{ "best_metric": null, "best_model_checkpoint": null, "epoch": 0.10908842980841345, "eval_steps": 500, "global_step": 100, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0, "learning_rate": 9.782608695652175e-07, "logits/chosen": -1.8868483304977417, "logits/rejected": -2.3036646842956543, "logps/chosen": -466.7117004394531, "logps/rejected": -99.3152084350586, "loss": 0.6931, "rewards/accuracies": 0.0, "rewards/chosen": 0.0, "rewards/margins": 0.0, "rewards/rejected": 0.0, "step": 1 }, { "epoch": 0.0, "learning_rate": 1.956521739130435e-06, "logits/chosen": -2.2296698093414307, "logits/rejected": -2.469517469406128, "logps/chosen": -335.50140380859375, "logps/rejected": -97.97496032714844, "loss": 0.6931, "rewards/accuracies": 0.0, "rewards/chosen": 0.0, "rewards/margins": 0.0, "rewards/rejected": 0.0, "step": 2 }, { "epoch": 0.0, "learning_rate": 2.9347826086956523e-06, "logits/chosen": -2.125047206878662, "logits/rejected": -2.445204019546509, "logps/chosen": -484.7939758300781, "logps/rejected": -125.93072509765625, "loss": 1.7268, "rewards/accuracies": 0.4375, "rewards/chosen": -1.0339633226394653, "rewards/margins": -0.9356753826141357, "rewards/rejected": -0.09828801453113556, "step": 3 }, { "epoch": 0.0, "learning_rate": 3.91304347826087e-06, "logits/chosen": -1.7763867378234863, "logits/rejected": -2.1961894035339355, "logps/chosen": -386.8011779785156, "logps/rejected": -106.04309844970703, "loss": 2.3806, "rewards/accuracies": 0.5625, "rewards/chosen": -0.9238373041152954, "rewards/margins": -1.2902063131332397, "rewards/rejected": 0.36636874079704285, "step": 4 }, { "epoch": 0.01, "learning_rate": 4.891304347826087e-06, "logits/chosen": -1.7257812023162842, "logits/rejected": -2.353095769882202, "logps/chosen": -524.7261962890625, "logps/rejected": -70.26946258544922, "loss": 0.7488, "rewards/accuracies": 0.625, "rewards/chosen": 2.197690010070801, "rewards/margins": 2.0008764266967773, "rewards/rejected": 0.19681359827518463, "step": 5 }, { "epoch": 0.01, "learning_rate": 5.869565217391305e-06, "logits/chosen": -2.1483359336853027, "logits/rejected": -2.4817819595336914, "logps/chosen": -283.6588134765625, "logps/rejected": -58.25455093383789, "loss": 0.8712, "rewards/accuracies": 0.5625, "rewards/chosen": 0.4915351867675781, "rewards/margins": 0.2005542814731598, "rewards/rejected": 0.2909809350967407, "step": 6 }, { "epoch": 0.01, "learning_rate": 6.847826086956523e-06, "logits/chosen": -2.1019060611724854, "logits/rejected": -2.300177812576294, "logps/chosen": -297.35235595703125, "logps/rejected": -175.5060272216797, "loss": 1.1392, "rewards/accuracies": 0.5625, "rewards/chosen": 0.6018204689025879, "rewards/margins": -0.036775827407836914, "rewards/rejected": 0.6385962963104248, "step": 7 }, { "epoch": 0.01, "learning_rate": 7.82608695652174e-06, "logits/chosen": -2.0169122219085693, "logits/rejected": -2.201702117919922, "logps/chosen": -342.04852294921875, "logps/rejected": -117.9761962890625, "loss": 1.1799, "rewards/accuracies": 0.5625, "rewards/chosen": 1.111680507659912, "rewards/margins": 0.17297500371932983, "rewards/rejected": 0.9387054443359375, "step": 8 }, { "epoch": 0.01, "learning_rate": 8.804347826086957e-06, "logits/chosen": -1.6254596710205078, "logits/rejected": -2.0724740028381348, "logps/chosen": -480.1983642578125, "logps/rejected": -108.05937194824219, "loss": 0.5203, "rewards/accuracies": 0.8125, "rewards/chosen": 2.161550283432007, "rewards/margins": 1.3515225648880005, "rewards/rejected": 0.8100277185440063, "step": 9 }, { "epoch": 0.01, "learning_rate": 9.782608695652175e-06, "logits/chosen": -1.4482485055923462, "logits/rejected": -2.1116294860839844, "logps/chosen": -451.5759582519531, "logps/rejected": -67.73944854736328, "loss": 0.355, "rewards/accuracies": 0.8125, "rewards/chosen": 6.18800163269043, "rewards/margins": 5.184471130371094, "rewards/rejected": 1.0035302639007568, "step": 10 }, { "epoch": 0.01, "learning_rate": 1.0760869565217392e-05, "logits/chosen": -1.2660062313079834, "logits/rejected": -1.8649226427078247, "logps/chosen": -265.8935546875, "logps/rejected": -57.866336822509766, "loss": 0.3178, "rewards/accuracies": 0.8125, "rewards/chosen": 5.78374719619751, "rewards/margins": 4.856299877166748, "rewards/rejected": 0.9274475574493408, "step": 11 }, { "epoch": 0.01, "learning_rate": 1.173913043478261e-05, "logits/chosen": -1.1804625988006592, "logits/rejected": -1.7009960412979126, "logps/chosen": -247.2069854736328, "logps/rejected": -79.96739196777344, "loss": 0.6617, "rewards/accuracies": 0.75, "rewards/chosen": 6.5535197257995605, "rewards/margins": 4.647799015045166, "rewards/rejected": 1.9057204723358154, "step": 12 }, { "epoch": 0.01, "learning_rate": 1.2717391304347827e-05, "logits/chosen": -0.9693112373352051, "logits/rejected": -1.6497248411178589, "logps/chosen": -372.8963623046875, "logps/rejected": -70.47633361816406, "loss": 0.2498, "rewards/accuracies": 0.8125, "rewards/chosen": 9.02272891998291, "rewards/margins": 7.325143814086914, "rewards/rejected": 1.6975862979888916, "step": 13 }, { "epoch": 0.02, "learning_rate": 1.3695652173913046e-05, "logits/chosen": -0.7723280787467957, "logits/rejected": -1.4176554679870605, "logps/chosen": -318.0392150878906, "logps/rejected": -79.94412994384766, "loss": 0.6401, "rewards/accuracies": 0.6875, "rewards/chosen": 6.698611259460449, "rewards/margins": 4.517416477203369, "rewards/rejected": 2.18119478225708, "step": 14 }, { "epoch": 0.02, "learning_rate": 1.4673913043478261e-05, "logits/chosen": -0.8084051609039307, "logits/rejected": -1.4428132772445679, "logps/chosen": -363.15869140625, "logps/rejected": -70.70079803466797, "loss": 0.4963, "rewards/accuracies": 0.75, "rewards/chosen": 7.02784538269043, "rewards/margins": 5.400259971618652, "rewards/rejected": 1.6275854110717773, "step": 15 }, { "epoch": 0.02, "learning_rate": 1.565217391304348e-05, "logits/chosen": -0.8056433200836182, "logits/rejected": -1.4629294872283936, "logps/chosen": -328.90435791015625, "logps/rejected": -77.36820983886719, "loss": 0.544, "rewards/accuracies": 0.75, "rewards/chosen": 9.417024612426758, "rewards/margins": 7.023505687713623, "rewards/rejected": 2.393519163131714, "step": 16 }, { "epoch": 0.02, "learning_rate": 1.6630434782608694e-05, "logits/chosen": -1.0918309688568115, "logits/rejected": -1.6613047122955322, "logps/chosen": -277.37982177734375, "logps/rejected": -56.75635528564453, "loss": 0.8958, "rewards/accuracies": 0.8125, "rewards/chosen": 6.844330310821533, "rewards/margins": 4.82295560836792, "rewards/rejected": 2.0213749408721924, "step": 17 }, { "epoch": 0.02, "learning_rate": 1.7608695652173915e-05, "logits/chosen": -0.9810999631881714, "logits/rejected": -1.975247859954834, "logps/chosen": -465.6856689453125, "logps/rejected": -56.440269470214844, "loss": 0.2275, "rewards/accuracies": 0.875, "rewards/chosen": 7.767542362213135, "rewards/margins": 6.572459697723389, "rewards/rejected": 1.1950825452804565, "step": 18 }, { "epoch": 0.02, "learning_rate": 1.8586956521739132e-05, "logits/chosen": -1.2428462505340576, "logits/rejected": -1.8064367771148682, "logps/chosen": -322.7100830078125, "logps/rejected": -108.71321105957031, "loss": 0.3006, "rewards/accuracies": 0.875, "rewards/chosen": 6.459216594696045, "rewards/margins": 3.9207711219787598, "rewards/rejected": 2.5384464263916016, "step": 19 }, { "epoch": 0.02, "learning_rate": 1.956521739130435e-05, "logits/chosen": -1.7773098945617676, "logits/rejected": -2.178387403488159, "logps/chosen": -357.8656921386719, "logps/rejected": -112.65576934814453, "loss": 0.3837, "rewards/accuracies": 0.9375, "rewards/chosen": 4.684590816497803, "rewards/margins": 4.16137170791626, "rewards/rejected": 0.5232191681861877, "step": 20 }, { "epoch": 0.02, "learning_rate": 2.0543478260869567e-05, "logits/chosen": -1.6802027225494385, "logits/rejected": -1.943274974822998, "logps/chosen": -327.85693359375, "logps/rejected": -87.53279113769531, "loss": 0.4428, "rewards/accuracies": 0.9375, "rewards/chosen": 2.681103229522705, "rewards/margins": 2.3518502712249756, "rewards/rejected": 0.32925307750701904, "step": 21 }, { "epoch": 0.02, "learning_rate": 2.1521739130434784e-05, "logits/chosen": -1.5789223909378052, "logits/rejected": -2.4593281745910645, "logps/chosen": -590.5732421875, "logps/rejected": -86.78819274902344, "loss": 0.252, "rewards/accuracies": 0.875, "rewards/chosen": 3.6503655910491943, "rewards/margins": 3.621542453765869, "rewards/rejected": 0.028822531923651695, "step": 22 }, { "epoch": 0.03, "learning_rate": 2.25e-05, "logits/chosen": -1.8142305612564087, "logits/rejected": -2.3507237434387207, "logps/chosen": -296.42266845703125, "logps/rejected": -60.37131118774414, "loss": 0.5602, "rewards/accuracies": 0.75, "rewards/chosen": 2.170171022415161, "rewards/margins": 2.5353941917419434, "rewards/rejected": -0.36522313952445984, "step": 23 }, { "epoch": 0.03, "learning_rate": 2.347826086956522e-05, "logits/chosen": -1.7050156593322754, "logits/rejected": -2.158708333969116, "logps/chosen": -523.41748046875, "logps/rejected": -81.6954345703125, "loss": 0.7359, "rewards/accuracies": 0.9375, "rewards/chosen": 2.3799068927764893, "rewards/margins": 2.963341236114502, "rewards/rejected": -0.583433985710144, "step": 24 }, { "epoch": 0.03, "learning_rate": 2.4456521739130436e-05, "logits/chosen": -1.704190731048584, "logits/rejected": -2.215942859649658, "logps/chosen": -457.6519775390625, "logps/rejected": -91.778076171875, "loss": 0.0865, "rewards/accuracies": 1.0, "rewards/chosen": 4.57167911529541, "rewards/margins": 4.6597795486450195, "rewards/rejected": -0.08810039609670639, "step": 25 }, { "epoch": 0.03, "learning_rate": 2.5434782608695653e-05, "logits/chosen": -1.4195502996444702, "logits/rejected": -2.109910488128662, "logps/chosen": -445.05908203125, "logps/rejected": -80.95398712158203, "loss": 0.1265, "rewards/accuracies": 0.9375, "rewards/chosen": 5.362960338592529, "rewards/margins": 5.412578582763672, "rewards/rejected": -0.049618080258369446, "step": 26 }, { "epoch": 0.03, "learning_rate": 2.6413043478260874e-05, "logits/chosen": -1.0274971723556519, "logits/rejected": -1.7681918144226074, "logps/chosen": -393.85870361328125, "logps/rejected": -77.79011535644531, "loss": 0.1702, "rewards/accuracies": 1.0, "rewards/chosen": 4.572484016418457, "rewards/margins": 4.946743488311768, "rewards/rejected": -0.374259889125824, "step": 27 }, { "epoch": 0.03, "learning_rate": 2.739130434782609e-05, "logits/chosen": -0.7365273237228394, "logits/rejected": -1.9988534450531006, "logps/chosen": -569.8936767578125, "logps/rejected": -79.71562194824219, "loss": 0.0465, "rewards/accuracies": 1.0, "rewards/chosen": 12.751686096191406, "rewards/margins": 12.474647521972656, "rewards/rejected": 0.2770393192768097, "step": 28 }, { "epoch": 0.03, "learning_rate": 2.836956521739131e-05, "logits/chosen": -0.6424872279167175, "logits/rejected": -1.6069284677505493, "logps/chosen": -368.68292236328125, "logps/rejected": -103.9141845703125, "loss": 0.0351, "rewards/accuracies": 1.0, "rewards/chosen": 10.144453048706055, "rewards/margins": 10.381340980529785, "rewards/rejected": -0.23688830435276031, "step": 29 }, { "epoch": 0.03, "learning_rate": 2.9347826086956523e-05, "logits/chosen": -0.6061868071556091, "logits/rejected": -1.4256294965744019, "logps/chosen": -360.4029846191406, "logps/rejected": -130.27224731445312, "loss": 0.0414, "rewards/accuracies": 1.0, "rewards/chosen": 12.649539947509766, "rewards/margins": 12.36604118347168, "rewards/rejected": 0.2834990620613098, "step": 30 }, { "epoch": 0.03, "learning_rate": 3.032608695652174e-05, "logits/chosen": -1.0075111389160156, "logits/rejected": -1.3918204307556152, "logps/chosen": -227.05450439453125, "logps/rejected": -172.67947387695312, "loss": 1.1145, "rewards/accuracies": 0.875, "rewards/chosen": 4.628115653991699, "rewards/margins": 3.9521055221557617, "rewards/rejected": 0.6760098934173584, "step": 31 }, { "epoch": 0.03, "learning_rate": 3.130434782608696e-05, "logits/chosen": -0.9060839414596558, "logits/rejected": -2.0123891830444336, "logps/chosen": -403.06781005859375, "logps/rejected": -75.35684967041016, "loss": 0.0319, "rewards/accuracies": 1.0, "rewards/chosen": 7.352475643157959, "rewards/margins": 8.778717041015625, "rewards/rejected": -1.4262410402297974, "step": 32 }, { "epoch": 0.04, "learning_rate": 3.228260869565217e-05, "logits/chosen": -1.2984271049499512, "logits/rejected": -1.9507720470428467, "logps/chosen": -306.7958984375, "logps/rejected": -107.01404571533203, "loss": 0.0795, "rewards/accuracies": 0.9375, "rewards/chosen": 5.223363399505615, "rewards/margins": 7.012631416320801, "rewards/rejected": -1.7892680168151855, "step": 33 }, { "epoch": 0.04, "learning_rate": 3.326086956521739e-05, "logits/chosen": -1.1749062538146973, "logits/rejected": -2.004805564880371, "logps/chosen": -437.9205322265625, "logps/rejected": -135.21376037597656, "loss": 0.0021, "rewards/accuracies": 1.0, "rewards/chosen": 9.3555908203125, "rewards/margins": 12.285521507263184, "rewards/rejected": -2.9299299716949463, "step": 34 }, { "epoch": 0.04, "learning_rate": 3.423913043478261e-05, "logits/chosen": -0.9383536577224731, "logits/rejected": -2.0366413593292236, "logps/chosen": -457.8018493652344, "logps/rejected": -87.7723617553711, "loss": 0.041, "rewards/accuracies": 1.0, "rewards/chosen": 6.836911678314209, "rewards/margins": 9.43667984008789, "rewards/rejected": -2.599769115447998, "step": 35 }, { "epoch": 0.04, "learning_rate": 3.521739130434783e-05, "logits/chosen": -1.120687484741211, "logits/rejected": -1.9933563470840454, "logps/chosen": -439.6186828613281, "logps/rejected": -94.24763488769531, "loss": 0.1906, "rewards/accuracies": 0.9375, "rewards/chosen": 6.565365314483643, "rewards/margins": 9.163484573364258, "rewards/rejected": -2.5981192588806152, "step": 36 }, { "epoch": 0.04, "learning_rate": 3.619565217391305e-05, "logits/chosen": -1.338189959526062, "logits/rejected": -1.9118335247039795, "logps/chosen": -286.96728515625, "logps/rejected": -127.22909545898438, "loss": 0.0279, "rewards/accuracies": 1.0, "rewards/chosen": 2.9341468811035156, "rewards/margins": 6.362301349639893, "rewards/rejected": -3.4281537532806396, "step": 37 }, { "epoch": 0.04, "learning_rate": 3.7173913043478264e-05, "logits/chosen": -1.2371840476989746, "logits/rejected": -1.7362310886383057, "logps/chosen": -319.52911376953125, "logps/rejected": -151.65084838867188, "loss": 0.0939, "rewards/accuracies": 0.9375, "rewards/chosen": 6.216956615447998, "rewards/margins": 9.073128700256348, "rewards/rejected": -2.856172561645508, "step": 38 }, { "epoch": 0.04, "learning_rate": 3.815217391304348e-05, "logits/chosen": -0.8356418609619141, "logits/rejected": -1.7252635955810547, "logps/chosen": -391.63275146484375, "logps/rejected": -154.64361572265625, "loss": 0.2483, "rewards/accuracies": 0.9375, "rewards/chosen": 8.802470207214355, "rewards/margins": 11.676048278808594, "rewards/rejected": -2.873577356338501, "step": 39 }, { "epoch": 0.04, "learning_rate": 3.91304347826087e-05, "logits/chosen": -0.9311200380325317, "logits/rejected": -1.8107151985168457, "logps/chosen": -374.2525939941406, "logps/rejected": -131.86715698242188, "loss": 0.0058, "rewards/accuracies": 1.0, "rewards/chosen": 8.586356163024902, "rewards/margins": 11.15583324432373, "rewards/rejected": -2.569479465484619, "step": 40 }, { "epoch": 0.04, "learning_rate": 4.0108695652173916e-05, "logits/chosen": -0.6861732006072998, "logits/rejected": -1.6295411586761475, "logps/chosen": -311.7944641113281, "logps/rejected": -132.4564666748047, "loss": 0.0124, "rewards/accuracies": 1.0, "rewards/chosen": 5.333560943603516, "rewards/margins": 10.023443222045898, "rewards/rejected": -4.689882755279541, "step": 41 }, { "epoch": 0.05, "learning_rate": 4.1086956521739134e-05, "logits/chosen": -1.1133558750152588, "logits/rejected": -2.0819039344787598, "logps/chosen": -282.22332763671875, "logps/rejected": -98.7315673828125, "loss": 0.0156, "rewards/accuracies": 1.0, "rewards/chosen": 4.930261611938477, "rewards/margins": 8.344033241271973, "rewards/rejected": -3.413771629333496, "step": 42 }, { "epoch": 0.05, "learning_rate": 4.206521739130435e-05, "logits/chosen": -1.0311042070388794, "logits/rejected": -2.0758776664733887, "logps/chosen": -369.7279052734375, "logps/rejected": -146.08364868164062, "loss": 0.0127, "rewards/accuracies": 1.0, "rewards/chosen": 7.7552289962768555, "rewards/margins": 13.249262809753418, "rewards/rejected": -5.494032859802246, "step": 43 }, { "epoch": 0.05, "learning_rate": 4.304347826086957e-05, "logits/chosen": -1.0191632509231567, "logits/rejected": -1.995982050895691, "logps/chosen": -367.692626953125, "logps/rejected": -152.65135192871094, "loss": 0.0004, "rewards/accuracies": 1.0, "rewards/chosen": 7.9517822265625, "rewards/margins": 13.869260787963867, "rewards/rejected": -5.917478084564209, "step": 44 }, { "epoch": 0.05, "learning_rate": 4.4021739130434786e-05, "logits/chosen": -1.3430471420288086, "logits/rejected": -2.0910849571228027, "logps/chosen": -334.8968811035156, "logps/rejected": -168.4519500732422, "loss": 0.0057, "rewards/accuracies": 1.0, "rewards/chosen": 5.113959312438965, "rewards/margins": 12.524876594543457, "rewards/rejected": -7.41091775894165, "step": 45 }, { "epoch": 0.05, "learning_rate": 4.5e-05, "logits/chosen": -0.8516018986701965, "logits/rejected": -1.9934579133987427, "logps/chosen": -386.7164001464844, "logps/rejected": -152.8460235595703, "loss": 0.0253, "rewards/accuracies": 1.0, "rewards/chosen": 7.0803022384643555, "rewards/margins": 13.202106475830078, "rewards/rejected": -6.121803283691406, "step": 46 }, { "epoch": 0.05, "learning_rate": 4.494827586206897e-05, "logits/chosen": -0.9460695385932922, "logits/rejected": -1.9993284940719604, "logps/chosen": -355.2873229980469, "logps/rejected": -188.5989532470703, "loss": 0.1008, "rewards/accuracies": 0.9375, "rewards/chosen": 6.138028144836426, "rewards/margins": 13.93982982635498, "rewards/rejected": -7.801800727844238, "step": 47 }, { "epoch": 0.05, "learning_rate": 4.489655172413793e-05, "logits/chosen": -1.212510108947754, "logits/rejected": -2.0202791690826416, "logps/chosen": -293.0318908691406, "logps/rejected": -194.92649841308594, "loss": 0.0058, "rewards/accuracies": 1.0, "rewards/chosen": 6.742659091949463, "rewards/margins": 17.209505081176758, "rewards/rejected": -10.466848373413086, "step": 48 }, { "epoch": 0.05, "learning_rate": 4.48448275862069e-05, "logits/chosen": -1.388157606124878, "logits/rejected": -2.1885251998901367, "logps/chosen": -330.6868591308594, "logps/rejected": -214.35423278808594, "loss": 0.0003, "rewards/accuracies": 1.0, "rewards/chosen": 5.653369903564453, "rewards/margins": 18.257244110107422, "rewards/rejected": -12.603873252868652, "step": 49 }, { "epoch": 0.05, "learning_rate": 4.479310344827587e-05, "logits/chosen": -1.020527958869934, "logits/rejected": -1.9433815479278564, "logps/chosen": -347.48394775390625, "logps/rejected": -191.15147399902344, "loss": 0.0001, "rewards/accuracies": 1.0, "rewards/chosen": 7.30579137802124, "rewards/margins": 18.734811782836914, "rewards/rejected": -11.429021835327148, "step": 50 }, { "epoch": 0.06, "learning_rate": 4.474137931034483e-05, "logits/chosen": -1.2260757684707642, "logits/rejected": -2.0841352939605713, "logps/chosen": -243.5459747314453, "logps/rejected": -188.00442504882812, "loss": 0.0002, "rewards/accuracies": 1.0, "rewards/chosen": 3.675260543823242, "rewards/margins": 14.774760246276855, "rewards/rejected": -11.099499702453613, "step": 51 }, { "epoch": 0.06, "learning_rate": 4.46896551724138e-05, "logits/chosen": -1.0378711223602295, "logits/rejected": -2.0174639225006104, "logps/chosen": -375.9520263671875, "logps/rejected": -182.97463989257812, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 7.18237829208374, "rewards/margins": 19.010208129882812, "rewards/rejected": -11.82783317565918, "step": 52 }, { "epoch": 0.06, "learning_rate": 4.4637931034482765e-05, "logits/chosen": -1.1728301048278809, "logits/rejected": -1.9667140245437622, "logps/chosen": -309.5382995605469, "logps/rejected": -241.13743591308594, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 6.681103706359863, "rewards/margins": 21.542236328125, "rewards/rejected": -14.861133575439453, "step": 53 }, { "epoch": 0.06, "learning_rate": 4.4586206896551726e-05, "logits/chosen": -1.1491725444793701, "logits/rejected": -1.760243535041809, "logps/chosen": -250.09036254882812, "logps/rejected": -226.221435546875, "loss": 0.0001, "rewards/accuracies": 1.0, "rewards/chosen": 2.3173880577087402, "rewards/margins": 16.417604446411133, "rewards/rejected": -14.100215911865234, "step": 54 }, { "epoch": 0.06, "learning_rate": 4.4534482758620694e-05, "logits/chosen": -1.2637637853622437, "logits/rejected": -1.8996949195861816, "logps/chosen": -401.9977722167969, "logps/rejected": -303.9198303222656, "loss": 0.0011, "rewards/accuracies": 1.0, "rewards/chosen": 8.595758438110352, "rewards/margins": 26.598325729370117, "rewards/rejected": -18.002567291259766, "step": 55 }, { "epoch": 0.06, "learning_rate": 4.4482758620689656e-05, "logits/chosen": -0.9855178594589233, "logits/rejected": -2.0190346240997314, "logps/chosen": -394.40625, "logps/rejected": -253.97720336914062, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 6.395060062408447, "rewards/margins": 23.680639266967773, "rewards/rejected": -17.28557777404785, "step": 56 }, { "epoch": 0.06, "learning_rate": 4.4431034482758624e-05, "logits/chosen": -1.049959421157837, "logits/rejected": -1.6531357765197754, "logps/chosen": -205.54058837890625, "logps/rejected": -250.66958618164062, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 4.874389171600342, "rewards/margins": 20.083953857421875, "rewards/rejected": -15.209564208984375, "step": 57 }, { "epoch": 0.06, "learning_rate": 4.4379310344827585e-05, "logits/chosen": -0.8982111811637878, "logits/rejected": -1.6001371145248413, "logps/chosen": -241.91790771484375, "logps/rejected": -252.10299682617188, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 6.044814109802246, "rewards/margins": 21.77715301513672, "rewards/rejected": -15.732340812683105, "step": 58 }, { "epoch": 0.06, "learning_rate": 4.432758620689655e-05, "logits/chosen": -0.753044068813324, "logits/rejected": -1.6325318813323975, "logps/chosen": -315.30230712890625, "logps/rejected": -245.91976928710938, "loss": 0.0078, "rewards/accuracies": 1.0, "rewards/chosen": 5.801758289337158, "rewards/margins": 20.689924240112305, "rewards/rejected": -14.888164520263672, "step": 59 }, { "epoch": 0.07, "learning_rate": 4.427586206896552e-05, "logits/chosen": -0.5443448424339294, "logits/rejected": -1.7224345207214355, "logps/chosen": -388.339599609375, "logps/rejected": -211.5500946044922, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 11.5880708694458, "rewards/margins": 24.889877319335938, "rewards/rejected": -13.301810264587402, "step": 60 }, { "epoch": 0.07, "learning_rate": 4.422413793103448e-05, "logits/chosen": -0.6921290755271912, "logits/rejected": -1.6139135360717773, "logps/chosen": -242.47312927246094, "logps/rejected": -214.73622131347656, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 11.329044342041016, "rewards/margins": 21.45581817626953, "rewards/rejected": -10.1267728805542, "step": 61 }, { "epoch": 0.07, "learning_rate": 4.417241379310345e-05, "logits/chosen": -0.4828271269798279, "logits/rejected": -1.649349570274353, "logps/chosen": -306.6004638671875, "logps/rejected": -172.2523956298828, "loss": 0.0003, "rewards/accuracies": 1.0, "rewards/chosen": 11.373737335205078, "rewards/margins": 20.521434783935547, "rewards/rejected": -9.147698402404785, "step": 62 }, { "epoch": 0.07, "learning_rate": 4.412068965517242e-05, "logits/chosen": -0.4676312506198883, "logits/rejected": -1.6393827199935913, "logps/chosen": -357.127685546875, "logps/rejected": -199.84524536132812, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 16.730457305908203, "rewards/margins": 26.409488677978516, "rewards/rejected": -9.679031372070312, "step": 63 }, { "epoch": 0.07, "learning_rate": 4.406896551724138e-05, "logits/chosen": -0.7774850726127625, "logits/rejected": -1.324657678604126, "logps/chosen": -193.75833129882812, "logps/rejected": -194.70233154296875, "loss": 0.0015, "rewards/accuracies": 1.0, "rewards/chosen": 5.371121406555176, "rewards/margins": 15.129293441772461, "rewards/rejected": -9.758172988891602, "step": 64 }, { "epoch": 0.07, "learning_rate": 4.401724137931035e-05, "logits/chosen": -0.5502160787582397, "logits/rejected": -1.466262936592102, "logps/chosen": -240.96836853027344, "logps/rejected": -174.6772003173828, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 7.808657169342041, "rewards/margins": 17.520769119262695, "rewards/rejected": -9.712109565734863, "step": 65 }, { "epoch": 0.07, "learning_rate": 4.3965517241379315e-05, "logits/chosen": -0.5615445375442505, "logits/rejected": -1.3243337869644165, "logps/chosen": -282.65484619140625, "logps/rejected": -211.33326721191406, "loss": 0.4403, "rewards/accuracies": 0.9375, "rewards/chosen": 9.213937759399414, "rewards/margins": 18.93706512451172, "rewards/rejected": -9.723125457763672, "step": 66 }, { "epoch": 0.07, "learning_rate": 4.3913793103448277e-05, "logits/chosen": -0.5882927775382996, "logits/rejected": -1.4038889408111572, "logps/chosen": -244.26333618164062, "logps/rejected": -192.05517578125, "loss": 0.0007, "rewards/accuracies": 1.0, "rewards/chosen": 5.999594688415527, "rewards/margins": 16.232702255249023, "rewards/rejected": -10.233107566833496, "step": 67 }, { "epoch": 0.07, "learning_rate": 4.3862068965517245e-05, "logits/chosen": -0.7710579633712769, "logits/rejected": -1.5811477899551392, "logps/chosen": -274.4286804199219, "logps/rejected": -189.31427001953125, "loss": 0.0003, "rewards/accuracies": 1.0, "rewards/chosen": 10.483807563781738, "rewards/margins": 22.807409286499023, "rewards/rejected": -12.323600769042969, "step": 68 }, { "epoch": 0.08, "learning_rate": 4.381034482758621e-05, "logits/chosen": -0.2001597136259079, "logits/rejected": -1.5477687120437622, "logps/chosen": -398.04296875, "logps/rejected": -208.8218536376953, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 11.122913360595703, "rewards/margins": 23.094343185424805, "rewards/rejected": -11.971429824829102, "step": 69 }, { "epoch": 0.08, "learning_rate": 4.3758620689655174e-05, "logits/chosen": -0.5891858339309692, "logits/rejected": -1.500455379486084, "logps/chosen": -223.07235717773438, "logps/rejected": -219.37405395507812, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 7.498288154602051, "rewards/margins": 19.65944480895996, "rewards/rejected": -12.16115665435791, "step": 70 }, { "epoch": 0.08, "learning_rate": 4.370689655172414e-05, "logits/chosen": -0.5222135186195374, "logits/rejected": -1.459835410118103, "logps/chosen": -264.3045349121094, "logps/rejected": -184.4599609375, "loss": 0.0006, "rewards/accuracies": 1.0, "rewards/chosen": 7.261447429656982, "rewards/margins": 17.37218475341797, "rewards/rejected": -10.110737800598145, "step": 71 }, { "epoch": 0.08, "learning_rate": 4.365517241379311e-05, "logits/chosen": -0.4974746108055115, "logits/rejected": -1.5458664894104004, "logps/chosen": -272.7602233886719, "logps/rejected": -181.00247192382812, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 10.925564765930176, "rewards/margins": 22.81020736694336, "rewards/rejected": -11.88464069366455, "step": 72 }, { "epoch": 0.08, "learning_rate": 4.360344827586207e-05, "logits/chosen": -0.3357059359550476, "logits/rejected": -1.4441150426864624, "logps/chosen": -414.8647766113281, "logps/rejected": -223.109375, "loss": 0.0004, "rewards/accuracies": 1.0, "rewards/chosen": 13.181656837463379, "rewards/margins": 27.527368545532227, "rewards/rejected": -14.345712661743164, "step": 73 }, { "epoch": 0.08, "learning_rate": 4.355172413793104e-05, "logits/chosen": -0.6531690359115601, "logits/rejected": -1.4611525535583496, "logps/chosen": -221.1251220703125, "logps/rejected": -226.30352783203125, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 7.807477951049805, "rewards/margins": 21.502174377441406, "rewards/rejected": -13.694696426391602, "step": 74 }, { "epoch": 0.08, "learning_rate": 4.35e-05, "logits/chosen": -0.5135932564735413, "logits/rejected": -1.4995931386947632, "logps/chosen": -332.56842041015625, "logps/rejected": -270.9891662597656, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 12.301117897033691, "rewards/margins": 26.830467224121094, "rewards/rejected": -14.529346466064453, "step": 75 }, { "epoch": 0.08, "learning_rate": 4.344827586206897e-05, "logits/chosen": -0.3770020604133606, "logits/rejected": -1.44219970703125, "logps/chosen": -317.96240234375, "logps/rejected": -250.3682098388672, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 7.8860931396484375, "rewards/margins": 23.53211784362793, "rewards/rejected": -15.646024703979492, "step": 76 }, { "epoch": 0.08, "learning_rate": 4.339655172413793e-05, "logits/chosen": -0.49628371000289917, "logits/rejected": -1.4045476913452148, "logps/chosen": -367.6162414550781, "logps/rejected": -265.2467346191406, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 10.148900985717773, "rewards/margins": 24.313669204711914, "rewards/rejected": -14.164766311645508, "step": 77 }, { "epoch": 0.09, "learning_rate": 4.33448275862069e-05, "logits/chosen": -0.5287263989448547, "logits/rejected": -1.3665097951889038, "logps/chosen": -341.87664794921875, "logps/rejected": -249.6795654296875, "loss": 0.1517, "rewards/accuracies": 0.9375, "rewards/chosen": 8.620326042175293, "rewards/margins": 19.850658416748047, "rewards/rejected": -11.230331420898438, "step": 78 }, { "epoch": 0.09, "learning_rate": 4.3293103448275865e-05, "logits/chosen": -0.2743966281414032, "logits/rejected": -1.4255130290985107, "logps/chosen": -334.3340759277344, "logps/rejected": -265.15032958984375, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 10.872364044189453, "rewards/margins": 27.00977325439453, "rewards/rejected": -16.13741111755371, "step": 79 }, { "epoch": 0.09, "learning_rate": 4.324137931034483e-05, "logits/chosen": -0.5563563108444214, "logits/rejected": -1.6934064626693726, "logps/chosen": -408.9612731933594, "logps/rejected": -220.00091552734375, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 10.237764358520508, "rewards/margins": 24.824390411376953, "rewards/rejected": -14.586627960205078, "step": 80 }, { "epoch": 0.09, "learning_rate": 4.3189655172413795e-05, "logits/chosen": -0.617451548576355, "logits/rejected": -1.640676736831665, "logps/chosen": -363.1540222167969, "logps/rejected": -226.42376708984375, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 10.672024726867676, "rewards/margins": 24.347591400146484, "rewards/rejected": -13.675565719604492, "step": 81 }, { "epoch": 0.09, "learning_rate": 4.313793103448276e-05, "logits/chosen": -0.45464372634887695, "logits/rejected": -1.4708175659179688, "logps/chosen": -328.35675048828125, "logps/rejected": -223.79766845703125, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 6.965122699737549, "rewards/margins": 21.22077178955078, "rewards/rejected": -14.25564956665039, "step": 82 }, { "epoch": 0.09, "learning_rate": 4.3086206896551724e-05, "logits/chosen": -0.6079340577125549, "logits/rejected": -1.5333082675933838, "logps/chosen": -219.4099578857422, "logps/rejected": -230.07948303222656, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 6.278050899505615, "rewards/margins": 20.45656967163086, "rewards/rejected": -14.178520202636719, "step": 83 }, { "epoch": 0.09, "learning_rate": 4.303448275862069e-05, "logits/chosen": -0.4874611794948578, "logits/rejected": -1.5135740041732788, "logps/chosen": -304.00274658203125, "logps/rejected": -263.0120849609375, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 9.557435989379883, "rewards/margins": 24.41012954711914, "rewards/rejected": -14.852693557739258, "step": 84 }, { "epoch": 0.09, "learning_rate": 4.298275862068966e-05, "logits/chosen": -0.4802761375904083, "logits/rejected": -1.4294252395629883, "logps/chosen": -293.41192626953125, "logps/rejected": -283.2821350097656, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 7.370968818664551, "rewards/margins": 24.32880401611328, "rewards/rejected": -16.957834243774414, "step": 85 }, { "epoch": 0.09, "learning_rate": 4.293103448275862e-05, "logits/chosen": -0.3580012321472168, "logits/rejected": -1.499201774597168, "logps/chosen": -350.5351257324219, "logps/rejected": -244.6523895263672, "loss": 0.0001, "rewards/accuracies": 1.0, "rewards/chosen": 12.237674713134766, "rewards/margins": 26.58084487915039, "rewards/rejected": -14.343170166015625, "step": 86 }, { "epoch": 0.09, "learning_rate": 4.287931034482759e-05, "logits/chosen": -0.6917211413383484, "logits/rejected": -1.4503819942474365, "logps/chosen": -217.5175323486328, "logps/rejected": -228.9481201171875, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 6.635045051574707, "rewards/margins": 20.42833709716797, "rewards/rejected": -13.793292999267578, "step": 87 }, { "epoch": 0.1, "learning_rate": 4.282758620689656e-05, "logits/chosen": -0.8255922794342041, "logits/rejected": -1.4748644828796387, "logps/chosen": -217.336669921875, "logps/rejected": -206.8959197998047, "loss": 0.0005, "rewards/accuracies": 1.0, "rewards/chosen": 5.22888708114624, "rewards/margins": 19.673564910888672, "rewards/rejected": -14.444679260253906, "step": 88 }, { "epoch": 0.1, "learning_rate": 4.2775862068965525e-05, "logits/chosen": -0.626908004283905, "logits/rejected": -1.6663663387298584, "logps/chosen": -328.56231689453125, "logps/rejected": -235.42843627929688, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 6.891221046447754, "rewards/margins": 21.77893829345703, "rewards/rejected": -14.887716293334961, "step": 89 }, { "epoch": 0.1, "learning_rate": 4.2724137931034486e-05, "logits/chosen": -0.5689429640769958, "logits/rejected": -1.618609070777893, "logps/chosen": -298.08770751953125, "logps/rejected": -237.06094360351562, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 9.328791618347168, "rewards/margins": 24.572046279907227, "rewards/rejected": -15.243253707885742, "step": 90 }, { "epoch": 0.1, "learning_rate": 4.2672413793103454e-05, "logits/chosen": -0.5474239587783813, "logits/rejected": -1.6142076253890991, "logps/chosen": -295.00494384765625, "logps/rejected": -268.9505615234375, "loss": 0.0287, "rewards/accuracies": 1.0, "rewards/chosen": 9.715471267700195, "rewards/margins": 23.968780517578125, "rewards/rejected": -14.25330924987793, "step": 91 }, { "epoch": 0.1, "learning_rate": 4.2620689655172416e-05, "logits/chosen": -0.5132906436920166, "logits/rejected": -1.6248977184295654, "logps/chosen": -314.5284423828125, "logps/rejected": -210.1342315673828, "loss": 0.0595, "rewards/accuracies": 0.9375, "rewards/chosen": 7.707674980163574, "rewards/margins": 20.18167495727539, "rewards/rejected": -12.474000930786133, "step": 92 }, { "epoch": 0.1, "learning_rate": 4.2568965517241384e-05, "logits/chosen": -0.45244526863098145, "logits/rejected": -1.7539148330688477, "logps/chosen": -394.9619140625, "logps/rejected": -230.32550048828125, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 11.32716178894043, "rewards/margins": 26.326129913330078, "rewards/rejected": -14.998967170715332, "step": 93 }, { "epoch": 0.1, "learning_rate": 4.2517241379310345e-05, "logits/chosen": -0.7538139224052429, "logits/rejected": -1.5796599388122559, "logps/chosen": -286.95281982421875, "logps/rejected": -266.7412414550781, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 8.003694534301758, "rewards/margins": 25.05299949645996, "rewards/rejected": -17.04930305480957, "step": 94 }, { "epoch": 0.1, "learning_rate": 4.246551724137931e-05, "logits/chosen": -0.44376400113105774, "logits/rejected": -1.5493687391281128, "logps/chosen": -406.93951416015625, "logps/rejected": -302.9401550292969, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 7.246862888336182, "rewards/margins": 27.937149047851562, "rewards/rejected": -20.69028663635254, "step": 95 }, { "epoch": 0.1, "learning_rate": 4.2413793103448274e-05, "logits/chosen": -0.7577255964279175, "logits/rejected": -1.77558171749115, "logps/chosen": -357.90838623046875, "logps/rejected": -246.73458862304688, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 7.267894744873047, "rewards/margins": 24.79804229736328, "rewards/rejected": -17.530147552490234, "step": 96 }, { "epoch": 0.11, "learning_rate": 4.236206896551724e-05, "logits/chosen": -0.8565713763237, "logits/rejected": -1.7760589122772217, "logps/chosen": -338.751220703125, "logps/rejected": -304.101806640625, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 9.238365173339844, "rewards/margins": 27.781978607177734, "rewards/rejected": -18.54361343383789, "step": 97 }, { "epoch": 0.11, "learning_rate": 4.231034482758621e-05, "logits/chosen": -0.40401750802993774, "logits/rejected": -1.7701493501663208, "logps/chosen": -537.5440063476562, "logps/rejected": -349.8548889160156, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 14.211512565612793, "rewards/margins": 37.746315002441406, "rewards/rejected": -23.53480339050293, "step": 98 }, { "epoch": 0.11, "learning_rate": 4.225862068965517e-05, "logits/chosen": -0.9799883365631104, "logits/rejected": -1.8936114311218262, "logps/chosen": -403.7567138671875, "logps/rejected": -257.6401672363281, "loss": 0.0001, "rewards/accuracies": 1.0, "rewards/chosen": 4.710639476776123, "rewards/margins": 23.923961639404297, "rewards/rejected": -19.213321685791016, "step": 99 }, { "epoch": 0.11, "learning_rate": 4.220689655172414e-05, "logits/chosen": -0.94623202085495, "logits/rejected": -1.7190004587173462, "logps/chosen": -360.8213806152344, "logps/rejected": -310.1994323730469, "loss": 0.0, "rewards/accuracies": 1.0, "rewards/chosen": 3.991710662841797, "rewards/margins": 26.59444808959961, "rewards/rejected": -22.60273551940918, "step": 100 } ], "logging_steps": 1, "max_steps": 916, "num_input_tokens_seen": 0, "num_train_epochs": 1, "save_steps": 100, "total_flos": 0.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }