| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9996728819103696, | |
| "eval_steps": 500, | |
| "global_step": 1528, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 3.267973856209151e-08, | |
| "logits/chosen": -1.591799020767212, | |
| "logits/rejected": -1.609328031539917, | |
| "logps/chosen": -327.9853210449219, | |
| "logps/rejected": -375.1213684082031, | |
| "loss": 0.0486, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 3.267973856209151e-07, | |
| "logits/chosen": -1.8375271558761597, | |
| "logits/rejected": -1.9081380367279053, | |
| "logps/chosen": -217.70407104492188, | |
| "logps/rejected": -291.03619384765625, | |
| "loss": 0.0481, | |
| "rewards/accuracies": 0.3888888955116272, | |
| "rewards/chosen": -0.00010180399840464815, | |
| "rewards/margins": -1.7401700461050496e-05, | |
| "rewards/rejected": -8.440230158157647e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 6.535947712418302e-07, | |
| "logits/chosen": -1.8446567058563232, | |
| "logits/rejected": -1.96743643283844, | |
| "logps/chosen": -188.84414672851562, | |
| "logps/rejected": -223.9464569091797, | |
| "loss": 0.0476, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.0001967716234503314, | |
| "rewards/margins": 0.0005254523130133748, | |
| "rewards/rejected": -0.0007222238346002996, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 9.80392156862745e-07, | |
| "logits/chosen": -2.0779905319213867, | |
| "logits/rejected": -1.7060468196868896, | |
| "logps/chosen": -239.03250122070312, | |
| "logps/rejected": -208.9864501953125, | |
| "loss": 0.0504, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.0017802307847887278, | |
| "rewards/margins": 0.001976901665329933, | |
| "rewards/rejected": -0.00375713175162673, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.3071895424836604e-06, | |
| "logits/chosen": -1.7227962017059326, | |
| "logits/rejected": -1.7102130651474, | |
| "logps/chosen": -361.55224609375, | |
| "logps/rejected": -365.0895690917969, | |
| "loss": 0.0376, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.019573379307985306, | |
| "rewards/margins": 0.011974343098700047, | |
| "rewards/rejected": -0.031547725200653076, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.6339869281045753e-06, | |
| "logits/chosen": -1.8763091564178467, | |
| "logits/rejected": -1.6960475444793701, | |
| "logps/chosen": -392.33709716796875, | |
| "logps/rejected": -353.79827880859375, | |
| "loss": 0.0283, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.06343773007392883, | |
| "rewards/margins": 0.023594681173563004, | |
| "rewards/rejected": -0.08703242242336273, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1.96078431372549e-06, | |
| "logits/chosen": -1.6577602624893188, | |
| "logits/rejected": -1.5122312307357788, | |
| "logps/chosen": -384.6730651855469, | |
| "logps/rejected": -499.12677001953125, | |
| "loss": 0.0236, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.1024140864610672, | |
| "rewards/margins": 0.13252298533916473, | |
| "rewards/rejected": -0.23493707180023193, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.2875816993464053e-06, | |
| "logits/chosen": -1.7868068218231201, | |
| "logits/rejected": -1.6683218479156494, | |
| "logps/chosen": -391.08642578125, | |
| "logps/rejected": -484.86737060546875, | |
| "loss": 0.0204, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.11379416286945343, | |
| "rewards/margins": 0.09717617928981781, | |
| "rewards/rejected": -0.21097035706043243, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.6143790849673208e-06, | |
| "logits/chosen": -1.7627947330474854, | |
| "logits/rejected": -1.526816725730896, | |
| "logps/chosen": -385.7118835449219, | |
| "logps/rejected": -395.92108154296875, | |
| "loss": 0.0225, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.10957636684179306, | |
| "rewards/margins": 0.08020367473363876, | |
| "rewards/rejected": -0.18978002667427063, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.9411764705882355e-06, | |
| "logits/chosen": -1.761884331703186, | |
| "logits/rejected": -1.7560245990753174, | |
| "logps/chosen": -301.0171813964844, | |
| "logps/rejected": -439.34906005859375, | |
| "loss": 0.0188, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.062449000775814056, | |
| "rewards/margins": 0.10943855345249176, | |
| "rewards/rejected": -0.17188753187656403, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.2679738562091506e-06, | |
| "logits/chosen": -1.6214962005615234, | |
| "logits/rejected": -1.6456191539764404, | |
| "logps/chosen": -258.74237060546875, | |
| "logps/rejected": -351.51641845703125, | |
| "loss": 0.0173, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.07586181163787842, | |
| "rewards/margins": 0.08772234618663788, | |
| "rewards/rejected": -0.1635841578245163, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.5947712418300657e-06, | |
| "logits/chosen": -1.6627695560455322, | |
| "logits/rejected": -1.5852539539337158, | |
| "logps/chosen": -331.30145263671875, | |
| "logps/rejected": -448.34124755859375, | |
| "loss": 0.0156, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.12424789369106293, | |
| "rewards/margins": 0.0813526064157486, | |
| "rewards/rejected": -0.20560050010681152, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.92156862745098e-06, | |
| "logits/chosen": -1.5744519233703613, | |
| "logits/rejected": -1.479768991470337, | |
| "logps/chosen": -462.4168395996094, | |
| "logps/rejected": -554.2327880859375, | |
| "loss": 0.0216, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.2043522298336029, | |
| "rewards/margins": 0.08921514451503754, | |
| "rewards/rejected": -0.29356735944747925, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.2483660130718954e-06, | |
| "logits/chosen": -1.918543815612793, | |
| "logits/rejected": -1.7244701385498047, | |
| "logps/chosen": -366.99761962890625, | |
| "logps/rejected": -438.06964111328125, | |
| "loss": 0.0303, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.11779004335403442, | |
| "rewards/margins": 0.08299387991428375, | |
| "rewards/rejected": -0.20078392326831818, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.5751633986928105e-06, | |
| "logits/chosen": -1.8474334478378296, | |
| "logits/rejected": -1.51116144657135, | |
| "logps/chosen": -443.89129638671875, | |
| "logps/rejected": -516.485107421875, | |
| "loss": 0.0285, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.1710434854030609, | |
| "rewards/margins": 0.08919176459312439, | |
| "rewards/rejected": -0.2602352499961853, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.901960784313726e-06, | |
| "logits/chosen": -1.946266770362854, | |
| "logits/rejected": -1.6877708435058594, | |
| "logps/chosen": -403.88897705078125, | |
| "logps/rejected": -492.328125, | |
| "loss": 0.0169, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.14448824524879456, | |
| "rewards/margins": 0.11163143813610077, | |
| "rewards/rejected": -0.25611966848373413, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999680264259825e-06, | |
| "logits/chosen": -1.8900115489959717, | |
| "logits/rejected": -1.774910569190979, | |
| "logps/chosen": -317.9911193847656, | |
| "logps/rejected": -480.47930908203125, | |
| "loss": 0.0222, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.12115655094385147, | |
| "rewards/margins": 0.13366180658340454, | |
| "rewards/rejected": -0.2548183798789978, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.998114408534616e-06, | |
| "logits/chosen": -1.8668384552001953, | |
| "logits/rejected": -1.594963788986206, | |
| "logps/chosen": -461.4814453125, | |
| "logps/rejected": -463.83062744140625, | |
| "loss": 0.022, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.13391847908496857, | |
| "rewards/margins": 0.09448060393333435, | |
| "rewards/rejected": -0.22839908301830292, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.995244522215781e-06, | |
| "logits/chosen": -1.4480278491973877, | |
| "logits/rejected": -1.2824418544769287, | |
| "logps/chosen": -398.3277587890625, | |
| "logps/rejected": -441.39776611328125, | |
| "loss": 0.0323, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.13857640326023102, | |
| "rewards/margins": 0.07922663539648056, | |
| "rewards/rejected": -0.21780303120613098, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.9910721034010655e-06, | |
| "logits/chosen": -1.8611648082733154, | |
| "logits/rejected": -1.5049927234649658, | |
| "logps/chosen": -442.27532958984375, | |
| "logps/rejected": -513.6070556640625, | |
| "loss": 0.0206, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.10470763593912125, | |
| "rewards/margins": 0.1342332810163498, | |
| "rewards/rejected": -0.23894090950489044, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.985599330117931e-06, | |
| "logits/chosen": -1.7740875482559204, | |
| "logits/rejected": -1.6385295391082764, | |
| "logps/chosen": -340.8525390625, | |
| "logps/rejected": -373.794677734375, | |
| "loss": 0.021, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.11725501716136932, | |
| "rewards/margins": 0.05946159362792969, | |
| "rewards/rejected": -0.176716610789299, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.978829059186611e-06, | |
| "logits/chosen": -1.6705108880996704, | |
| "logits/rejected": -1.5942919254302979, | |
| "logps/chosen": -390.35748291015625, | |
| "logps/rejected": -478.62713623046875, | |
| "loss": 0.0167, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.1433432400226593, | |
| "rewards/margins": 0.08838480710983276, | |
| "rewards/rejected": -0.23172803223133087, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.97076482472884e-06, | |
| "logits/chosen": -1.8241045475006104, | |
| "logits/rejected": -1.6329164505004883, | |
| "logps/chosen": -475.8793029785156, | |
| "logps/rejected": -482.19854736328125, | |
| "loss": 0.0205, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.14704802632331848, | |
| "rewards/margins": 0.05757030099630356, | |
| "rewards/rejected": -0.20461833477020264, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.961410836323014e-06, | |
| "logits/chosen": -1.5026575326919556, | |
| "logits/rejected": -1.4128906726837158, | |
| "logps/chosen": -374.00555419921875, | |
| "logps/rejected": -541.5623168945312, | |
| "loss": 0.023, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.15363118052482605, | |
| "rewards/margins": 0.13868795335292816, | |
| "rewards/rejected": -0.292319118976593, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.950771976806769e-06, | |
| "logits/chosen": -1.7081775665283203, | |
| "logits/rejected": -1.6031742095947266, | |
| "logps/chosen": -385.6658935546875, | |
| "logps/rejected": -442.20697021484375, | |
| "loss": 0.0199, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.145090714097023, | |
| "rewards/margins": 0.05894991010427475, | |
| "rewards/rejected": -0.20404061675071716, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.938853799728112e-06, | |
| "logits/chosen": -1.7450878620147705, | |
| "logits/rejected": -1.6923738718032837, | |
| "logps/chosen": -437.3587341308594, | |
| "logps/rejected": -505.5862731933594, | |
| "loss": 0.0166, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.16524454951286316, | |
| "rewards/margins": 0.10024659335613251, | |
| "rewards/rejected": -0.2654911279678345, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.925662526446431e-06, | |
| "logits/chosen": -1.5717720985412598, | |
| "logits/rejected": -1.779036521911621, | |
| "logps/chosen": -432.2626037597656, | |
| "logps/rejected": -544.822509765625, | |
| "loss": 0.0236, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.1972414255142212, | |
| "rewards/margins": 0.06645989418029785, | |
| "rewards/rejected": -0.26370128989219666, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.911205042884912e-06, | |
| "logits/chosen": -1.7134885787963867, | |
| "logits/rejected": -1.5530506372451782, | |
| "logps/chosen": -467.70001220703125, | |
| "logps/rejected": -521.0452880859375, | |
| "loss": 0.0153, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.16030625998973846, | |
| "rewards/margins": 0.11603019386529922, | |
| "rewards/rejected": -0.2763364017009735, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.895488895936047e-06, | |
| "logits/chosen": -1.7987966537475586, | |
| "logits/rejected": -1.7680168151855469, | |
| "logps/chosen": -449.7742614746094, | |
| "logps/rejected": -502.37359619140625, | |
| "loss": 0.0212, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.11571071296930313, | |
| "rewards/margins": 0.0691571980714798, | |
| "rewards/rejected": -0.18486790359020233, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.8785222895221075e-06, | |
| "logits/chosen": -1.8355566263198853, | |
| "logits/rejected": -1.8976913690567017, | |
| "logps/chosen": -426.9527282714844, | |
| "logps/rejected": -486.2103576660156, | |
| "loss": 0.0153, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.11018826067447662, | |
| "rewards/margins": 0.06115338206291199, | |
| "rewards/rejected": -0.17134162783622742, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.860314080312651e-06, | |
| "logits/chosen": -1.716607689857483, | |
| "logits/rejected": -1.781498908996582, | |
| "logps/chosen": -279.833740234375, | |
| "logps/rejected": -413.525146484375, | |
| "loss": 0.0228, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.09433625638484955, | |
| "rewards/margins": 0.10549966245889664, | |
| "rewards/rejected": -0.19983592629432678, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.840873773101287e-06, | |
| "logits/chosen": -1.9624496698379517, | |
| "logits/rejected": -1.6330894231796265, | |
| "logps/chosen": -339.2301940917969, | |
| "logps/rejected": -369.59686279296875, | |
| "loss": 0.0292, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.06697295606136322, | |
| "rewards/margins": 0.07780499756336212, | |
| "rewards/rejected": -0.14477796852588654, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.820211515844116e-06, | |
| "logits/chosen": -1.9243313074111938, | |
| "logits/rejected": -1.6555219888687134, | |
| "logps/chosen": -368.7879943847656, | |
| "logps/rejected": -404.6373596191406, | |
| "loss": 0.0257, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.10985688120126724, | |
| "rewards/margins": 0.07864276319742203, | |
| "rewards/rejected": -0.18849964439868927, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.798338094362439e-06, | |
| "logits/chosen": -1.7449579238891602, | |
| "logits/rejected": -1.4674043655395508, | |
| "logps/chosen": -423.55255126953125, | |
| "logps/rejected": -456.8785095214844, | |
| "loss": 0.0183, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.12963487207889557, | |
| "rewards/margins": 0.09587951004505157, | |
| "rewards/rejected": -0.22551438212394714, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.775264926712489e-06, | |
| "logits/chosen": -1.6274173259735107, | |
| "logits/rejected": -1.6101815700531006, | |
| "logps/chosen": -323.04388427734375, | |
| "logps/rejected": -481.54949951171875, | |
| "loss": 0.0308, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.1510605365037918, | |
| "rewards/margins": 0.11264171451330185, | |
| "rewards/rejected": -0.26370224356651306, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.751004057225147e-06, | |
| "logits/chosen": -1.719301462173462, | |
| "logits/rejected": -1.6450971364974976, | |
| "logps/chosen": -572.3885498046875, | |
| "logps/rejected": -655.76953125, | |
| "loss": 0.0197, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.21081407368183136, | |
| "rewards/margins": 0.08522544801235199, | |
| "rewards/rejected": -0.29603949189186096, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.725568150218719e-06, | |
| "logits/chosen": -1.7547239065170288, | |
| "logits/rejected": -1.3802999258041382, | |
| "logps/chosen": -414.50445556640625, | |
| "logps/rejected": -463.7474670410156, | |
| "loss": 0.0225, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.11397633701562881, | |
| "rewards/margins": 0.10422484576702118, | |
| "rewards/rejected": -0.21820120513439178, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.6989704833880936e-06, | |
| "logits/chosen": -1.653165578842163, | |
| "logits/rejected": -1.7715513706207275, | |
| "logps/chosen": -407.423583984375, | |
| "logps/rejected": -510.4872131347656, | |
| "loss": 0.0211, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.1670456975698471, | |
| "rewards/margins": 0.0627368614077568, | |
| "rewards/rejected": -0.22978255152702332, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.671224940873704e-06, | |
| "logits/chosen": -1.8084532022476196, | |
| "logits/rejected": -1.6008380651474, | |
| "logps/chosen": -411.17547607421875, | |
| "logps/rejected": -478.0519104003906, | |
| "loss": 0.0215, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.1790023148059845, | |
| "rewards/margins": 0.0856521874666214, | |
| "rewards/rejected": -0.2646545171737671, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.642346006013925e-06, | |
| "logits/chosen": -1.7398563623428345, | |
| "logits/rejected": -1.7137126922607422, | |
| "logps/chosen": -489.26483154296875, | |
| "logps/rejected": -501.8857421875, | |
| "loss": 0.0207, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.15890449285507202, | |
| "rewards/margins": 0.04155914857983589, | |
| "rewards/rejected": -0.20046362280845642, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.612348753784682e-06, | |
| "logits/chosen": -1.628105878829956, | |
| "logits/rejected": -1.6889688968658447, | |
| "logps/chosen": -348.6719055175781, | |
| "logps/rejected": -442.29046630859375, | |
| "loss": 0.0213, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.1342150866985321, | |
| "rewards/margins": 0.07590731233358383, | |
| "rewards/rejected": -0.21012239158153534, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5812488429302245e-06, | |
| "logits/chosen": -1.8471730947494507, | |
| "logits/rejected": -1.5871822834014893, | |
| "logps/chosen": -460.2723693847656, | |
| "logps/rejected": -453.3777770996094, | |
| "loss": 0.0122, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.08317364007234573, | |
| "rewards/margins": 0.09449958801269531, | |
| "rewards/rejected": -0.17767322063446045, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.54906250778917e-06, | |
| "logits/chosen": -2.0099849700927734, | |
| "logits/rejected": -1.8552219867706299, | |
| "logps/chosen": -278.762451171875, | |
| "logps/rejected": -330.7545166015625, | |
| "loss": 0.0158, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.054166488349437714, | |
| "rewards/margins": 0.0891014039516449, | |
| "rewards/rejected": -0.1432678997516632, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.515806549820084e-06, | |
| "logits/chosen": -1.8863036632537842, | |
| "logits/rejected": -1.7360244989395142, | |
| "logps/chosen": -385.07623291015625, | |
| "logps/rejected": -412.04766845703125, | |
| "loss": 0.025, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.10509233176708221, | |
| "rewards/margins": 0.06545521318912506, | |
| "rewards/rejected": -0.17054755985736847, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.48149832883101e-06, | |
| "logits/chosen": -1.9307434558868408, | |
| "logits/rejected": -1.6779987812042236, | |
| "logps/chosen": -407.74383544921875, | |
| "logps/rejected": -515.69873046875, | |
| "loss": 0.0205, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.10622302442789078, | |
| "rewards/margins": 0.12537193298339844, | |
| "rewards/rejected": -0.23159494996070862, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.446155753917559e-06, | |
| "logits/chosen": -1.9925124645233154, | |
| "logits/rejected": -1.9259634017944336, | |
| "logps/chosen": -374.60223388671875, | |
| "logps/rejected": -428.49969482421875, | |
| "loss": 0.0217, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.12692932784557343, | |
| "rewards/margins": 0.06185678765177727, | |
| "rewards/rejected": -0.18878613412380219, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.409797274114245e-06, | |
| "logits/chosen": -2.01132869720459, | |
| "logits/rejected": -1.7584896087646484, | |
| "logps/chosen": -459.43756103515625, | |
| "logps/rejected": -480.87799072265625, | |
| "loss": 0.0143, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.12262854725122452, | |
| "rewards/margins": 0.06466259062290192, | |
| "rewards/rejected": -0.18729116022586823, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.372441868763981e-06, | |
| "logits/chosen": -1.6305385828018188, | |
| "logits/rejected": -1.368896245956421, | |
| "logps/chosen": -360.996826171875, | |
| "logps/rejected": -434.44097900390625, | |
| "loss": 0.0273, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.1023339256644249, | |
| "rewards/margins": 0.1018548384308815, | |
| "rewards/rejected": -0.2041887789964676, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.334109037610757e-06, | |
| "logits/chosen": -1.9129701852798462, | |
| "logits/rejected": -1.6556217670440674, | |
| "logps/chosen": -380.2204284667969, | |
| "logps/rejected": -451.8375549316406, | |
| "loss": 0.0136, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.09523477405309677, | |
| "rewards/margins": 0.0824359580874443, | |
| "rewards/rejected": -0.17767074704170227, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.294818790620644e-06, | |
| "logits/chosen": -1.6334559917449951, | |
| "logits/rejected": -1.5937422513961792, | |
| "logps/chosen": -355.0863342285156, | |
| "logps/rejected": -435.58770751953125, | |
| "loss": 0.0269, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.13321994245052338, | |
| "rewards/margins": 0.054788071662187576, | |
| "rewards/rejected": -0.18800799548625946, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.2545916375364835e-06, | |
| "logits/chosen": -1.7967026233673096, | |
| "logits/rejected": -1.6936085224151611, | |
| "logps/chosen": -297.0005798339844, | |
| "logps/rejected": -450.44097900390625, | |
| "loss": 0.0191, | |
| "rewards/accuracies": 0.8500000238418579, | |
| "rewards/chosen": -0.10439346730709076, | |
| "rewards/margins": 0.13930422067642212, | |
| "rewards/rejected": -0.24369768798351288, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.213448577171676e-06, | |
| "logits/chosen": -1.8098952770233154, | |
| "logits/rejected": -1.6606693267822266, | |
| "logps/chosen": -403.1839294433594, | |
| "logps/rejected": -525.0617065429688, | |
| "loss": 0.0235, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.1185646504163742, | |
| "rewards/margins": 0.13588686287403107, | |
| "rewards/rejected": -0.25445154309272766, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.171411086448674e-06, | |
| "logits/chosen": -1.695320725440979, | |
| "logits/rejected": -1.7887874841690063, | |
| "logps/chosen": -334.51434326171875, | |
| "logps/rejected": -463.96844482421875, | |
| "loss": 0.0278, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.10001678764820099, | |
| "rewards/margins": 0.10297022759914398, | |
| "rewards/rejected": -0.20298703014850616, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.128501109187903e-06, | |
| "logits/chosen": -1.88802170753479, | |
| "logits/rejected": -1.5964090824127197, | |
| "logps/chosen": -340.66851806640625, | |
| "logps/rejected": -431.9608459472656, | |
| "loss": 0.0225, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.09433753788471222, | |
| "rewards/margins": 0.09536565840244293, | |
| "rewards/rejected": -0.18970318138599396, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.084741044652956e-06, | |
| "logits/chosen": -1.8472003936767578, | |
| "logits/rejected": -1.7370185852050781, | |
| "logps/chosen": -400.97662353515625, | |
| "logps/rejected": -482.4466247558594, | |
| "loss": 0.0173, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.13318313658237457, | |
| "rewards/margins": 0.09391706436872482, | |
| "rewards/rejected": -0.2271002233028412, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.040153735858041e-06, | |
| "logits/chosen": -1.7910776138305664, | |
| "logits/rejected": -1.4051631689071655, | |
| "logps/chosen": -472.873291015625, | |
| "logps/rejected": -538.3961791992188, | |
| "loss": 0.0215, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.20243683457374573, | |
| "rewards/margins": 0.11980029195547104, | |
| "rewards/rejected": -0.322237104177475, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.9947624576437975e-06, | |
| "logits/chosen": -1.8611959218978882, | |
| "logits/rejected": -1.5981895923614502, | |
| "logps/chosen": -379.29010009765625, | |
| "logps/rejected": -485.96649169921875, | |
| "loss": 0.0163, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.125335693359375, | |
| "rewards/margins": 0.11214844882488251, | |
| "rewards/rejected": -0.2374841421842575, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.948590904527689e-06, | |
| "logits/chosen": -1.879133939743042, | |
| "logits/rejected": -1.824703574180603, | |
| "logps/chosen": -372.97430419921875, | |
| "logps/rejected": -450.3955078125, | |
| "loss": 0.0187, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.11131548881530762, | |
| "rewards/margins": 0.07067877054214478, | |
| "rewards/rejected": -0.1819942742586136, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.901663178335318e-06, | |
| "logits/chosen": -1.7198855876922607, | |
| "logits/rejected": -1.7563533782958984, | |
| "logps/chosen": -334.19659423828125, | |
| "logps/rejected": -416.8275451660156, | |
| "loss": 0.024, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.1272047609090805, | |
| "rewards/margins": 0.06788735836744308, | |
| "rewards/rejected": -0.195092111825943, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.854003775619142e-06, | |
| "logits/chosen": -1.9564396142959595, | |
| "logits/rejected": -1.7008472681045532, | |
| "logps/chosen": -400.6410827636719, | |
| "logps/rejected": -497.25433349609375, | |
| "loss": 0.0167, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.1666775941848755, | |
| "rewards/margins": 0.10049338638782501, | |
| "rewards/rejected": -0.2671709656715393, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.805637574871115e-06, | |
| "logits/chosen": -1.6202869415283203, | |
| "logits/rejected": -1.6643507480621338, | |
| "logps/chosen": -424.1668395996094, | |
| "logps/rejected": -568.9484252929688, | |
| "loss": 0.0161, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.1753002405166626, | |
| "rewards/margins": 0.11582837998867035, | |
| "rewards/rejected": -0.29112863540649414, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7565898235359717e-06, | |
| "logits/chosen": -1.801203966140747, | |
| "logits/rejected": -1.4760253429412842, | |
| "logps/chosen": -466.33258056640625, | |
| "logps/rejected": -551.4314575195312, | |
| "loss": 0.0156, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.1938503533601761, | |
| "rewards/margins": 0.11616852134466171, | |
| "rewards/rejected": -0.3100188672542572, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.7068861248319127e-06, | |
| "logits/chosen": -1.7439998388290405, | |
| "logits/rejected": -1.5612130165100098, | |
| "logps/chosen": -405.24310302734375, | |
| "logps/rejected": -510.33258056640625, | |
| "loss": 0.0156, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.1456567347049713, | |
| "rewards/margins": 0.11440464109182358, | |
| "rewards/rejected": -0.2600614130496979, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.6565524243855695e-06, | |
| "logits/chosen": -1.8692495822906494, | |
| "logits/rejected": -1.547742247581482, | |
| "logps/chosen": -383.50909423828125, | |
| "logps/rejected": -442.6529846191406, | |
| "loss": 0.0148, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.11476878821849823, | |
| "rewards/margins": 0.10948886722326279, | |
| "rewards/rejected": -0.2242576777935028, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.6056149966882325e-06, | |
| "logits/chosen": -1.894252061843872, | |
| "logits/rejected": -1.7816390991210938, | |
| "logps/chosen": -362.1920471191406, | |
| "logps/rejected": -425.6106872558594, | |
| "loss": 0.0233, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.11883090436458588, | |
| "rewards/margins": 0.07852541655302048, | |
| "rewards/rejected": -0.19735632836818695, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.554100431380414e-06, | |
| "logits/chosen": -1.7054016590118408, | |
| "logits/rejected": -1.615083932876587, | |
| "logps/chosen": -408.5704345703125, | |
| "logps/rejected": -545.007568359375, | |
| "loss": 0.0212, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.12973299622535706, | |
| "rewards/margins": 0.11624966561794281, | |
| "rewards/rejected": -0.24598267674446106, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.5020356193718934e-06, | |
| "logits/chosen": -1.8543107509613037, | |
| "logits/rejected": -1.5251867771148682, | |
| "logps/chosen": -361.87066650390625, | |
| "logps/rejected": -412.02947998046875, | |
| "loss": 0.0176, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.10308834165334702, | |
| "rewards/margins": 0.07799768447875977, | |
| "rewards/rejected": -0.18108603358268738, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4494477388045035e-06, | |
| "logits/chosen": -1.6554620265960693, | |
| "logits/rejected": -1.5981006622314453, | |
| "logps/chosen": -466.32623291015625, | |
| "logps/rejected": -572.7677612304688, | |
| "loss": 0.0317, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.16580899059772491, | |
| "rewards/margins": 0.10259596258401871, | |
| "rewards/rejected": -0.26840493083000183, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.3963642408649783e-06, | |
| "logits/chosen": -1.904425024986267, | |
| "logits/rejected": -1.8261890411376953, | |
| "logps/chosen": -458.0430603027344, | |
| "logps/rejected": -514.0013427734375, | |
| "loss": 0.0147, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.12817558646202087, | |
| "rewards/margins": 0.08143512904644012, | |
| "rewards/rejected": -0.2096107006072998, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3428128354552727e-06, | |
| "logits/chosen": -1.8188130855560303, | |
| "logits/rejected": -1.619821548461914, | |
| "logps/chosen": -363.2087097167969, | |
| "logps/rejected": -493.71844482421875, | |
| "loss": 0.0282, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.09969594329595566, | |
| "rewards/margins": 0.11581007391214371, | |
| "rewards/rejected": -0.21550603210926056, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2888214767278246e-06, | |
| "logits/chosen": -1.7933197021484375, | |
| "logits/rejected": -1.7083864212036133, | |
| "logps/chosen": -396.96832275390625, | |
| "logps/rejected": -464.05462646484375, | |
| "loss": 0.0156, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.1241300106048584, | |
| "rewards/margins": 0.06391225755214691, | |
| "rewards/rejected": -0.1880422681570053, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2344183484933247e-06, | |
| "logits/chosen": -1.863983392715454, | |
| "logits/rejected": -1.502881646156311, | |
| "logps/chosen": -404.3202209472656, | |
| "logps/rejected": -506.7728576660156, | |
| "loss": 0.0139, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.1348416805267334, | |
| "rewards/margins": 0.10185597836971283, | |
| "rewards/rejected": -0.23669764399528503, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.179631849508597e-06, | |
| "logits/chosen": -1.732456922531128, | |
| "logits/rejected": -1.5833830833435059, | |
| "logps/chosen": -370.2289123535156, | |
| "logps/rejected": -396.13299560546875, | |
| "loss": 0.0196, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.10813581943511963, | |
| "rewards/margins": 0.049150481820106506, | |
| "rewards/rejected": -0.15728630125522614, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1244905786522796e-06, | |
| "logits/chosen": -1.7448046207427979, | |
| "logits/rejected": -1.7952296733856201, | |
| "logps/chosen": -374.96612548828125, | |
| "logps/rejected": -485.88427734375, | |
| "loss": 0.0221, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.12416452169418335, | |
| "rewards/margins": 0.07095464318990707, | |
| "rewards/rejected": -0.19511918723583221, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.0690233199960393e-06, | |
| "logits/chosen": -1.8050892353057861, | |
| "logits/rejected": -1.51162588596344, | |
| "logps/chosen": -443.24102783203125, | |
| "logps/rejected": -499.46417236328125, | |
| "loss": 0.0134, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.14171595871448517, | |
| "rewards/margins": 0.12594462931156158, | |
| "rewards/rejected": -0.26766055822372437, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0132590277791163e-06, | |
| "logits/chosen": -2.1219112873077393, | |
| "logits/rejected": -1.8001441955566406, | |
| "logps/chosen": -450.737060546875, | |
| "logps/rejected": -471.3848571777344, | |
| "loss": 0.0161, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.16006456315517426, | |
| "rewards/margins": 0.05473906546831131, | |
| "rewards/rejected": -0.21480365097522736, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9572268112940354e-06, | |
| "logits/chosen": -1.9351593255996704, | |
| "logits/rejected": -1.6238371133804321, | |
| "logps/chosen": -431.12701416015625, | |
| "logps/rejected": -458.659423828125, | |
| "loss": 0.0207, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.1289026290178299, | |
| "rewards/margins": 0.08564122021198273, | |
| "rewards/rejected": -0.21454384922981262, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9009559196913882e-06, | |
| "logits/chosen": -1.8791773319244385, | |
| "logits/rejected": -1.7700687646865845, | |
| "logps/chosen": -350.3512268066406, | |
| "logps/rejected": -448.70513916015625, | |
| "loss": 0.0153, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.13599291443824768, | |
| "rewards/margins": 0.08505571633577347, | |
| "rewards/rejected": -0.22104862332344055, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.844475726711595e-06, | |
| "logits/chosen": -1.8688617944717407, | |
| "logits/rejected": -1.6239147186279297, | |
| "logps/chosen": -379.91534423828125, | |
| "logps/rejected": -455.30792236328125, | |
| "loss": 0.0186, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.1507028043270111, | |
| "rewards/margins": 0.09271397441625595, | |
| "rewards/rejected": -0.24341675639152527, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7878157153516446e-06, | |
| "logits/chosen": -1.8980302810668945, | |
| "logits/rejected": -1.4565205574035645, | |
| "logps/chosen": -368.5240478515625, | |
| "logps/rejected": -475.89013671875, | |
| "loss": 0.0203, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.12418484687805176, | |
| "rewards/margins": 0.12127965688705444, | |
| "rewards/rejected": -0.2454645186662674, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.731005462474787e-06, | |
| "logits/chosen": -2.0655064582824707, | |
| "logits/rejected": -1.778184175491333, | |
| "logps/chosen": -362.5508728027344, | |
| "logps/rejected": -415.6459045410156, | |
| "loss": 0.0307, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.11265746504068375, | |
| "rewards/margins": 0.09785433113574982, | |
| "rewards/rejected": -0.21051180362701416, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.67407462337124e-06, | |
| "logits/chosen": -1.669070839881897, | |
| "logits/rejected": -1.6044425964355469, | |
| "logps/chosen": -367.7320556640625, | |
| "logps/rejected": -497.50970458984375, | |
| "loss": 0.0199, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.12311694771051407, | |
| "rewards/margins": 0.12384019047021866, | |
| "rewards/rejected": -0.24695713818073273, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.617052916277952e-06, | |
| "logits/chosen": -2.0958499908447266, | |
| "logits/rejected": -1.8438915014266968, | |
| "logps/chosen": -428.4202575683594, | |
| "logps/rejected": -517.01025390625, | |
| "loss": 0.0196, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.11365973949432373, | |
| "rewards/margins": 0.0948016420006752, | |
| "rewards/rejected": -0.20846137404441833, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5599701068654985e-06, | |
| "logits/chosen": -1.7646852731704712, | |
| "logits/rejected": -1.435537338256836, | |
| "logps/chosen": -422.580810546875, | |
| "logps/rejected": -500.8734436035156, | |
| "loss": 0.0201, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.14953741431236267, | |
| "rewards/margins": 0.11128922551870346, | |
| "rewards/rejected": -0.2608266770839691, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.5028559927002326e-06, | |
| "logits/chosen": -2.0240190029144287, | |
| "logits/rejected": -1.542840600013733, | |
| "logps/chosen": -451.66876220703125, | |
| "logps/rejected": -516.1002197265625, | |
| "loss": 0.0174, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.1411553919315338, | |
| "rewards/margins": 0.13159087300300598, | |
| "rewards/rejected": -0.2727462351322174, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.4457403876897756e-06, | |
| "logits/chosen": -1.6757514476776123, | |
| "logits/rejected": -1.4124213457107544, | |
| "logps/chosen": -414.86883544921875, | |
| "logps/rejected": -541.7101440429688, | |
| "loss": 0.0145, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.15753154456615448, | |
| "rewards/margins": 0.1300159990787506, | |
| "rewards/rejected": -0.2875475585460663, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.388653106519975e-06, | |
| "logits/chosen": -1.888702630996704, | |
| "logits/rejected": -1.7808609008789062, | |
| "logps/chosen": -413.09295654296875, | |
| "logps/rejected": -521.931884765625, | |
| "loss": 0.0194, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.15543514490127563, | |
| "rewards/margins": 0.09803062677383423, | |
| "rewards/rejected": -0.2534657418727875, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.331623949091467e-06, | |
| "logits/chosen": -1.8230218887329102, | |
| "logits/rejected": -1.629063367843628, | |
| "logps/chosen": -441.42034912109375, | |
| "logps/rejected": -485.918212890625, | |
| "loss": 0.0239, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.12905916571617126, | |
| "rewards/margins": 0.10947714000940323, | |
| "rewards/rejected": -0.2385363131761551, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2746826849639513e-06, | |
| "logits/chosen": -2.148059129714966, | |
| "logits/rejected": -1.7364435195922852, | |
| "logps/chosen": -456.3299255371094, | |
| "logps/rejected": -466.22515869140625, | |
| "loss": 0.0189, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.1590055525302887, | |
| "rewards/margins": 0.06117180734872818, | |
| "rewards/rejected": -0.22017733752727509, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2178590378162957e-06, | |
| "logits/chosen": -1.765479326248169, | |
| "logits/rejected": -1.6112077236175537, | |
| "logps/chosen": -437.97381591796875, | |
| "logps/rejected": -574.2666015625, | |
| "loss": 0.0227, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.1820591539144516, | |
| "rewards/margins": 0.1335388422012329, | |
| "rewards/rejected": -0.3155979812145233, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1611826699306104e-06, | |
| "logits/chosen": -1.8870223760604858, | |
| "logits/rejected": -1.520627737045288, | |
| "logps/chosen": -420.63323974609375, | |
| "logps/rejected": -466.6917419433594, | |
| "loss": 0.0155, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.15777216851711273, | |
| "rewards/margins": 0.05299310013651848, | |
| "rewards/rejected": -0.2107652723789215, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.1046831667083483e-06, | |
| "logits/chosen": -1.8595775365829468, | |
| "logits/rejected": -1.6033302545547485, | |
| "logps/chosen": -453.5380859375, | |
| "logps/rejected": -464.77398681640625, | |
| "loss": 0.0291, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.16525158286094666, | |
| "rewards/margins": 0.07691897451877594, | |
| "rewards/rejected": -0.2421705275774002, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.048390021226559e-06, | |
| "logits/chosen": -1.9039933681488037, | |
| "logits/rejected": -1.672969102859497, | |
| "logps/chosen": -395.4986572265625, | |
| "logps/rejected": -436.7999572753906, | |
| "loss": 0.0178, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.1351805031299591, | |
| "rewards/margins": 0.08283836394548416, | |
| "rewards/rejected": -0.21801885962486267, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.9923326188423212e-06, | |
| "logits/chosen": -1.8809058666229248, | |
| "logits/rejected": -1.6939058303833008, | |
| "logps/chosen": -359.4674987792969, | |
| "logps/rejected": -431.02752685546875, | |
| "loss": 0.0137, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.11502976715564728, | |
| "rewards/margins": 0.09757871180772781, | |
| "rewards/rejected": -0.2126084864139557, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.936540221853415e-06, | |
| "logits/chosen": -1.687034249305725, | |
| "logits/rejected": -1.7980533838272095, | |
| "logps/chosen": -385.8014221191406, | |
| "logps/rejected": -513.2298583984375, | |
| "loss": 0.0179, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.12754614651203156, | |
| "rewards/margins": 0.10141207277774811, | |
| "rewards/rejected": -0.22895821928977966, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8810419542232245e-06, | |
| "logits/chosen": -1.832686424255371, | |
| "logits/rejected": -1.7866284847259521, | |
| "logps/chosen": -375.6191711425781, | |
| "logps/rejected": -473.59747314453125, | |
| "loss": 0.0143, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.13123920559883118, | |
| "rewards/margins": 0.07887652516365051, | |
| "rewards/rejected": -0.2101157158613205, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.8258667863778573e-06, | |
| "logits/chosen": -1.8020689487457275, | |
| "logits/rejected": -1.5563127994537354, | |
| "logps/chosen": -391.3179931640625, | |
| "logps/rejected": -437.33990478515625, | |
| "loss": 0.0253, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.15354005992412567, | |
| "rewards/margins": 0.07208997011184692, | |
| "rewards/rejected": -0.2256300449371338, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.7710435200834126e-06, | |
| "logits/chosen": -1.8209571838378906, | |
| "logits/rejected": -1.6538410186767578, | |
| "logps/chosen": -382.71331787109375, | |
| "logps/rejected": -439.09765625, | |
| "loss": 0.0171, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.13624554872512817, | |
| "rewards/margins": 0.08001069724559784, | |
| "rewards/rejected": -0.2162562608718872, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7166007734112808e-06, | |
| "logits/chosen": -1.6404680013656616, | |
| "logits/rejected": -1.4822646379470825, | |
| "logps/chosen": -394.8628845214844, | |
| "logps/rejected": -502.18524169921875, | |
| "loss": 0.0203, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.15191176533699036, | |
| "rewards/margins": 0.10607035458087921, | |
| "rewards/rejected": -0.2579821050167084, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6625669657993483e-06, | |
| "logits/chosen": -1.8440077304840088, | |
| "logits/rejected": -1.6627609729766846, | |
| "logps/chosen": -379.8057556152344, | |
| "logps/rejected": -463.258544921875, | |
| "loss": 0.0367, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.11155873537063599, | |
| "rewards/margins": 0.09738429635763168, | |
| "rewards/rejected": -0.20894300937652588, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6089703032168736e-06, | |
| "logits/chosen": -1.7317651510238647, | |
| "logits/rejected": -1.6905380487442017, | |
| "logps/chosen": -416.31622314453125, | |
| "logps/rejected": -504.07415771484375, | |
| "loss": 0.0191, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.13056668639183044, | |
| "rewards/margins": 0.1028464064002037, | |
| "rewards/rejected": -0.23341310024261475, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.55583876344081e-06, | |
| "logits/chosen": -1.8492292165756226, | |
| "logits/rejected": -1.7530572414398193, | |
| "logps/chosen": -425.2975158691406, | |
| "logps/rejected": -527.4212646484375, | |
| "loss": 0.0189, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.14273019134998322, | |
| "rewards/margins": 0.0867033302783966, | |
| "rewards/rejected": -0.22943353652954102, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.5032000814512372e-06, | |
| "logits/chosen": -1.8472225666046143, | |
| "logits/rejected": -1.8299192190170288, | |
| "logps/chosen": -367.6811828613281, | |
| "logps/rejected": -439.68975830078125, | |
| "loss": 0.0232, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.14622075855731964, | |
| "rewards/margins": 0.07023419439792633, | |
| "rewards/rejected": -0.21645493805408478, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.4510817349535323e-06, | |
| "logits/chosen": -1.8069846630096436, | |
| "logits/rejected": -1.733162522315979, | |
| "logps/chosen": -352.75054931640625, | |
| "logps/rejected": -439.2781677246094, | |
| "loss": 0.023, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.12202004343271255, | |
| "rewards/margins": 0.07520598918199539, | |
| "rewards/rejected": -0.19722604751586914, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.3995109300348537e-06, | |
| "logits/chosen": -1.7219600677490234, | |
| "logits/rejected": -1.5683410167694092, | |
| "logps/chosen": -397.7275085449219, | |
| "logps/rejected": -526.4014282226562, | |
| "loss": 0.0233, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.15109822154045105, | |
| "rewards/margins": 0.12963151931762695, | |
| "rewards/rejected": -0.280729740858078, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.348514586962389e-06, | |
| "logits/chosen": -1.9068981409072876, | |
| "logits/rejected": -1.6539924144744873, | |
| "logps/chosen": -451.5171813964844, | |
| "logps/rejected": -513.7973022460938, | |
| "loss": 0.0224, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.18701951205730438, | |
| "rewards/margins": 0.063180111348629, | |
| "rewards/rejected": -0.25019964575767517, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.2981193261308284e-06, | |
| "logits/chosen": -1.601199746131897, | |
| "logits/rejected": -1.4806240797042847, | |
| "logps/chosen": -401.56243896484375, | |
| "logps/rejected": -576.2053833007812, | |
| "loss": 0.0208, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.1684935986995697, | |
| "rewards/margins": 0.1360427886247635, | |
| "rewards/rejected": -0.304536372423172, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2483514541663501e-06, | |
| "logits/chosen": -1.7805836200714111, | |
| "logits/rejected": -1.636907935142517, | |
| "logps/chosen": -396.3351135253906, | |
| "logps/rejected": -458.74725341796875, | |
| "loss": 0.0119, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.13784000277519226, | |
| "rewards/margins": 0.08524910360574722, | |
| "rewards/rejected": -0.22308912873268127, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1992369501944096e-06, | |
| "logits/chosen": -1.8869998455047607, | |
| "logits/rejected": -1.8042247295379639, | |
| "logps/chosen": -363.39788818359375, | |
| "logps/rejected": -539.3111572265625, | |
| "loss": 0.0213, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.14907002449035645, | |
| "rewards/margins": 0.14250054955482483, | |
| "rewards/rejected": -0.29157060384750366, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1508014522784803e-06, | |
| "logits/chosen": -1.8102096319198608, | |
| "logits/rejected": -1.695726752281189, | |
| "logps/chosen": -443.92327880859375, | |
| "logps/rejected": -519.6880493164062, | |
| "loss": 0.0199, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.1804676353931427, | |
| "rewards/margins": 0.09004662185907364, | |
| "rewards/rejected": -0.27051427960395813, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.1030702440368319e-06, | |
| "logits/chosen": -1.5621861219406128, | |
| "logits/rejected": -1.5408356189727783, | |
| "logps/chosen": -452.7931213378906, | |
| "logps/rejected": -742.4008178710938, | |
| "loss": 0.0272, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.18929113447666168, | |
| "rewards/margins": 0.12937882542610168, | |
| "rewards/rejected": -0.31866997480392456, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0560682414443315e-06, | |
| "logits/chosen": -1.8824927806854248, | |
| "logits/rejected": -1.7471214532852173, | |
| "logps/chosen": -460.31640625, | |
| "logps/rejected": -539.2716064453125, | |
| "loss": 0.0114, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.1245940774679184, | |
| "rewards/margins": 0.12403732538223267, | |
| "rewards/rejected": -0.24863143265247345, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.009819979826156e-06, | |
| "logits/chosen": -1.9311294555664062, | |
| "logits/rejected": -1.5975712537765503, | |
| "logps/chosen": -442.8894958496094, | |
| "logps/rejected": -503.81048583984375, | |
| "loss": 0.0304, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.1375322788953781, | |
| "rewards/margins": 0.08977995812892914, | |
| "rewards/rejected": -0.22731228172779083, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.643496010502054e-07, | |
| "logits/chosen": -1.8107373714447021, | |
| "logits/rejected": -1.6052519083023071, | |
| "logps/chosen": -426.5626525878906, | |
| "logps/rejected": -484.64971923828125, | |
| "loss": 0.0149, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.13215045630931854, | |
| "rewards/margins": 0.09739480167627335, | |
| "rewards/rejected": -0.2295452356338501, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 9.196808409249086e-07, | |
| "logits/chosen": -1.901611566543579, | |
| "logits/rejected": -1.9235025644302368, | |
| "logps/chosen": -372.97808837890625, | |
| "logps/rejected": -540.9254150390625, | |
| "loss": 0.0176, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.11672798544168472, | |
| "rewards/margins": 0.11135835945606232, | |
| "rewards/rejected": -0.22808635234832764, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.758370168089797e-07, | |
| "logits/chosen": -1.8253875970840454, | |
| "logits/rejected": -1.6909106969833374, | |
| "logps/chosen": -321.0955505371094, | |
| "logps/rejected": -385.1728515625, | |
| "loss": 0.0159, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.12681838870048523, | |
| "rewards/margins": 0.06274305284023285, | |
| "rewards/rejected": -0.18956145644187927, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.328410154396318e-07, | |
| "logits/chosen": -1.8118044137954712, | |
| "logits/rejected": -1.8118581771850586, | |
| "logps/chosen": -379.05615234375, | |
| "logps/rejected": -470.08184814453125, | |
| "loss": 0.0176, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.12152254581451416, | |
| "rewards/margins": 0.06952939927577972, | |
| "rewards/rejected": -0.19105194509029388, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.907152809855529e-07, | |
| "logits/chosen": -1.877508521080017, | |
| "logits/rejected": -1.6081901788711548, | |
| "logps/chosen": -366.0731201171875, | |
| "logps/rejected": -417.31683349609375, | |
| "loss": 0.0202, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.12633651494979858, | |
| "rewards/margins": 0.06398525089025497, | |
| "rewards/rejected": -0.19032175838947296, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.494818033309207e-07, | |
| "logits/chosen": -1.7752797603607178, | |
| "logits/rejected": -1.6302067041397095, | |
| "logps/chosen": -360.30023193359375, | |
| "logps/rejected": -443.12908935546875, | |
| "loss": 0.0261, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.12191496789455414, | |
| "rewards/margins": 0.07757655531167984, | |
| "rewards/rejected": -0.19949153065681458, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.091621065965521e-07, | |
| "logits/chosen": -1.7687801122665405, | |
| "logits/rejected": -1.4826685190200806, | |
| "logps/chosen": -366.3409118652344, | |
| "logps/rejected": -412.891357421875, | |
| "loss": 0.0155, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.1527003049850464, | |
| "rewards/margins": 0.08991221338510513, | |
| "rewards/rejected": -0.24261251091957092, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.697772379041823e-07, | |
| "logits/chosen": -2.0650315284729004, | |
| "logits/rejected": -1.6713231801986694, | |
| "logps/chosen": -448.5660095214844, | |
| "logps/rejected": -507.5440368652344, | |
| "loss": 0.0299, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.14657150208950043, | |
| "rewards/margins": 0.1087670549750328, | |
| "rewards/rejected": -0.255338579416275, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.313477563897466e-07, | |
| "logits/chosen": -1.8000714778900146, | |
| "logits/rejected": -1.6599605083465576, | |
| "logps/chosen": -451.50140380859375, | |
| "logps/rejected": -568.575927734375, | |
| "loss": 0.0186, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.18340735137462616, | |
| "rewards/margins": 0.10143464803695679, | |
| "rewards/rejected": -0.28484198451042175, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.9389372247138e-07, | |
| "logits/chosen": -1.7278544902801514, | |
| "logits/rejected": -1.328420639038086, | |
| "logps/chosen": -498.06280517578125, | |
| "logps/rejected": -581.17041015625, | |
| "loss": 0.0193, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.1877037137746811, | |
| "rewards/margins": 0.12751582264900208, | |
| "rewards/rejected": -0.315219521522522, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.574346873777714e-07, | |
| "logits/chosen": -1.5999044179916382, | |
| "logits/rejected": -1.5874285697937012, | |
| "logps/chosen": -431.46185302734375, | |
| "logps/rejected": -516.9653930664062, | |
| "loss": 0.0167, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.16077804565429688, | |
| "rewards/margins": 0.07463011145591736, | |
| "rewards/rejected": -0.23540815711021423, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.219896829422927e-07, | |
| "logits/chosen": -1.6012115478515625, | |
| "logits/rejected": -1.5557236671447754, | |
| "logps/chosen": -308.88348388671875, | |
| "logps/rejected": -417.7029724121094, | |
| "loss": 0.0222, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.14151635766029358, | |
| "rewards/margins": 0.10681549459695816, | |
| "rewards/rejected": -0.24833182990550995, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.875772116682817e-07, | |
| "logits/chosen": -1.9544742107391357, | |
| "logits/rejected": -1.5280301570892334, | |
| "logps/chosen": -457.99078369140625, | |
| "logps/rejected": -516.620849609375, | |
| "loss": 0.0222, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.13625690340995789, | |
| "rewards/margins": 0.0871158018708229, | |
| "rewards/rejected": -0.22337272763252258, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.542152370706149e-07, | |
| "logits/chosen": -1.8728841543197632, | |
| "logits/rejected": -1.6816173791885376, | |
| "logps/chosen": -441.79693603515625, | |
| "logps/rejected": -553.9568481445312, | |
| "loss": 0.0127, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.16504457592964172, | |
| "rewards/margins": 0.11639133840799332, | |
| "rewards/rejected": -0.28143590688705444, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.2192117429865067e-07, | |
| "logits/chosen": -2.061262845993042, | |
| "logits/rejected": -1.6205766201019287, | |
| "logps/chosen": -463.18408203125, | |
| "logps/rejected": -434.10662841796875, | |
| "loss": 0.0173, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.15750975906848907, | |
| "rewards/margins": 0.07638247311115265, | |
| "rewards/rejected": -0.23389224708080292, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.907118810454172e-07, | |
| "logits/chosen": -1.746106505393982, | |
| "logits/rejected": -1.5782015323638916, | |
| "logps/chosen": -492.3348083496094, | |
| "logps/rejected": -519.789306640625, | |
| "loss": 0.0184, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.17052312195301056, | |
| "rewards/margins": 0.0672903060913086, | |
| "rewards/rejected": -0.23781342804431915, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.6060364874779455e-07, | |
| "logits/chosen": -1.9814525842666626, | |
| "logits/rejected": -1.5197412967681885, | |
| "logps/chosen": -486.7601623535156, | |
| "logps/rejected": -499.59320068359375, | |
| "loss": 0.0193, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.15768514573574066, | |
| "rewards/margins": 0.09786301851272583, | |
| "rewards/rejected": -0.2555481493473053, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.3161219408229026e-07, | |
| "logits/chosen": -1.8380473852157593, | |
| "logits/rejected": -1.5035631656646729, | |
| "logps/chosen": -428.0066833496094, | |
| "logps/rejected": -499.0686950683594, | |
| "loss": 0.0252, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.15843157470226288, | |
| "rewards/margins": 0.11070273071527481, | |
| "rewards/rejected": -0.2691343128681183, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.0375265076083796e-07, | |
| "logits/chosen": -1.7933199405670166, | |
| "logits/rejected": -1.6641219854354858, | |
| "logps/chosen": -398.84716796875, | |
| "logps/rejected": -450.40704345703125, | |
| "loss": 0.0219, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.1423635184764862, | |
| "rewards/margins": 0.06088986247777939, | |
| "rewards/rejected": -0.2032533884048462, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.7703956163091153e-07, | |
| "logits/chosen": -1.6546907424926758, | |
| "logits/rejected": -1.491092562675476, | |
| "logps/chosen": -442.4610290527344, | |
| "logps/rejected": -543.3446044921875, | |
| "loss": 0.0214, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.1697242110967636, | |
| "rewards/margins": 0.11211826652288437, | |
| "rewards/rejected": -0.28184249997138977, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.514868710840723e-07, | |
| "logits/chosen": -1.7892532348632812, | |
| "logits/rejected": -1.673905611038208, | |
| "logps/chosen": -385.91119384765625, | |
| "logps/rejected": -513.5677490234375, | |
| "loss": 0.0256, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.13928866386413574, | |
| "rewards/margins": 0.10927131026983261, | |
| "rewards/rejected": -0.24855998158454895, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.271079177769117e-07, | |
| "logits/chosen": -1.801029920578003, | |
| "logits/rejected": -1.7066704034805298, | |
| "logps/chosen": -447.7923889160156, | |
| "logps/rejected": -510.22088623046875, | |
| "loss": 0.016, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.14237356185913086, | |
| "rewards/margins": 0.09682431071996689, | |
| "rewards/rejected": -0.23919788002967834, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.0391542766819456e-07, | |
| "logits/chosen": -1.6222654581069946, | |
| "logits/rejected": -1.5031228065490723, | |
| "logps/chosen": -444.59112548828125, | |
| "logps/rejected": -575.5186157226562, | |
| "loss": 0.0206, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.17282214760780334, | |
| "rewards/margins": 0.13414433598518372, | |
| "rewards/rejected": -0.30696648359298706, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.8192150737583264e-07, | |
| "logits/chosen": -1.8754476308822632, | |
| "logits/rejected": -1.4290292263031006, | |
| "logps/chosen": -465.01922607421875, | |
| "logps/rejected": -613.7750244140625, | |
| "loss": 0.0151, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": -0.17955392599105835, | |
| "rewards/margins": 0.1857127696275711, | |
| "rewards/rejected": -0.3652666509151459, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.61137637857158e-07, | |
| "logits/chosen": -1.8329002857208252, | |
| "logits/rejected": -1.6551681756973267, | |
| "logps/chosen": -404.65765380859375, | |
| "logps/rejected": -481.35931396484375, | |
| "loss": 0.0176, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.15663662552833557, | |
| "rewards/margins": 0.08110956102609634, | |
| "rewards/rejected": -0.2377461940050125, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.415746684157951e-07, | |
| "logits/chosen": -1.815999984741211, | |
| "logits/rejected": -1.5772883892059326, | |
| "logps/chosen": -508.98248291015625, | |
| "logps/rejected": -616.5827026367188, | |
| "loss": 0.0129, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": -0.16146263480186462, | |
| "rewards/margins": 0.15336495637893677, | |
| "rewards/rejected": -0.3148276209831238, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.232428110382586e-07, | |
| "logits/chosen": -1.705453634262085, | |
| "logits/rejected": -1.7126257419586182, | |
| "logps/chosen": -443.9436950683594, | |
| "logps/rejected": -507.04681396484375, | |
| "loss": 0.0185, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.14823515713214874, | |
| "rewards/margins": 0.06383778899908066, | |
| "rewards/rejected": -0.2120729386806488, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 1.0615163506323856e-07, | |
| "logits/chosen": -1.734545111656189, | |
| "logits/rejected": -1.7701743841171265, | |
| "logps/chosen": -467.75225830078125, | |
| "logps/rejected": -597.7060546875, | |
| "loss": 0.0213, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.17752481997013092, | |
| "rewards/margins": 0.08124741166830063, | |
| "rewards/rejected": -0.25877219438552856, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.031006218634342e-08, | |
| "logits/chosen": -1.8693500757217407, | |
| "logits/rejected": -1.7246434688568115, | |
| "logps/chosen": -405.7983703613281, | |
| "logps/rejected": -492.0767517089844, | |
| "loss": 0.0167, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.14290982484817505, | |
| "rewards/margins": 0.08994929492473602, | |
| "rewards/rejected": -0.23285913467407227, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 7.572636180292831e-08, | |
| "logits/chosen": -1.9143638610839844, | |
| "logits/rejected": -1.8357505798339844, | |
| "logps/chosen": -359.0616149902344, | |
| "logps/rejected": -476.2394104003906, | |
| "loss": 0.0251, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.14140944182872772, | |
| "rewards/margins": 0.09983612596988678, | |
| "rewards/rejected": -0.2412455528974533, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.240814669141559e-08, | |
| "logits/chosen": -1.6147396564483643, | |
| "logits/rejected": -1.5888813734054565, | |
| "logps/chosen": -426.96929931640625, | |
| "logps/rejected": -460.4154357910156, | |
| "loss": 0.023, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.1609814316034317, | |
| "rewards/margins": 0.04549731686711311, | |
| "rewards/rejected": -0.20647874474525452, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.036236903938285e-08, | |
| "logits/chosen": -1.7797266244888306, | |
| "logits/rejected": -1.4338808059692383, | |
| "logps/chosen": -383.97021484375, | |
| "logps/rejected": -458.4593811035156, | |
| "loss": 0.0131, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.14667093753814697, | |
| "rewards/margins": 0.11462962627410889, | |
| "rewards/rejected": -0.26130059361457825, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.959531681447859e-08, | |
| "logits/chosen": -1.8102188110351562, | |
| "logits/rejected": -1.7282259464263916, | |
| "logps/chosen": -432.2884826660156, | |
| "logps/rejected": -555.3939208984375, | |
| "loss": 0.0139, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.18346509337425232, | |
| "rewards/margins": 0.11375024169683456, | |
| "rewards/rejected": -0.2972153425216675, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 3.0112610482064544e-08, | |
| "logits/chosen": -1.774643898010254, | |
| "logits/rejected": -1.499586582183838, | |
| "logps/chosen": -427.5147399902344, | |
| "logps/rejected": -525.1546630859375, | |
| "loss": 0.0193, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.1594984531402588, | |
| "rewards/margins": 0.09911755472421646, | |
| "rewards/rejected": -0.25861600041389465, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.1919200071301715e-08, | |
| "logits/chosen": -1.7066888809204102, | |
| "logits/rejected": -1.6820920705795288, | |
| "logps/chosen": -494.503662109375, | |
| "logps/rejected": -650.3525390625, | |
| "loss": 0.0154, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.19812996685504913, | |
| "rewards/margins": 0.12928268313407898, | |
| "rewards/rejected": -0.3274126648902893, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.50193625912029e-08, | |
| "logits/chosen": -1.6772607564926147, | |
| "logits/rejected": -1.3972876071929932, | |
| "logps/chosen": -391.4365539550781, | |
| "logps/rejected": -551.2503662109375, | |
| "loss": 0.0233, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.17442724108695984, | |
| "rewards/margins": 0.14638181030750275, | |
| "rewards/rejected": -0.3208090364933014, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 9.416699798010521e-09, | |
| "logits/chosen": -1.8336042165756226, | |
| "logits/rejected": -1.3778246641159058, | |
| "logps/chosen": -412.4085388183594, | |
| "logps/rejected": -431.26690673828125, | |
| "loss": 0.0284, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.14829722046852112, | |
| "rewards/margins": 0.08893366903066635, | |
| "rewards/rejected": -0.23723092675209045, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 5.114136315058083e-09, | |
| "logits/chosen": -1.5414822101593018, | |
| "logits/rejected": -1.5957967042922974, | |
| "logps/chosen": -371.25933837890625, | |
| "logps/rejected": -498.68695068359375, | |
| "loss": 0.0276, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.15921153128147125, | |
| "rewards/margins": 0.08158004283905029, | |
| "rewards/rejected": -0.24079158902168274, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2.113918106098345e-09, | |
| "logits/chosen": -1.774679183959961, | |
| "logits/rejected": -1.5650732517242432, | |
| "logps/chosen": -484.24627685546875, | |
| "logps/rejected": -497.2215270996094, | |
| "loss": 0.0197, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.16433146595954895, | |
| "rewards/margins": 0.07303442806005478, | |
| "rewards/rejected": -0.23736588656902313, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 4.176113028983575e-10, | |
| "logits/chosen": -2.0026144981384277, | |
| "logits/rejected": -1.8274720907211304, | |
| "logps/chosen": -379.2737731933594, | |
| "logps/rejected": -428.299072265625, | |
| "loss": 0.0196, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.13032202422618866, | |
| "rewards/margins": 0.08789823949337006, | |
| "rewards/rejected": -0.2182202786207199, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1528, | |
| "total_flos": 0.0, | |
| "train_loss": 0.021083176277421845, | |
| "train_runtime": 50303.8354, | |
| "train_samples_per_second": 0.365, | |
| "train_steps_per_second": 0.03 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1528, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |