewards/margins": 0.003211658913642168, | |
| "rewards/rejected": -0.0012303430121392012, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 1.718213058419244e-07, | |
| "logits/chosen": -3.0111169815063477, | |
| "logits/rejected": -3.006265640258789, | |
| "logps/chosen": -267.73577880859375, | |
| "logps/rejected": -222.9344482421875, | |
| "loss": 1.0009, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.005559581331908703, | |
| "rewards/margins": 0.004569889511913061, | |
| "rewards/rejected": 0.000989692285656929, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 1.8900343642611682e-07, | |
| "logits/chosen": -3.0228209495544434, | |
| "logits/rejected": -2.9778640270233154, | |
| "logps/chosen": -269.3376770019531, | |
| "logps/rejected": -230.95877075195312, | |
| "loss": 0.9971, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.0013723246520385146, | |
| "rewards/margins": 0.0024911228101700544, | |
| "rewards/rejected": -0.0011187975760549307, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.0618556701030925e-07, | |
| "logits/chosen": -3.0604119300842285, | |
| "logits/rejected": -3.0274159908294678, | |
| "logps/chosen": -310.97454833984375, | |
| "logps/rejected": -232.7030029296875, | |
| "loss": 0.9997, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": 0.0033910819329321384, | |
| "rewards/margins": 0.0011182299349457026, | |
| "rewards/rejected": 0.002272851997986436, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 2.2336769759450173e-07, | |
| "logits/chosen": -3.1223320960998535, | |
| "logits/rejected": -3.0860095024108887, | |
| "logps/chosen": -286.6527099609375, | |
| "logps/rejected": -241.933349609375, | |
| "loss": 0.996, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": 0.0031617667991667986, | |
| "rewards/margins": 0.005714719649404287, | |
| "rewards/rejected": -0.0025529528502374887, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 2.405498281786942e-07, | |
| "logits/chosen": -3.045279026031494, | |
| "logits/rejected": -3.040912628173828, | |
| "logps/chosen": -292.4465637207031, | |
| "logps/rejected": -234.72903442382812, | |
| "loss": 0.9968, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.0010984055697917938, | |
| "rewards/margins": 0.002134153386577964, | |
| "rewards/rejected": -0.00103574781678617, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 2.5773195876288655e-07, | |
| "logits/chosen": -3.0267093181610107, | |
| "logits/rejected": -3.0170674324035645, | |
| "logps/chosen": -275.6455993652344, | |
| "logps/rejected": -256.4563903808594, | |
| "loss": 0.9998, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.0011621230514720082, | |
| "rewards/margins": -2.7875230443896726e-05, | |
| "rewards/rejected": -0.001134247868321836, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 2.7491408934707903e-07, | |
| "logits/chosen": -3.08597731590271, | |
| "logits/rejected": -3.0747408866882324, | |
| "logps/chosen": -278.91754150390625, | |
| "logps/rejected": -218.88558959960938, | |
| "loss": 0.9949, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.004570486024022102, | |
| "rewards/margins": 0.01147634070366621, | |
| "rewards/rejected": -0.006905855145305395, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 2.9209621993127146e-07, | |
| "logits/chosen": -3.0499391555786133, | |
| "logits/rejected": -3.036341905593872, | |
| "logps/chosen": -292.2102966308594, | |
| "logps/rejected": -236.95703125, | |
| "loss": 0.9925, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.0019660559482872486, | |
| "rewards/margins": 0.009962075389921665, | |
| "rewards/rejected": -0.00799601897597313, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 3.0927835051546394e-07, | |
| "logits/chosen": -3.091757297515869, | |
| "logits/rejected": -3.080824375152588, | |
| "logps/chosen": -248.6970672607422, | |
| "logps/rejected": -211.7375946044922, | |
| "loss": 0.993, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.0008403388783335686, | |
| "rewards/margins": 0.004896977450698614, | |
| "rewards/rejected": -0.004056639038026333, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 3.2646048109965636e-07, | |
| "logits/chosen": -3.0578689575195312, | |
| "logits/rejected": -3.033844232559204, | |
| "logps/chosen": -246.07040405273438, | |
| "logps/rejected": -200.9595184326172, | |
| "loss": 0.9978, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.0035698451101779938, | |
| "rewards/margins": 0.012628579512238503, | |
| "rewards/rejected": -0.009058734402060509, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 3.436426116838488e-07, | |
| "logits/chosen": -3.08526349067688, | |
| "logits/rejected": -3.063560724258423, | |
| "logps/chosen": -252.265869140625, | |
| "logps/rejected": -192.21331787109375, | |
| "loss": 0.99, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.009008857421576977, | |
| "rewards/margins": 0.016440508887171745, | |
| "rewards/rejected": -0.007431652396917343, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 3.608247422680412e-07, | |
| "logits/chosen": -2.9606690406799316, | |
| "logits/rejected": -2.906953811645508, | |
| "logps/chosen": -292.0260925292969, | |
| "logps/rejected": -215.34036254882812, | |
| "loss": 0.9855, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.0046757906675338745, | |
| "rewards/margins": 0.013047484681010246, | |
| "rewards/rejected": -0.008371694944798946, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 3.7800687285223364e-07, | |
| "logits/chosen": -2.990668773651123, | |
| "logits/rejected": -2.986696481704712, | |
| "logps/chosen": -260.21832275390625, | |
| "logps/rejected": -237.1192626953125, | |
| "loss": 0.9839, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": 0.006101504433900118, | |
| "rewards/margins": 0.011856775730848312, | |
| "rewards/rejected": -0.005755270831286907, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 3.9518900343642607e-07, | |
| "logits/chosen": -3.0254712104797363, | |
| "logits/rejected": -3.0170722007751465, | |
| "logps/chosen": -263.41680908203125, | |
| "logps/rejected": -240.377685546875, | |
| "loss": 0.9888, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.006732765585184097, | |
| "rewards/margins": 0.014450904913246632, | |
| "rewards/rejected": -0.007718136068433523, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.123711340206185e-07, | |
| "logits/chosen": -3.078042507171631, | |
| "logits/rejected": -3.0471174716949463, | |
| "logps/chosen": -299.16107177734375, | |
| "logps/rejected": -214.18759155273438, | |
| "loss": 0.9867, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.006704004947096109, | |
| "rewards/margins": 0.01422835886478424, | |
| "rewards/rejected": -0.007524352520704269, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.2955326460481097e-07, | |
| "logits/chosen": -3.026909589767456, | |
| "logits/rejected": -3.018611431121826, | |
| "logps/chosen": -264.2486572265625, | |
| "logps/rejected": -233.31826782226562, | |
| "loss": 0.9832, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.011163066141307354, | |
| "rewards/margins": 0.019244546070694923, | |
| "rewards/rejected": -0.008081478998064995, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.4673539518900345e-07, | |
| "logits/chosen": -3.0170772075653076, | |
| "logits/rejected": -3.0285823345184326, | |
| "logps/chosen": -263.9449157714844, | |
| "logps/rejected": -219.4688262939453, | |
| "loss": 0.9828, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.010579807683825493, | |
| "rewards/margins": 0.024952661246061325, | |
| "rewards/rejected": -0.014372853562235832, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.639175257731959e-07, | |
| "logits/chosen": -3.0661990642547607, | |
| "logits/rejected": -3.0676910877227783, | |
| "logps/chosen": -274.22003173828125, | |
| "logps/rejected": -229.6044158935547, | |
| "loss": 0.9799, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": 0.004864403046667576, | |
| "rewards/margins": 0.016584355384111404, | |
| "rewards/rejected": -0.011719951406121254, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.810996563573884e-07, | |
| "logits/chosen": -3.031026840209961, | |
| "logits/rejected": -3.0113377571105957, | |
| "logps/chosen": -283.8157653808594, | |
| "logps/rejected": -235.0233612060547, | |
| "loss": 0.9754, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.013933306559920311, | |
| "rewards/margins": 0.027073601260781288, | |
| "rewards/rejected": -0.013140290975570679, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.982817869415807e-07, | |
| "logits/chosen": -3.094398021697998, | |
| "logits/rejected": -3.0440070629119873, | |
| "logps/chosen": -270.22052001953125, | |
| "logps/rejected": -223.65493774414062, | |
| "loss": 0.9774, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.008421173319220543, | |
| "rewards/margins": 0.019475247710943222, | |
| "rewards/rejected": -0.01105407439172268, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.982778415614236e-07, | |
| "logits/chosen": -3.023087978363037, | |
| "logits/rejected": -2.9992034435272217, | |
| "logps/chosen": -244.30337524414062, | |
| "logps/rejected": -218.7770538330078, | |
| "loss": 0.9735, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": 0.008720096200704575, | |
| "rewards/margins": 0.029811996966600418, | |
| "rewards/rejected": -0.021091898903250694, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.963643321852277e-07, | |
| "logits/chosen": -3.0621352195739746, | |
| "logits/rejected": -3.0481069087982178, | |
| "logps/chosen": -299.58758544921875, | |
| "logps/rejected": -257.4301452636719, | |
| "loss": 0.9719, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.011654629372060299, | |
| "rewards/margins": 0.02595471777021885, | |
| "rewards/rejected": -0.01430008839815855, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.944508228090318e-07, | |
| "logits/chosen": -3.026646137237549, | |
| "logits/rejected": -3.0066604614257812, | |
| "logps/chosen": -242.5664825439453, | |
| "logps/rejected": -187.6553497314453, | |
| "loss": 0.9641, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": 0.014026440680027008, | |
| "rewards/margins": 0.0420592799782753, | |
| "rewards/rejected": -0.02803283929824829, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.925373134328357e-07, | |
| "logits/chosen": -3.0689878463745117, | |
| "logits/rejected": -3.052264928817749, | |
| "logps/chosen": -303.94036865234375, | |
| "logps/rejected": -238.4488067626953, | |
| "loss": 0.9634, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.02398153766989708, | |
| "rewards/margins": 0.042572326958179474, | |
| "rewards/rejected": -0.018590793013572693, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.906238040566398e-07, | |
| "logits/chosen": -2.995407819747925, | |
| "logits/rejected": -2.9780545234680176, | |
| "logps/chosen": -235.09848022460938, | |
| "logps/rejected": -236.380859375, | |
| "loss": 0.9616, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.006325787398964167, | |
| "rewards/margins": 0.03254387527704239, | |
| "rewards/rejected": -0.02621809020638466, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.887102946804438e-07, | |
| "logits/chosen": -3.066584348678589, | |
| "logits/rejected": -3.03863263130188, | |
| "logps/chosen": -306.0690612792969, | |
| "logps/rejected": -245.27407836914062, | |
| "loss": 0.9508, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.02622169815003872, | |
| "rewards/margins": 0.059129487723112106, | |
| "rewards/rejected": -0.032907791435718536, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 4.867967853042479e-07, | |
| "logits/chosen": -3.0108275413513184, | |
| "logits/rejected": -3.008779525756836, | |
| "logps/chosen": -294.0123596191406, | |
| "logps/rejected": -248.9111785888672, | |
| "loss": 0.9496, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.008144224062561989, | |
| "rewards/margins": 0.045391060411930084, | |
| "rewards/rejected": -0.037246834486722946, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 4.84883275928052e-07, | |
| "logits/chosen": -3.048269510269165, | |
| "logits/rejected": -3.011050224304199, | |
| "logps/chosen": -303.9602355957031, | |
| "logps/rejected": -269.4437561035156, | |
| "loss": 0.9473, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.021487019956111908, | |
| "rewards/margins": 0.06678290665149689, | |
| "rewards/rejected": -0.04529587924480438, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 4.82969766551856e-07, | |
| "logits/chosen": -3.0777668952941895, | |
| "logits/rejected": -3.068040370941162, | |
| "logps/chosen": -282.12713623046875, | |
| "logps/rejected": -236.7052459716797, | |
| "loss": 0.9515, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": 0.02360449731349945, | |
| "rewards/margins": 0.05431375652551651, | |
| "rewards/rejected": -0.03070926107466221, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 4.810562571756601e-07, | |
| "logits/chosen": -2.9710302352905273, | |
| "logits/rejected": -2.983682155609131, | |
| "logps/chosen": -272.12713623046875, | |
| "logps/rejected": -235.8425750732422, | |
| "loss": 0.9475, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": 0.013235519640147686, | |
| "rewards/margins": 0.05444386601448059, | |
| "rewards/rejected": -0.04120834544301033, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 4.791427477994642e-07, | |
| "logits/chosen": -3.0254111289978027, | |
| "logits/rejected": -3.006087303161621, | |
| "logps/chosen": -254.69107055664062, | |
| "logps/rejected": -210.39474487304688, | |
| "loss": 0.9402, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.022726301103830338, | |
| "rewards/margins": 0.06853805482387543, | |
| "rewards/rejected": -0.04581175372004509, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 4.772292384232682e-07, | |
| "logits/chosen": -3.0280232429504395, | |
| "logits/rejected": -2.9936630725860596, | |
| "logps/chosen": -261.80731201171875, | |
| "logps/rejected": -251.14950561523438, | |
| "loss": 0.9398, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": 0.014309520833194256, | |
| "rewards/margins": 0.050246305763721466, | |
| "rewards/rejected": -0.03593678027391434, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 4.753157290470723e-07, | |
| "logits/chosen": -3.0422613620758057, | |
| "logits/rejected": -3.004459857940674, | |
| "logps/chosen": -245.8754425048828, | |
| "logps/rejected": -202.38157653808594, | |
| "loss": 0.9391, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": 0.005309578962624073, | |
| "rewards/margins": 0.0503067672252655, | |
| "rewards/rejected": -0.0449971929192543, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 4.7340221967087635e-07, | |
| "logits/chosen": -3.061112880706787, | |
| "logits/rejected": -3.045253038406372, | |
| "logps/chosen": -257.9686584472656, | |
| "logps/rejected": -239.047119140625, | |
| "loss": 0.9323, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.012545446865260601, | |
| "rewards/margins": 0.07196511328220367, | |
| "rewards/rejected": -0.0594196543097496, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 4.714887102946804e-07, | |
| "logits/chosen": -3.0016167163848877, | |
| "logits/rejected": -2.9664511680603027, | |
| "logps/chosen": -278.2596740722656, | |
| "logps/rejected": -246.8672637939453, | |
| "loss": 0.9194, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.00037018657894805074, | |
| "rewards/margins": 0.0750364139676094, | |
| "rewards/rejected": -0.07466623187065125, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.6957520091848447e-07, | |
| "logits/chosen": -3.0514495372772217, | |
| "logits/rejected": -3.0419204235076904, | |
| "logps/chosen": -272.3708801269531, | |
| "logps/rejected": -227.4208221435547, | |
| "loss": 0.9356, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": 0.01730353757739067, | |
| "rewards/margins": 0.059229202568531036, | |
| "rewards/rejected": -0.04192566126585007, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 4.6766169154228853e-07, | |
| "logits/chosen": -3.0307857990264893, | |
| "logits/rejected": -3.0139756202697754, | |
| "logps/chosen": -264.16168212890625, | |
| "logps/rejected": -219.8409881591797, | |
| "loss": 0.9122, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.007891577668488026, | |
| "rewards/margins": 0.07121269404888153, | |
| "rewards/rejected": -0.06332111358642578, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 4.657481821660926e-07, | |
| "logits/chosen": -3.016810178756714, | |
| "logits/rejected": -3.019348621368408, | |
| "logps/chosen": -238.8083038330078, | |
| "logps/rejected": -212.76193237304688, | |
| "loss": 0.9028, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.028794366866350174, | |
| "rewards/margins": 0.08307679742574692, | |
| "rewards/rejected": -0.054282426834106445, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 4.6383467278989666e-07, | |
| "logits/chosen": -3.044045925140381, | |
| "logits/rejected": -3.0309886932373047, | |
| "logps/chosen": -282.59814453125, | |
| "logps/rejected": -229.63858032226562, | |
| "loss": 0.9063, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.022610556334257126, | |
| "rewards/margins": 0.09912824630737305, | |
| "rewards/rejected": -0.07651769369840622, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 4.6192116341370067e-07, | |
| "logits/chosen": -3.0251693725585938, | |
| "logits/rejected": -3.040748119354248, | |
| "logps/chosen": -289.87896728515625, | |
| "logps/rejected": -243.85952758789062, | |
| "loss": 0.8948, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": 0.03339407593011856, | |
| "rewards/margins": 0.1147690862417221, | |
| "rewards/rejected": -0.08137501776218414, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 4.6000765403750473e-07, | |
| "logits/chosen": -3.0105278491973877, | |
| "logits/rejected": -2.9718642234802246, | |
| "logps/chosen": -244.4474334716797, | |
| "logps/rejected": -221.8011474609375, | |
| "loss": 0.8829, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": 0.017204025760293007, | |
| "rewards/margins": 0.0894416943192482, | |
| "rewards/rejected": -0.07223766297101974, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 4.580941446613088e-07, | |
| "logits/chosen": -2.9197583198547363, | |
| "logits/rejected": -2.915809392929077, | |
| "logps/chosen": -304.55645751953125, | |
| "logps/rejected": -253.95828247070312, | |
| "loss": 0.8834, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.02509082481265068, | |
| "rewards/margins": 0.11200642585754395, | |
| "rewards/rejected": -0.08691558986902237, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 4.5618063528511285e-07, | |
| "logits/chosen": -3.0240163803100586, | |
| "logits/rejected": -2.998610019683838, | |
| "logps/chosen": -270.31378173828125, | |
| "logps/rejected": -242.4883270263672, | |
| "loss": 0.8813, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.02489570900797844, | |
| "rewards/margins": 0.13008132576942444, | |
| "rewards/rejected": -0.1051856279373169, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 4.542671259089169e-07, | |
| "logits/chosen": -3.0373079776763916, | |
| "logits/rejected": -3.0401182174682617, | |
| "logps/chosen": -235.19534301757812, | |
| "logps/rejected": -221.75363159179688, | |
| "loss": 0.8933, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.00027992590912617743, | |
| "rewards/margins": 0.09569151699542999, | |
| "rewards/rejected": -0.09541159123182297, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 4.52353616532721e-07, | |
| "logits/chosen": -3.0647318363189697, | |
| "logits/rejected": -3.054605484008789, | |
| "logps/chosen": -280.6692810058594, | |
| "logps/rejected": -237.3024444580078, | |
| "loss": 0.8826, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.03587502986192703, | |
| "rewards/margins": 0.12422885000705719, | |
| "rewards/rejected": -0.08835381269454956, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 4.5044010715652504e-07, | |
| "logits/chosen": -3.0048820972442627, | |
| "logits/rejected": -3.0057966709136963, | |
| "logps/chosen": -256.4432373046875, | |
| "logps/rejected": -229.13198852539062, | |
| "loss": 0.878, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.039617545902729034, | |
| "rewards/margins": 0.11413271725177765, | |
| "rewards/rejected": -0.07451517134904861, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 4.485265977803291e-07, | |
| "logits/chosen": -3.0452122688293457, | |
| "logits/rejected": -3.041473865509033, | |
| "logps/chosen": -293.8966369628906, | |
| "logps/rejected": -251.0624542236328, | |
| "loss": 0.8656, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.04036722332239151, | |
| "rewards/margins": 0.1490786224603653, | |
| "rewards/rejected": -0.10871138423681259, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 4.4661308840413316e-07, | |
| "logits/chosen": -3.0647714138031006, | |
| "logits/rejected": -3.01206636428833, | |
| "logps/chosen": -260.71343994140625, | |
| "logps/rejected": -221.20632934570312, | |
| "loss": 0.8666, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.011453949846327305, | |
| "rewards/margins": 0.13622693717479706, | |
| "rewards/rejected": -0.12477298080921173, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 4.446995790279372e-07, | |
| "logits/chosen": -3.034562349319458, | |
| "logits/rejected": -2.989062547683716, | |
| "logps/chosen": -291.01446533203125, | |
| "logps/rejected": -241.4232177734375, | |
| "loss": 0.8519, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": 0.035387031733989716, | |
| "rewards/margins": 0.18669307231903076, | |
| "rewards/rejected": -0.15130606293678284, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 4.4278606965174123e-07, | |
| "logits/chosen": -3.012864351272583, | |
| "logits/rejected": -3.005479097366333, | |
| "logps/chosen": -243.08975219726562, | |
| "logps/rejected": -216.92074584960938, | |
| "loss": 0.8492, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.002515086205676198, | |
| "rewards/margins": 0.121725894510746, | |
| "rewards/rejected": -0.11921081691980362, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 4.408725602755453e-07, | |
| "logits/chosen": -3.013345718383789, | |
| "logits/rejected": -3.0360682010650635, | |
| "logps/chosen": -248.50326538085938, | |
| "logps/rejected": -203.6788787841797, | |
| "loss": 0.8617, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": 0.0035158656537532806, | |
| "rewards/margins": 0.14544394612312317, | |
| "rewards/rejected": -0.1419280618429184, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 4.3895905089934936e-07, | |
| "logits/chosen": -3.047393321990967, | |
| "logits/rejected": -3.0524885654449463, | |
| "logps/chosen": -255.6022491455078, | |
| "logps/rejected": -219.2570343017578, | |
| "loss": 0.8256, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.03486616909503937, | |
| "rewards/margins": 0.1496470421552658, | |
| "rewards/rejected": -0.11478086560964584, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 4.370455415231534e-07, | |
| "logits/chosen": -3.066741466522217, | |
| "logits/rejected": -3.046435832977295, | |
| "logps/chosen": -305.30712890625, | |
| "logps/rejected": -249.01968383789062, | |
| "loss": 0.8203, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": 0.04162443429231644, | |
| "rewards/margins": 0.18165114521980286, | |
| "rewards/rejected": -0.14002671837806702, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 4.351320321469575e-07, | |
| "logits/chosen": -2.990051746368408, | |
| "logits/rejected": -2.992987632751465, | |
| "logps/chosen": -251.0054168701172, | |
| "logps/rejected": -238.8704376220703, | |
| "loss": 0.8282, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.028837282210588455, | |
| "rewards/margins": 0.1591511368751526, | |
| "rewards/rejected": -0.13031385838985443, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 4.3321852277076154e-07, | |
| "logits/chosen": -3.0534205436706543, | |
| "logits/rejected": -3.0434939861297607, | |
| "logps/chosen": -256.14019775390625, | |
| "logps/rejected": -216.6497802734375, | |
| "loss": 0.8083, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": 0.034832023084163666, | |
| "rewards/margins": 0.20265790820121765, | |
| "rewards/rejected": -0.16782590746879578, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 4.313050133945656e-07, | |
| "logits/chosen": -3.0195870399475098, | |
| "logits/rejected": -3.027886390686035, | |
| "logps/chosen": -277.5207214355469, | |
| "logps/rejected": -232.2840576171875, | |
| "loss": 0.8119, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.05330047011375427, | |
| "rewards/margins": 0.21206972002983093, | |
| "rewards/rejected": -0.15876924991607666, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 4.2939150401836967e-07, | |
| "logits/chosen": -3.0407612323760986, | |
| "logits/rejected": -3.014266014099121, | |
| "logps/chosen": -271.2508850097656, | |
| "logps/rejected": -233.294189453125, | |
| "loss": 0.792, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.02551809512078762, | |
| "rewards/margins": 0.21874013543128967, | |
| "rewards/rejected": -0.1932220160961151, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 4.2747799464217373e-07, | |
| "logits/chosen": -2.994800567626953, | |
| "logits/rejected": -2.9698376655578613, | |
| "logps/chosen": -287.5726623535156, | |
| "logps/rejected": -237.8695068359375, | |
| "loss": 0.8042, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 0.006596171762794256, | |
| "rewards/margins": 0.18577079474925995, | |
| "rewards/rejected": -0.17917463183403015, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 4.255644852659778e-07, | |
| "logits/chosen": -3.0441243648529053, | |
| "logits/rejected": -3.053039073944092, | |
| "logps/chosen": -281.0970153808594, | |
| "logps/rejected": -227.56851196289062, | |
| "loss": 0.8389, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.022974971681833267, | |
| "rewards/margins": 0.15209509432315826, | |
| "rewards/rejected": -0.1291201412677765, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 4.236509758897818e-07, | |
| "logits/chosen": -3.0892813205718994, | |
| "logits/rejected": -3.0816287994384766, | |
| "logps/chosen": -258.57781982421875, | |
| "logps/rejected": -230.27615356445312, | |
| "loss": 0.8064, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": 0.006377282552421093, | |
| "rewards/margins": 0.21148601174354553, | |
| "rewards/rejected": -0.20510873198509216, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 4.2173746651358586e-07, | |
| "logits/chosen": -3.0129265785217285, | |
| "logits/rejected": -2.9877090454101562, | |
| "logps/chosen": -275.82568359375, | |
| "logps/rejected": -219.6047821044922, | |
| "loss": 0.8098, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": 0.04388806223869324, | |
| "rewards/margins": 0.2091568410396576, | |
| "rewards/rejected": -0.16526879370212555, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 4.198239571373899e-07, | |
| "logits/chosen": -2.9762985706329346, | |
| "logits/rejected": -2.986323833465576, | |
| "logps/chosen": -262.74371337890625, | |
| "logps/rejected": -207.13418579101562, | |
| "loss": 0.8056, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.02364834025502205, | |
| "rewards/margins": 0.2192881554365158, | |
| "rewards/rejected": -0.19563981890678406, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 4.17910447761194e-07, | |
| "logits/chosen": -3.0574042797088623, | |
| "logits/rejected": -3.0168094635009766, | |
| "logps/chosen": -296.2746276855469, | |
| "logps/rejected": -256.06854248046875, | |
| "loss": 0.8097, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.013486223295331001, | |
| "rewards/margins": 0.14934802055358887, | |
| "rewards/rejected": -0.13586178421974182, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 4.1599693838499805e-07, | |
| "logits/chosen": -3.0353150367736816, | |
| "logits/rejected": -3.0224339962005615, | |
| "logps/chosen": -261.10992431640625, | |
| "logps/rejected": -251.72293090820312, | |
| "loss": 0.7899, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.007094231434166431, | |
| "rewards/margins": 0.16344769299030304, | |
| "rewards/rejected": -0.1705418974161148, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 4.140834290088021e-07, | |
| "logits/chosen": -3.008420944213867, | |
| "logits/rejected": -2.9702980518341064, | |
| "logps/chosen": -269.81903076171875, | |
| "logps/rejected": -199.1494903564453, | |
| "loss": 0.7627, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.01843448355793953, | |
| "rewards/margins": 0.2920437455177307, | |
| "rewards/rejected": -0.2736092209815979, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 4.121699196326062e-07, | |
| "logits/chosen": -3.004692792892456, | |
| "logits/rejected": -2.9878294467926025, | |
| "logps/chosen": -262.6126708984375, | |
| "logps/rejected": -220.02096557617188, | |
| "loss": 0.7653, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.001659922068938613, | |
| "rewards/margins": 0.25197452306747437, | |
| "rewards/rejected": -0.2503146231174469, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 4.1025641025641024e-07, | |
| "logits/chosen": -2.9879119396209717, | |
| "logits/rejected": -2.980886459350586, | |
| "logps/chosen": -262.13287353515625, | |
| "logps/rejected": -249.5322723388672, | |
| "loss": 0.782, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.036843255162239075, | |
| "rewards/margins": 0.17954595386981964, | |
| "rewards/rejected": -0.21638920903205872, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 4.083429008802143e-07, | |
| "logits/chosen": -3.0227646827697754, | |
| "logits/rejected": -3.0109972953796387, | |
| "logps/chosen": -266.5917663574219, | |
| "logps/rejected": -237.83108520507812, | |
| "loss": 0.7486, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.03099486604332924, | |
| "rewards/margins": 0.23689258098602295, | |
| "rewards/rejected": -0.20589768886566162, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 4.0642939150401836e-07, | |
| "logits/chosen": -2.9976096153259277, | |
| "logits/rejected": -2.9851810932159424, | |
| "logps/chosen": -283.94677734375, | |
| "logps/rejected": -237.71841430664062, | |
| "loss": 0.7486, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.020093750208616257, | |
| "rewards/margins": 0.27248382568359375, | |
| "rewards/rejected": -0.2925775945186615, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.0451588212782237e-07, | |
| "logits/chosen": -3.031033754348755, | |
| "logits/rejected": -3.024602174758911, | |
| "logps/chosen": -261.4927978515625, | |
| "logps/rejected": -222.33651733398438, | |
| "loss": 0.7052, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": 0.016252126544713974, | |
| "rewards/margins": 0.25844720005989075, | |
| "rewards/rejected": -0.24219508469104767, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.0260237275162643e-07, | |
| "logits/chosen": -3.017972946166992, | |
| "logits/rejected": -2.993112564086914, | |
| "logps/chosen": -277.76177978515625, | |
| "logps/rejected": -241.6676483154297, | |
| "loss": 0.765, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": 0.010549797676503658, | |
| "rewards/margins": 0.2188224494457245, | |
| "rewards/rejected": -0.20827265083789825, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 4.006888633754305e-07, | |
| "logits/chosen": -3.0501556396484375, | |
| "logits/rejected": -3.039248466491699, | |
| "logps/chosen": -261.3684387207031, | |
| "logps/rejected": -231.1971893310547, | |
| "loss": 0.6973, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": 0.002077583223581314, | |
| "rewards/margins": 0.3389395475387573, | |
| "rewards/rejected": -0.3368619680404663, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.9877535399923456e-07, | |
| "logits/chosen": -3.063708782196045, | |
| "logits/rejected": -3.0399699211120605, | |
| "logps/chosen": -294.5484924316406, | |
| "logps/rejected": -228.28854370117188, | |
| "loss": 0.7335, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.037904877215623856, | |
| "rewards/margins": 0.2777422368526459, | |
| "rewards/rejected": -0.2398373782634735, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.968618446230386e-07, | |
| "logits/chosen": -3.044889211654663, | |
| "logits/rejected": -3.0049989223480225, | |
| "logps/chosen": -280.3213195800781, | |
| "logps/rejected": -223.97787475585938, | |
| "loss": 0.6679, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.06323406845331192, | |
| "rewards/margins": 0.37538376450538635, | |
| "rewards/rejected": -0.31214970350265503, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 3.949483352468427e-07, | |
| "logits/chosen": -3.0053951740264893, | |
| "logits/rejected": -2.98051118850708, | |
| "logps/chosen": -252.00387573242188, | |
| "logps/rejected": -218.14602661132812, | |
| "loss": 0.7566, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.040188662707805634, | |
| "rewards/margins": 0.23952028155326843, | |
| "rewards/rejected": -0.2797089219093323, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 3.9303482587064674e-07, | |
| "logits/chosen": -3.0364222526550293, | |
| "logits/rejected": -3.0096614360809326, | |
| "logps/chosen": -253.96923828125, | |
| "logps/rejected": -229.8144989013672, | |
| "loss": 0.7089, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.007532055489718914, | |
| "rewards/margins": 0.29781442880630493, | |
| "rewards/rejected": -0.29028236865997314, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 3.911213164944508e-07, | |
| "logits/chosen": -3.0369744300842285, | |
| "logits/rejected": -3.0353286266326904, | |
| "logps/chosen": -290.00042724609375, | |
| "logps/rejected": -273.43267822265625, | |
| "loss": 0.7019, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.018222318962216377, | |
| "rewards/margins": 0.31897154450416565, | |
| "rewards/rejected": -0.3007492423057556, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 3.8920780711825487e-07, | |
| "logits/chosen": -2.9672696590423584, | |
| "logits/rejected": -2.9645469188690186, | |
| "logps/chosen": -307.1138610839844, | |
| "logps/rejected": -237.83786010742188, | |
| "loss": 0.6937, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": 0.005494369193911552, | |
| "rewards/margins": 0.31811192631721497, | |
| "rewards/rejected": -0.31261754035949707, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 3.8729429774205893e-07, | |
| "logits/chosen": -3.0339221954345703, | |
| "logits/rejected": -3.0139739513397217, | |
| "logps/chosen": -284.10772705078125, | |
| "logps/rejected": -243.3460235595703, | |
| "loss": 0.6766, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.01381886936724186, | |
| "rewards/margins": 0.38050729036331177, | |
| "rewards/rejected": -0.39432623982429504, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 3.8538078836586294e-07, | |
| "logits/chosen": -3.0409202575683594, | |
| "logits/rejected": -3.017521381378174, | |
| "logps/chosen": -274.10675048828125, | |
| "logps/rejected": -235.25332641601562, | |
| "loss": 0.7214, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.045437611639499664, | |
| "rewards/margins": 0.3723521828651428, | |
| "rewards/rejected": -0.32691454887390137, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 3.83467278989667e-07, | |
| "logits/chosen": -3.0040934085845947, | |
| "logits/rejected": -2.9863858222961426, | |
| "logps/chosen": -264.234130859375, | |
| "logps/rejected": -212.1512908935547, | |
| "loss": 0.6845, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": 0.005959497299045324, | |
| "rewards/margins": 0.3572615683078766, | |
| "rewards/rejected": -0.3513020873069763, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 3.8155376961347106e-07, | |
| "logits/chosen": -3.0645086765289307, | |
| "logits/rejected": -3.007006883621216, | |
| "logps/chosen": -230.1156005859375, | |
| "logps/rejected": -207.70779418945312, | |
| "loss": 0.6515, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.018168695271015167, | |
| "rewards/margins": 0.38268524408340454, | |
| "rewards/rejected": -0.4008539617061615, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.796402602372751e-07, | |
| "logits/chosen": -3.0344386100769043, | |
| "logits/rejected": -3.020028829574585, | |
| "logps/chosen": -273.54412841796875, | |
| "logps/rejected": -222.2449493408203, | |
| "loss": 0.6707, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.026546839624643326, | |
| "rewards/margins": 0.2530173659324646, | |
| "rewards/rejected": -0.27956423163414, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 3.777267508610792e-07, | |
| "logits/chosen": -2.9949254989624023, | |
| "logits/rejected": -3.0048325061798096, | |
| "logps/chosen": -247.9457244873047, | |
| "logps/rejected": -203.7804412841797, | |
| "loss": 0.6833, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.05914358049631119, | |
| "rewards/margins": 0.3101976811885834, | |
| "rewards/rejected": -0.3693412244319916, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 3.7581324148488325e-07, | |
| "logits/chosen": -3.0347888469696045, | |
| "logits/rejected": -2.994046688079834, | |
| "logps/chosen": -263.569091796875, | |
| "logps/rejected": -231.2215118408203, | |
| "loss": 0.6597, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.021738069131970406, | |
| "rewards/margins": 0.3827739953994751, | |
| "rewards/rejected": -0.36103588342666626, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 3.738997321086873e-07, | |
| "logits/chosen": -3.0062592029571533, | |
| "logits/rejected": -2.993868589401245, | |
| "logps/chosen": -276.2633056640625, | |
| "logps/rejected": -232.21493530273438, | |
| "loss": 0.6677, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.03908178210258484, | |
| "rewards/margins": 0.2943773865699768, | |
| "rewards/rejected": -0.33345913887023926, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 3.7198622273249137e-07, | |
| "logits/chosen": -2.9820735454559326, | |
| "logits/rejected": -3.0040230751037598, | |
| "logps/chosen": -275.2674865722656, | |
| "logps/rejected": -240.7872314453125, | |
| "loss": 0.6137, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": 0.020224738866090775, | |
| "rewards/margins": 0.3282891809940338, | |
| "rewards/rejected": -0.30806440114974976, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_logits/chosen": -3.0153579711914062, | |
| "eval_logits/rejected": -2.9988856315612793, | |
| "eval_logps/chosen": -271.4433288574219, | |
| "eval_logps/rejected": -232.6822967529297, | |
| "eval_loss": 0.6276752948760986, | |
| "eval_rewards/accuracies": 0.7039999961853027, | |
| "eval_rewards/chosen": -0.028680188581347466, | |
| "eval_rewards/margins": 0.3904646337032318, | |
| "eval_rewards/rejected": -0.4191448390483856, | |
| "eval_runtime": 449.0184, | |
| "eval_samples_per_second": 4.454, | |
| "eval_steps_per_second": 0.278, | |
| "step": 968 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 3.7007271335629544e-07, | |
| "logits/chosen": -3.0252368450164795, | |
| "logits/rejected": -3.0065531730651855, | |
| "logps/chosen": -275.3318176269531, | |
| "logps/rejected": -244.02072143554688, | |
| "loss": 0.6284, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.04684365913271904, | |
| "rewards/margins": 0.3803178668022156, | |
| "rewards/rejected": -0.42716145515441895, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 3.681592039800995e-07, | |
| "logits/chosen": -3.0262959003448486, | |
| "logits/rejected": -3.0287508964538574, | |
| "logps/chosen": -271.2191162109375, | |
| "logps/rejected": -233.7063446044922, | |
| "loss": 0.6202, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.028351956978440285, | |
| "rewards/margins": 0.39485400915145874, | |
| "rewards/rejected": -0.4232059419155121, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "learning_rate": 3.662456946039035e-07, | |
| "logits/chosen": -3.0115294456481934, | |
| "logits/rejected": -3.0146679878234863, | |
| "logps/chosen": -251.9388885498047, | |
| "logps/rejected": -229.016357421875, | |
| "loss": 0.6954, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.07080521434545517, | |
| "rewards/margins": 0.35120025277137756, | |
| "rewards/rejected": -0.42200547456741333, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 3.6433218522770757e-07, | |
| "logits/chosen": -3.007986068725586, | |
| "logits/rejected": -2.9926083087921143, | |
| "logps/chosen": -262.0426940917969, | |
| "logps/rejected": -256.99188232421875, | |
| "loss": 0.6533, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.070511594414711, | |
| "rewards/margins": 0.3111626207828522, | |
| "rewards/rejected": -0.381674200296402, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "learning_rate": 3.6241867585151163e-07, | |
| "logits/chosen": -2.9853007793426514, | |
| "logits/rejected": -3.0082547664642334, | |
| "logps/chosen": -270.3041076660156, | |
| "logps/rejected": -236.4681854248047, | |
| "loss": 0.6359, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.022838518023490906, | |
| "rewards/margins": 0.3421871066093445, | |
| "rewards/rejected": -0.3650256097316742, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "learning_rate": 3.605051664753157e-07, | |
| "logits/chosen": -2.9447665214538574, | |
| "logits/rejected": -2.933103084564209, | |
| "logps/chosen": -266.8084411621094, | |
| "logps/rejected": -215.50308227539062, | |
| "loss": 0.538, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.04661910608410835, | |
| "rewards/margins": 0.4469257891178131, | |
| "rewards/rejected": -0.49354487657546997, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "learning_rate": 3.5859165709911975e-07, | |
| "logits/chosen": -3.0248591899871826, | |
| "logits/rejected": -2.98162579536438, | |
| "logps/chosen": -277.4466552734375, | |
| "logps/rejected": -250.49893188476562, | |
| "loss": 0.5931, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.004877015482634306, | |
| "rewards/margins": 0.4228130280971527, | |
| "rewards/rejected": -0.4276900887489319, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 3.566781477229238e-07, | |
| "logits/chosen": -3.0339090824127197, | |
| "logits/rejected": -2.9957573413848877, | |
| "logps/chosen": -275.76910400390625, | |
| "logps/rejected": -235.79830932617188, | |
| "loss": 0.6007, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.08596549183130264, | |
| "rewards/margins": 0.29382139444351196, | |
| "rewards/rejected": -0.3797869086265564, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "learning_rate": 3.547646383467279e-07, | |
| "logits/chosen": -2.9797756671905518, | |
| "logits/rejected": -2.9523847103118896, | |
| "logps/chosen": -284.53839111328125, | |
| "logps/rejected": -214.0476531982422, | |
| "loss": 0.5431, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.07950444519519806, | |
| "rewards/margins": 0.4745730459690094, | |
| "rewards/rejected": -0.5540775060653687, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "learning_rate": 3.5285112897053194e-07, | |
| "logits/chosen": -2.946882963180542, | |
| "logits/rejected": -2.9203898906707764, | |
| "logps/chosen": -257.57611083984375, | |
| "logps/rejected": -244.93075561523438, | |
| "loss": 0.5401, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.020254051312804222, | |
| "rewards/margins": 0.4997115135192871, | |
| "rewards/rejected": -0.5199655294418335, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 3.50937619594336e-07, | |
| "logits/chosen": -2.991321563720703, | |
| "logits/rejected": -3.0015194416046143, | |
| "logps/chosen": -279.9203186035156, | |
| "logps/rejected": -243.14889526367188, | |
| "loss": 0.5657, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.08546855300664902, | |
| "rewards/margins": 0.4243212342262268, | |
| "rewards/rejected": -0.5097898244857788, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 3.4902411021814007e-07, | |
| "logits/chosen": -3.034787893295288, | |
| "logits/rejected": -2.9858031272888184, | |
| "logps/chosen": -259.6640319824219, | |
| "logps/rejected": -238.2606201171875, | |
| "loss": 0.5362, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.11487703025341034, | |
| "rewards/margins": 0.41897639632225037, | |
| "rewards/rejected": -0.5338534116744995, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "learning_rate": 3.4711060084194413e-07, | |
| "logits/chosen": -2.951054096221924, | |
| "logits/rejected": -2.953270673751831, | |
| "logps/chosen": -288.8511962890625, | |
| "logps/rejected": -241.43490600585938, | |
| "loss": 0.5902, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.041309602558612823, | |
| "rewards/margins": 0.47613269090652466, | |
| "rewards/rejected": -0.5174422860145569, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "learning_rate": 3.4519709146574814e-07, | |
| "logits/chosen": -3.0297837257385254, | |
| "logits/rejected": -2.9690465927124023, | |
| "logps/chosen": -249.8524932861328, | |
| "logps/rejected": -211.01486206054688, | |
| "loss": 0.579, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.15785838663578033, | |
| "rewards/margins": 0.3855217397212982, | |
| "rewards/rejected": -0.5433801412582397, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "learning_rate": 3.432835820895522e-07, | |
| "logits/chosen": -3.023965358734131, | |
| "logits/rejected": -3.012373685836792, | |
| "logps/chosen": -296.8121643066406, | |
| "logps/rejected": -269.64410400390625, | |
| "loss": 0.576, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.013493712060153484, | |
| "rewards/margins": 0.45747965574264526, | |
| "rewards/rejected": -0.47097334265708923, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 3.4137007271335626e-07, | |
| "logits/chosen": -2.9927139282226562, | |
| "logits/rejected": -2.995068073272705, | |
| "logps/chosen": -238.1875, | |
| "logps/rejected": -238.9112548828125, | |
| "loss": 0.5405, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.07948704063892365, | |
| "rewards/margins": 0.4464842677116394, | |
| "rewards/rejected": -0.5259712934494019, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "learning_rate": 3.394565633371603e-07, | |
| "logits/chosen": -3.00760555267334, | |
| "logits/rejected": -2.992061138153076, | |
| "logps/chosen": -304.2665100097656, | |
| "logps/rejected": -251.5317840576172, | |
| "loss": 0.5381, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.1150180920958519, | |
| "rewards/margins": 0.46640753746032715, | |
| "rewards/rejected": -0.581425666809082, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "learning_rate": 3.375430539609644e-07, | |
| "logits/chosen": -3.0006496906280518, | |
| "logits/rejected": -3.0168707370758057, | |
| "logps/chosen": -260.1852111816406, | |
| "logps/rejected": -252.5244598388672, | |
| "loss": 0.4865, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.16791771352291107, | |
| "rewards/margins": 0.4443301260471344, | |
| "rewards/rejected": -0.6122478246688843, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 3.3562954458476845e-07, | |
| "logits/chosen": -3.023045063018799, | |
| "logits/rejected": -2.9913861751556396, | |
| "logps/chosen": -295.21502685546875, | |
| "logps/rejected": -226.15426635742188, | |
| "loss": 0.5825, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.09654710441827774, | |
| "rewards/margins": 0.504024088382721, | |
| "rewards/rejected": -0.6005711555480957, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 3.337160352085725e-07, | |
| "logits/chosen": -2.9731438159942627, | |
| "logits/rejected": -2.9854893684387207, | |
| "logps/chosen": -260.6565856933594, | |
| "logps/rejected": -235.22476196289062, | |
| "loss": 0.5259, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.0705583468079567, | |
| "rewards/margins": 0.5393149852752686, | |
| "rewards/rejected": -0.6098732948303223, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "learning_rate": 3.3180252583237657e-07, | |
| "logits/chosen": -3.015777111053467, | |
| "logits/rejected": -2.9891409873962402, | |
| "logps/chosen": -270.416748046875, | |
| "logps/rejected": -232.32089233398438, | |
| "loss": 0.5407, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.13838811218738556, | |
| "rewards/margins": 0.4489242136478424, | |
| "rewards/rejected": -0.5873123407363892, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "learning_rate": 3.2988901645618063e-07, | |
| "logits/chosen": -2.980250597000122, | |
| "logits/rejected": -2.9568262100219727, | |
| "logps/chosen": -257.48504638671875, | |
| "logps/rejected": -231.20101928710938, | |
| "loss": 0.5238, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.26241081953048706, | |
| "rewards/margins": 0.4477602541446686, | |
| "rewards/rejected": -0.7101710438728333, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "learning_rate": 3.279755070799847e-07, | |
| "logits/chosen": -2.987285614013672, | |
| "logits/rejected": -2.9756875038146973, | |
| "logps/chosen": -250.1853790283203, | |
| "logps/rejected": -234.43313598632812, | |
| "loss": 0.4216, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.1883806735277176, | |
| "rewards/margins": 0.47481974959373474, | |
| "rewards/rejected": -0.6632004380226135, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 3.260619977037887e-07, | |
| "logits/chosen": -2.9940743446350098, | |
| "logits/rejected": -2.9866111278533936, | |
| "logps/chosen": -261.4402770996094, | |
| "logps/rejected": -236.15670776367188, | |
| "loss": 0.4046, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.11559490859508514, | |
| "rewards/margins": 0.5072110891342163, | |
| "rewards/rejected": -0.6228059530258179, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "learning_rate": 3.2414848832759277e-07, | |
| "logits/chosen": -3.0217831134796143, | |
| "logits/rejected": -2.9786057472229004, | |
| "logps/chosen": -263.88592529296875, | |
| "logps/rejected": -227.23828125, | |
| "loss": 0.4809, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.1470792442560196, | |
| "rewards/margins": 0.5500885844230652, | |
| "rewards/rejected": -0.6971677541732788, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "learning_rate": 3.2223497895139683e-07, | |
| "logits/chosen": -2.983037233352661, | |
| "logits/rejected": -2.9867138862609863, | |
| "logps/chosen": -270.10089111328125, | |
| "logps/rejected": -262.2214660644531, | |
| "loss": 0.5297, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.19438357651233673, | |
| "rewards/margins": 0.3766476809978485, | |
| "rewards/rejected": -0.5710312128067017, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "learning_rate": 3.203214695752009e-07, | |
| "logits/chosen": -2.923466920852661, | |
| "logits/rejected": -2.9358396530151367, | |
| "logps/chosen": -291.2686767578125, | |
| "logps/rejected": -251.558837890625, | |
| "loss": 0.5026, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.16642309725284576, | |
| "rewards/margins": 0.3848158121109009, | |
| "rewards/rejected": -0.5512388944625854, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 3.1840796019900495e-07, | |
| "logits/chosen": -3.007678508758545, | |
| "logits/rejected": -2.988664388656616, | |
| "logps/chosen": -291.9914245605469, | |
| "logps/rejected": -263.6539611816406, | |
| "loss": 0.4669, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.1000869870185852, | |
| "rewards/margins": 0.597449004650116, | |
| "rewards/rejected": -0.6975361108779907, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 3.16494450822809e-07, | |
| "logits/chosen": -3.0168919563293457, | |
| "logits/rejected": -2.987907886505127, | |
| "logps/chosen": -249.50704956054688, | |
| "logps/rejected": -224.2857208251953, | |
| "loss": 0.4447, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.16220508515834808, | |
| "rewards/margins": 0.5248938798904419, | |
| "rewards/rejected": -0.6870989203453064, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "learning_rate": 3.145809414466131e-07, | |
| "logits/chosen": -3.0371620655059814, | |
| "logits/rejected": -3.0139615535736084, | |
| "logps/chosen": -273.0102233886719, | |
| "logps/rejected": -234.0775909423828, | |
| "loss": 0.439, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.13673745095729828, | |
| "rewards/margins": 0.6153367161750793, | |
| "rewards/rejected": -0.752074122428894, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "learning_rate": 3.1266743207041714e-07, | |
| "logits/chosen": -3.0206820964813232, | |
| "logits/rejected": -3.0111212730407715, | |
| "logps/chosen": -309.1382751464844, | |
| "logps/rejected": -276.2449645996094, | |
| "loss": 0.4547, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.1778876632452011, | |
| "rewards/margins": 0.6206526756286621, | |
| "rewards/rejected": -0.7985404133796692, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 3.107539226942212e-07, | |
| "logits/chosen": -3.007878303527832, | |
| "logits/rejected": -3.0043411254882812, | |
| "logps/chosen": -238.5860137939453, | |
| "logps/rejected": -227.81234741210938, | |
| "loss": 0.5382, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.1700361967086792, | |
| "rewards/margins": 0.43853870034217834, | |
| "rewards/rejected": -0.6085748672485352, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 3.0884041331802526e-07, | |
| "logits/chosen": -2.970362901687622, | |
| "logits/rejected": -2.969972848892212, | |
| "logps/chosen": -257.76812744140625, | |
| "logps/rejected": -230.56820678710938, | |
| "loss": 0.4622, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.14981916546821594, | |
| "rewards/margins": 0.5685046315193176, | |
| "rewards/rejected": -0.718323826789856, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "learning_rate": 3.0692690394182927e-07, | |
| "logits/chosen": -3.0062780380249023, | |
| "logits/rejected": -2.9634342193603516, | |
| "logps/chosen": -278.4779968261719, | |
| "logps/rejected": -237.66873168945312, | |
| "loss": 0.3758, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.20812156796455383, | |
| "rewards/margins": 0.5646450519561768, | |
| "rewards/rejected": -0.7727665305137634, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "learning_rate": 3.0501339456563334e-07, | |
| "logits/chosen": -2.974381923675537, | |
| "logits/rejected": -2.9355669021606445, | |
| "logps/chosen": -286.21453857421875, | |
| "logps/rejected": -236.1365203857422, | |
| "loss": 0.378, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.08342897891998291, | |
| "rewards/margins": 0.6408315896987915, | |
| "rewards/rejected": -0.7242605090141296, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 3.030998851894374e-07, | |
| "logits/chosen": -3.02203369140625, | |
| "logits/rejected": -3.0106561183929443, | |
| "logps/chosen": -244.7522430419922, | |
| "logps/rejected": -214.9333038330078, | |
| "loss": 0.3807, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.24016663432121277, | |
| "rewards/margins": 0.5892859697341919, | |
| "rewards/rejected": -0.829452633857727, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 3.0118637581324146e-07, | |
| "logits/chosen": -2.9896299839019775, | |
| "logits/rejected": -2.9215025901794434, | |
| "logps/chosen": -263.9308166503906, | |
| "logps/rejected": -236.4021453857422, | |
| "loss": 0.4104, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.14813081920146942, | |
| "rewards/margins": 0.6776358485221863, | |
| "rewards/rejected": -0.8257666826248169, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "learning_rate": 2.992728664370455e-07, | |
| "logits/chosen": -3.0315442085266113, | |
| "logits/rejected": -3.035961866378784, | |
| "logps/chosen": -261.2752990722656, | |
| "logps/rejected": -235.47705078125, | |
| "loss": 0.3998, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.2596365213394165, | |
| "rewards/margins": 0.658948540687561, | |
| "rewards/rejected": -0.9185851812362671, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "learning_rate": 2.973593570608496e-07, | |
| "logits/chosen": -3.0204250812530518, | |
| "logits/rejected": -3.015413761138916, | |
| "logps/chosen": -246.57290649414062, | |
| "logps/rejected": -231.91238403320312, | |
| "loss": 0.4193, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.3196146488189697, | |
| "rewards/margins": 0.4760715365409851, | |
| "rewards/rejected": -0.7956861257553101, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 2.9544584768465365e-07, | |
| "logits/chosen": -3.016695022583008, | |
| "logits/rejected": -2.9959843158721924, | |
| "logps/chosen": -275.1457824707031, | |
| "logps/rejected": -220.0662078857422, | |
| "loss": 0.4373, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.33876991271972656, | |
| "rewards/margins": 0.457784503698349, | |
| "rewards/rejected": -0.796554446220398, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "learning_rate": 2.935323383084577e-07, | |
| "logits/chosen": -3.059567451477051, | |
| "logits/rejected": -2.9867048263549805, | |
| "logps/chosen": -287.88153076171875, | |
| "logps/rejected": -235.1467742919922, | |
| "loss": 0.3824, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.264303594827652, | |
| "rewards/margins": 0.6922810673713684, | |
| "rewards/rejected": -0.956584632396698, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "learning_rate": 2.9161882893226177e-07, | |
| "logits/chosen": -2.975642442703247, | |
| "logits/rejected": -2.9326682090759277, | |
| "logps/chosen": -251.0466766357422, | |
| "logps/rejected": -223.5084686279297, | |
| "loss": 0.4333, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.1487085521221161, | |
| "rewards/margins": 0.689045250415802, | |
| "rewards/rejected": -0.8377537727355957, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "learning_rate": 2.8970531955606583e-07, | |
| "logits/chosen": -3.0104973316192627, | |
| "logits/rejected": -2.9991507530212402, | |
| "logps/chosen": -296.5973205566406, | |
| "logps/rejected": -226.55111694335938, | |
| "loss": 0.3786, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.20407001674175262, | |
| "rewards/margins": 0.6605352163314819, | |
| "rewards/rejected": -0.8646053075790405, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 2.8779181017986984e-07, | |
| "logits/chosen": -2.9817981719970703, | |
| "logits/rejected": -2.988542079925537, | |
| "logps/chosen": -285.8086853027344, | |
| "logps/rejected": -253.98934936523438, | |
| "loss": 0.4188, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.25984063744544983, | |
| "rewards/margins": 0.5409084558486938, | |
| "rewards/rejected": -0.8007491230964661, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 2.858783008036739e-07, | |
| "logits/chosen": -2.99088978767395, | |
| "logits/rejected": -2.9880213737487793, | |
| "logps/chosen": -269.12982177734375, | |
| "logps/rejected": -245.9726104736328, | |
| "loss": 0.3797, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.3341473937034607, | |
| "rewards/margins": 0.6310809254646301, | |
| "rewards/rejected": -0.9652281999588013, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "learning_rate": 2.8396479142747797e-07, | |
| "logits/chosen": -3.031111717224121, | |
| "logits/rejected": -3.011475086212158, | |
| "logps/chosen": -265.3778381347656, | |
| "logps/rejected": -236.4595947265625, | |
| "loss": 0.3353, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.23791718482971191, | |
| "rewards/margins": 0.7006685733795166, | |
| "rewards/rejected": -0.938585638999939, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "learning_rate": 2.8205128205128203e-07, | |
| "logits/chosen": -2.972764492034912, | |
| "logits/rejected": -2.9663169384002686, | |
| "logps/chosen": -261.602294921875, | |
| "logps/rejected": -261.4017639160156, | |
| "loss": 0.3131, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.2686450779438019, | |
| "rewards/margins": 0.671106219291687, | |
| "rewards/rejected": -0.9397512674331665, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 2.801377726750861e-07, | |
| "logits/chosen": -3.018611192703247, | |
| "logits/rejected": -2.9920201301574707, | |
| "logps/chosen": -253.02285766601562, | |
| "logps/rejected": -222.1746826171875, | |
| "loss": 0.3765, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.35335665941238403, | |
| "rewards/margins": 0.5318818092346191, | |
| "rewards/rejected": -0.8852384686470032, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 2.7822426329889015e-07, | |
| "logits/chosen": -2.9993600845336914, | |
| "logits/rejected": -2.953437328338623, | |
| "logps/chosen": -272.62347412109375, | |
| "logps/rejected": -223.8088836669922, | |
| "loss": 0.3979, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.3333335518836975, | |
| "rewards/margins": 0.6397331953048706, | |
| "rewards/rejected": -0.9730666875839233, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "learning_rate": 2.763107539226942e-07, | |
| "logits/chosen": -3.0080342292785645, | |
| "logits/rejected": -2.9877243041992188, | |
| "logps/chosen": -285.1849670410156, | |
| "logps/rejected": -218.37954711914062, | |
| "loss": 0.3865, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.35278716683387756, | |
| "rewards/margins": 0.6462265253067017, | |
| "rewards/rejected": -0.9990137815475464, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "learning_rate": 2.743972445464983e-07, | |
| "logits/chosen": -2.989492177963257, | |
| "logits/rejected": -2.9886727333068848, | |
| "logps/chosen": -276.1953430175781, | |
| "logps/rejected": -228.0172882080078, | |
| "loss": 0.4094, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.3594445586204529, | |
| "rewards/margins": 0.6459834575653076, | |
| "rewards/rejected": -1.0054280757904053, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 2.7248373517030234e-07, | |
| "logits/chosen": -2.9692909717559814, | |
| "logits/rejected": -2.978031635284424, | |
| "logps/chosen": -260.3178405761719, | |
| "logps/rejected": -236.1215362548828, | |
| "loss": 0.3159, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.33140963315963745, | |
| "rewards/margins": 0.5777403116226196, | |
| "rewards/rejected": -0.9091499447822571, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "learning_rate": 2.705702257941064e-07, | |
| "logits/chosen": -2.9462051391601562, | |
| "logits/rejected": -2.9286155700683594, | |
| "logps/chosen": -289.8208923339844, | |
| "logps/rejected": -247.3198699951172, | |
| "loss": 0.3091, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.3897753953933716, | |
| "rewards/margins": 0.5884745717048645, | |
| "rewards/rejected": -0.978249728679657, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 2.686567164179104e-07, | |
| "logits/chosen": -3.018886089324951, | |
| "logits/rejected": -2.9545936584472656, | |
| "logps/chosen": -265.15240478515625, | |
| "logps/rejected": -227.3736114501953, | |
| "loss": 0.3156, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.2435026615858078, | |
| "rewards/margins": 0.8812161684036255, | |
| "rewards/rejected": -1.1247189044952393, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "learning_rate": 2.6674320704171447e-07, | |
| "logits/chosen": -3.019289016723633, | |
| "logits/rejected": -2.9821600914001465, | |
| "logps/chosen": -288.81646728515625, | |
| "logps/rejected": -226.24755859375, | |
| "loss": 0.262, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.3103589713573456, | |
| "rewards/margins": 0.6744868159294128, | |
| "rewards/rejected": -0.9848458170890808, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 2.6482969766551853e-07, | |
| "logits/chosen": -2.918996572494507, | |
| "logits/rejected": -2.9260575771331787, | |
| "logps/chosen": -229.50146484375, | |
| "logps/rejected": -233.02432250976562, | |
| "loss": 0.3571, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.3969925045967102, | |
| "rewards/margins": 0.5756487250328064, | |
| "rewards/rejected": -0.9726413488388062, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "learning_rate": 2.629161882893226e-07, | |
| "logits/chosen": -3.0124337673187256, | |
| "logits/rejected": -2.9816842079162598, | |
| "logps/chosen": -272.998779296875, | |
| "logps/rejected": -232.07638549804688, | |
| "loss": 0.2496, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.3978312909603119, | |
| "rewards/margins": 0.6203471422195435, | |
| "rewards/rejected": -1.0181784629821777, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "learning_rate": 2.6100267891312666e-07, | |
| "logits/chosen": -3.0306756496429443, | |
| "logits/rejected": -3.016284465789795, | |
| "logps/chosen": -293.61407470703125, | |
| "logps/rejected": -241.7076873779297, | |
| "loss": 0.2565, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.3279929757118225, | |
| "rewards/margins": 0.6071383953094482, | |
| "rewards/rejected": -0.9351313710212708, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 2.590891695369307e-07, | |
| "logits/chosen": -2.9170010089874268, | |
| "logits/rejected": -2.904465436935425, | |
| "logps/chosen": -251.91354370117188, | |
| "logps/rejected": -234.70751953125, | |
| "loss": 0.2908, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.33537206053733826, | |
| "rewards/margins": 0.6658948063850403, | |
| "rewards/rejected": -1.0012669563293457, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 2.571756601607348e-07, | |
| "logits/chosen": -3.0143392086029053, | |
| "logits/rejected": -2.9897260665893555, | |
| "logps/chosen": -301.2428894042969, | |
| "logps/rejected": -244.96145629882812, | |
| "loss": 0.2721, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.3570322096347809, | |
| "rewards/margins": 0.7769542336463928, | |
| "rewards/rejected": -1.133986473083496, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "learning_rate": 2.5526215078453884e-07, | |
| "logits/chosen": -3.022491216659546, | |
| "logits/rejected": -2.9913883209228516, | |
| "logps/chosen": -284.7639465332031, | |
| "logps/rejected": -248.6448211669922, | |
| "loss": 0.2595, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.3536800742149353, | |
| "rewards/margins": 0.7084030508995056, | |
| "rewards/rejected": -1.0620832443237305, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "learning_rate": 2.533486414083429e-07, | |
| "logits/chosen": -3.0248942375183105, | |
| "logits/rejected": -2.9973325729370117, | |
| "logps/chosen": -284.40472412109375, | |
| "logps/rejected": -241.6700897216797, | |
| "loss": 0.1589, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.3532782196998596, | |
| "rewards/margins": 0.7998963594436646, | |
| "rewards/rejected": -1.153174638748169, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 2.5143513203214697e-07, | |
| "logits/chosen": -2.9913268089294434, | |
| "logits/rejected": -2.971450090408325, | |
| "logps/chosen": -319.8451232910156, | |
| "logps/rejected": -254.3859100341797, | |
| "loss": 0.3029, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.31969529390335083, | |
| "rewards/margins": 0.671107292175293, | |
| "rewards/rejected": -0.9908025860786438, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 2.49521622655951e-07, | |
| "logits/chosen": -2.959479331970215, | |
| "logits/rejected": -2.908963441848755, | |
| "logps/chosen": -272.8207092285156, | |
| "logps/rejected": -235.93020629882812, | |
| "loss": 0.2796, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.34670349955558777, | |
| "rewards/margins": 0.7621387243270874, | |
| "rewards/rejected": -1.108842134475708, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "learning_rate": 2.4760811327975504e-07, | |
| "logits/chosen": -2.9981067180633545, | |
| "logits/rejected": -2.9972851276397705, | |
| "logps/chosen": -278.3575744628906, | |
| "logps/rejected": -239.57223510742188, | |
| "loss": 0.2067, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.5294784307479858, | |
| "rewards/margins": 0.6004038453102112, | |
| "rewards/rejected": -1.1298822164535522, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "learning_rate": 2.456946039035591e-07, | |
| "logits/chosen": -3.030169725418091, | |
| "logits/rejected": -2.9914770126342773, | |
| "logps/chosen": -289.63116455078125, | |
| "logps/rejected": -249.74649047851562, | |
| "loss": 0.3592, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.44787079095840454, | |
| "rewards/margins": 0.63862544298172, | |
| "rewards/rejected": -1.086496114730835, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "learning_rate": 2.4378109452736316e-07, | |
| "logits/chosen": -2.9960074424743652, | |
| "logits/rejected": -2.980121374130249, | |
| "logps/chosen": -267.38446044921875, | |
| "logps/rejected": -244.11593627929688, | |
| "loss": 0.2287, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.34468650817871094, | |
| "rewards/margins": 0.8173492550849915, | |
| "rewards/rejected": -1.1620357036590576, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 2.418675851511672e-07, | |
| "logits/chosen": -2.913470506668091, | |
| "logits/rejected": -2.90181040763855, | |
| "logps/chosen": -268.4359436035156, | |
| "logps/rejected": -238.85806274414062, | |
| "loss": 0.2227, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.43024927377700806, | |
| "rewards/margins": 0.583633303642273, | |
| "rewards/rejected": -1.0138825178146362, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "learning_rate": 2.399540757749713e-07, | |
| "logits/chosen": -2.9995226860046387, | |
| "logits/rejected": -2.971139430999756, | |
| "logps/chosen": -276.39031982421875, | |
| "logps/rejected": -242.7020263671875, | |
| "loss": 0.2584, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.5093586444854736, | |
| "rewards/margins": 0.657143771648407, | |
| "rewards/rejected": -1.1665024757385254, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 2.3804056639877535e-07, | |
| "logits/chosen": -2.95731258392334, | |
| "logits/rejected": -2.9587581157684326, | |
| "logps/chosen": -248.0506134033203, | |
| "logps/rejected": -226.7265625, | |
| "loss": 0.2349, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.3466361165046692, | |
| "rewards/margins": 0.8473577499389648, | |
| "rewards/rejected": -1.1939939260482788, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "learning_rate": 2.361270570225794e-07, | |
| "logits/chosen": -3.0136914253234863, | |
| "logits/rejected": -2.9863269329071045, | |
| "logps/chosen": -285.3213195800781, | |
| "logps/rejected": -233.94619750976562, | |
| "loss": 0.1567, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.3250366151332855, | |
| "rewards/margins": 0.8868409991264343, | |
| "rewards/rejected": -1.2118775844573975, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "learning_rate": 2.3421354764638345e-07, | |
| "logits/chosen": -2.9790773391723633, | |
| "logits/rejected": -2.952543020248413, | |
| "logps/chosen": -276.2936706542969, | |
| "logps/rejected": -257.79718017578125, | |
| "loss": 0.1955, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.4911643862724304, | |
| "rewards/margins": 0.6523909568786621, | |
| "rewards/rejected": -1.1435552835464478, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 2.323000382701875e-07, | |
| "logits/chosen": -2.9914324283599854, | |
| "logits/rejected": -2.98734974861145, | |
| "logps/chosen": -322.0937805175781, | |
| "logps/rejected": -261.5361633300781, | |
| "loss": 0.1863, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.3622845411300659, | |
| "rewards/margins": 0.9477508664131165, | |
| "rewards/rejected": -1.3100353479385376, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "learning_rate": 2.3038652889399157e-07, | |
| "logits/chosen": -2.9850406646728516, | |
| "logits/rejected": -2.972008228302002, | |
| "logps/chosen": -262.46197509765625, | |
| "logps/rejected": -237.19375610351562, | |
| "loss": 0.1809, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.53041011095047, | |
| "rewards/margins": 0.7410814762115479, | |
| "rewards/rejected": -1.2714916467666626, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "learning_rate": 2.2847301951779563e-07, | |
| "logits/chosen": -2.983722448348999, | |
| "logits/rejected": -2.956228256225586, | |
| "logps/chosen": -272.92156982421875, | |
| "logps/rejected": -256.91107177734375, | |
| "loss": 0.1089, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.5018629431724548, | |
| "rewards/margins": 0.8895123600959778, | |
| "rewards/rejected": -1.391375184059143, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 2.265595101415997e-07, | |
| "logits/chosen": -2.9394657611846924, | |
| "logits/rejected": -2.92991304397583, | |
| "logps/chosen": -273.16815185546875, | |
| "logps/rejected": -262.08917236328125, | |
| "loss": 0.1644, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.5341050624847412, | |
| "rewards/margins": 0.7611141800880432, | |
| "rewards/rejected": -1.2952191829681396, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "learning_rate": 2.2464600076540373e-07, | |
| "logits/chosen": -2.946911334991455, | |
| "logits/rejected": -2.931849718093872, | |
| "logps/chosen": -308.5828552246094, | |
| "logps/rejected": -247.25338745117188, | |
| "loss": 0.1064, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.5544737577438354, | |
| "rewards/margins": 0.9120496511459351, | |
| "rewards/rejected": -1.46652352809906, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "learning_rate": 2.227324913892078e-07, | |
| "logits/chosen": -2.951788902282715, | |
| "logits/rejected": -2.909658908843994, | |
| "logps/chosen": -295.8861389160156, | |
| "logps/rejected": -243.9182891845703, | |
| "loss": 0.0733, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.5971349477767944, | |
| "rewards/margins": 0.7489143013954163, | |
| "rewards/rejected": -1.3460490703582764, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "learning_rate": 2.2081898201301186e-07, | |
| "logits/chosen": -3.023937702178955, | |
| "logits/rejected": -2.9923651218414307, | |
| "logps/chosen": -301.7521057128906, | |
| "logps/rejected": -257.3502197265625, | |
| "loss": 0.1576, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.5676156282424927, | |
| "rewards/margins": 0.8253081440925598, | |
| "rewards/rejected": -1.3929237127304077, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 2.1890547263681592e-07, | |
| "logits/chosen": -2.979418992996216, | |
| "logits/rejected": -2.979773998260498, | |
| "logps/chosen": -278.7250061035156, | |
| "logps/rejected": -281.96417236328125, | |
| "loss": 0.1553, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.5305044651031494, | |
| "rewards/margins": 0.8628344535827637, | |
| "rewards/rejected": -1.393338918685913, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 2.1699196326061998e-07, | |
| "logits/chosen": -2.93343448638916, | |
| "logits/rejected": -2.9477057456970215, | |
| "logps/chosen": -259.96685791015625, | |
| "logps/rejected": -260.68145751953125, | |
| "loss": 0.2389, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.5586804151535034, | |
| "rewards/margins": 0.8505626916885376, | |
| "rewards/rejected": -1.409243106842041, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "learning_rate": 2.1507845388442402e-07, | |
| "logits/chosen": -3.001537799835205, | |
| "logits/rejected": -3.000216007232666, | |
| "logps/chosen": -253.8172607421875, | |
| "logps/rejected": -229.71261596679688, | |
| "loss": 0.117, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.6867496371269226, | |
| "rewards/margins": 0.839769184589386, | |
| "rewards/rejected": -1.5265188217163086, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "learning_rate": 2.1316494450822808e-07, | |
| "logits/chosen": -2.9645118713378906, | |
| "logits/rejected": -2.921504497528076, | |
| "logps/chosen": -249.779296875, | |
| "logps/rejected": -217.2705078125, | |
| "loss": 0.187, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.6474353075027466, | |
| "rewards/margins": 0.7807954549789429, | |
| "rewards/rejected": -1.428230881690979, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 2.1125143513203214e-07, | |
| "logits/chosen": -2.995269298553467, | |
| "logits/rejected": -2.9834041595458984, | |
| "logps/chosen": -293.59716796875, | |
| "logps/rejected": -229.48092651367188, | |
| "loss": 0.0751, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.6189619302749634, | |
| "rewards/margins": 0.7833685874938965, | |
| "rewards/rejected": -1.4023306369781494, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 2.093379257558362e-07, | |
| "logits/chosen": -2.9907360076904297, | |
| "logits/rejected": -2.963219404220581, | |
| "logps/chosen": -274.832275390625, | |
| "logps/rejected": -237.86355590820312, | |
| "loss": 0.2011, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.7243114709854126, | |
| "rewards/margins": 0.5377382636070251, | |
| "rewards/rejected": -1.2620497941970825, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "learning_rate": 2.0742441637964026e-07, | |
| "logits/chosen": -2.9230685234069824, | |
| "logits/rejected": -2.88169264793396, | |
| "logps/chosen": -309.228515625, | |
| "logps/rejected": -246.1265411376953, | |
| "loss": -0.039, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.5196012258529663, | |
| "rewards/margins": 1.0259116888046265, | |
| "rewards/rejected": -1.5455129146575928, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "learning_rate": 2.055109070034443e-07, | |
| "logits/chosen": -2.9902000427246094, | |
| "logits/rejected": -2.9472696781158447, | |
| "logps/chosen": -300.9866027832031, | |
| "logps/rejected": -249.76412963867188, | |
| "loss": 0.1368, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.6899394989013672, | |
| "rewards/margins": 0.7042897343635559, | |
| "rewards/rejected": -1.3942292928695679, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 2.0359739762724836e-07, | |
| "logits/chosen": -3.0022895336151123, | |
| "logits/rejected": -2.973423719406128, | |
| "logps/chosen": -313.137939453125, | |
| "logps/rejected": -254.30709838867188, | |
| "loss": 0.0727, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.5544772744178772, | |
| "rewards/margins": 1.1537253856658936, | |
| "rewards/rejected": -1.7082027196884155, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "learning_rate": 2.0168388825105242e-07, | |
| "logits/chosen": -2.976815700531006, | |
| "logits/rejected": -2.958930492401123, | |
| "logps/chosen": -285.1069641113281, | |
| "logps/rejected": -275.8247985839844, | |
| "loss": 0.046, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.5782260894775391, | |
| "rewards/margins": 0.8343210220336914, | |
| "rewards/rejected": -1.41254723072052, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 1.997703788748565e-07, | |
| "logits/chosen": -2.997415065765381, | |
| "logits/rejected": -2.95662784576416, | |
| "logps/chosen": -302.427734375, | |
| "logps/rejected": -215.8485565185547, | |
| "loss": 0.0356, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.4519409239292145, | |
| "rewards/margins": 1.185298204421997, | |
| "rewards/rejected": -1.6372392177581787, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 1.9785686949866055e-07, | |
| "logits/chosen": -2.943821668624878, | |
| "logits/rejected": -2.9366466999053955, | |
| "logps/chosen": -267.79620361328125, | |
| "logps/rejected": -246.8135223388672, | |
| "loss": 0.158, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.6127734184265137, | |
| "rewards/margins": 0.7575459480285645, | |
| "rewards/rejected": -1.3703193664550781, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 1.9594336012246458e-07, | |
| "logits/chosen": -2.941462278366089, | |
| "logits/rejected": -2.9433281421661377, | |
| "logps/chosen": -253.16256713867188, | |
| "logps/rejected": -241.6164093017578, | |
| "loss": 0.0191, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.5904272198677063, | |
| "rewards/margins": 0.8854333758354187, | |
| "rewards/rejected": -1.475860595703125, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "learning_rate": 1.9402985074626865e-07, | |
| "logits/chosen": -2.990996837615967, | |
| "logits/rejected": -2.9396445751190186, | |
| "logps/chosen": -286.08404541015625, | |
| "logps/rejected": -246.9237823486328, | |
| "loss": 0.0619, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.7161077260971069, | |
| "rewards/margins": 0.9890392422676086, | |
| "rewards/rejected": -1.7051467895507812, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "learning_rate": 1.921163413700727e-07, | |
| "logits/chosen": -2.9665896892547607, | |
| "logits/rejected": -2.985443353652954, | |
| "logps/chosen": -273.56964111328125, | |
| "logps/rejected": -255.6343536376953, | |
| "loss": 0.1713, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.7685881853103638, | |
| "rewards/margins": 0.8922961354255676, | |
| "rewards/rejected": -1.6608844995498657, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "learning_rate": 1.9020283199387677e-07, | |
| "logits/chosen": -2.993872880935669, | |
| "logits/rejected": -2.9838452339172363, | |
| "logps/chosen": -268.5338134765625, | |
| "logps/rejected": -235.5106658935547, | |
| "loss": 0.1585, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.5740541219711304, | |
| "rewards/margins": 0.8445619344711304, | |
| "rewards/rejected": -1.4186161756515503, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 1.8828932261768083e-07, | |
| "logits/chosen": -2.918458938598633, | |
| "logits/rejected": -2.924969434738159, | |
| "logps/chosen": -275.4915466308594, | |
| "logps/rejected": -230.2731170654297, | |
| "loss": -0.0004, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.717719316482544, | |
| "rewards/margins": 1.1288950443267822, | |
| "rewards/rejected": -1.8466142416000366, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "learning_rate": 1.8637581324148487e-07, | |
| "logits/chosen": -3.0128183364868164, | |
| "logits/rejected": -3.0055766105651855, | |
| "logps/chosen": -273.6071472167969, | |
| "logps/rejected": -248.68997192382812, | |
| "loss": 0.0705, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.7629455327987671, | |
| "rewards/margins": 1.0152510404586792, | |
| "rewards/rejected": -1.7781963348388672, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_logits/chosen": -2.983034372329712, | |
| "eval_logits/rejected": -2.960925817489624, | |
| "eval_logps/chosen": -277.86474609375, | |
| "eval_logps/rejected": -245.16693115234375, | |
| "eval_loss": 0.05700839310884476, | |
| "eval_rewards/accuracies": 0.6959999799728394, | |
| "eval_rewards/chosen": -0.6708189845085144, | |
| "eval_rewards/margins": 0.9967920184135437, | |
| "eval_rewards/rejected": -1.6676111221313477, | |
| "eval_runtime": 448.2745, | |
| "eval_samples_per_second": 4.462, | |
| "eval_steps_per_second": 0.279, | |
| "step": 1937 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "learning_rate": 1.8446230386528893e-07, | |
| "logits/chosen": -2.99873423576355, | |
| "logits/rejected": -2.9760866165161133, | |
| "logps/chosen": -270.5379943847656, | |
| "logps/rejected": -233.87255859375, | |
| "loss": 0.104, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.6196664571762085, | |
| "rewards/margins": 0.8855097889900208, | |
| "rewards/rejected": -1.505176305770874, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 1.82548794489093e-07, | |
| "logits/chosen": -2.972491502761841, | |
| "logits/rejected": -2.9654781818389893, | |
| "logps/chosen": -272.2441711425781, | |
| "logps/rejected": -247.496337890625, | |
| "loss": 0.0695, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.8281705975532532, | |
| "rewards/margins": 0.8387929201126099, | |
| "rewards/rejected": -1.6669635772705078, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 1.8063528511289706e-07, | |
| "logits/chosen": -2.9022774696350098, | |
| "logits/rejected": -2.8986656665802, | |
| "logps/chosen": -272.8119201660156, | |
| "logps/rejected": -286.4425964355469, | |
| "loss": 0.0376, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.6966532468795776, | |
| "rewards/margins": 1.0244591236114502, | |
| "rewards/rejected": -1.7211124897003174, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 2.03, | |
| "learning_rate": 1.7872177573670112e-07, | |
| "logits/chosen": -3.0008738040924072, | |
| "logits/rejected": -2.9414420127868652, | |
| "logps/chosen": -269.3557434082031, | |
| "logps/rejected": -228.98892211914062, | |
| "loss": 0.1134, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.6135867238044739, | |
| "rewards/margins": 1.0139942169189453, | |
| "rewards/rejected": -1.627581000328064, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 2.04, | |
| "learning_rate": 1.7680826636050515e-07, | |
| "logits/chosen": -2.97796893119812, | |
| "logits/rejected": -2.9451377391815186, | |
| "logps/chosen": -290.0522155761719, | |
| "logps/rejected": -247.7688446044922, | |
| "loss": 0.0676, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.7555999755859375, | |
| "rewards/margins": 0.919529139995575, | |
| "rewards/rejected": -1.6751289367675781, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "learning_rate": 1.7489475698430921e-07, | |
| "logits/chosen": -2.9556636810302734, | |
| "logits/rejected": -2.9225146770477295, | |
| "logps/chosen": -281.0018615722656, | |
| "logps/rejected": -257.39471435546875, | |
| "loss": -0.0304, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.7463997006416321, | |
| "rewards/margins": 0.9945909380912781, | |
| "rewards/rejected": -1.7409906387329102, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 1.7298124760811328e-07, | |
| "logits/chosen": -2.9892733097076416, | |
| "logits/rejected": -2.9609320163726807, | |
| "logps/chosen": -265.5694885253906, | |
| "logps/rejected": -227.8389434814453, | |
| "loss": 0.068, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.8352034687995911, | |
| "rewards/margins": 1.0397050380706787, | |
| "rewards/rejected": -1.874908685684204, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "learning_rate": 1.7106773823191734e-07, | |
| "logits/chosen": -2.9671616554260254, | |
| "logits/rejected": -2.937986373901367, | |
| "logps/chosen": -281.4359436035156, | |
| "logps/rejected": -241.79080200195312, | |
| "loss": -0.0155, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.6781337857246399, | |
| "rewards/margins": 1.3058927059173584, | |
| "rewards/rejected": -1.984026312828064, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 2.09, | |
| "learning_rate": 1.691542288557214e-07, | |
| "logits/chosen": -2.9833037853240967, | |
| "logits/rejected": -2.943441152572632, | |
| "logps/chosen": -295.3065490722656, | |
| "logps/rejected": -253.66360473632812, | |
| "loss": 0.0664, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.7511407136917114, | |
| "rewards/margins": 0.9544817209243774, | |
| "rewards/rejected": -1.7056224346160889, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 1.6724071947952544e-07, | |
| "logits/chosen": -2.9694552421569824, | |
| "logits/rejected": -2.9318954944610596, | |
| "logps/chosen": -254.4512939453125, | |
| "logps/rejected": -235.69595336914062, | |
| "loss": 0.0632, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.7740479111671448, | |
| "rewards/margins": 0.9885379672050476, | |
| "rewards/rejected": -1.762585997581482, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 1.653272101033295e-07, | |
| "logits/chosen": -2.9927315711975098, | |
| "logits/rejected": -2.9617929458618164, | |
| "logps/chosen": -286.9222412109375, | |
| "logps/rejected": -262.54559326171875, | |
| "loss": -0.0387, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.7180510759353638, | |
| "rewards/margins": 1.0184663534164429, | |
| "rewards/rejected": -1.7365175485610962, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "learning_rate": 1.6341370072713356e-07, | |
| "logits/chosen": -2.9783358573913574, | |
| "logits/rejected": -2.9684910774230957, | |
| "logps/chosen": -288.4617614746094, | |
| "logps/rejected": -269.187255859375, | |
| "loss": -0.0243, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.75580233335495, | |
| "rewards/margins": 0.9117294549942017, | |
| "rewards/rejected": -1.6675317287445068, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 1.6150019135093762e-07, | |
| "logits/chosen": -3.0292530059814453, | |
| "logits/rejected": -2.9775586128234863, | |
| "logps/chosen": -298.93585205078125, | |
| "logps/rejected": -247.0900115966797, | |
| "loss": -0.0146, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.6352591514587402, | |
| "rewards/margins": 1.129183053970337, | |
| "rewards/rejected": -1.7644420862197876, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 1.5958668197474169e-07, | |
| "logits/chosen": -3.0193440914154053, | |
| "logits/rejected": -2.99593186378479, | |
| "logps/chosen": -263.82464599609375, | |
| "logps/rejected": -277.60418701171875, | |
| "loss": 0.0702, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.6851625442504883, | |
| "rewards/margins": 0.9985010027885437, | |
| "rewards/rejected": -1.6836636066436768, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 1.5767317259854572e-07, | |
| "logits/chosen": -3.0217273235321045, | |
| "logits/rejected": -2.970132350921631, | |
| "logps/chosen": -307.5477600097656, | |
| "logps/rejected": -248.69448852539062, | |
| "loss": -0.024, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.65143883228302, | |
| "rewards/margins": 1.1803841590881348, | |
| "rewards/rejected": -1.8318227529525757, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "learning_rate": 1.5575966322234978e-07, | |
| "logits/chosen": -2.931065559387207, | |
| "logits/rejected": -2.9219722747802734, | |
| "logps/chosen": -276.0883483886719, | |
| "logps/rejected": -247.0487060546875, | |
| "loss": -0.052, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.64451664686203, | |
| "rewards/margins": 1.1405279636383057, | |
| "rewards/rejected": -1.7850444316864014, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 2.17, | |
| "learning_rate": 1.5384615384615385e-07, | |
| "logits/chosen": -3.0177485942840576, | |
| "logits/rejected": -3.0004374980926514, | |
| "logps/chosen": -263.8681640625, | |
| "logps/rejected": -226.38119506835938, | |
| "loss": -0.047, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.8407020568847656, | |
| "rewards/margins": 1.1751278638839722, | |
| "rewards/rejected": -2.0158302783966064, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 2.18, | |
| "learning_rate": 1.519326444699579e-07, | |
| "logits/chosen": -2.9868111610412598, | |
| "logits/rejected": -2.961195468902588, | |
| "logps/chosen": -294.43402099609375, | |
| "logps/rejected": -237.2691650390625, | |
| "loss": -0.1127, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.678774356842041, | |
| "rewards/margins": 1.201111078262329, | |
| "rewards/rejected": -1.8798853158950806, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 1.5001913509376197e-07, | |
| "logits/chosen": -3.0007481575012207, | |
| "logits/rejected": -2.9801931381225586, | |
| "logps/chosen": -290.3331298828125, | |
| "logps/rejected": -263.9862976074219, | |
| "loss": -0.071, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.908436119556427, | |
| "rewards/margins": 0.9944915771484375, | |
| "rewards/rejected": -1.9029273986816406, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 2.2, | |
| "learning_rate": 1.4810562571756603e-07, | |
| "logits/chosen": -3.011329174041748, | |
| "logits/rejected": -2.980834484100342, | |
| "logps/chosen": -259.76031494140625, | |
| "logps/rejected": -239.0861053466797, | |
| "loss": 0.0556, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.774993360042572, | |
| "rewards/margins": 1.0526189804077148, | |
| "rewards/rejected": -1.8276125192642212, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "learning_rate": 1.4619211634137007e-07, | |
| "logits/chosen": -2.9631924629211426, | |
| "logits/rejected": -2.954556703567505, | |
| "logps/chosen": -248.605224609375, | |
| "logps/rejected": -215.9401397705078, | |
| "loss": -0.0457, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.8637608289718628, | |
| "rewards/margins": 1.0648829936981201, | |
| "rewards/rejected": -1.9286441802978516, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 2.22, | |
| "learning_rate": 1.4427860696517413e-07, | |
| "logits/chosen": -2.950859308242798, | |
| "logits/rejected": -2.961940288543701, | |
| "logps/chosen": -328.16314697265625, | |
| "logps/rejected": -277.8406982421875, | |
| "loss": 0.0939, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.7997218370437622, | |
| "rewards/margins": 1.051257848739624, | |
| "rewards/rejected": -1.8509798049926758, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 1.423650975889782e-07, | |
| "logits/chosen": -3.0021846294403076, | |
| "logits/rejected": -2.9544687271118164, | |
| "logps/chosen": -284.53570556640625, | |
| "logps/rejected": -254.57559204101562, | |
| "loss": -0.1516, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.8068240880966187, | |
| "rewards/margins": 1.2634921073913574, | |
| "rewards/rejected": -2.0703163146972656, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "learning_rate": 1.4045158821278225e-07, | |
| "logits/chosen": -2.968425750732422, | |
| "logits/rejected": -2.9926857948303223, | |
| "logps/chosen": -295.5940246582031, | |
| "logps/rejected": -265.8125, | |
| "loss": -0.1379, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.7207796573638916, | |
| "rewards/margins": 1.3109276294708252, | |
| "rewards/rejected": -2.0317070484161377, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 2.25, | |
| "learning_rate": 1.3853807883658632e-07, | |
| "logits/chosen": -3.0091452598571777, | |
| "logits/rejected": -2.9704108238220215, | |
| "logps/chosen": -298.40948486328125, | |
| "logps/rejected": -241.05337524414062, | |
| "loss": 0.1214, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.9363592863082886, | |
| "rewards/margins": 0.984167754650116, | |
| "rewards/rejected": -1.9205271005630493, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 2.26, | |
| "learning_rate": 1.3662456946039035e-07, | |
| "logits/chosen": -2.924776554107666, | |
| "logits/rejected": -2.920172691345215, | |
| "logps/chosen": -259.19580078125, | |
| "logps/rejected": -247.6489715576172, | |
| "loss": 0.0587, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.8932672739028931, | |
| "rewards/margins": 0.9966402053833008, | |
| "rewards/rejected": -1.8899074792861938, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 1.3471106008419441e-07, | |
| "logits/chosen": -2.9807791709899902, | |
| "logits/rejected": -2.9884274005889893, | |
| "logps/chosen": -244.3290252685547, | |
| "logps/rejected": -231.4929656982422, | |
| "loss": 0.0254, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.9376288652420044, | |
| "rewards/margins": 1.005837082862854, | |
| "rewards/rejected": -1.9434658288955688, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 1.3279755070799848e-07, | |
| "logits/chosen": -2.9636523723602295, | |
| "logits/rejected": -2.898196220397949, | |
| "logps/chosen": -302.7827453613281, | |
| "logps/rejected": -254.11367797851562, | |
| "loss": -0.0107, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0612095594406128, | |
| "rewards/margins": 0.9029472470283508, | |
| "rewards/rejected": -1.9641568660736084, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "learning_rate": 1.3088404133180254e-07, | |
| "logits/chosen": -2.962622880935669, | |
| "logits/rejected": -2.9335360527038574, | |
| "logps/chosen": -321.44342041015625, | |
| "logps/rejected": -250.2067108154297, | |
| "loss": -0.2537, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.8546531796455383, | |
| "rewards/margins": 1.267183780670166, | |
| "rewards/rejected": -2.1218371391296387, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "learning_rate": 1.289705319556066e-07, | |
| "logits/chosen": -2.961509943008423, | |
| "logits/rejected": -2.923830509185791, | |
| "logps/chosen": -294.05279541015625, | |
| "logps/rejected": -256.3243103027344, | |
| "loss": -0.0748, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.8142738342285156, | |
| "rewards/margins": 1.3003737926483154, | |
| "rewards/rejected": -2.114647626876831, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 1.2705702257941064e-07, | |
| "logits/chosen": -2.9922921657562256, | |
| "logits/rejected": -2.9666550159454346, | |
| "logps/chosen": -281.14788818359375, | |
| "logps/rejected": -261.50445556640625, | |
| "loss": -0.103, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.889809787273407, | |
| "rewards/margins": 1.2682868242263794, | |
| "rewards/rejected": -2.1580967903137207, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "learning_rate": 1.251435132032147e-07, | |
| "logits/chosen": -3.009636878967285, | |
| "logits/rejected": -2.9912712574005127, | |
| "logps/chosen": -262.1191101074219, | |
| "logps/rejected": -254.8567352294922, | |
| "loss": 0.0064, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.9876982569694519, | |
| "rewards/margins": 0.9436071515083313, | |
| "rewards/rejected": -1.9313055276870728, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 2.33, | |
| "learning_rate": 1.2323000382701873e-07, | |
| "logits/chosen": -3.0256762504577637, | |
| "logits/rejected": -3.006740093231201, | |
| "logps/chosen": -309.2908630371094, | |
| "logps/rejected": -252.59323120117188, | |
| "loss": 0.0352, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.9725021123886108, | |
| "rewards/margins": 1.1422688961029053, | |
| "rewards/rejected": -2.1147711277008057, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "learning_rate": 1.213164944508228e-07, | |
| "logits/chosen": -2.9900639057159424, | |
| "logits/rejected": -2.9898974895477295, | |
| "logps/chosen": -299.51177978515625, | |
| "logps/rejected": -267.8308410644531, | |
| "loss": -0.1218, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.0168074369430542, | |
| "rewards/margins": 1.148054838180542, | |
| "rewards/rejected": -2.1648621559143066, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 1.1940298507462686e-07, | |
| "logits/chosen": -2.986194610595703, | |
| "logits/rejected": -2.9787726402282715, | |
| "logps/chosen": -285.0082702636719, | |
| "logps/rejected": -241.463134765625, | |
| "loss": -0.0902, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.0312423706054688, | |
| "rewards/margins": 1.1112394332885742, | |
| "rewards/rejected": -2.142481803894043, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 1.1748947569843092e-07, | |
| "logits/chosen": -2.949153184890747, | |
| "logits/rejected": -2.939624786376953, | |
| "logps/chosen": -294.7286682128906, | |
| "logps/rejected": -233.32864379882812, | |
| "loss": -0.2292, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.8459509015083313, | |
| "rewards/margins": 1.125945806503296, | |
| "rewards/rejected": -1.9718964099884033, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "learning_rate": 1.1557596632223497e-07, | |
| "logits/chosen": -3.0164811611175537, | |
| "logits/rejected": -3.015223741531372, | |
| "logps/chosen": -283.99652099609375, | |
| "logps/rejected": -248.39602661132812, | |
| "loss": -0.175, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.068522334098816, | |
| "rewards/margins": 0.9977639317512512, | |
| "rewards/rejected": -2.066286325454712, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 2.39, | |
| "learning_rate": 1.1366245694603903e-07, | |
| "logits/chosen": -2.9585719108581543, | |
| "logits/rejected": -2.9644482135772705, | |
| "logps/chosen": -282.3656921386719, | |
| "logps/rejected": -257.11041259765625, | |
| "loss": -0.04, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -1.0202432870864868, | |
| "rewards/margins": 1.0421122312545776, | |
| "rewards/rejected": -2.0623555183410645, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 1.1174894756984308e-07, | |
| "logits/chosen": -2.976578712463379, | |
| "logits/rejected": -2.958589792251587, | |
| "logps/chosen": -243.4874725341797, | |
| "logps/rejected": -247.5500030517578, | |
| "loss": -0.1438, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.131103277206421, | |
| "rewards/margins": 1.045607328414917, | |
| "rewards/rejected": -2.176710605621338, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "learning_rate": 1.0983543819364714e-07, | |
| "logits/chosen": -2.966090440750122, | |
| "logits/rejected": -2.963219165802002, | |
| "logps/chosen": -265.3279113769531, | |
| "logps/rejected": -245.90011596679688, | |
| "loss": -0.3008, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1297423839569092, | |
| "rewards/margins": 1.0481914281845093, | |
| "rewards/rejected": -2.177933931350708, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 2.42, | |
| "learning_rate": 1.079219288174512e-07, | |
| "logits/chosen": -2.8891921043395996, | |
| "logits/rejected": -2.8980748653411865, | |
| "logps/chosen": -245.95260620117188, | |
| "logps/rejected": -274.68511962890625, | |
| "loss": -0.0161, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.029191255569458, | |
| "rewards/margins": 1.0524277687072754, | |
| "rewards/rejected": -2.0816187858581543, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 2.43, | |
| "learning_rate": 1.0600841944125525e-07, | |
| "logits/chosen": -2.954099178314209, | |
| "logits/rejected": -2.9609456062316895, | |
| "logps/chosen": -305.5159606933594, | |
| "logps/rejected": -262.00189208984375, | |
| "loss": 0.0565, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0684045553207397, | |
| "rewards/margins": 1.0408602952957153, | |
| "rewards/rejected": -2.109265089035034, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 1.0409491006505931e-07, | |
| "logits/chosen": -2.95751690864563, | |
| "logits/rejected": -2.9540932178497314, | |
| "logps/chosen": -272.47802734375, | |
| "logps/rejected": -252.9427490234375, | |
| "loss": -0.1484, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.9607385396957397, | |
| "rewards/margins": 1.1778024435043335, | |
| "rewards/rejected": -2.1385409832000732, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "learning_rate": 1.0218140068886336e-07, | |
| "logits/chosen": -2.9053263664245605, | |
| "logits/rejected": -2.8971660137176514, | |
| "logps/chosen": -288.0937805175781, | |
| "logps/rejected": -249.78091430664062, | |
| "loss": -0.0786, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.9809226989746094, | |
| "rewards/margins": 1.1126328706741333, | |
| "rewards/rejected": -2.0935556888580322, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 2.46, | |
| "learning_rate": 1.0026789131266743e-07, | |
| "logits/chosen": -2.902618169784546, | |
| "logits/rejected": -2.8822264671325684, | |
| "logps/chosen": -282.31939697265625, | |
| "logps/rejected": -234.1844940185547, | |
| "loss": -0.0294, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0446959733963013, | |
| "rewards/margins": 1.026447057723999, | |
| "rewards/rejected": -2.0711429119110107, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 9.835438193647149e-08, | |
| "logits/chosen": -2.979097366333008, | |
| "logits/rejected": -2.939335584640503, | |
| "logps/chosen": -292.04156494140625, | |
| "logps/rejected": -249.23880004882812, | |
| "loss": -0.0569, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.1423760652542114, | |
| "rewards/margins": 1.2263530492782593, | |
| "rewards/rejected": -2.3687291145324707, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 9.644087256027554e-08, | |
| "logits/chosen": -2.9777472019195557, | |
| "logits/rejected": -2.9249491691589355, | |
| "logps/chosen": -300.14605712890625, | |
| "logps/rejected": -261.2189025878906, | |
| "loss": -0.1677, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.053171992301941, | |
| "rewards/margins": 1.085288166999817, | |
| "rewards/rejected": -2.1384599208831787, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "learning_rate": 9.45273631840796e-08, | |
| "logits/chosen": -2.9415881633758545, | |
| "logits/rejected": -2.920991897583008, | |
| "logps/chosen": -283.33758544921875, | |
| "logps/rejected": -262.96820068359375, | |
| "loss": -0.1276, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.8991168737411499, | |
| "rewards/margins": 1.2601044178009033, | |
| "rewards/rejected": -2.1592211723327637, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "learning_rate": 9.261385380788366e-08, | |
| "logits/chosen": -2.9756853580474854, | |
| "logits/rejected": -2.946068525314331, | |
| "logps/chosen": -280.66241455078125, | |
| "logps/rejected": -240.4520263671875, | |
| "loss": -0.126, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0988253355026245, | |
| "rewards/margins": 0.9154653549194336, | |
| "rewards/rejected": -2.0142908096313477, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 2.51, | |
| "learning_rate": 9.070034443168771e-08, | |
| "logits/chosen": -2.981102466583252, | |
| "logits/rejected": -2.961693286895752, | |
| "logps/chosen": -277.573974609375, | |
| "logps/rejected": -287.72515869140625, | |
| "loss": -0.0153, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.1206835508346558, | |
| "rewards/margins": 1.1004483699798584, | |
| "rewards/rejected": -2.2211318016052246, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 8.878683505549177e-08, | |
| "logits/chosen": -2.969409465789795, | |
| "logits/rejected": -2.9719467163085938, | |
| "logps/chosen": -269.2553405761719, | |
| "logps/rejected": -276.8422546386719, | |
| "loss": -0.2463, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.0086815357208252, | |
| "rewards/margins": 1.2824007272720337, | |
| "rewards/rejected": -2.2910819053649902, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 2.53, | |
| "learning_rate": 8.687332567929582e-08, | |
| "logits/chosen": -2.9511923789978027, | |
| "logits/rejected": -2.945295810699463, | |
| "logps/chosen": -259.01507568359375, | |
| "logps/rejected": -261.44781494140625, | |
| "loss": -0.1105, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.238194465637207, | |
| "rewards/margins": 0.8837486505508423, | |
| "rewards/rejected": -2.1219429969787598, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 2.54, | |
| "learning_rate": 8.495981630309988e-08, | |
| "logits/chosen": -2.9451937675476074, | |
| "logits/rejected": -2.9014077186584473, | |
| "logps/chosen": -306.05902099609375, | |
| "logps/rejected": -251.77072143554688, | |
| "loss": -0.0642, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.9629167318344116, | |
| "rewards/margins": 1.1698607206344604, | |
| "rewards/rejected": -2.132777452468872, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 2.55, | |
| "learning_rate": 8.304630692690395e-08, | |
| "logits/chosen": -2.968233108520508, | |
| "logits/rejected": -2.9597315788269043, | |
| "logps/chosen": -272.21258544921875, | |
| "logps/rejected": -263.3721008300781, | |
| "loss": -0.1204, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0844926834106445, | |
| "rewards/margins": 1.1088446378707886, | |
| "rewards/rejected": -2.1933372020721436, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 8.1132797550708e-08, | |
| "logits/chosen": -3.0234591960906982, | |
| "logits/rejected": -2.9642250537872314, | |
| "logps/chosen": -313.2720947265625, | |
| "logps/rejected": -251.24020385742188, | |
| "loss": -0.1713, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.7700778245925903, | |
| "rewards/margins": 1.3790977001190186, | |
| "rewards/rejected": -2.1491756439208984, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 7.921928817451206e-08, | |
| "logits/chosen": -2.968733549118042, | |
| "logits/rejected": -2.9718379974365234, | |
| "logps/chosen": -277.2424621582031, | |
| "logps/rejected": -242.42172241210938, | |
| "loss": -0.1563, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.095190167427063, | |
| "rewards/margins": 1.008971095085144, | |
| "rewards/rejected": -2.104161262512207, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 2.58, | |
| "learning_rate": 7.73057787983161e-08, | |
| "logits/chosen": -2.9798028469085693, | |
| "logits/rejected": -2.9650533199310303, | |
| "logps/chosen": -290.0262756347656, | |
| "logps/rejected": -264.9449157714844, | |
| "loss": -0.2883, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -1.0747302770614624, | |
| "rewards/margins": 1.2478855848312378, | |
| "rewards/rejected": -2.3226161003112793, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 2.59, | |
| "learning_rate": 7.539226942212017e-08, | |
| "logits/chosen": -2.945732593536377, | |
| "logits/rejected": -2.94990611076355, | |
| "logps/chosen": -305.3632507324219, | |
| "logps/rejected": -274.59722900390625, | |
| "loss": -0.2619, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.0448615550994873, | |
| "rewards/margins": 1.2908263206481934, | |
| "rewards/rejected": -2.3356876373291016, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 7.347876004592423e-08, | |
| "logits/chosen": -2.9896063804626465, | |
| "logits/rejected": -2.9576222896575928, | |
| "logps/chosen": -272.3258361816406, | |
| "logps/rejected": -214.65298461914062, | |
| "loss": -0.1643, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.9911131858825684, | |
| "rewards/margins": 1.24199640750885, | |
| "rewards/rejected": -2.23310923576355, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 2.61, | |
| "learning_rate": 7.156525066972828e-08, | |
| "logits/chosen": -2.9311654567718506, | |
| "logits/rejected": -2.9140172004699707, | |
| "logps/chosen": -268.74444580078125, | |
| "logps/rejected": -236.22988891601562, | |
| "loss": -0.0763, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.111546277999878, | |
| "rewards/margins": 1.021420955657959, | |
| "rewards/rejected": -2.132966995239258, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 2.62, | |
| "learning_rate": 6.965174129353234e-08, | |
| "logits/chosen": -2.9497828483581543, | |
| "logits/rejected": -2.9311363697052, | |
| "logps/chosen": -305.40240478515625, | |
| "logps/rejected": -264.26177978515625, | |
| "loss": -0.1631, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.7808889746665955, | |
| "rewards/margins": 1.4971510171890259, | |
| "rewards/rejected": -2.2780401706695557, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "learning_rate": 6.773823191733639e-08, | |
| "logits/chosen": -2.8948826789855957, | |
| "logits/rejected": -2.8769936561584473, | |
| "logps/chosen": -272.2402648925781, | |
| "logps/rejected": -246.4182891845703, | |
| "loss": -0.2377, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.8501654863357544, | |
| "rewards/margins": 1.4621152877807617, | |
| "rewards/rejected": -2.3122806549072266, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 6.582472254114045e-08, | |
| "logits/chosen": -2.9354665279388428, | |
| "logits/rejected": -2.8916828632354736, | |
| "logps/chosen": -272.43682861328125, | |
| "logps/rejected": -274.33819580078125, | |
| "loss": -0.0427, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.1567806005477905, | |
| "rewards/margins": 0.9909998178482056, | |
| "rewards/rejected": -2.147780179977417, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 6.391121316494451e-08, | |
| "logits/chosen": -2.9822142124176025, | |
| "logits/rejected": -2.982553482055664, | |
| "logps/chosen": -282.8048400878906, | |
| "logps/rejected": -264.55401611328125, | |
| "loss": -0.2413, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.2675011157989502, | |
| "rewards/margins": 1.1336963176727295, | |
| "rewards/rejected": -2.4011974334716797, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 2.66, | |
| "learning_rate": 6.199770378874856e-08, | |
| "logits/chosen": -2.946916103363037, | |
| "logits/rejected": -2.921715497970581, | |
| "logps/chosen": -295.14581298828125, | |
| "logps/rejected": -264.893798828125, | |
| "loss": -0.1178, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0069832801818848, | |
| "rewards/margins": 1.1966971158981323, | |
| "rewards/rejected": -2.2036805152893066, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 6.008419441255262e-08, | |
| "logits/chosen": -2.996048927307129, | |
| "logits/rejected": -2.9277799129486084, | |
| "logps/chosen": -288.4164733886719, | |
| "logps/rejected": -240.5204315185547, | |
| "loss": -0.1271, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.2253119945526123, | |
| "rewards/margins": 0.9576913714408875, | |
| "rewards/rejected": -2.1830034255981445, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "learning_rate": 5.817068503635668e-08, | |
| "logits/chosen": -2.979665756225586, | |
| "logits/rejected": -3.006110668182373, | |
| "logps/chosen": -309.9973449707031, | |
| "logps/rejected": -266.8681945800781, | |
| "loss": -0.1954, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.072455644607544, | |
| "rewards/margins": 1.285042643547058, | |
| "rewards/rejected": -2.3574986457824707, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 5.6257175660160735e-08, | |
| "logits/chosen": -2.995903968811035, | |
| "logits/rejected": -2.985555410385132, | |
| "logps/chosen": -255.1632537841797, | |
| "logps/rejected": -262.6217346191406, | |
| "loss": -0.1921, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.135871171951294, | |
| "rewards/margins": 1.1645699739456177, | |
| "rewards/rejected": -2.300441026687622, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "learning_rate": 5.4343666283964784e-08, | |
| "logits/chosen": -2.947361946105957, | |
| "logits/rejected": -2.9222099781036377, | |
| "logps/chosen": -283.39483642578125, | |
| "logps/rejected": -236.3388214111328, | |
| "loss": -0.2522, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -1.0431753396987915, | |
| "rewards/margins": 1.4626045227050781, | |
| "rewards/rejected": -2.50577974319458, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 2.72, | |
| "learning_rate": 5.243015690776884e-08, | |
| "logits/chosen": -2.934537649154663, | |
| "logits/rejected": -2.903463840484619, | |
| "logps/chosen": -297.0821228027344, | |
| "logps/rejected": -241.7076416015625, | |
| "loss": -0.1882, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.178007960319519, | |
| "rewards/margins": 1.0751577615737915, | |
| "rewards/rejected": -2.2531659603118896, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 5.05166475315729e-08, | |
| "logits/chosen": -2.9361720085144043, | |
| "logits/rejected": -2.882931709289551, | |
| "logps/chosen": -270.83148193359375, | |
| "logps/rejected": -255.24337768554688, | |
| "loss": -0.1525, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.2923060655593872, | |
| "rewards/margins": 1.1520329713821411, | |
| "rewards/rejected": -2.444338798522949, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 4.860313815537696e-08, | |
| "logits/chosen": -2.9840073585510254, | |
| "logits/rejected": -2.9296116828918457, | |
| "logps/chosen": -283.15643310546875, | |
| "logps/rejected": -250.57608032226562, | |
| "loss": -0.1079, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.0855220556259155, | |
| "rewards/margins": 1.3181664943695068, | |
| "rewards/rejected": -2.4036881923675537, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 2.75, | |
| "learning_rate": 4.668962877918101e-08, | |
| "logits/chosen": -2.9590346813201904, | |
| "logits/rejected": -2.9196345806121826, | |
| "logps/chosen": -256.04315185546875, | |
| "logps/rejected": -279.31787109375, | |
| "loss": -0.1673, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.3440836668014526, | |
| "rewards/margins": 1.061953067779541, | |
| "rewards/rejected": -2.406036853790283, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "learning_rate": 4.477611940298507e-08, | |
| "logits/chosen": -2.9351706504821777, | |
| "logits/rejected": -2.9165854454040527, | |
| "logps/chosen": -302.4515075683594, | |
| "logps/rejected": -260.8388366699219, | |
| "loss": -0.4567, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.1327465772628784, | |
| "rewards/margins": 1.2759268283843994, | |
| "rewards/rejected": -2.4086735248565674, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 4.2862610026789124e-08, | |
| "logits/chosen": -3.013035297393799, | |
| "logits/rejected": -2.989163398742676, | |
| "logps/chosen": -270.7334899902344, | |
| "logps/rejected": -254.8367156982422, | |
| "loss": -0.1713, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0498313903808594, | |
| "rewards/margins": 1.175312876701355, | |
| "rewards/rejected": -2.225144147872925, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "learning_rate": 4.0949100650593186e-08, | |
| "logits/chosen": -2.880460023880005, | |
| "logits/rejected": -2.861250400543213, | |
| "logps/chosen": -269.2752990722656, | |
| "logps/rejected": -265.5714111328125, | |
| "loss": -0.0863, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.1616452932357788, | |
| "rewards/margins": 1.2222042083740234, | |
| "rewards/rejected": -2.383849620819092, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 2.79, | |
| "learning_rate": 3.903559127439724e-08, | |
| "logits/chosen": -2.9040732383728027, | |
| "logits/rejected": -2.920821189880371, | |
| "logps/chosen": -267.18621826171875, | |
| "logps/rejected": -227.2675018310547, | |
| "loss": -0.3545, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.007155179977417, | |
| "rewards/margins": 1.408747911453247, | |
| "rewards/rejected": -2.415903091430664, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "learning_rate": 3.71220818982013e-08, | |
| "logits/chosen": -2.9555585384368896, | |
| "logits/rejected": -2.891514778137207, | |
| "logps/chosen": -293.35809326171875, | |
| "logps/rejected": -241.7179412841797, | |
| "loss": -0.0836, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.3558762073516846, | |
| "rewards/margins": 0.7381645441055298, | |
| "rewards/rejected": -2.094040632247925, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 3.520857252200535e-08, | |
| "logits/chosen": -2.9766342639923096, | |
| "logits/rejected": -2.9516689777374268, | |
| "logps/chosen": -288.6914978027344, | |
| "logps/rejected": -282.8217468261719, | |
| "loss": -0.079, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.2634086608886719, | |
| "rewards/margins": 0.9370874166488647, | |
| "rewards/rejected": -2.200496196746826, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "learning_rate": 3.3295063145809414e-08, | |
| "logits/chosen": -2.9203312397003174, | |
| "logits/rejected": -2.923161268234253, | |
| "logps/chosen": -278.013916015625, | |
| "logps/rejected": -267.54168701171875, | |
| "loss": -0.1356, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.0634254217147827, | |
| "rewards/margins": 1.14893639087677, | |
| "rewards/rejected": -2.212362051010132, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 3.138155376961347e-08, | |
| "logits/chosen": -3.015432834625244, | |
| "logits/rejected": -2.9883949756622314, | |
| "logps/chosen": -306.4672546386719, | |
| "logps/rejected": -284.0825500488281, | |
| "loss": -0.0924, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.2742321491241455, | |
| "rewards/margins": 1.3397479057312012, | |
| "rewards/rejected": -2.6139800548553467, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "learning_rate": 2.9468044393417525e-08, | |
| "logits/chosen": -2.9456729888916016, | |
| "logits/rejected": -2.9101481437683105, | |
| "logps/chosen": -281.6629638671875, | |
| "logps/rejected": -247.2931671142578, | |
| "loss": -0.2122, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.8153074383735657, | |
| "rewards/margins": 1.5913559198379517, | |
| "rewards/rejected": -2.406663417816162, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 2.755453501722158e-08, | |
| "logits/chosen": -2.8928332328796387, | |
| "logits/rejected": -2.8488776683807373, | |
| "logps/chosen": -262.3460693359375, | |
| "logps/rejected": -253.6611328125, | |
| "loss": -0.3582, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.2655181884765625, | |
| "rewards/margins": 1.1255066394805908, | |
| "rewards/rejected": -2.3910248279571533, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "learning_rate": 2.564102564102564e-08, | |
| "logits/chosen": -2.9541738033294678, | |
| "logits/rejected": -2.9483609199523926, | |
| "logps/chosen": -281.6482849121094, | |
| "logps/rejected": -263.13116455078125, | |
| "loss": -0.2229, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -1.135956048965454, | |
| "rewards/margins": 1.4537389278411865, | |
| "rewards/rejected": -2.5896952152252197, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 2.87, | |
| "learning_rate": 2.3727516264829695e-08, | |
| "logits/chosen": -2.96710467338562, | |
| "logits/rejected": -2.935039520263672, | |
| "logps/chosen": -314.2068176269531, | |
| "logps/rejected": -288.13592529296875, | |
| "loss": -0.0912, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.2961633205413818, | |
| "rewards/margins": 1.0811083316802979, | |
| "rewards/rejected": -2.3772716522216797, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "learning_rate": 2.1814006888633754e-08, | |
| "logits/chosen": -2.9459080696105957, | |
| "logits/rejected": -2.89158296585083, | |
| "logps/chosen": -286.0101318359375, | |
| "logps/rejected": -239.49160766601562, | |
| "loss": -0.115, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.2027372121810913, | |
| "rewards/margins": 1.0859975814819336, | |
| "rewards/rejected": -2.2887349128723145, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 1.990049751243781e-08, | |
| "logits/chosen": -2.9392828941345215, | |
| "logits/rejected": -2.9443650245666504, | |
| "logps/chosen": -291.0428771972656, | |
| "logps/rejected": -242.24203491210938, | |
| "loss": -0.2681, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.150545358657837, | |
| "rewards/margins": 1.183664083480835, | |
| "rewards/rejected": -2.334209680557251, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 2.9, | |
| "learning_rate": 1.7986988136241865e-08, | |
| "logits/chosen": -2.936471462249756, | |
| "logits/rejected": -2.891524076461792, | |
| "logps/chosen": -287.06024169921875, | |
| "logps/rejected": -288.4886169433594, | |
| "loss": -0.2539, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.1829683780670166, | |
| "rewards/margins": 1.063999891281128, | |
| "rewards/rejected": -2.2469685077667236, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 2.91, | |
| "learning_rate": 1.6073478760045924e-08, | |
| "logits/chosen": -2.9530060291290283, | |
| "logits/rejected": -2.9200854301452637, | |
| "logps/chosen": -283.59716796875, | |
| "logps/rejected": -236.90316772460938, | |
| "loss": -0.1698, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.1723743677139282, | |
| "rewards/margins": 1.1659984588623047, | |
| "rewards/rejected": -2.3383727073669434, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 1.4159969383849981e-08, | |
| "logits/chosen": -2.9582297801971436, | |
| "logits/rejected": -2.9348087310791016, | |
| "logps/chosen": -269.26531982421875, | |
| "logps/rejected": -241.8841094970703, | |
| "loss": -0.3465, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.1937872171401978, | |
| "rewards/margins": 1.0226060152053833, | |
| "rewards/rejected": -2.216393232345581, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 1.2246460007654037e-08, | |
| "logits/chosen": -2.985964775085449, | |
| "logits/rejected": -2.9667985439300537, | |
| "logps/chosen": -297.0205078125, | |
| "logps/rejected": -251.09805297851562, | |
| "loss": -0.1413, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.407981276512146, | |
| "rewards/margins": 0.9606062173843384, | |
| "rewards/rejected": -2.3685877323150635, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 2.94, | |
| "learning_rate": 1.0332950631458094e-08, | |
| "logits/chosen": -2.9180400371551514, | |
| "logits/rejected": -2.9097437858581543, | |
| "logps/chosen": -291.8419494628906, | |
| "logps/rejected": -265.8045349121094, | |
| "loss": -0.116, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.1430332660675049, | |
| "rewards/margins": 1.2068586349487305, | |
| "rewards/rejected": -2.3498916625976562, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 2.95, | |
| "learning_rate": 8.419441255262151e-09, | |
| "logits/chosen": -2.9281063079833984, | |
| "logits/rejected": -2.8641605377197266, | |
| "logps/chosen": -270.81256103515625, | |
| "logps/rejected": -246.7694549560547, | |
| "loss": -0.2278, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.335326910018921, | |
| "rewards/margins": 1.1297729015350342, | |
| "rewards/rejected": -2.465099811553955, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "learning_rate": 6.505931879066207e-09, | |
| "logits/chosen": -2.948106050491333, | |
| "logits/rejected": -2.921739339828491, | |
| "logps/chosen": -276.6488952636719, | |
| "logps/rejected": -248.07608032226562, | |
| "loss": -0.1698, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.1927988529205322, | |
| "rewards/margins": 1.302239179611206, | |
| "rewards/rejected": -2.4950382709503174, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "learning_rate": 4.592422502870264e-09, | |
| "logits/chosen": -2.9207305908203125, | |
| "logits/rejected": -2.8921730518341064, | |
| "logps/chosen": -288.192138671875, | |
| "logps/rejected": -244.62954711914062, | |
| "loss": -0.1019, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.206674337387085, | |
| "rewards/margins": 0.8675975799560547, | |
| "rewards/rejected": -2.0742716789245605, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 2.98, | |
| "learning_rate": 2.6789131266743202e-09, | |
| "logits/chosen": -3.0061116218566895, | |
| "logits/rejected": -2.972447156906128, | |
| "logps/chosen": -306.70977783203125, | |
| "logps/rejected": -261.82244873046875, | |
| "loss": -0.1622, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -1.2506418228149414, | |
| "rewards/margins": 1.2404968738555908, | |
| "rewards/rejected": -2.4911389350891113, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "learning_rate": 7.654037504783773e-10, | |
| "logits/chosen": -2.974818706512451, | |
| "logits/rejected": -2.963379383087158, | |
| "logps/chosen": -263.7333679199219, | |
| "logps/rejected": -262.7867126464844, | |
| "loss": -0.2602, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.150649070739746, | |
| "rewards/margins": 1.2455689907073975, | |
| "rewards/rejected": -2.3962180614471436, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_logits/chosen": -2.965512990951538, | |
| "eval_logits/rejected": -2.9399757385253906, | |
| "eval_logps/chosen": -282.7847900390625, | |
| "eval_logps/rejected": -252.9479217529297, | |
| "eval_loss": -0.203842431306839, | |
| "eval_rewards/accuracies": 0.6840000152587891, | |
| "eval_rewards/chosen": -1.1628247499465942, | |
| "eval_rewards/margins": 1.2828813791275024, | |
| "eval_rewards/rejected": -2.4457061290740967, | |
| "eval_runtime": 446.4813, | |
| "eval_samples_per_second": 4.479, | |
| "eval_steps_per_second": 0.28, | |
| "step": 2904 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 2904, | |
| "total_flos": 0.0, | |
| "train_loss": 0.36701411283100355, | |
| "train_runtime": 84636.1866, | |
| "train_samples_per_second": 2.196, | |
| "train_steps_per_second": 0.034 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2904, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 0.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |