| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9992254066615027, | |
| "eval_steps": 100, | |
| "global_step": 726, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 6.84931506849315e-09, | |
| "logits/chosen": -2.3491616249084473, | |
| "logits/rejected": -2.418564796447754, | |
| "logps/chosen": -271.3881530761719, | |
| "logps/rejected": -208.9749298095703, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 6.84931506849315e-08, | |
| "logits/chosen": -2.4231245517730713, | |
| "logits/rejected": -2.3566601276397705, | |
| "logps/chosen": -293.38800048828125, | |
| "logps/rejected": -226.29283142089844, | |
| "loss": 0.6933, | |
| "rewards/accuracies": 0.4548611044883728, | |
| "rewards/chosen": 0.003188559552654624, | |
| "rewards/margins": 0.0021638227626681328, | |
| "rewards/rejected": 0.0010247372556477785, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 1.36986301369863e-07, | |
| "logits/chosen": -2.431933879852295, | |
| "logits/rejected": -2.405198574066162, | |
| "logps/chosen": -278.5166931152344, | |
| "logps/rejected": -216.7791290283203, | |
| "loss": 0.6942, | |
| "rewards/accuracies": 0.504687488079071, | |
| "rewards/chosen": -0.000816329091321677, | |
| "rewards/margins": 0.0019420869648456573, | |
| "rewards/rejected": -0.0027584161143749952, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 2.054794520547945e-07, | |
| "logits/chosen": -2.389657497406006, | |
| "logits/rejected": -2.348972797393799, | |
| "logps/chosen": -252.9993438720703, | |
| "logps/rejected": -207.1633758544922, | |
| "loss": 0.6935, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.0014556010719388723, | |
| "rewards/margins": 0.0006705918349325657, | |
| "rewards/rejected": -0.0021261931397020817, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 2.73972602739726e-07, | |
| "logits/chosen": -2.460561752319336, | |
| "logits/rejected": -2.414844036102295, | |
| "logps/chosen": -283.7592468261719, | |
| "logps/rejected": -216.4773712158203, | |
| "loss": 0.6909, | |
| "rewards/accuracies": 0.5484374761581421, | |
| "rewards/chosen": 0.004375931341201067, | |
| "rewards/margins": 0.00752140861004591, | |
| "rewards/rejected": -0.0031454775016754866, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 3.424657534246575e-07, | |
| "logits/chosen": -2.4366953372955322, | |
| "logits/rejected": -2.3972277641296387, | |
| "logps/chosen": -267.2607727050781, | |
| "logps/rejected": -223.6705780029297, | |
| "loss": 0.6897, | |
| "rewards/accuracies": 0.5406249761581421, | |
| "rewards/chosen": 0.004865794442594051, | |
| "rewards/margins": 0.007948420941829681, | |
| "rewards/rejected": -0.0030826255679130554, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.10958904109589e-07, | |
| "logits/chosen": -2.412304639816284, | |
| "logits/rejected": -2.3892178535461426, | |
| "logps/chosen": -266.85028076171875, | |
| "logps/rejected": -214.7494659423828, | |
| "loss": 0.6873, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.00616841483861208, | |
| "rewards/margins": 0.012422902509570122, | |
| "rewards/rejected": -0.006254489067941904, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.794520547945205e-07, | |
| "logits/chosen": -2.390881299972534, | |
| "logits/rejected": -2.3977627754211426, | |
| "logps/chosen": -254.04043579101562, | |
| "logps/rejected": -214.6400909423828, | |
| "loss": 0.6847, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.009625923819839954, | |
| "rewards/margins": 0.01777799427509308, | |
| "rewards/rejected": -0.008152070455253124, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.946401225114854e-07, | |
| "logits/chosen": -2.4295055866241455, | |
| "logits/rejected": -2.37807035446167, | |
| "logps/chosen": -265.05718994140625, | |
| "logps/rejected": -218.77059936523438, | |
| "loss": 0.6795, | |
| "rewards/accuracies": 0.6015625, | |
| "rewards/chosen": 0.013735203072428703, | |
| "rewards/margins": 0.028081998229026794, | |
| "rewards/rejected": -0.014346795156598091, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 4.869831546707504e-07, | |
| "logits/chosen": -2.4787497520446777, | |
| "logits/rejected": -2.422356128692627, | |
| "logps/chosen": -271.730224609375, | |
| "logps/rejected": -224.4839324951172, | |
| "loss": 0.6734, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": 0.022287212312221527, | |
| "rewards/margins": 0.04716240242123604, | |
| "rewards/rejected": -0.02487519010901451, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 4.793261868300153e-07, | |
| "logits/chosen": -2.417426586151123, | |
| "logits/rejected": -2.4039013385772705, | |
| "logps/chosen": -273.8762512207031, | |
| "logps/rejected": -227.797607421875, | |
| "loss": 0.6692, | |
| "rewards/accuracies": 0.659375011920929, | |
| "rewards/chosen": 0.025275733321905136, | |
| "rewards/margins": 0.057271964848041534, | |
| "rewards/rejected": -0.0319962315261364, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 4.7166921898928023e-07, | |
| "logits/chosen": -2.4509871006011963, | |
| "logits/rejected": -2.3908462524414062, | |
| "logps/chosen": -252.6064453125, | |
| "logps/rejected": -222.64639282226562, | |
| "loss": 0.6628, | |
| "rewards/accuracies": 0.667187511920929, | |
| "rewards/chosen": 0.024279529228806496, | |
| "rewards/margins": 0.06397499889135361, | |
| "rewards/rejected": -0.03969546779990196, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 4.640122511485451e-07, | |
| "logits/chosen": -2.412073850631714, | |
| "logits/rejected": -2.4017536640167236, | |
| "logps/chosen": -256.89813232421875, | |
| "logps/rejected": -219.8057098388672, | |
| "loss": 0.6581, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.029207896441221237, | |
| "rewards/margins": 0.07588861137628555, | |
| "rewards/rejected": -0.04668071120977402, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 4.563552833078101e-07, | |
| "logits/chosen": -2.4562458992004395, | |
| "logits/rejected": -2.3951973915100098, | |
| "logps/chosen": -265.9952697753906, | |
| "logps/rejected": -225.94125366210938, | |
| "loss": 0.6504, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.025836100801825523, | |
| "rewards/margins": 0.09206128865480423, | |
| "rewards/rejected": -0.06622518599033356, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 4.4869831546707505e-07, | |
| "logits/chosen": -2.460266590118408, | |
| "logits/rejected": -2.401520013809204, | |
| "logps/chosen": -270.29888916015625, | |
| "logps/rejected": -230.37539672851562, | |
| "loss": 0.6495, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": 0.03395112603902817, | |
| "rewards/margins": 0.11508414894342422, | |
| "rewards/rejected": -0.08113302290439606, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 4.4104134762633994e-07, | |
| "logits/chosen": -2.4852752685546875, | |
| "logits/rejected": -2.4318604469299316, | |
| "logps/chosen": -265.973388671875, | |
| "logps/rejected": -226.55484008789062, | |
| "loss": 0.6433, | |
| "rewards/accuracies": 0.692187488079071, | |
| "rewards/chosen": 0.03317371755838394, | |
| "rewards/margins": 0.12171275913715363, | |
| "rewards/rejected": -0.08853904157876968, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 4.333843797856049e-07, | |
| "logits/chosen": -2.4467227458953857, | |
| "logits/rejected": -2.4029695987701416, | |
| "logps/chosen": -279.33648681640625, | |
| "logps/rejected": -239.00009155273438, | |
| "loss": 0.6343, | |
| "rewards/accuracies": 0.6796875, | |
| "rewards/chosen": 0.04138711839914322, | |
| "rewards/margins": 0.1465190351009369, | |
| "rewards/rejected": -0.10513193905353546, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 4.257274119448698e-07, | |
| "logits/chosen": -2.4798355102539062, | |
| "logits/rejected": -2.452397108078003, | |
| "logps/chosen": -270.727783203125, | |
| "logps/rejected": -239.09780883789062, | |
| "loss": 0.6335, | |
| "rewards/accuracies": 0.653124988079071, | |
| "rewards/chosen": 0.024355659261345863, | |
| "rewards/margins": 0.1352422684431076, | |
| "rewards/rejected": -0.11088661849498749, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 4.180704441041347e-07, | |
| "logits/chosen": -2.4359121322631836, | |
| "logits/rejected": -2.388683795928955, | |
| "logps/chosen": -256.79022216796875, | |
| "logps/rejected": -226.1436767578125, | |
| "loss": 0.6367, | |
| "rewards/accuracies": 0.6640625, | |
| "rewards/chosen": 0.016556020826101303, | |
| "rewards/margins": 0.147763192653656, | |
| "rewards/rejected": -0.1312071532011032, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 4.1041347626339966e-07, | |
| "logits/chosen": -2.4478251934051514, | |
| "logits/rejected": -2.4065427780151367, | |
| "logps/chosen": -261.63702392578125, | |
| "logps/rejected": -213.1779327392578, | |
| "loss": 0.6269, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": 0.02061801217496395, | |
| "rewards/margins": 0.17272573709487915, | |
| "rewards/rejected": -0.15210774540901184, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.027565084226646e-07, | |
| "logits/chosen": -2.4714255332946777, | |
| "logits/rejected": -2.414602279663086, | |
| "logps/chosen": -262.29486083984375, | |
| "logps/rejected": -218.0116424560547, | |
| "loss": 0.6175, | |
| "rewards/accuracies": 0.6890624761581421, | |
| "rewards/chosen": 0.027450546622276306, | |
| "rewards/margins": 0.19447624683380127, | |
| "rewards/rejected": -0.16702571511268616, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 3.9509954058192954e-07, | |
| "logits/chosen": -2.4752840995788574, | |
| "logits/rejected": -2.4354655742645264, | |
| "logps/chosen": -283.89959716796875, | |
| "logps/rejected": -231.7078399658203, | |
| "loss": 0.6161, | |
| "rewards/accuracies": 0.653124988079071, | |
| "rewards/chosen": 0.024292152374982834, | |
| "rewards/margins": 0.21237091720104218, | |
| "rewards/rejected": -0.18807876110076904, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 3.874425727411945e-07, | |
| "logits/chosen": -2.422091007232666, | |
| "logits/rejected": -2.40881609916687, | |
| "logps/chosen": -276.7785339355469, | |
| "logps/rejected": -229.2734832763672, | |
| "loss": 0.6144, | |
| "rewards/accuracies": 0.660937488079071, | |
| "rewards/chosen": 0.017709506675601006, | |
| "rewards/margins": 0.21948948502540588, | |
| "rewards/rejected": -0.20178000628948212, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.797856049004594e-07, | |
| "logits/chosen": -2.4441866874694824, | |
| "logits/rejected": -2.38869571685791, | |
| "logps/chosen": -264.10430908203125, | |
| "logps/rejected": -228.3271484375, | |
| "loss": 0.6117, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.029474353417754173, | |
| "rewards/margins": 0.22035422921180725, | |
| "rewards/rejected": -0.19087985157966614, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 3.7212863705972436e-07, | |
| "logits/chosen": -2.4633097648620605, | |
| "logits/rejected": -2.4186224937438965, | |
| "logps/chosen": -271.4654235839844, | |
| "logps/rejected": -222.46841430664062, | |
| "loss": 0.6125, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": 0.024074096232652664, | |
| "rewards/margins": 0.2165375053882599, | |
| "rewards/rejected": -0.19246339797973633, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_logits/chosen": -2.144517421722412, | |
| "eval_logits/rejected": -2.0242087841033936, | |
| "eval_logps/chosen": -264.5997619628906, | |
| "eval_logps/rejected": -221.6983184814453, | |
| "eval_loss": 0.6057174205780029, | |
| "eval_rewards/accuracies": 0.6759999990463257, | |
| "eval_rewards/chosen": 0.007874858565628529, | |
| "eval_rewards/margins": 0.24786852300167084, | |
| "eval_rewards/rejected": -0.23999367654323578, | |
| "eval_runtime": 278.9133, | |
| "eval_samples_per_second": 7.171, | |
| "eval_steps_per_second": 0.448, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "learning_rate": 3.6447166921898925e-07, | |
| "logits/chosen": -2.400252103805542, | |
| "logits/rejected": -2.3472890853881836, | |
| "logps/chosen": -257.4571838378906, | |
| "logps/rejected": -210.4391326904297, | |
| "loss": 0.6164, | |
| "rewards/accuracies": 0.6734374761581421, | |
| "rewards/chosen": 0.004229591693729162, | |
| "rewards/margins": 0.22420725226402283, | |
| "rewards/rejected": -0.21997769176959991, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 3.568147013782542e-07, | |
| "logits/chosen": -2.415523052215576, | |
| "logits/rejected": -2.3758111000061035, | |
| "logps/chosen": -261.9351501464844, | |
| "logps/rejected": -226.16259765625, | |
| "loss": 0.6006, | |
| "rewards/accuracies": 0.7265625, | |
| "rewards/chosen": 0.023142099380493164, | |
| "rewards/margins": 0.2655286490917206, | |
| "rewards/rejected": -0.24238653481006622, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "learning_rate": 3.4915773353751913e-07, | |
| "logits/chosen": -2.429934024810791, | |
| "logits/rejected": -2.365861415863037, | |
| "logps/chosen": -278.4029846191406, | |
| "logps/rejected": -236.08688354492188, | |
| "loss": 0.5925, | |
| "rewards/accuracies": 0.7406250238418579, | |
| "rewards/chosen": 0.02966948226094246, | |
| "rewards/margins": 0.33913469314575195, | |
| "rewards/rejected": -0.3094651699066162, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "learning_rate": 3.41500765696784e-07, | |
| "logits/chosen": -2.4358582496643066, | |
| "logits/rejected": -2.396267890930176, | |
| "logps/chosen": -251.093017578125, | |
| "logps/rejected": -225.80685424804688, | |
| "loss": 0.6036, | |
| "rewards/accuracies": 0.692187488079071, | |
| "rewards/chosen": 0.010072538629174232, | |
| "rewards/margins": 0.24589493870735168, | |
| "rewards/rejected": -0.2358224093914032, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "learning_rate": 3.33843797856049e-07, | |
| "logits/chosen": -2.408804416656494, | |
| "logits/rejected": -2.394888401031494, | |
| "logps/chosen": -283.15380859375, | |
| "logps/rejected": -228.33767700195312, | |
| "loss": 0.5915, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": 0.008710218593478203, | |
| "rewards/margins": 0.3084966242313385, | |
| "rewards/rejected": -0.29978638887405396, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "learning_rate": 3.2618683001531396e-07, | |
| "logits/chosen": -2.4084572792053223, | |
| "logits/rejected": -2.337435722351074, | |
| "logps/chosen": -261.3924865722656, | |
| "logps/rejected": -227.77651977539062, | |
| "loss": 0.588, | |
| "rewards/accuracies": 0.698437511920929, | |
| "rewards/chosen": 0.008477389812469482, | |
| "rewards/margins": 0.298746258020401, | |
| "rewards/rejected": -0.29026883840560913, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 3.1852986217457885e-07, | |
| "logits/chosen": -2.4575297832489014, | |
| "logits/rejected": -2.373924493789673, | |
| "logps/chosen": -261.287109375, | |
| "logps/rejected": -228.5553741455078, | |
| "loss": 0.5972, | |
| "rewards/accuracies": 0.6953125, | |
| "rewards/chosen": 0.002108477521687746, | |
| "rewards/margins": 0.2948620915412903, | |
| "rewards/rejected": -0.2927536368370056, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "learning_rate": 3.108728943338438e-07, | |
| "logits/chosen": -2.4443328380584717, | |
| "logits/rejected": -2.4351658821105957, | |
| "logps/chosen": -252.80996704101562, | |
| "logps/rejected": -237.87631225585938, | |
| "loss": 0.5943, | |
| "rewards/accuracies": 0.684374988079071, | |
| "rewards/chosen": -0.015406561084091663, | |
| "rewards/margins": 0.2515925168991089, | |
| "rewards/rejected": -0.2669990658760071, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "learning_rate": 3.0321592649310873e-07, | |
| "logits/chosen": -2.424647092819214, | |
| "logits/rejected": -2.357273578643799, | |
| "logps/chosen": -253.7325897216797, | |
| "logps/rejected": -224.3144073486328, | |
| "loss": 0.587, | |
| "rewards/accuracies": 0.6859375238418579, | |
| "rewards/chosen": -0.0076437839306890965, | |
| "rewards/margins": 0.2989902198314667, | |
| "rewards/rejected": -0.30663400888442993, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "learning_rate": 2.955589586523736e-07, | |
| "logits/chosen": -2.4427545070648193, | |
| "logits/rejected": -2.3824856281280518, | |
| "logps/chosen": -265.68939208984375, | |
| "logps/rejected": -226.4335174560547, | |
| "loss": 0.592, | |
| "rewards/accuracies": 0.7015625238418579, | |
| "rewards/chosen": -0.015530401840806007, | |
| "rewards/margins": 0.3260301351547241, | |
| "rewards/rejected": -0.3415605425834656, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "learning_rate": 2.8790199081163856e-07, | |
| "logits/chosen": -2.439944267272949, | |
| "logits/rejected": -2.3695976734161377, | |
| "logps/chosen": -266.065673828125, | |
| "logps/rejected": -225.2880859375, | |
| "loss": 0.5939, | |
| "rewards/accuracies": 0.6656249761581421, | |
| "rewards/chosen": -0.026788845658302307, | |
| "rewards/margins": 0.28384846448898315, | |
| "rewards/rejected": -0.31063732504844666, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "learning_rate": 2.802450229709035e-07, | |
| "logits/chosen": -2.399728298187256, | |
| "logits/rejected": -2.3489761352539062, | |
| "logps/chosen": -254.9022216796875, | |
| "logps/rejected": -213.33193969726562, | |
| "loss": 0.5847, | |
| "rewards/accuracies": 0.667187511920929, | |
| "rewards/chosen": -0.04024948924779892, | |
| "rewards/margins": 0.29891303181648254, | |
| "rewards/rejected": -0.33916252851486206, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "learning_rate": 2.725880551301684e-07, | |
| "logits/chosen": -2.462254047393799, | |
| "logits/rejected": -2.406602621078491, | |
| "logps/chosen": -274.6975402832031, | |
| "logps/rejected": -232.84591674804688, | |
| "loss": 0.5929, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.02199350856244564, | |
| "rewards/margins": 0.31067317724227905, | |
| "rewards/rejected": -0.33266669511795044, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "learning_rate": 2.649310872894334e-07, | |
| "logits/chosen": -2.4482955932617188, | |
| "logits/rejected": -2.4154446125030518, | |
| "logps/chosen": -275.00775146484375, | |
| "logps/rejected": -223.1331787109375, | |
| "loss": 0.5816, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.020305102691054344, | |
| "rewards/margins": 0.37037259340286255, | |
| "rewards/rejected": -0.3906777501106262, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "learning_rate": 2.572741194486983e-07, | |
| "logits/chosen": -2.448878765106201, | |
| "logits/rejected": -2.393206834793091, | |
| "logps/chosen": -273.81109619140625, | |
| "logps/rejected": -208.37985229492188, | |
| "loss": 0.5799, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.029856573790311813, | |
| "rewards/margins": 0.3645634055137634, | |
| "rewards/rejected": -0.3944200277328491, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "learning_rate": 2.496171516079632e-07, | |
| "logits/chosen": -2.4658501148223877, | |
| "logits/rejected": -2.399857521057129, | |
| "logps/chosen": -293.2225341796875, | |
| "logps/rejected": -239.4982452392578, | |
| "loss": 0.5813, | |
| "rewards/accuracies": 0.734375, | |
| "rewards/chosen": 0.0061371102929115295, | |
| "rewards/margins": 0.399463027715683, | |
| "rewards/rejected": -0.39332595467567444, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "learning_rate": 2.4196018376722816e-07, | |
| "logits/chosen": -2.429685115814209, | |
| "logits/rejected": -2.4006247520446777, | |
| "logps/chosen": -278.5813903808594, | |
| "logps/rejected": -228.4702911376953, | |
| "loss": 0.5864, | |
| "rewards/accuracies": 0.6859375238418579, | |
| "rewards/chosen": -0.02374974638223648, | |
| "rewards/margins": 0.37017589807510376, | |
| "rewards/rejected": -0.39392566680908203, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 2.343032159264931e-07, | |
| "logits/chosen": -2.403900146484375, | |
| "logits/rejected": -2.3333194255828857, | |
| "logps/chosen": -268.872802734375, | |
| "logps/rejected": -224.37728881835938, | |
| "loss": 0.579, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.020599449053406715, | |
| "rewards/margins": 0.431951105594635, | |
| "rewards/rejected": -0.4525505602359772, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "learning_rate": 2.26646248085758e-07, | |
| "logits/chosen": -2.383470058441162, | |
| "logits/rejected": -2.3353710174560547, | |
| "logps/chosen": -259.7237854003906, | |
| "logps/rejected": -217.79946899414062, | |
| "loss": 0.573, | |
| "rewards/accuracies": 0.723437488079071, | |
| "rewards/chosen": -0.0417955107986927, | |
| "rewards/margins": 0.39140504598617554, | |
| "rewards/rejected": -0.43320053815841675, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "learning_rate": 2.1898928024502298e-07, | |
| "logits/chosen": -2.4446728229522705, | |
| "logits/rejected": -2.3874154090881348, | |
| "logps/chosen": -263.4950256347656, | |
| "logps/rejected": -221.4724578857422, | |
| "loss": 0.5753, | |
| "rewards/accuracies": 0.7015625238418579, | |
| "rewards/chosen": -0.02173582836985588, | |
| "rewards/margins": 0.39518997073173523, | |
| "rewards/rejected": -0.4169258177280426, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "learning_rate": 2.113323124042879e-07, | |
| "logits/chosen": -2.4275262355804443, | |
| "logits/rejected": -2.3907971382141113, | |
| "logps/chosen": -271.2684326171875, | |
| "logps/rejected": -231.44381713867188, | |
| "loss": 0.5744, | |
| "rewards/accuracies": 0.721875011920929, | |
| "rewards/chosen": -0.020908143371343613, | |
| "rewards/margins": 0.41243448853492737, | |
| "rewards/rejected": -0.4333426058292389, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "learning_rate": 2.036753445635528e-07, | |
| "logits/chosen": -2.4363036155700684, | |
| "logits/rejected": -2.4147400856018066, | |
| "logps/chosen": -284.01824951171875, | |
| "logps/rejected": -238.273681640625, | |
| "loss": 0.569, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.018130071461200714, | |
| "rewards/margins": 0.4541456699371338, | |
| "rewards/rejected": -0.4722757339477539, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "learning_rate": 1.9601837672281775e-07, | |
| "logits/chosen": -2.4180634021759033, | |
| "logits/rejected": -2.3854622840881348, | |
| "logps/chosen": -270.1515808105469, | |
| "logps/rejected": -236.3723907470703, | |
| "loss": 0.576, | |
| "rewards/accuracies": 0.7046874761581421, | |
| "rewards/chosen": -0.05924994498491287, | |
| "rewards/margins": 0.37609511613845825, | |
| "rewards/rejected": -0.4353450834751129, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "learning_rate": 1.883614088820827e-07, | |
| "logits/chosen": -2.4381699562072754, | |
| "logits/rejected": -2.391515016555786, | |
| "logps/chosen": -268.735595703125, | |
| "logps/rejected": -224.8667755126953, | |
| "loss": 0.5849, | |
| "rewards/accuracies": 0.723437488079071, | |
| "rewards/chosen": -0.04678649455308914, | |
| "rewards/margins": 0.3817201852798462, | |
| "rewards/rejected": -0.42850667238235474, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_logits/chosen": -2.127939224243164, | |
| "eval_logits/rejected": -2.007131576538086, | |
| "eval_logps/chosen": -265.25634765625, | |
| "eval_logps/rejected": -224.01229858398438, | |
| "eval_loss": 0.5730655789375305, | |
| "eval_rewards/accuracies": 0.6899999976158142, | |
| "eval_rewards/chosen": -0.05778134614229202, | |
| "eval_rewards/margins": 0.4136123061180115, | |
| "eval_rewards/rejected": -0.4713936746120453, | |
| "eval_runtime": 277.7189, | |
| "eval_samples_per_second": 7.202, | |
| "eval_steps_per_second": 0.45, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 2.02, | |
| "learning_rate": 1.807044410413476e-07, | |
| "logits/chosen": -2.4097964763641357, | |
| "logits/rejected": -2.3763108253479004, | |
| "logps/chosen": -257.9292297363281, | |
| "logps/rejected": -236.3641815185547, | |
| "loss": 0.5772, | |
| "rewards/accuracies": 0.703125, | |
| "rewards/chosen": -0.05177872255444527, | |
| "rewards/margins": 0.39789050817489624, | |
| "rewards/rejected": -0.4496693015098572, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.07, | |
| "learning_rate": 1.7304747320061255e-07, | |
| "logits/chosen": -2.4072229862213135, | |
| "logits/rejected": -2.4033942222595215, | |
| "logps/chosen": -263.5710754394531, | |
| "logps/rejected": -230.6610107421875, | |
| "loss": 0.5772, | |
| "rewards/accuracies": 0.707812488079071, | |
| "rewards/chosen": -0.019381705671548843, | |
| "rewards/margins": 0.4000469744205475, | |
| "rewards/rejected": -0.41942867636680603, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.11, | |
| "learning_rate": 1.6539050535987747e-07, | |
| "logits/chosen": -2.4798319339752197, | |
| "logits/rejected": -2.370913028717041, | |
| "logps/chosen": -270.12432861328125, | |
| "logps/rejected": -225.058349609375, | |
| "loss": 0.5712, | |
| "rewards/accuracies": 0.7093750238418579, | |
| "rewards/chosen": -0.039138875901699066, | |
| "rewards/margins": 0.4365014135837555, | |
| "rewards/rejected": -0.47564029693603516, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.15, | |
| "learning_rate": 1.5773353751914243e-07, | |
| "logits/chosen": -2.4861385822296143, | |
| "logits/rejected": -2.425265312194824, | |
| "logps/chosen": -284.8677673339844, | |
| "logps/rejected": -229.98681640625, | |
| "loss": 0.5715, | |
| "rewards/accuracies": 0.7359374761581421, | |
| "rewards/chosen": -0.024127285927534103, | |
| "rewards/margins": 0.4679562449455261, | |
| "rewards/rejected": -0.4920835494995117, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 1.5007656967840735e-07, | |
| "logits/chosen": -2.383533000946045, | |
| "logits/rejected": -2.3430206775665283, | |
| "logps/chosen": -254.0509796142578, | |
| "logps/rejected": -230.5810089111328, | |
| "loss": 0.5677, | |
| "rewards/accuracies": 0.721875011920929, | |
| "rewards/chosen": -0.039885733276605606, | |
| "rewards/margins": 0.4442899823188782, | |
| "rewards/rejected": -0.4841756820678711, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "learning_rate": 1.4241960183767226e-07, | |
| "logits/chosen": -2.4291586875915527, | |
| "logits/rejected": -2.372559070587158, | |
| "logps/chosen": -282.87982177734375, | |
| "logps/rejected": -235.8987274169922, | |
| "loss": 0.573, | |
| "rewards/accuracies": 0.692187488079071, | |
| "rewards/chosen": -0.05947133153676987, | |
| "rewards/margins": 0.41908422112464905, | |
| "rewards/rejected": -0.4785555303096771, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "learning_rate": 1.347626339969372e-07, | |
| "logits/chosen": -2.423152208328247, | |
| "logits/rejected": -2.3877062797546387, | |
| "logps/chosen": -270.82269287109375, | |
| "logps/rejected": -242.1062469482422, | |
| "loss": 0.5759, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.01524378638714552, | |
| "rewards/margins": 0.40917444229125977, | |
| "rewards/rejected": -0.42441821098327637, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.31, | |
| "learning_rate": 1.2710566615620215e-07, | |
| "logits/chosen": -2.3735625743865967, | |
| "logits/rejected": -2.327951431274414, | |
| "logps/chosen": -274.332763671875, | |
| "logps/rejected": -225.1637420654297, | |
| "loss": 0.5594, | |
| "rewards/accuracies": 0.707812488079071, | |
| "rewards/chosen": -0.03534569963812828, | |
| "rewards/margins": 0.4445571005344391, | |
| "rewards/rejected": -0.47990283370018005, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.35, | |
| "learning_rate": 1.1944869831546706e-07, | |
| "logits/chosen": -2.3997702598571777, | |
| "logits/rejected": -2.3793346881866455, | |
| "logps/chosen": -267.025390625, | |
| "logps/rejected": -238.75692749023438, | |
| "loss": 0.5724, | |
| "rewards/accuracies": 0.6968749761581421, | |
| "rewards/chosen": -0.07640588283538818, | |
| "rewards/margins": 0.4082149565219879, | |
| "rewards/rejected": -0.4846208095550537, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 1.11791730474732e-07, | |
| "logits/chosen": -2.4132089614868164, | |
| "logits/rejected": -2.3745548725128174, | |
| "logps/chosen": -262.74658203125, | |
| "logps/rejected": -226.48898315429688, | |
| "loss": 0.5658, | |
| "rewards/accuracies": 0.7359374761581421, | |
| "rewards/chosen": -0.07082664221525192, | |
| "rewards/margins": 0.48243194818496704, | |
| "rewards/rejected": -0.5532585382461548, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.44, | |
| "learning_rate": 1.0413476263399694e-07, | |
| "logits/chosen": -2.451371669769287, | |
| "logits/rejected": -2.407169818878174, | |
| "logps/chosen": -269.4725646972656, | |
| "logps/rejected": -219.11929321289062, | |
| "loss": 0.5664, | |
| "rewards/accuracies": 0.729687511920929, | |
| "rewards/chosen": -0.07148631662130356, | |
| "rewards/margins": 0.4374977946281433, | |
| "rewards/rejected": -0.5089840888977051, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "learning_rate": 9.647779479326186e-08, | |
| "logits/chosen": -2.3929615020751953, | |
| "logits/rejected": -2.3882527351379395, | |
| "logps/chosen": -251.06576538085938, | |
| "logps/rejected": -224.4808807373047, | |
| "loss": 0.5768, | |
| "rewards/accuracies": 0.667187511920929, | |
| "rewards/chosen": -0.06799022853374481, | |
| "rewards/margins": 0.3539872467517853, | |
| "rewards/rejected": -0.42197751998901367, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "learning_rate": 8.88208269525268e-08, | |
| "logits/chosen": -2.3815102577209473, | |
| "logits/rejected": -2.3912739753723145, | |
| "logps/chosen": -260.7129821777344, | |
| "logps/rejected": -223.4461212158203, | |
| "loss": 0.566, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.06933742761611938, | |
| "rewards/margins": 0.41847410798072815, | |
| "rewards/rejected": -0.48781150579452515, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 8.116385911179173e-08, | |
| "logits/chosen": -2.3711135387420654, | |
| "logits/rejected": -2.3626675605773926, | |
| "logps/chosen": -279.54461669921875, | |
| "logps/rejected": -219.48974609375, | |
| "loss": 0.5701, | |
| "rewards/accuracies": 0.734375, | |
| "rewards/chosen": -0.023505648598074913, | |
| "rewards/margins": 0.4627605378627777, | |
| "rewards/rejected": -0.4862661361694336, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "learning_rate": 7.350689127105667e-08, | |
| "logits/chosen": -2.4528985023498535, | |
| "logits/rejected": -2.3787388801574707, | |
| "logps/chosen": -272.67572021484375, | |
| "logps/rejected": -232.7178192138672, | |
| "loss": 0.5603, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.0497988685965538, | |
| "rewards/margins": 0.46719294786453247, | |
| "rewards/rejected": -0.5169917941093445, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "learning_rate": 6.584992343032159e-08, | |
| "logits/chosen": -2.355626106262207, | |
| "logits/rejected": -2.342153549194336, | |
| "logps/chosen": -263.1979675292969, | |
| "logps/rejected": -229.1007843017578, | |
| "loss": 0.5752, | |
| "rewards/accuracies": 0.690625011920929, | |
| "rewards/chosen": -0.04885732755064964, | |
| "rewards/margins": 0.42579683661460876, | |
| "rewards/rejected": -0.4746541380882263, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.69, | |
| "learning_rate": 5.819295558958652e-08, | |
| "logits/chosen": -2.3994874954223633, | |
| "logits/rejected": -2.34912109375, | |
| "logps/chosen": -288.324462890625, | |
| "logps/rejected": -222.3997344970703, | |
| "loss": 0.5646, | |
| "rewards/accuracies": 0.7484375238418579, | |
| "rewards/chosen": -0.021456807851791382, | |
| "rewards/margins": 0.5244570374488831, | |
| "rewards/rejected": -0.545913815498352, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.73, | |
| "learning_rate": 5.0535987748851455e-08, | |
| "logits/chosen": -2.4647653102874756, | |
| "logits/rejected": -2.4115538597106934, | |
| "logps/chosen": -275.77947998046875, | |
| "logps/rejected": -231.9734649658203, | |
| "loss": 0.5621, | |
| "rewards/accuracies": 0.7015625238418579, | |
| "rewards/chosen": -0.07127931714057922, | |
| "rewards/margins": 0.41370710730552673, | |
| "rewards/rejected": -0.48498645424842834, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.77, | |
| "learning_rate": 4.287901990811638e-08, | |
| "logits/chosen": -2.4428927898406982, | |
| "logits/rejected": -2.3569588661193848, | |
| "logps/chosen": -268.5105285644531, | |
| "logps/rejected": -244.9532928466797, | |
| "loss": 0.5636, | |
| "rewards/accuracies": 0.7109375, | |
| "rewards/chosen": -0.048072461038827896, | |
| "rewards/margins": 0.46722808480262756, | |
| "rewards/rejected": -0.5153006315231323, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "learning_rate": 3.522205206738132e-08, | |
| "logits/chosen": -2.4071362018585205, | |
| "logits/rejected": -2.3730602264404297, | |
| "logps/chosen": -275.3606872558594, | |
| "logps/rejected": -230.1616668701172, | |
| "loss": 0.5682, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.076107919216156, | |
| "rewards/margins": 0.44682103395462036, | |
| "rewards/rejected": -0.5229289531707764, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "learning_rate": 2.7565084226646246e-08, | |
| "logits/chosen": -2.3889846801757812, | |
| "logits/rejected": -2.376112937927246, | |
| "logps/chosen": -264.30804443359375, | |
| "logps/rejected": -236.71640014648438, | |
| "loss": 0.5747, | |
| "rewards/accuracies": 0.7109375, | |
| "rewards/chosen": -0.04645932838320732, | |
| "rewards/margins": 0.4625559449195862, | |
| "rewards/rejected": -0.5090152621269226, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "learning_rate": 1.9908116385911178e-08, | |
| "logits/chosen": -2.4064643383026123, | |
| "logits/rejected": -2.388768434524536, | |
| "logps/chosen": -260.59393310546875, | |
| "logps/rejected": -227.6177215576172, | |
| "loss": 0.5718, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": -0.09053535759449005, | |
| "rewards/margins": 0.3823908865451813, | |
| "rewards/rejected": -0.47292619943618774, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 1.225114854517611e-08, | |
| "logits/chosen": -2.436859130859375, | |
| "logits/rejected": -2.3719522953033447, | |
| "logps/chosen": -280.87774658203125, | |
| "logps/rejected": -221.83944702148438, | |
| "loss": 0.5696, | |
| "rewards/accuracies": 0.7093750238418579, | |
| "rewards/chosen": -0.06765580177307129, | |
| "rewards/margins": 0.4343256950378418, | |
| "rewards/rejected": -0.5019814968109131, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.97, | |
| "learning_rate": 4.594180704441042e-09, | |
| "logits/chosen": -2.4007372856140137, | |
| "logits/rejected": -2.378627300262451, | |
| "logps/chosen": -257.9851989746094, | |
| "logps/rejected": -225.0294189453125, | |
| "loss": 0.5671, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.053695209324359894, | |
| "rewards/margins": 0.4172247052192688, | |
| "rewards/rejected": -0.4709199070930481, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_logits/chosen": -2.1214964389801025, | |
| "eval_logits/rejected": -2.000164031982422, | |
| "eval_logps/chosen": -265.42364501953125, | |
| "eval_logps/rejected": -224.6269989013672, | |
| "eval_loss": 0.5655443072319031, | |
| "eval_rewards/accuracies": 0.699999988079071, | |
| "eval_rewards/chosen": -0.07451467216014862, | |
| "eval_rewards/margins": 0.4583480656147003, | |
| "eval_rewards/rejected": -0.5328627228736877, | |
| "eval_runtime": 278.3863, | |
| "eval_samples_per_second": 7.184, | |
| "eval_steps_per_second": 0.449, | |
| "step": 726 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 726, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6037390495627379, | |
| "train_runtime": 36278.6969, | |
| "train_samples_per_second": 5.124, | |
| "train_steps_per_second": 0.02 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 726, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 0.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |