| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9994353472614342, | |
| "eval_steps": 1000, | |
| "global_step": 885, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.001129305477131564, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.1235955056179774e-08, | |
| "logits/chosen": -2.8912107944488525, | |
| "logits/rejected": -2.9116690158843994, | |
| "logps/chosen": -205.43255615234375, | |
| "logps/rejected": -277.7913818359375, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01129305477131564, | |
| "grad_norm": 11.5, | |
| "learning_rate": 1.1235955056179774e-07, | |
| "logits/chosen": -2.9240243434906006, | |
| "logits/rejected": -2.895505666732788, | |
| "logps/chosen": -286.3099365234375, | |
| "logps/rejected": -234.07386779785156, | |
| "loss": 0.6929, | |
| "rewards/accuracies": 0.3888888955116272, | |
| "rewards/chosen": 0.00452011963352561, | |
| "rewards/margins": 0.0007920749485492706, | |
| "rewards/rejected": 0.0037280453834682703, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02258610954263128, | |
| "grad_norm": 8.125, | |
| "learning_rate": 2.2471910112359549e-07, | |
| "logits/chosen": -2.9115793704986572, | |
| "logits/rejected": -2.8940234184265137, | |
| "logps/chosen": -242.3488311767578, | |
| "logps/rejected": -230.69912719726562, | |
| "loss": 0.6965, | |
| "rewards/accuracies": 0.40625, | |
| "rewards/chosen": -0.006699255667626858, | |
| "rewards/margins": -0.006314368452876806, | |
| "rewards/rejected": -0.0003848885535262525, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03387916431394692, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 3.3707865168539325e-07, | |
| "logits/chosen": -2.933945894241333, | |
| "logits/rejected": -2.9031498432159424, | |
| "logps/chosen": -227.9883270263672, | |
| "logps/rejected": -206.83169555664062, | |
| "loss": 0.6914, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": 0.0010180685203522444, | |
| "rewards/margins": 0.003906534984707832, | |
| "rewards/rejected": -0.0028884666971862316, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04517221908526256, | |
| "grad_norm": 10.5, | |
| "learning_rate": 4.4943820224719097e-07, | |
| "logits/chosen": -2.940782308578491, | |
| "logits/rejected": -2.903311252593994, | |
| "logps/chosen": -239.7572021484375, | |
| "logps/rejected": -223.0744171142578, | |
| "loss": 0.6943, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.001310513005591929, | |
| "rewards/margins": -0.0019447545055299997, | |
| "rewards/rejected": 0.0006342411506921053, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05646527385657821, | |
| "grad_norm": 11.25, | |
| "learning_rate": 5.617977528089887e-07, | |
| "logits/chosen": -2.926548719406128, | |
| "logits/rejected": -2.891540050506592, | |
| "logps/chosen": -252.4969024658203, | |
| "logps/rejected": -208.90573120117188, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.002666835905984044, | |
| "rewards/margins": 0.0014006331330165267, | |
| "rewards/rejected": -0.004067468456923962, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06775832862789384, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 6.741573033707865e-07, | |
| "logits/chosen": -2.951221227645874, | |
| "logits/rejected": -2.9214468002319336, | |
| "logps/chosen": -230.9364013671875, | |
| "logps/rejected": -222.9247283935547, | |
| "loss": 0.696, | |
| "rewards/accuracies": 0.42500001192092896, | |
| "rewards/chosen": -0.008724420331418514, | |
| "rewards/margins": -0.0053501129150390625, | |
| "rewards/rejected": -0.003374306485056877, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07905138339920949, | |
| "grad_norm": 15.75, | |
| "learning_rate": 7.865168539325843e-07, | |
| "logits/chosen": -2.915444850921631, | |
| "logits/rejected": -2.8806443214416504, | |
| "logps/chosen": -249.3389434814453, | |
| "logps/rejected": -235.85330200195312, | |
| "loss": 0.6919, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.006868250668048859, | |
| "rewards/margins": 0.003010864369571209, | |
| "rewards/rejected": -0.009879115037620068, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.09034443817052512, | |
| "grad_norm": 13.375, | |
| "learning_rate": 8.988764044943819e-07, | |
| "logits/chosen": -2.893014430999756, | |
| "logits/rejected": -2.8584141731262207, | |
| "logps/chosen": -244.0336456298828, | |
| "logps/rejected": -241.87136840820312, | |
| "loss": 0.6888, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.01003711111843586, | |
| "rewards/margins": 0.00937909446656704, | |
| "rewards/rejected": -0.0194162018597126, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.10163749294184077, | |
| "grad_norm": 6.90625, | |
| "learning_rate": 9.99996105846605e-07, | |
| "logits/chosen": -2.912137031555176, | |
| "logits/rejected": -2.8917088508605957, | |
| "logps/chosen": -230.53872680664062, | |
| "logps/rejected": -230.29702758789062, | |
| "loss": 0.6905, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.01865602657198906, | |
| "rewards/margins": 0.006008770316839218, | |
| "rewards/rejected": -0.024664796888828278, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.11293054771315642, | |
| "grad_norm": 8.0625, | |
| "learning_rate": 9.99528880830604e-07, | |
| "logits/chosen": -2.954880952835083, | |
| "logits/rejected": -2.9029383659362793, | |
| "logps/chosen": -230.21240234375, | |
| "logps/rejected": -242.08786010742188, | |
| "loss": 0.6888, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.03368181362748146, | |
| "rewards/margins": 0.010173236019909382, | |
| "rewards/rejected": -0.04385504871606827, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12422360248447205, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 9.982836589657042e-07, | |
| "logits/chosen": -2.893187999725342, | |
| "logits/rejected": -2.855039596557617, | |
| "logps/chosen": -253.92001342773438, | |
| "logps/rejected": -249.1540069580078, | |
| "loss": 0.6881, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.04104708880186081, | |
| "rewards/margins": 0.011361517943441868, | |
| "rewards/rejected": -0.0524086132645607, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.13551665725578768, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 9.962623796366428e-07, | |
| "logits/chosen": -2.9364559650421143, | |
| "logits/rejected": -2.921806812286377, | |
| "logps/chosen": -233.13034057617188, | |
| "logps/rejected": -215.0299072265625, | |
| "loss": 0.6917, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.0493025965988636, | |
| "rewards/margins": 0.004110801964998245, | |
| "rewards/rejected": -0.05341339856386185, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.14680971202710333, | |
| "grad_norm": 10.875, | |
| "learning_rate": 9.934681909075434e-07, | |
| "logits/chosen": -2.9207844734191895, | |
| "logits/rejected": -2.894355297088623, | |
| "logps/chosen": -246.37173461914062, | |
| "logps/rejected": -252.4733428955078, | |
| "loss": 0.6846, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.056733161211013794, | |
| "rewards/margins": 0.019008487462997437, | |
| "rewards/rejected": -0.07574164122343063, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.15810276679841898, | |
| "grad_norm": 7.75, | |
| "learning_rate": 9.899054446189302e-07, | |
| "logits/chosen": -2.9487192630767822, | |
| "logits/rejected": -2.9129204750061035, | |
| "logps/chosen": -248.5870361328125, | |
| "logps/rejected": -237.9082489013672, | |
| "loss": 0.6859, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.06868256628513336, | |
| "rewards/margins": 0.016593391075730324, | |
| "rewards/rejected": -0.08527596294879913, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.16939582156973462, | |
| "grad_norm": 10.75, | |
| "learning_rate": 9.855796896099044e-07, | |
| "logits/chosen": -2.915994167327881, | |
| "logits/rejected": -2.8766446113586426, | |
| "logps/chosen": -257.65740966796875, | |
| "logps/rejected": -226.9933624267578, | |
| "loss": 0.6854, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.07911563664674759, | |
| "rewards/margins": 0.018480664119124413, | |
| "rewards/rejected": -0.09759628772735596, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.18068887634105024, | |
| "grad_norm": 9.375, | |
| "learning_rate": 9.804976630760418e-07, | |
| "logits/chosen": -2.918802261352539, | |
| "logits/rejected": -2.87815260887146, | |
| "logps/chosen": -219.70858764648438, | |
| "logps/rejected": -212.9764404296875, | |
| "loss": 0.6801, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.07974254339933395, | |
| "rewards/margins": 0.029045332223176956, | |
| "rewards/rejected": -0.10878787934780121, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.1919819311123659, | |
| "grad_norm": 58.5, | |
| "learning_rate": 9.746672800764734e-07, | |
| "logits/chosen": -2.9482781887054443, | |
| "logits/rejected": -2.911285400390625, | |
| "logps/chosen": -248.09774780273438, | |
| "logps/rejected": -235.35397338867188, | |
| "loss": 0.6792, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.09282518178224564, | |
| "rewards/margins": 0.03139440342783928, | |
| "rewards/rejected": -0.12421958148479462, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.20327498588368154, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 9.680976212064874e-07, | |
| "logits/chosen": -2.9101614952087402, | |
| "logits/rejected": -2.865682363510132, | |
| "logps/chosen": -244.28964233398438, | |
| "logps/rejected": -202.06594848632812, | |
| "loss": 0.6829, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.10991182178258896, | |
| "rewards/margins": 0.024404151365160942, | |
| "rewards/rejected": -0.13431595265865326, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.21456804065499718, | |
| "grad_norm": 9.5, | |
| "learning_rate": 9.607989184548542e-07, | |
| "logits/chosen": -2.928318738937378, | |
| "logits/rejected": -2.9046478271484375, | |
| "logps/chosen": -249.5603790283203, | |
| "logps/rejected": -234.435791015625, | |
| "loss": 0.6893, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.11975125968456268, | |
| "rewards/margins": 0.012782303616404533, | |
| "rewards/rejected": -0.13253356516361237, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.22586109542631283, | |
| "grad_norm": 10.0625, | |
| "learning_rate": 9.52782539267901e-07, | |
| "logits/chosen": -2.9089949131011963, | |
| "logits/rejected": -2.8771417140960693, | |
| "logps/chosen": -238.6053009033203, | |
| "logps/rejected": -244.843994140625, | |
| "loss": 0.6775, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.12132523953914642, | |
| "rewards/margins": 0.03615022823214531, | |
| "rewards/rejected": -0.15747547149658203, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.23715415019762845, | |
| "grad_norm": 10.75, | |
| "learning_rate": 9.44060968845156e-07, | |
| "logits/chosen": -2.957009792327881, | |
| "logits/rejected": -2.917191982269287, | |
| "logps/chosen": -254.15109252929688, | |
| "logps/rejected": -213.61257934570312, | |
| "loss": 0.6785, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.12494392693042755, | |
| "rewards/margins": 0.034127604216337204, | |
| "rewards/rejected": -0.15907153487205505, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2484472049689441, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 9.346477906941331e-07, | |
| "logits/chosen": -2.9415574073791504, | |
| "logits/rejected": -2.908458709716797, | |
| "logps/chosen": -273.68597412109375, | |
| "logps/rejected": -242.5727996826172, | |
| "loss": 0.6812, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.14751294255256653, | |
| "rewards/margins": 0.03075326606631279, | |
| "rewards/rejected": -0.17826619744300842, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.2597402597402597, | |
| "grad_norm": 10.75, | |
| "learning_rate": 9.245576654745471e-07, | |
| "logits/chosen": -2.961373805999756, | |
| "logits/rejected": -2.9270126819610596, | |
| "logps/chosen": -271.6230163574219, | |
| "logps/rejected": -242.1273956298828, | |
| "loss": 0.6785, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.15283231437206268, | |
| "rewards/margins": 0.035538919270038605, | |
| "rewards/rejected": -0.18837125599384308, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.27103331451157536, | |
| "grad_norm": 11.4375, | |
| "learning_rate": 9.13806308164905e-07, | |
| "logits/chosen": -2.9348835945129395, | |
| "logits/rejected": -2.8914520740509033, | |
| "logps/chosen": -231.8206024169922, | |
| "logps/rejected": -218.68106079101562, | |
| "loss": 0.6784, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.1429007351398468, | |
| "rewards/margins": 0.037423450499773026, | |
| "rewards/rejected": -0.18032419681549072, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.282326369282891, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 9.024104635870367e-07, | |
| "logits/chosen": -2.9596941471099854, | |
| "logits/rejected": -2.9176220893859863, | |
| "logps/chosen": -234.1117706298828, | |
| "logps/rejected": -221.8756866455078, | |
| "loss": 0.6625, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.16116580367088318, | |
| "rewards/margins": 0.07054189592599869, | |
| "rewards/rejected": -0.23170769214630127, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.29361942405420666, | |
| "grad_norm": 10.875, | |
| "learning_rate": 8.90387880326684e-07, | |
| "logits/chosen": -2.9382996559143066, | |
| "logits/rejected": -2.9059150218963623, | |
| "logps/chosen": -236.49020385742188, | |
| "logps/rejected": -239.8851318359375, | |
| "loss": 0.6765, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.15588422119617462, | |
| "rewards/margins": 0.04047652706503868, | |
| "rewards/rejected": -0.1963607370853424, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3049124788255223, | |
| "grad_norm": 9.0, | |
| "learning_rate": 8.777572830907684e-07, | |
| "logits/chosen": -2.950742721557617, | |
| "logits/rejected": -2.935871124267578, | |
| "logps/chosen": -235.8411407470703, | |
| "logps/rejected": -237.037109375, | |
| "loss": 0.6797, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.18154451251029968, | |
| "rewards/margins": 0.03585205227136612, | |
| "rewards/rejected": -0.2173965722322464, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.31620553359683795, | |
| "grad_norm": 9.5, | |
| "learning_rate": 8.645383435443851e-07, | |
| "logits/chosen": -2.944791793823242, | |
| "logits/rejected": -2.9193873405456543, | |
| "logps/chosen": -258.33331298828125, | |
| "logps/rejected": -272.0276184082031, | |
| "loss": 0.6716, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.1910783350467682, | |
| "rewards/margins": 0.051886141300201416, | |
| "rewards/rejected": -0.242964506149292, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3274985883681536, | |
| "grad_norm": 11.1875, | |
| "learning_rate": 8.507516496729493e-07, | |
| "logits/chosen": -2.9242568016052246, | |
| "logits/rejected": -2.9168529510498047, | |
| "logps/chosen": -251.5631103515625, | |
| "logps/rejected": -245.46444702148438, | |
| "loss": 0.703, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.23275713622570038, | |
| "rewards/margins": -0.012155926786363125, | |
| "rewards/rejected": -0.22060124576091766, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.33879164313946925, | |
| "grad_norm": 11.6875, | |
| "learning_rate": 8.364186737172068e-07, | |
| "logits/chosen": -2.9377264976501465, | |
| "logits/rejected": -2.9192919731140137, | |
| "logps/chosen": -272.1338806152344, | |
| "logps/rejected": -252.2882537841797, | |
| "loss": 0.6809, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.19352756440639496, | |
| "rewards/margins": 0.03381720930337906, | |
| "rewards/rejected": -0.2273447960615158, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.3500846979107849, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 8.215617387310522e-07, | |
| "logits/chosen": -2.961489200592041, | |
| "logits/rejected": -2.9303905963897705, | |
| "logps/chosen": -255.98681640625, | |
| "logps/rejected": -215.4890594482422, | |
| "loss": 0.6801, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.19549933075904846, | |
| "rewards/margins": 0.035571686923503876, | |
| "rewards/rejected": -0.23107102513313293, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.3613777526821005, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 8.062039838142401e-07, | |
| "logits/chosen": -2.9097650051116943, | |
| "logits/rejected": -2.879265546798706, | |
| "logps/chosen": -233.59078979492188, | |
| "logps/rejected": -255.03677368164062, | |
| "loss": 0.6642, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.1973254680633545, | |
| "rewards/margins": 0.06654740124940872, | |
| "rewards/rejected": -0.2638728618621826, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.37267080745341613, | |
| "grad_norm": 9.5, | |
| "learning_rate": 7.903693280741331e-07, | |
| "logits/chosen": -2.932699203491211, | |
| "logits/rejected": -2.8970255851745605, | |
| "logps/chosen": -245.6687469482422, | |
| "logps/rejected": -264.0479736328125, | |
| "loss": 0.6635, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.1542760580778122, | |
| "rewards/margins": 0.07145790755748749, | |
| "rewards/rejected": -0.22573396563529968, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3839638622247318, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 7.740824333726213e-07, | |
| "logits/chosen": -2.974498748779297, | |
| "logits/rejected": -2.9419915676116943, | |
| "logps/chosen": -254.0312957763672, | |
| "logps/rejected": -241.7189178466797, | |
| "loss": 0.6755, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.19228370487689972, | |
| "rewards/margins": 0.04604003205895424, | |
| "rewards/rejected": -0.23832373321056366, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.3952569169960474, | |
| "grad_norm": 12.375, | |
| "learning_rate": 7.573686659162293e-07, | |
| "logits/chosen": -2.959407329559326, | |
| "logits/rejected": -2.937516212463379, | |
| "logps/chosen": -246.8042449951172, | |
| "logps/rejected": -242.3812713623047, | |
| "loss": 0.6706, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.20731408894062042, | |
| "rewards/margins": 0.061324410140514374, | |
| "rewards/rejected": -0.2686384916305542, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.40654997176736307, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 7.402540567492336e-07, | |
| "logits/chosen": -2.9905340671539307, | |
| "logits/rejected": -2.95076060295105, | |
| "logps/chosen": -243.1343536376953, | |
| "logps/rejected": -226.71694946289062, | |
| "loss": 0.6703, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.22629483044147491, | |
| "rewards/margins": 0.06099146604537964, | |
| "rewards/rejected": -0.28728628158569336, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.4178430265386787, | |
| "grad_norm": 10.125, | |
| "learning_rate": 7.227652612113213e-07, | |
| "logits/chosen": -2.8893442153930664, | |
| "logits/rejected": -2.852905035018921, | |
| "logps/chosen": -266.652099609375, | |
| "logps/rejected": -236.76708984375, | |
| "loss": 0.6789, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.2726292014122009, | |
| "rewards/margins": 0.046290718019008636, | |
| "rewards/rejected": -0.31891992688179016, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.42913608130999437, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 7.049295174229328e-07, | |
| "logits/chosen": -2.9600396156311035, | |
| "logits/rejected": -2.916116237640381, | |
| "logps/chosen": -279.30035400390625, | |
| "logps/rejected": -240.49032592773438, | |
| "loss": 0.6625, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.2531939148902893, | |
| "rewards/margins": 0.0769796222448349, | |
| "rewards/rejected": -0.3301735520362854, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.44042913608131, | |
| "grad_norm": 11.8125, | |
| "learning_rate": 6.867746038629462e-07, | |
| "logits/chosen": -2.9024429321289062, | |
| "logits/rejected": -2.8742191791534424, | |
| "logps/chosen": -242.5234375, | |
| "logps/rejected": -207.02584838867188, | |
| "loss": 0.6728, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.26158207654953003, | |
| "rewards/margins": 0.05569303780794144, | |
| "rewards/rejected": -0.31727513670921326, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.45172219085262566, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 6.683287961047741e-07, | |
| "logits/chosen": -2.9038491249084473, | |
| "logits/rejected": -2.8683886528015137, | |
| "logps/chosen": -287.3917541503906, | |
| "logps/rejected": -234.09072875976562, | |
| "loss": 0.6692, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.23289862275123596, | |
| "rewards/margins": 0.06903888285160065, | |
| "rewards/rejected": -0.3019375205039978, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.46301524562394125, | |
| "grad_norm": 13.125, | |
| "learning_rate": 6.496208227782556e-07, | |
| "logits/chosen": -2.8886725902557373, | |
| "logits/rejected": -2.8595938682556152, | |
| "logps/chosen": -253.18283081054688, | |
| "logps/rejected": -246.322509765625, | |
| "loss": 0.6765, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.27256402373313904, | |
| "rewards/margins": 0.05145202949643135, | |
| "rewards/rejected": -0.3240160644054413, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4743083003952569, | |
| "grad_norm": 10.375, | |
| "learning_rate": 6.306798208259297e-07, | |
| "logits/chosen": -2.9606752395629883, | |
| "logits/rejected": -2.9306626319885254, | |
| "logps/chosen": -253.44491577148438, | |
| "logps/rejected": -240.0795135498047, | |
| "loss": 0.6778, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.2641372084617615, | |
| "rewards/margins": 0.0465705581009388, | |
| "rewards/rejected": -0.31070777773857117, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.48560135516657255, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 6.115352901233778e-07, | |
| "logits/chosen": -2.9563944339752197, | |
| "logits/rejected": -2.9226722717285156, | |
| "logps/chosen": -275.2793273925781, | |
| "logps/rejected": -250.719482421875, | |
| "loss": 0.6735, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.27404916286468506, | |
| "rewards/margins": 0.056207090616226196, | |
| "rewards/rejected": -0.33025628328323364, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4968944099378882, | |
| "grad_norm": 8.9375, | |
| "learning_rate": 5.922170475343124e-07, | |
| "logits/chosen": -2.9221410751342773, | |
| "logits/rejected": -2.891664505004883, | |
| "logps/chosen": -252.1197509765625, | |
| "logps/rejected": -239.79946899414062, | |
| "loss": 0.6643, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.2773807644844055, | |
| "rewards/margins": 0.08373435586690903, | |
| "rewards/rejected": -0.36111515760421753, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.5081874647092038, | |
| "grad_norm": 8.0, | |
| "learning_rate": 5.727551804719693e-07, | |
| "logits/chosen": -2.9440019130706787, | |
| "logits/rejected": -2.9178504943847656, | |
| "logps/chosen": -240.7253875732422, | |
| "logps/rejected": -245.9892120361328, | |
| "loss": 0.6604, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.2647743225097656, | |
| "rewards/margins": 0.08178414404392242, | |
| "rewards/rejected": -0.34655848145484924, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.5194805194805194, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 5.531800000391275e-07, | |
| "logits/chosen": -2.9381301403045654, | |
| "logits/rejected": -2.899775743484497, | |
| "logps/chosen": -253.4866485595703, | |
| "logps/rejected": -227.26168823242188, | |
| "loss": 0.6578, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.2723321318626404, | |
| "rewards/margins": 0.09097979962825775, | |
| "rewards/rejected": -0.3633119463920593, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.5307735742518351, | |
| "grad_norm": 9.625, | |
| "learning_rate": 5.335219938197445e-07, | |
| "logits/chosen": -2.9399991035461426, | |
| "logits/rejected": -2.903505802154541, | |
| "logps/chosen": -246.1412353515625, | |
| "logps/rejected": -221.0483856201172, | |
| "loss": 0.6679, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.24423566460609436, | |
| "rewards/margins": 0.06885796785354614, | |
| "rewards/rejected": -0.3130936622619629, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5420666290231507, | |
| "grad_norm": 8.8125, | |
| "learning_rate": 5.138117783957261e-07, | |
| "logits/chosen": -2.9549431800842285, | |
| "logits/rejected": -2.9279115200042725, | |
| "logps/chosen": -258.9583435058594, | |
| "logps/rejected": -230.93106079101562, | |
| "loss": 0.6651, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.28696444630622864, | |
| "rewards/margins": 0.07692141085863113, | |
| "rewards/rejected": -0.3638857901096344, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5533596837944664, | |
| "grad_norm": 11.25, | |
| "learning_rate": 4.940800516627885e-07, | |
| "logits/chosen": -2.9042859077453613, | |
| "logits/rejected": -2.8782927989959717, | |
| "logps/chosen": -259.2146911621094, | |
| "logps/rejected": -238.78005981445312, | |
| "loss": 0.6506, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.25122493505477905, | |
| "rewards/margins": 0.10116108506917953, | |
| "rewards/rejected": -0.3523860573768616, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.564652738565782, | |
| "grad_norm": 8.125, | |
| "learning_rate": 4.743575450196772e-07, | |
| "logits/chosen": -2.9411609172821045, | |
| "logits/rejected": -2.9064881801605225, | |
| "logps/chosen": -245.3742218017578, | |
| "logps/rejected": -242.74740600585938, | |
| "loss": 0.6671, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.29283052682876587, | |
| "rewards/margins": 0.07133854180574417, | |
| "rewards/rejected": -0.36416909098625183, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5759457933370977, | |
| "grad_norm": 9.0, | |
| "learning_rate": 4.5467497550520505e-07, | |
| "logits/chosen": -2.894622325897217, | |
| "logits/rejected": -2.8733580112457275, | |
| "logps/chosen": -222.1239471435547, | |
| "logps/rejected": -213.3123779296875, | |
| "loss": 0.6705, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.2835238575935364, | |
| "rewards/margins": 0.060517072677612305, | |
| "rewards/rejected": -0.3440409302711487, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5872388481084133, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 4.350629979576569e-07, | |
| "logits/chosen": -2.9461731910705566, | |
| "logits/rejected": -2.8937017917633057, | |
| "logps/chosen": -232.8549346923828, | |
| "logps/rejected": -213.99911499023438, | |
| "loss": 0.6405, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.27412381768226624, | |
| "rewards/margins": 0.13067342340946198, | |
| "rewards/rejected": -0.404797226190567, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.598531902879729, | |
| "grad_norm": 10.0, | |
| "learning_rate": 4.155521572710684e-07, | |
| "logits/chosen": -2.9578232765197754, | |
| "logits/rejected": -2.920867919921875, | |
| "logps/chosen": -230.7373809814453, | |
| "logps/rejected": -210.01296997070312, | |
| "loss": 0.6765, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.27560561895370483, | |
| "rewards/margins": 0.05005854368209839, | |
| "rewards/rejected": -0.325664222240448, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.6098249576510446, | |
| "grad_norm": 10.375, | |
| "learning_rate": 3.9617284082273836e-07, | |
| "logits/chosen": -2.9541661739349365, | |
| "logits/rejected": -2.905620574951172, | |
| "logps/chosen": -233.76467895507812, | |
| "logps/rejected": -225.522705078125, | |
| "loss": 0.6498, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.2639801800251007, | |
| "rewards/margins": 0.10957106202840805, | |
| "rewards/rejected": -0.37355121970176697, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.6211180124223602, | |
| "grad_norm": 10.6875, | |
| "learning_rate": 3.7695523114606836e-07, | |
| "logits/chosen": -2.936779499053955, | |
| "logits/rejected": -2.9056625366210938, | |
| "logps/chosen": -244.0745086669922, | |
| "logps/rejected": -241.7302703857422, | |
| "loss": 0.6858, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.3084772527217865, | |
| "rewards/margins": 0.04285974055528641, | |
| "rewards/rejected": -0.3513370156288147, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.6324110671936759, | |
| "grad_norm": 10.75, | |
| "learning_rate": 3.579292589224374e-07, | |
| "logits/chosen": -2.9537434577941895, | |
| "logits/rejected": -2.9143412113189697, | |
| "logps/chosen": -263.7715759277344, | |
| "logps/rejected": -249.03793334960938, | |
| "loss": 0.6506, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.28075623512268066, | |
| "rewards/margins": 0.10373449325561523, | |
| "rewards/rejected": -0.3844907283782959, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6437041219649915, | |
| "grad_norm": 9.4375, | |
| "learning_rate": 3.391245563653276e-07, | |
| "logits/chosen": -2.957152843475342, | |
| "logits/rejected": -2.9082839488983154, | |
| "logps/chosen": -252.3236846923828, | |
| "logps/rejected": -221.5681610107422, | |
| "loss": 0.6473, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.2503497004508972, | |
| "rewards/margins": 0.11741576343774796, | |
| "rewards/rejected": -0.36776548624038696, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6549971767363072, | |
| "grad_norm": 11.25, | |
| "learning_rate": 3.20570411069301e-07, | |
| "logits/chosen": -2.947129487991333, | |
| "logits/rejected": -2.910438060760498, | |
| "logps/chosen": -247.65579223632812, | |
| "logps/rejected": -219.4207763671875, | |
| "loss": 0.6602, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.34559428691864014, | |
| "rewards/margins": 0.08174686133861542, | |
| "rewards/rejected": -0.4273412227630615, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6662902315076228, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 3.0229572039570826e-07, | |
| "logits/chosen": -2.9612464904785156, | |
| "logits/rejected": -2.912680149078369, | |
| "logps/chosen": -237.4712677001953, | |
| "logps/rejected": -232.04373168945312, | |
| "loss": 0.653, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.2624071538448334, | |
| "rewards/margins": 0.10756198316812515, | |
| "rewards/rejected": -0.3699691891670227, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6775832862789385, | |
| "grad_norm": 11.125, | |
| "learning_rate": 2.8432894646616886e-07, | |
| "logits/chosen": -2.9047579765319824, | |
| "logits/rejected": -2.882847547531128, | |
| "logps/chosen": -231.67147827148438, | |
| "logps/rejected": -236.0974884033203, | |
| "loss": 0.6688, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.3304700553417206, | |
| "rewards/margins": 0.07191035896539688, | |
| "rewards/rejected": -0.40238040685653687, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6888763410502541, | |
| "grad_norm": 7.4375, | |
| "learning_rate": 2.6669807183392105e-07, | |
| "logits/chosen": -2.92600679397583, | |
| "logits/rejected": -2.8905367851257324, | |
| "logps/chosen": -236.2364044189453, | |
| "logps/rejected": -206.38174438476562, | |
| "loss": 0.6684, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.28464430570602417, | |
| "rewards/margins": 0.07026463001966476, | |
| "rewards/rejected": -0.35490894317626953, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.7001693958215698, | |
| "grad_norm": 10.8125, | |
| "learning_rate": 2.494305559020822e-07, | |
| "logits/chosen": -2.9433751106262207, | |
| "logits/rejected": -2.907177686691284, | |
| "logps/chosen": -240.6074676513672, | |
| "logps/rejected": -226.85989379882812, | |
| "loss": 0.6727, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.3261820673942566, | |
| "rewards/margins": 0.06915761530399323, | |
| "rewards/rejected": -0.3953397274017334, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.7114624505928854, | |
| "grad_norm": 11.0, | |
| "learning_rate": 2.3255329215669184e-07, | |
| "logits/chosen": -2.950742244720459, | |
| "logits/rejected": -2.920048952102661, | |
| "logps/chosen": -244.8520965576172, | |
| "logps/rejected": -253.5353546142578, | |
| "loss": 0.6715, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.2790936529636383, | |
| "rewards/margins": 0.06808016449213028, | |
| "rewards/rejected": -0.3471737802028656, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.722755505364201, | |
| "grad_norm": 11.0625, | |
| "learning_rate": 2.1609256628115312e-07, | |
| "logits/chosen": -2.9226107597351074, | |
| "logits/rejected": -2.8994510173797607, | |
| "logps/chosen": -228.3145751953125, | |
| "logps/rejected": -198.56753540039062, | |
| "loss": 0.6619, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.284686803817749, | |
| "rewards/margins": 0.08213174343109131, | |
| "rewards/rejected": -0.3668185770511627, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.7340485601355167, | |
| "grad_norm": 13.5625, | |
| "learning_rate": 2.000740152172986e-07, | |
| "logits/chosen": -2.9056789875030518, | |
| "logits/rejected": -2.8870625495910645, | |
| "logps/chosen": -245.57168579101562, | |
| "logps/rejected": -241.86618041992188, | |
| "loss": 0.6735, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.31977060437202454, | |
| "rewards/margins": 0.060752321034669876, | |
| "rewards/rejected": -0.3805229663848877, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.7453416149068323, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.8452258723684995e-07, | |
| "logits/chosen": -2.9322152137756348, | |
| "logits/rejected": -2.8824894428253174, | |
| "logps/chosen": -228.6182098388672, | |
| "logps/rejected": -198.11158752441406, | |
| "loss": 0.6541, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.30365481972694397, | |
| "rewards/margins": 0.10447897762060165, | |
| "rewards/rejected": -0.4081338346004486, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.756634669678148, | |
| "grad_norm": 9.8125, | |
| "learning_rate": 1.6946250308545124e-07, | |
| "logits/chosen": -2.9027180671691895, | |
| "logits/rejected": -2.8488707542419434, | |
| "logps/chosen": -238.1764678955078, | |
| "logps/rejected": -241.59326171875, | |
| "loss": 0.647, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.2668387293815613, | |
| "rewards/margins": 0.11640632152557373, | |
| "rewards/rejected": -0.383245050907135, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7679277244494636, | |
| "grad_norm": 11.3125, | |
| "learning_rate": 1.54917218259799e-07, | |
| "logits/chosen": -2.941662549972534, | |
| "logits/rejected": -2.894787073135376, | |
| "logps/chosen": -256.3278503417969, | |
| "logps/rejected": -216.21945190429688, | |
| "loss": 0.667, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.3198787271976471, | |
| "rewards/margins": 0.07660157978534698, | |
| "rewards/rejected": -0.3964803218841553, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.7792207792207793, | |
| "grad_norm": 10.375, | |
| "learning_rate": 1.409093864766146e-07, | |
| "logits/chosen": -2.9099855422973633, | |
| "logits/rejected": -2.8785247802734375, | |
| "logps/chosen": -238.84249877929688, | |
| "logps/rejected": -240.7554473876953, | |
| "loss": 0.6672, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.3247315585613251, | |
| "rewards/margins": 0.06632229685783386, | |
| "rewards/rejected": -0.3910538852214813, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7905138339920948, | |
| "grad_norm": 9.0625, | |
| "learning_rate": 1.2746082439036114e-07, | |
| "logits/chosen": -2.9594428539276123, | |
| "logits/rejected": -2.939582347869873, | |
| "logps/chosen": -244.93490600585938, | |
| "logps/rejected": -220.2227020263672, | |
| "loss": 0.6717, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.2836626172065735, | |
| "rewards/margins": 0.060529135167598724, | |
| "rewards/rejected": -0.3441917896270752, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.8018068887634106, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 1.1459247761464907e-07, | |
| "logits/chosen": -2.9392457008361816, | |
| "logits/rejected": -2.9211068153381348, | |
| "logps/chosen": -238.0613555908203, | |
| "logps/rejected": -216.3132781982422, | |
| "loss": 0.6608, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.28733736276626587, | |
| "rewards/margins": 0.0887908786535263, | |
| "rewards/rejected": -0.37612825632095337, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.8130999435347261, | |
| "grad_norm": 8.125, | |
| "learning_rate": 1.0232438810025728e-07, | |
| "logits/chosen": -2.929105043411255, | |
| "logits/rejected": -2.8989386558532715, | |
| "logps/chosen": -238.7790985107422, | |
| "logps/rejected": -196.06771850585938, | |
| "loss": 0.6614, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.2947736084461212, | |
| "rewards/margins": 0.08014187216758728, | |
| "rewards/rejected": -0.3749154806137085, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.8243929983060417, | |
| "grad_norm": 8.25, | |
| "learning_rate": 9.067566292057083e-08, | |
| "logits/chosen": -2.927602767944336, | |
| "logits/rejected": -2.899686098098755, | |
| "logps/chosen": -226.2371826171875, | |
| "logps/rejected": -250.6548309326172, | |
| "loss": 0.6538, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.2974669933319092, | |
| "rewards/margins": 0.09935982525348663, | |
| "rewards/rejected": -0.3968268036842346, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.8356860530773574, | |
| "grad_norm": 8.5625, | |
| "learning_rate": 7.966444451305727e-08, | |
| "logits/chosen": -2.9271061420440674, | |
| "logits/rejected": -2.8855221271514893, | |
| "logps/chosen": -230.8705596923828, | |
| "logps/rejected": -248.7850341796875, | |
| "loss": 0.655, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.32860639691352844, | |
| "rewards/margins": 0.10086989402770996, | |
| "rewards/rejected": -0.429476261138916, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.846979107848673, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 6.930788242312252e-08, | |
| "logits/chosen": -2.9216761589050293, | |
| "logits/rejected": -2.8618099689483643, | |
| "logps/chosen": -219.88839721679688, | |
| "logps/rejected": -189.08688354492188, | |
| "loss": 0.644, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.31415051221847534, | |
| "rewards/margins": 0.1270967572927475, | |
| "rewards/rejected": -0.44124728441238403, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8582721626199887, | |
| "grad_norm": 10.4375, | |
| "learning_rate": 5.96221065943609e-08, | |
| "logits/chosen": -2.9630675315856934, | |
| "logits/rejected": -2.941108226776123, | |
| "logps/chosen": -223.5160369873047, | |
| "logps/rejected": -232.03451538085938, | |
| "loss": 0.6592, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.30072498321533203, | |
| "rewards/margins": 0.09193964302539825, | |
| "rewards/rejected": -0.3926646113395691, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 9.6875, | |
| "learning_rate": 5.062220224679276e-08, | |
| "logits/chosen": -2.936577320098877, | |
| "logits/rejected": -2.9232354164123535, | |
| "logps/chosen": -229.6407012939453, | |
| "logps/rejected": -229.1260223388672, | |
| "loss": 0.6702, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.27426886558532715, | |
| "rewards/margins": 0.06261973083019257, | |
| "rewards/rejected": -0.3368886113166809, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.88085827216262, | |
| "grad_norm": 10.375, | |
| "learning_rate": 4.2322186382220295e-08, | |
| "logits/chosen": -2.913611650466919, | |
| "logits/rejected": -2.8943240642547607, | |
| "logps/chosen": -227.52316284179688, | |
| "logps/rejected": -237.62997436523438, | |
| "loss": 0.6853, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.3313713073730469, | |
| "rewards/margins": 0.03858897089958191, | |
| "rewards/rejected": -0.3699602782726288, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8921513269339356, | |
| "grad_norm": 8.1875, | |
| "learning_rate": 3.4734985953290774e-08, | |
| "logits/chosen": -2.9326188564300537, | |
| "logits/rejected": -2.8966469764709473, | |
| "logps/chosen": -255.98684692382812, | |
| "logps/rejected": -265.226806640625, | |
| "loss": 0.644, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.2588121294975281, | |
| "rewards/margins": 0.12674114108085632, | |
| "rewards/rejected": -0.3855533003807068, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.9034443817052513, | |
| "grad_norm": 9.5625, | |
| "learning_rate": 2.7872417730269327e-08, | |
| "logits/chosen": -2.933467388153076, | |
| "logits/rejected": -2.898324728012085, | |
| "logps/chosen": -246.2470245361328, | |
| "logps/rejected": -263.4474792480469, | |
| "loss": 0.6672, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.29884645342826843, | |
| "rewards/margins": 0.07485932111740112, | |
| "rewards/rejected": -0.37370574474334717, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.9147374364765669, | |
| "grad_norm": 8.375, | |
| "learning_rate": 2.1745169896878412e-08, | |
| "logits/chosen": -2.971459150314331, | |
| "logits/rejected": -2.9422390460968018, | |
| "logps/chosen": -257.4516296386719, | |
| "logps/rejected": -244.69961547851562, | |
| "loss": 0.6585, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.2682439386844635, | |
| "rewards/margins": 0.09004910290241241, | |
| "rewards/rejected": -0.3582930564880371, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.9260304912478825, | |
| "grad_norm": 10.5625, | |
| "learning_rate": 1.6362785403865488e-08, | |
| "logits/chosen": -2.9577174186706543, | |
| "logits/rejected": -2.9316866397857666, | |
| "logps/chosen": -236.38302612304688, | |
| "logps/rejected": -231.70822143554688, | |
| "loss": 0.6696, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.3118794858455658, | |
| "rewards/margins": 0.07028525322675705, | |
| "rewards/rejected": -0.38216471672058105, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.9373235460191982, | |
| "grad_norm": 12.0, | |
| "learning_rate": 1.1733647106228373e-08, | |
| "logits/chosen": -2.965700149536133, | |
| "logits/rejected": -2.935534715652466, | |
| "logps/chosen": -241.2149658203125, | |
| "logps/rejected": -219.77786254882812, | |
| "loss": 0.6709, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.3273731470108032, | |
| "rewards/margins": 0.060812730342149734, | |
| "rewards/rejected": -0.38818588852882385, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.9486166007905138, | |
| "grad_norm": 11.625, | |
| "learning_rate": 7.864964707243071e-09, | |
| "logits/chosen": -2.9139742851257324, | |
| "logits/rejected": -2.8788352012634277, | |
| "logps/chosen": -244.0455322265625, | |
| "logps/rejected": -233.3373565673828, | |
| "loss": 0.6784, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.3300473690032959, | |
| "rewards/margins": 0.05264363810420036, | |
| "rewards/rejected": -0.382690966129303, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.9599096555618295, | |
| "grad_norm": 9.75, | |
| "learning_rate": 4.762763529631342e-09, | |
| "logits/chosen": -2.956047773361206, | |
| "logits/rejected": -2.936790943145752, | |
| "logps/chosen": -240.4857177734375, | |
| "logps/rejected": -235.99301147460938, | |
| "loss": 0.6877, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.31692594289779663, | |
| "rewards/margins": 0.042497653514146805, | |
| "rewards/rejected": -0.35942360758781433, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9712027103331451, | |
| "grad_norm": 10.1875, | |
| "learning_rate": 2.431875131354011e-09, | |
| "logits/chosen": -2.970977544784546, | |
| "logits/rejected": -2.9481375217437744, | |
| "logps/chosen": -258.04437255859375, | |
| "logps/rejected": -255.94552612304688, | |
| "loss": 0.664, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.3285280168056488, | |
| "rewards/margins": 0.08176346123218536, | |
| "rewards/rejected": -0.41029149293899536, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9824957651044608, | |
| "grad_norm": 9.9375, | |
| "learning_rate": 8.759297806469335e-10, | |
| "logits/chosen": -2.9342379570007324, | |
| "logits/rejected": -2.8985533714294434, | |
| "logps/chosen": -260.99530029296875, | |
| "logps/rejected": -243.7850799560547, | |
| "loss": 0.6639, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.28514906764030457, | |
| "rewards/margins": 0.0856957957148552, | |
| "rewards/rejected": -0.37084484100341797, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9937888198757764, | |
| "grad_norm": 9.25, | |
| "learning_rate": 9.735080201922486e-11, | |
| "logits/chosen": -2.933637857437134, | |
| "logits/rejected": -2.88521146774292, | |
| "logps/chosen": -285.87249755859375, | |
| "logps/rejected": -235.1495819091797, | |
| "loss": 0.6537, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.32956260442733765, | |
| "rewards/margins": 0.10394410043954849, | |
| "rewards/rejected": -0.43350666761398315, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9994353472614342, | |
| "step": 885, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6718437324135991, | |
| "train_runtime": 9969.8232, | |
| "train_samples_per_second": 1.421, | |
| "train_steps_per_second": 0.089 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 885, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |