| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9997382884061764, | |
| "eval_steps": 100, | |
| "global_step": 955, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 5.208333333333333e-09, | |
| "logits/chosen": -2.570180892944336, | |
| "logits/rejected": -2.5666794776916504, | |
| "logps/chosen": -302.8643798828125, | |
| "logps/rejected": -232.7855682373047, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": -2.547344207763672, | |
| "logits/rejected": -2.532263994216919, | |
| "logps/chosen": -294.40087890625, | |
| "logps/rejected": -276.11126708984375, | |
| "loss": 0.6879, | |
| "rewards/accuracies": 0.4340277910232544, | |
| "rewards/chosen": 0.0004416291194502264, | |
| "rewards/margins": -0.0003637511981651187, | |
| "rewards/rejected": 0.000805380754172802, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.3985350131988525, | |
| "logits/rejected": -2.426626682281494, | |
| "logps/chosen": -244.53689575195312, | |
| "logps/rejected": -248.93624877929688, | |
| "loss": 0.6882, | |
| "rewards/accuracies": 0.49687498807907104, | |
| "rewards/chosen": -0.002964673563838005, | |
| "rewards/margins": 0.00020738500461447984, | |
| "rewards/rejected": -0.0031720586121082306, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": -2.5088179111480713, | |
| "logits/rejected": -2.4950203895568848, | |
| "logps/chosen": -265.425537109375, | |
| "logps/rejected": -244.62191772460938, | |
| "loss": 0.6868, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -0.003862012643367052, | |
| "rewards/margins": -0.0030291248112916946, | |
| "rewards/rejected": -0.0008328882977366447, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.472740888595581, | |
| "logits/rejected": -2.498608112335205, | |
| "logps/chosen": -264.92767333984375, | |
| "logps/rejected": -260.81536865234375, | |
| "loss": 0.6883, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": 0.002251622499898076, | |
| "rewards/margins": 0.002210150007158518, | |
| "rewards/rejected": 4.1472725570201874e-05, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.604166666666667e-07, | |
| "logits/chosen": -2.511932611465454, | |
| "logits/rejected": -2.4823100566864014, | |
| "logps/chosen": -255.1762237548828, | |
| "logps/rejected": -249.55859375, | |
| "loss": 0.6856, | |
| "rewards/accuracies": 0.5406249761581421, | |
| "rewards/chosen": 0.0019304720917716622, | |
| "rewards/margins": 0.0034704413264989853, | |
| "rewards/rejected": -0.0015399687690660357, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.4945309162139893, | |
| "logits/rejected": -2.5101318359375, | |
| "logps/chosen": -288.4325256347656, | |
| "logps/rejected": -261.18310546875, | |
| "loss": 0.6867, | |
| "rewards/accuracies": 0.503125011920929, | |
| "rewards/chosen": -0.002425801707431674, | |
| "rewards/margins": 0.002398231066763401, | |
| "rewards/rejected": -0.004824032541364431, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.645833333333333e-07, | |
| "logits/chosen": -2.521496534347534, | |
| "logits/rejected": -2.506330966949463, | |
| "logps/chosen": -275.6085510253906, | |
| "logps/rejected": -259.58392333984375, | |
| "loss": 0.6867, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.0025966810062527657, | |
| "rewards/margins": 0.007796707563102245, | |
| "rewards/rejected": -0.005200026091188192, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.456026315689087, | |
| "logits/rejected": -2.4563703536987305, | |
| "logps/chosen": -295.45941162109375, | |
| "logps/rejected": -263.0380859375, | |
| "loss": 0.6859, | |
| "rewards/accuracies": 0.4593749940395355, | |
| "rewards/chosen": -2.993037924170494e-05, | |
| "rewards/margins": -0.000812772021163255, | |
| "rewards/rejected": 0.0007828413508832455, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": -2.4610087871551514, | |
| "logits/rejected": -2.4633374214172363, | |
| "logps/chosen": -282.82684326171875, | |
| "logps/rejected": -245.01785278320312, | |
| "loss": 0.6867, | |
| "rewards/accuracies": 0.4906249940395355, | |
| "rewards/chosen": -0.003978222608566284, | |
| "rewards/margins": -0.003358404617756605, | |
| "rewards/rejected": -0.0006198181654326618, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.976717112922002e-07, | |
| "logits/chosen": -2.468822479248047, | |
| "logits/rejected": -2.4451956748962402, | |
| "logps/chosen": -264.9376525878906, | |
| "logps/rejected": -259.81182861328125, | |
| "loss": 0.6865, | |
| "rewards/accuracies": 0.528124988079071, | |
| "rewards/chosen": 0.004453591071069241, | |
| "rewards/margins": 0.004475563298910856, | |
| "rewards/rejected": -2.197279900428839e-05, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.918509895227007e-07, | |
| "logits/chosen": -2.487288236618042, | |
| "logits/rejected": -2.4807934761047363, | |
| "logps/chosen": -263.60906982421875, | |
| "logps/rejected": -244.46047973632812, | |
| "loss": 0.6852, | |
| "rewards/accuracies": 0.534375011920929, | |
| "rewards/chosen": 0.00421065092086792, | |
| "rewards/margins": 0.007397785782814026, | |
| "rewards/rejected": -0.003187134861946106, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.860302677532014e-07, | |
| "logits/chosen": -2.4685282707214355, | |
| "logits/rejected": -2.497313976287842, | |
| "logps/chosen": -277.75543212890625, | |
| "logps/rejected": -245.1427764892578, | |
| "loss": 0.6852, | |
| "rewards/accuracies": 0.5531250238418579, | |
| "rewards/chosen": 0.0019862186163663864, | |
| "rewards/margins": 0.0036893251817673445, | |
| "rewards/rejected": -0.0017031064489856362, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.802095459837019e-07, | |
| "logits/chosen": -2.512012481689453, | |
| "logits/rejected": -2.498793840408325, | |
| "logps/chosen": -283.26959228515625, | |
| "logps/rejected": -265.85369873046875, | |
| "loss": 0.6838, | |
| "rewards/accuracies": 0.534375011920929, | |
| "rewards/chosen": 0.006660536862909794, | |
| "rewards/margins": 0.007559786085039377, | |
| "rewards/rejected": -0.0008992499788291752, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.743888242142026e-07, | |
| "logits/chosen": -2.471496105194092, | |
| "logits/rejected": -2.4844748973846436, | |
| "logps/chosen": -255.3807373046875, | |
| "logps/rejected": -242.02658081054688, | |
| "loss": 0.6857, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": 0.0008136004325933754, | |
| "rewards/margins": 4.014703517896123e-05, | |
| "rewards/rejected": 0.0007734533282928169, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.685681024447031e-07, | |
| "logits/chosen": -2.471660614013672, | |
| "logits/rejected": -2.463984966278076, | |
| "logps/chosen": -287.92523193359375, | |
| "logps/rejected": -260.02313232421875, | |
| "loss": 0.6837, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": 0.00581919914111495, | |
| "rewards/margins": 0.008211213164031506, | |
| "rewards/rejected": -0.0023920147214084864, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.627473806752037e-07, | |
| "logits/chosen": -2.520613193511963, | |
| "logits/rejected": -2.4651389122009277, | |
| "logps/chosen": -279.1468505859375, | |
| "logps/rejected": -264.69915771484375, | |
| "loss": 0.682, | |
| "rewards/accuracies": 0.5531250238418579, | |
| "rewards/chosen": 0.008319775573909283, | |
| "rewards/margins": 0.011747308075428009, | |
| "rewards/rejected": -0.003427532035857439, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.5692665890570433e-07, | |
| "logits/chosen": -2.5029516220092773, | |
| "logits/rejected": -2.496436595916748, | |
| "logps/chosen": -279.58843994140625, | |
| "logps/rejected": -243.44253540039062, | |
| "loss": 0.6817, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.008326916955411434, | |
| "rewards/margins": 0.014212583191692829, | |
| "rewards/rejected": -0.00588566530495882, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.5110593713620486e-07, | |
| "logits/chosen": -2.4697937965393066, | |
| "logits/rejected": -2.4516141414642334, | |
| "logps/chosen": -236.89675903320312, | |
| "logps/rejected": -253.85598754882812, | |
| "loss": 0.6832, | |
| "rewards/accuracies": 0.5531250238418579, | |
| "rewards/chosen": 0.00583054032176733, | |
| "rewards/margins": 0.012070106342434883, | |
| "rewards/rejected": -0.00623956648632884, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.4528521536670544e-07, | |
| "logits/chosen": -2.5360610485076904, | |
| "logits/rejected": -2.505577802658081, | |
| "logps/chosen": -291.94586181640625, | |
| "logps/rejected": -260.87054443359375, | |
| "loss": 0.6801, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.011800072155892849, | |
| "rewards/margins": 0.02227787859737873, | |
| "rewards/rejected": -0.010477803647518158, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.3946449359720607e-07, | |
| "logits/chosen": -2.4782238006591797, | |
| "logits/rejected": -2.508026599884033, | |
| "logps/chosen": -258.73895263671875, | |
| "logps/rejected": -252.0211639404297, | |
| "loss": 0.6809, | |
| "rewards/accuracies": 0.6031249761581421, | |
| "rewards/chosen": 0.009132781066000462, | |
| "rewards/margins": 0.0161266028881073, | |
| "rewards/rejected": -0.006993822753429413, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.336437718277066e-07, | |
| "logits/chosen": -2.4700803756713867, | |
| "logits/rejected": -2.479269504547119, | |
| "logps/chosen": -272.45904541015625, | |
| "logps/rejected": -256.2876281738281, | |
| "loss": 0.6783, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.016153430566191673, | |
| "rewards/margins": 0.023166943341493607, | |
| "rewards/rejected": -0.007013511843979359, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.278230500582072e-07, | |
| "logits/chosen": -2.4889426231384277, | |
| "logits/rejected": -2.5127673149108887, | |
| "logps/chosen": -267.21966552734375, | |
| "logps/rejected": -258.29205322265625, | |
| "loss": 0.6793, | |
| "rewards/accuracies": 0.6031249761581421, | |
| "rewards/chosen": 0.005786339286714792, | |
| "rewards/margins": 0.01781514100730419, | |
| "rewards/rejected": -0.012028800323605537, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.220023282887078e-07, | |
| "logits/chosen": -2.4950485229492188, | |
| "logits/rejected": -2.4980530738830566, | |
| "logps/chosen": -295.85028076171875, | |
| "logps/rejected": -259.36529541015625, | |
| "loss": 0.6777, | |
| "rewards/accuracies": 0.596875011920929, | |
| "rewards/chosen": 0.007827522233128548, | |
| "rewards/margins": 0.014727133326232433, | |
| "rewards/rejected": -0.006899611093103886, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.1618160651920834e-07, | |
| "logits/chosen": -2.466526508331299, | |
| "logits/rejected": -2.4315407276153564, | |
| "logps/chosen": -282.66192626953125, | |
| "logps/rejected": -248.1639862060547, | |
| "loss": 0.6778, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": 0.01533576101064682, | |
| "rewards/margins": 0.017889009788632393, | |
| "rewards/rejected": -0.0025532490108162165, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.103608847497089e-07, | |
| "logits/chosen": -2.4848718643188477, | |
| "logits/rejected": -2.443408250808716, | |
| "logps/chosen": -288.2550964355469, | |
| "logps/rejected": -271.9901123046875, | |
| "loss": 0.6761, | |
| "rewards/accuracies": 0.621874988079071, | |
| "rewards/chosen": 0.013339035212993622, | |
| "rewards/margins": 0.023220960050821304, | |
| "rewards/rejected": -0.009881924837827682, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.0454016298020956e-07, | |
| "logits/chosen": -2.4762511253356934, | |
| "logits/rejected": -2.446619749069214, | |
| "logps/chosen": -268.5465393066406, | |
| "logps/rejected": -251.9985809326172, | |
| "loss": 0.6763, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.006427633110433817, | |
| "rewards/margins": 0.015916740521788597, | |
| "rewards/rejected": -0.009489107877016068, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 3.987194412107101e-07, | |
| "logits/chosen": -2.4372754096984863, | |
| "logits/rejected": -2.450491189956665, | |
| "logps/chosen": -263.63262939453125, | |
| "logps/rejected": -252.0638427734375, | |
| "loss": 0.6735, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": 0.013371949084103107, | |
| "rewards/margins": 0.023976340889930725, | |
| "rewards/rejected": -0.010604391805827618, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 3.9289871944121066e-07, | |
| "logits/chosen": -2.5451080799102783, | |
| "logits/rejected": -2.516885995864868, | |
| "logps/chosen": -282.615234375, | |
| "logps/rejected": -242.2822265625, | |
| "loss": 0.6722, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": 0.0192705187946558, | |
| "rewards/margins": 0.029350727796554565, | |
| "rewards/rejected": -0.010080209001898766, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 3.870779976717113e-07, | |
| "logits/chosen": -2.4943687915802, | |
| "logits/rejected": -2.5217864513397217, | |
| "logps/chosen": -277.5215759277344, | |
| "logps/rejected": -229.96298217773438, | |
| "loss": 0.6722, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": 0.015035443007946014, | |
| "rewards/margins": 0.024382654577493668, | |
| "rewards/rejected": -0.009347210638225079, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 3.812572759022118e-07, | |
| "logits/chosen": -2.49491810798645, | |
| "logits/rejected": -2.501307249069214, | |
| "logps/chosen": -281.02484130859375, | |
| "logps/rejected": -237.4232940673828, | |
| "loss": 0.6726, | |
| "rewards/accuracies": 0.628125011920929, | |
| "rewards/chosen": 0.018051721155643463, | |
| "rewards/margins": 0.032371118664741516, | |
| "rewards/rejected": -0.014319395646452904, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 3.754365541327124e-07, | |
| "logits/chosen": -2.4540202617645264, | |
| "logits/rejected": -2.435542583465576, | |
| "logps/chosen": -258.79193115234375, | |
| "logps/rejected": -253.71694946289062, | |
| "loss": 0.673, | |
| "rewards/accuracies": 0.6468750238418579, | |
| "rewards/chosen": 0.015593824908137321, | |
| "rewards/margins": 0.0299003217369318, | |
| "rewards/rejected": -0.014306495897471905, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 3.6961583236321304e-07, | |
| "logits/chosen": -2.488478422164917, | |
| "logits/rejected": -2.471566677093506, | |
| "logps/chosen": -287.9583435058594, | |
| "logps/rejected": -256.38427734375, | |
| "loss": 0.6681, | |
| "rewards/accuracies": 0.653124988079071, | |
| "rewards/chosen": 0.01864878088235855, | |
| "rewards/margins": 0.03612912446260452, | |
| "rewards/rejected": -0.01748034544289112, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 3.637951105937136e-07, | |
| "logits/chosen": -2.469587802886963, | |
| "logits/rejected": -2.4525370597839355, | |
| "logps/chosen": -249.15280151367188, | |
| "logps/rejected": -235.3613739013672, | |
| "loss": 0.6664, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": 0.013197916559875011, | |
| "rewards/margins": 0.030393391847610474, | |
| "rewards/rejected": -0.01719547249376774, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 3.579743888242142e-07, | |
| "logits/chosen": -2.486199378967285, | |
| "logits/rejected": -2.491894006729126, | |
| "logps/chosen": -279.8724060058594, | |
| "logps/rejected": -249.26123046875, | |
| "loss": 0.6691, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.016388490796089172, | |
| "rewards/margins": 0.02631198987364769, | |
| "rewards/rejected": -0.009923500940203667, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.521536670547148e-07, | |
| "logits/chosen": -2.501917839050293, | |
| "logits/rejected": -2.500124216079712, | |
| "logps/chosen": -277.4295349121094, | |
| "logps/rejected": -261.6769104003906, | |
| "loss": 0.665, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.019429894164204597, | |
| "rewards/margins": 0.030850976705551147, | |
| "rewards/rejected": -0.011421086266636848, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.4633294528521536e-07, | |
| "logits/chosen": -2.5457377433776855, | |
| "logits/rejected": -2.5352020263671875, | |
| "logps/chosen": -280.8336486816406, | |
| "logps/rejected": -259.36895751953125, | |
| "loss": 0.6614, | |
| "rewards/accuracies": 0.6156250238418579, | |
| "rewards/chosen": 0.01814255118370056, | |
| "rewards/margins": 0.03848281502723694, | |
| "rewards/rejected": -0.02034026011824608, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.4051222351571594e-07, | |
| "logits/chosen": -2.5180299282073975, | |
| "logits/rejected": -2.4479191303253174, | |
| "logps/chosen": -274.1128845214844, | |
| "logps/rejected": -259.3582763671875, | |
| "loss": 0.6664, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": 0.023864779621362686, | |
| "rewards/margins": 0.041657593101263046, | |
| "rewards/rejected": -0.01779281720519066, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.346915017462165e-07, | |
| "logits/chosen": -2.504103899002075, | |
| "logits/rejected": -2.4852070808410645, | |
| "logps/chosen": -269.26409912109375, | |
| "logps/rejected": -253.5343780517578, | |
| "loss": 0.663, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": 0.021114524453878403, | |
| "rewards/margins": 0.042601972818374634, | |
| "rewards/rejected": -0.021487446501851082, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.288707799767171e-07, | |
| "logits/chosen": -2.4515504837036133, | |
| "logits/rejected": -2.447246551513672, | |
| "logps/chosen": -259.05401611328125, | |
| "logps/rejected": -240.0911407470703, | |
| "loss": 0.663, | |
| "rewards/accuracies": 0.628125011920929, | |
| "rewards/chosen": 0.014576256275177002, | |
| "rewards/margins": 0.04004546254873276, | |
| "rewards/rejected": -0.025469202548265457, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.230500582072177e-07, | |
| "logits/chosen": -2.509997844696045, | |
| "logits/rejected": -2.522238254547119, | |
| "logps/chosen": -282.1490478515625, | |
| "logps/rejected": -243.8563690185547, | |
| "loss": 0.6624, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.025715211406350136, | |
| "rewards/margins": 0.043169617652893066, | |
| "rewards/rejected": -0.01745440624654293, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.1722933643771827e-07, | |
| "logits/chosen": -2.498192548751831, | |
| "logits/rejected": -2.4921789169311523, | |
| "logps/chosen": -271.2725830078125, | |
| "logps/rejected": -244.8346710205078, | |
| "loss": 0.6594, | |
| "rewards/accuracies": 0.6468750238418579, | |
| "rewards/chosen": 0.02634851261973381, | |
| "rewards/margins": 0.0405191034078598, | |
| "rewards/rejected": -0.014170585200190544, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.1140861466821885e-07, | |
| "logits/chosen": -2.4811172485351562, | |
| "logits/rejected": -2.4510648250579834, | |
| "logps/chosen": -253.666015625, | |
| "logps/rejected": -243.52877807617188, | |
| "loss": 0.6617, | |
| "rewards/accuracies": 0.628125011920929, | |
| "rewards/chosen": 0.024721184745430946, | |
| "rewards/margins": 0.03779374435544014, | |
| "rewards/rejected": -0.013072559610009193, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.0558789289871943e-07, | |
| "logits/chosen": -2.5075089931488037, | |
| "logits/rejected": -2.4863505363464355, | |
| "logps/chosen": -279.38818359375, | |
| "logps/rejected": -254.1600799560547, | |
| "loss": 0.6609, | |
| "rewards/accuracies": 0.596875011920929, | |
| "rewards/chosen": 0.02350917086005211, | |
| "rewards/margins": 0.03294364735484123, | |
| "rewards/rejected": -0.009434476494789124, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 2.9976717112922e-07, | |
| "logits/chosen": -2.5201539993286133, | |
| "logits/rejected": -2.511667013168335, | |
| "logps/chosen": -272.9944152832031, | |
| "logps/rejected": -233.72854614257812, | |
| "loss": 0.6589, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.020865267142653465, | |
| "rewards/margins": 0.04071135073900223, | |
| "rewards/rejected": -0.019846081733703613, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 2.939464493597206e-07, | |
| "logits/chosen": -2.4186248779296875, | |
| "logits/rejected": -2.3882439136505127, | |
| "logps/chosen": -244.801025390625, | |
| "logps/rejected": -241.0476837158203, | |
| "loss": 0.6588, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.019629117101430893, | |
| "rewards/margins": 0.037906430661678314, | |
| "rewards/rejected": -0.01827731356024742, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 2.8812572759022117e-07, | |
| "logits/chosen": -2.5232837200164795, | |
| "logits/rejected": -2.5000243186950684, | |
| "logps/chosen": -294.4573669433594, | |
| "logps/rejected": -256.6082458496094, | |
| "loss": 0.6559, | |
| "rewards/accuracies": 0.6156250238418579, | |
| "rewards/chosen": 0.02450462244451046, | |
| "rewards/margins": 0.0395827516913414, | |
| "rewards/rejected": -0.015078130178153515, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.8230500582072175e-07, | |
| "logits/chosen": -2.4724531173706055, | |
| "logits/rejected": -2.4754977226257324, | |
| "logps/chosen": -283.3002624511719, | |
| "logps/rejected": -251.4059295654297, | |
| "loss": 0.6557, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.024192675948143005, | |
| "rewards/margins": 0.05236636474728584, | |
| "rewards/rejected": -0.028173688799142838, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.7648428405122233e-07, | |
| "logits/chosen": -2.410810947418213, | |
| "logits/rejected": -2.4008870124816895, | |
| "logps/chosen": -242.271484375, | |
| "logps/rejected": -250.8043212890625, | |
| "loss": 0.6593, | |
| "rewards/accuracies": 0.6031249761581421, | |
| "rewards/chosen": 0.012235969305038452, | |
| "rewards/margins": 0.02636186219751835, | |
| "rewards/rejected": -0.01412589568644762, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.706635622817229e-07, | |
| "logits/chosen": -2.4645464420318604, | |
| "logits/rejected": -2.469316244125366, | |
| "logps/chosen": -272.09796142578125, | |
| "logps/rejected": -238.6561737060547, | |
| "loss": 0.6533, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": 0.024095356464385986, | |
| "rewards/margins": 0.05057717487215996, | |
| "rewards/rejected": -0.02648181840777397, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.648428405122235e-07, | |
| "logits/chosen": -2.508164167404175, | |
| "logits/rejected": -2.4578819274902344, | |
| "logps/chosen": -277.1492004394531, | |
| "logps/rejected": -256.0687561035156, | |
| "loss": 0.6531, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": 0.03192506358027458, | |
| "rewards/margins": 0.05114533379673958, | |
| "rewards/rejected": -0.019220268353819847, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.590221187427241e-07, | |
| "logits/chosen": -2.5105607509613037, | |
| "logits/rejected": -2.4753470420837402, | |
| "logps/chosen": -263.40966796875, | |
| "logps/rejected": -244.8257598876953, | |
| "loss": 0.6522, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": 0.027492288500070572, | |
| "rewards/margins": 0.04991314187645912, | |
| "rewards/rejected": -0.022420858964323997, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.5320139697322466e-07, | |
| "logits/chosen": -2.5335867404937744, | |
| "logits/rejected": -2.468331813812256, | |
| "logps/chosen": -272.64385986328125, | |
| "logps/rejected": -244.65286254882812, | |
| "loss": 0.651, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.03846021741628647, | |
| "rewards/margins": 0.06949218362569809, | |
| "rewards/rejected": -0.03103196993470192, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.4738067520372524e-07, | |
| "logits/chosen": -2.505664110183716, | |
| "logits/rejected": -2.512172222137451, | |
| "logps/chosen": -280.70819091796875, | |
| "logps/rejected": -255.12228393554688, | |
| "loss": 0.6487, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": 0.035160940140485764, | |
| "rewards/margins": 0.06470004469156265, | |
| "rewards/rejected": -0.029539108276367188, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.415599534342258e-07, | |
| "logits/chosen": -2.5341105461120605, | |
| "logits/rejected": -2.4813685417175293, | |
| "logps/chosen": -280.51751708984375, | |
| "logps/rejected": -252.9778289794922, | |
| "loss": 0.6471, | |
| "rewards/accuracies": 0.653124988079071, | |
| "rewards/chosen": 0.03973756358027458, | |
| "rewards/margins": 0.06431148201227188, | |
| "rewards/rejected": -0.02457391656935215, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.3573923166472642e-07, | |
| "logits/chosen": -2.506739377975464, | |
| "logits/rejected": -2.487623691558838, | |
| "logps/chosen": -272.10675048828125, | |
| "logps/rejected": -258.69647216796875, | |
| "loss": 0.6479, | |
| "rewards/accuracies": 0.6343749761581421, | |
| "rewards/chosen": 0.028482938185334206, | |
| "rewards/margins": 0.057554639875888824, | |
| "rewards/rejected": -0.02907169796526432, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.2991850989522698e-07, | |
| "logits/chosen": -2.442371129989624, | |
| "logits/rejected": -2.4590439796447754, | |
| "logps/chosen": -280.783935546875, | |
| "logps/rejected": -227.5161895751953, | |
| "loss": 0.6513, | |
| "rewards/accuracies": 0.628125011920929, | |
| "rewards/chosen": 0.02295442670583725, | |
| "rewards/margins": 0.05334927886724472, | |
| "rewards/rejected": -0.03039485774934292, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.2409778812572759e-07, | |
| "logits/chosen": -2.470194101333618, | |
| "logits/rejected": -2.441986560821533, | |
| "logps/chosen": -255.4265594482422, | |
| "logps/rejected": -241.7792205810547, | |
| "loss": 0.6487, | |
| "rewards/accuracies": 0.6343749761581421, | |
| "rewards/chosen": 0.024136796593666077, | |
| "rewards/margins": 0.04882895201444626, | |
| "rewards/rejected": -0.02469216100871563, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.1827706635622817e-07, | |
| "logits/chosen": -2.5127058029174805, | |
| "logits/rejected": -2.496788263320923, | |
| "logps/chosen": -262.003662109375, | |
| "logps/rejected": -251.4923858642578, | |
| "loss": 0.6508, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": 0.025721842423081398, | |
| "rewards/margins": 0.056790102273225784, | |
| "rewards/rejected": -0.031068259850144386, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 2.1245634458672875e-07, | |
| "logits/chosen": -2.4597954750061035, | |
| "logits/rejected": -2.4486401081085205, | |
| "logps/chosen": -272.46258544921875, | |
| "logps/rejected": -250.1220703125, | |
| "loss": 0.6465, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": 0.03377040475606918, | |
| "rewards/margins": 0.05980812385678291, | |
| "rewards/rejected": -0.02603771723806858, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 2.0663562281722933e-07, | |
| "logits/chosen": -2.468313217163086, | |
| "logits/rejected": -2.4899404048919678, | |
| "logps/chosen": -280.9910888671875, | |
| "logps/rejected": -279.2497863769531, | |
| "loss": 0.6493, | |
| "rewards/accuracies": 0.6468750238418579, | |
| "rewards/chosen": 0.03328691050410271, | |
| "rewards/margins": 0.05028299614787102, | |
| "rewards/rejected": -0.016996093094348907, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 2.008149010477299e-07, | |
| "logits/chosen": -2.515079975128174, | |
| "logits/rejected": -2.471588611602783, | |
| "logps/chosen": -284.3990478515625, | |
| "logps/rejected": -246.21835327148438, | |
| "loss": 0.6456, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.03540956228971481, | |
| "rewards/margins": 0.062371380627155304, | |
| "rewards/rejected": -0.026961814612150192, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.949941792782305e-07, | |
| "logits/chosen": -2.5138742923736572, | |
| "logits/rejected": -2.451676845550537, | |
| "logps/chosen": -272.3465881347656, | |
| "logps/rejected": -240.0901641845703, | |
| "loss": 0.6492, | |
| "rewards/accuracies": 0.6343749761581421, | |
| "rewards/chosen": 0.03495832532644272, | |
| "rewards/margins": 0.05197330191731453, | |
| "rewards/rejected": -0.017014967277646065, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.8917345750873107e-07, | |
| "logits/chosen": -2.4583332538604736, | |
| "logits/rejected": -2.468278169631958, | |
| "logps/chosen": -264.42913818359375, | |
| "logps/rejected": -256.8759765625, | |
| "loss": 0.6451, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.02500128373503685, | |
| "rewards/margins": 0.04210829734802246, | |
| "rewards/rejected": -0.01710701361298561, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.8335273573923165e-07, | |
| "logits/chosen": -2.5137476921081543, | |
| "logits/rejected": -2.473841667175293, | |
| "logps/chosen": -252.57705688476562, | |
| "logps/rejected": -230.48471069335938, | |
| "loss": 0.6434, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": 0.03589923307299614, | |
| "rewards/margins": 0.06454572081565857, | |
| "rewards/rejected": -0.028646480292081833, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.7753201396973226e-07, | |
| "logits/chosen": -2.45261549949646, | |
| "logits/rejected": -2.4653515815734863, | |
| "logps/chosen": -274.7919006347656, | |
| "logps/rejected": -261.8819580078125, | |
| "loss": 0.6407, | |
| "rewards/accuracies": 0.6468750238418579, | |
| "rewards/chosen": 0.03197002038359642, | |
| "rewards/margins": 0.07201725244522095, | |
| "rewards/rejected": -0.040047239512205124, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.7171129220023281e-07, | |
| "logits/chosen": -2.4911041259765625, | |
| "logits/rejected": -2.4684813022613525, | |
| "logps/chosen": -267.3164978027344, | |
| "logps/rejected": -232.7432403564453, | |
| "loss": 0.6489, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.022824671119451523, | |
| "rewards/margins": 0.04568014293909073, | |
| "rewards/rejected": -0.022855471819639206, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.658905704307334e-07, | |
| "logits/chosen": -2.4661662578582764, | |
| "logits/rejected": -2.429795026779175, | |
| "logps/chosen": -267.6933898925781, | |
| "logps/rejected": -238.0891876220703, | |
| "loss": 0.6461, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.029722299426794052, | |
| "rewards/margins": 0.061273299157619476, | |
| "rewards/rejected": -0.031550996005535126, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.60069848661234e-07, | |
| "logits/chosen": -2.494511127471924, | |
| "logits/rejected": -2.505267381668091, | |
| "logps/chosen": -287.6944274902344, | |
| "logps/rejected": -259.6459045410156, | |
| "loss": 0.6466, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": 0.041329581290483475, | |
| "rewards/margins": 0.05584716796875, | |
| "rewards/rejected": -0.014517592266201973, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.5424912689173456e-07, | |
| "logits/chosen": -2.4804582595825195, | |
| "logits/rejected": -2.462634325027466, | |
| "logps/chosen": -270.22698974609375, | |
| "logps/rejected": -227.7113800048828, | |
| "loss": 0.6407, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": 0.037567246705293655, | |
| "rewards/margins": 0.0671004056930542, | |
| "rewards/rejected": -0.02953316643834114, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.4842840512223514e-07, | |
| "logits/chosen": -2.468341112136841, | |
| "logits/rejected": -2.4682674407958984, | |
| "logps/chosen": -257.55096435546875, | |
| "logps/rejected": -232.13906860351562, | |
| "loss": 0.647, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": 0.03749538213014603, | |
| "rewards/margins": 0.0758819431066513, | |
| "rewards/rejected": -0.03838656097650528, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 1.4260768335273574e-07, | |
| "logits/chosen": -2.5157179832458496, | |
| "logits/rejected": -2.497849702835083, | |
| "logps/chosen": -264.9339904785156, | |
| "logps/rejected": -229.9503936767578, | |
| "loss": 0.6375, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": 0.03981786221265793, | |
| "rewards/margins": 0.06821642816066742, | |
| "rewards/rejected": -0.02839856967329979, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 1.3678696158323632e-07, | |
| "logits/chosen": -2.4586503505706787, | |
| "logits/rejected": -2.4603028297424316, | |
| "logps/chosen": -262.70880126953125, | |
| "logps/rejected": -237.71212768554688, | |
| "loss": 0.645, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": 0.025259777903556824, | |
| "rewards/margins": 0.0490126870572567, | |
| "rewards/rejected": -0.023752911016345024, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 1.3096623981373688e-07, | |
| "logits/chosen": -2.502094268798828, | |
| "logits/rejected": -2.4800729751586914, | |
| "logps/chosen": -282.3802185058594, | |
| "logps/rejected": -236.8638153076172, | |
| "loss": 0.6438, | |
| "rewards/accuracies": 0.6656249761581421, | |
| "rewards/chosen": 0.04312821850180626, | |
| "rewards/margins": 0.07856440544128418, | |
| "rewards/rejected": -0.03543618693947792, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 1.2514551804423749e-07, | |
| "logits/chosen": -2.4671387672424316, | |
| "logits/rejected": -2.514324903488159, | |
| "logps/chosen": -285.3311462402344, | |
| "logps/rejected": -263.767822265625, | |
| "loss": 0.642, | |
| "rewards/accuracies": 0.684374988079071, | |
| "rewards/chosen": 0.04080774262547493, | |
| "rewards/margins": 0.07315204292535782, | |
| "rewards/rejected": -0.03234430402517319, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 1.1932479627473807e-07, | |
| "logits/chosen": -2.490058183670044, | |
| "logits/rejected": -2.4621422290802, | |
| "logps/chosen": -251.80148315429688, | |
| "logps/rejected": -249.9593048095703, | |
| "loss": 0.6395, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.028191978111863136, | |
| "rewards/margins": 0.06258732825517654, | |
| "rewards/rejected": -0.03439534455537796, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 1.1350407450523865e-07, | |
| "logits/chosen": -2.470336437225342, | |
| "logits/rejected": -2.5136280059814453, | |
| "logps/chosen": -281.42767333984375, | |
| "logps/rejected": -254.0972137451172, | |
| "loss": 0.6419, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.036293573677539825, | |
| "rewards/margins": 0.06266774237155914, | |
| "rewards/rejected": -0.02637416496872902, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 1.0768335273573923e-07, | |
| "logits/chosen": -2.5531864166259766, | |
| "logits/rejected": -2.545531749725342, | |
| "logps/chosen": -305.70111083984375, | |
| "logps/rejected": -276.01019287109375, | |
| "loss": 0.64, | |
| "rewards/accuracies": 0.621874988079071, | |
| "rewards/chosen": 0.046415045857429504, | |
| "rewards/margins": 0.07657970488071442, | |
| "rewards/rejected": -0.030164653435349464, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 1.0186263096623981e-07, | |
| "logits/chosen": -2.484541416168213, | |
| "logits/rejected": -2.4959442615509033, | |
| "logps/chosen": -272.62445068359375, | |
| "logps/rejected": -249.4247589111328, | |
| "loss": 0.6416, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": 0.03473570942878723, | |
| "rewards/margins": 0.057318903505802155, | |
| "rewards/rejected": -0.022583190351724625, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 9.604190919674039e-08, | |
| "logits/chosen": -2.480217218399048, | |
| "logits/rejected": -2.4997220039367676, | |
| "logps/chosen": -253.22061157226562, | |
| "logps/rejected": -248.5176239013672, | |
| "loss": 0.6447, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.02552936039865017, | |
| "rewards/margins": 0.05032141134142876, | |
| "rewards/rejected": -0.024792049080133438, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 9.022118742724097e-08, | |
| "logits/chosen": -2.4872233867645264, | |
| "logits/rejected": -2.461895704269409, | |
| "logps/chosen": -271.40191650390625, | |
| "logps/rejected": -261.29241943359375, | |
| "loss": 0.6352, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": 0.035524625331163406, | |
| "rewards/margins": 0.0665384978055954, | |
| "rewards/rejected": -0.031013870611786842, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 8.440046565774157e-08, | |
| "logits/chosen": -2.462954521179199, | |
| "logits/rejected": -2.472172498703003, | |
| "logps/chosen": -272.635986328125, | |
| "logps/rejected": -229.35366821289062, | |
| "loss": 0.6432, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.03845527023077011, | |
| "rewards/margins": 0.05895563215017319, | |
| "rewards/rejected": -0.020500360056757927, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 7.857974388824213e-08, | |
| "logits/chosen": -2.481274127960205, | |
| "logits/rejected": -2.4623420238494873, | |
| "logps/chosen": -282.552734375, | |
| "logps/rejected": -246.048095703125, | |
| "loss": 0.6457, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": 0.04651721939444542, | |
| "rewards/margins": 0.07211866974830627, | |
| "rewards/rejected": -0.025601446628570557, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 7.275902211874273e-08, | |
| "logits/chosen": -2.495702028274536, | |
| "logits/rejected": -2.4765870571136475, | |
| "logps/chosen": -249.3843536376953, | |
| "logps/rejected": -278.4346923828125, | |
| "loss": 0.644, | |
| "rewards/accuracies": 0.684374988079071, | |
| "rewards/chosen": 0.021697301417589188, | |
| "rewards/margins": 0.059917084872722626, | |
| "rewards/rejected": -0.038219790905714035, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 6.693830034924331e-08, | |
| "logits/chosen": -2.4697625637054443, | |
| "logits/rejected": -2.4722721576690674, | |
| "logps/chosen": -265.5938415527344, | |
| "logps/rejected": -257.2969665527344, | |
| "loss": 0.6376, | |
| "rewards/accuracies": 0.653124988079071, | |
| "rewards/chosen": 0.028097212314605713, | |
| "rewards/margins": 0.046820152550935745, | |
| "rewards/rejected": -0.018722938373684883, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 6.111757857974389e-08, | |
| "logits/chosen": -2.4835047721862793, | |
| "logits/rejected": -2.5235934257507324, | |
| "logps/chosen": -263.8602294921875, | |
| "logps/rejected": -241.2860870361328, | |
| "loss": 0.6443, | |
| "rewards/accuracies": 0.640625, | |
| "rewards/chosen": 0.03554076701402664, | |
| "rewards/margins": 0.065843865275383, | |
| "rewards/rejected": -0.030303100124001503, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 5.529685681024446e-08, | |
| "logits/chosen": -2.446396589279175, | |
| "logits/rejected": -2.447554111480713, | |
| "logps/chosen": -278.2039489746094, | |
| "logps/rejected": -267.393798828125, | |
| "loss": 0.6433, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": 0.02485939860343933, | |
| "rewards/margins": 0.04789874702692032, | |
| "rewards/rejected": -0.023039352148771286, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 4.947613504074505e-08, | |
| "logits/chosen": -2.451906442642212, | |
| "logits/rejected": -2.4401917457580566, | |
| "logps/chosen": -254.4597625732422, | |
| "logps/rejected": -240.5189666748047, | |
| "loss": 0.6406, | |
| "rewards/accuracies": 0.596875011920929, | |
| "rewards/chosen": 0.03511255979537964, | |
| "rewards/margins": 0.05534617230296135, | |
| "rewards/rejected": -0.020233619958162308, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 4.365541327124563e-08, | |
| "logits/chosen": -2.507020950317383, | |
| "logits/rejected": -2.4901695251464844, | |
| "logps/chosen": -260.8884582519531, | |
| "logps/rejected": -251.20938110351562, | |
| "loss": 0.6417, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": 0.04014205187559128, | |
| "rewards/margins": 0.060112785547971725, | |
| "rewards/rejected": -0.01997072994709015, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 3.783469150174622e-08, | |
| "logits/chosen": -2.5302183628082275, | |
| "logits/rejected": -2.5431013107299805, | |
| "logps/chosen": -254.4590606689453, | |
| "logps/rejected": -259.6826477050781, | |
| "loss": 0.6409, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": 0.02364221215248108, | |
| "rewards/margins": 0.04719501733779907, | |
| "rewards/rejected": -0.023552805185317993, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 3.20139697322468e-08, | |
| "logits/chosen": -2.4807305335998535, | |
| "logits/rejected": -2.5058140754699707, | |
| "logps/chosen": -264.16021728515625, | |
| "logps/rejected": -249.19570922851562, | |
| "loss": 0.6452, | |
| "rewards/accuracies": 0.628125011920929, | |
| "rewards/chosen": 0.029653768986463547, | |
| "rewards/margins": 0.05728424713015556, | |
| "rewards/rejected": -0.027630474418401718, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 2.619324796274738e-08, | |
| "logits/chosen": -2.48866605758667, | |
| "logits/rejected": -2.5024585723876953, | |
| "logps/chosen": -289.67401123046875, | |
| "logps/rejected": -261.8035583496094, | |
| "loss": 0.6424, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": 0.041562773287296295, | |
| "rewards/margins": 0.06672655045986176, | |
| "rewards/rejected": -0.025163773447275162, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.037252619324796e-08, | |
| "logits/chosen": -2.4807181358337402, | |
| "logits/rejected": -2.4651541709899902, | |
| "logps/chosen": -270.3717346191406, | |
| "logps/rejected": -248.5556182861328, | |
| "loss": 0.6415, | |
| "rewards/accuracies": 0.653124988079071, | |
| "rewards/chosen": 0.022484585642814636, | |
| "rewards/margins": 0.05588115006685257, | |
| "rewards/rejected": -0.033396560698747635, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.4551804423748545e-08, | |
| "logits/chosen": -2.4763240814208984, | |
| "logits/rejected": -2.4676055908203125, | |
| "logps/chosen": -272.7355041503906, | |
| "logps/rejected": -254.0940704345703, | |
| "loss": 0.6432, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.046184636652469635, | |
| "rewards/margins": 0.0702408105134964, | |
| "rewards/rejected": -0.02405618131160736, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 8.731082654249125e-09, | |
| "logits/chosen": -2.4857518672943115, | |
| "logits/rejected": -2.5030598640441895, | |
| "logps/chosen": -275.62445068359375, | |
| "logps/rejected": -264.8072509765625, | |
| "loss": 0.6407, | |
| "rewards/accuracies": 0.671875, | |
| "rewards/chosen": 0.036477264016866684, | |
| "rewards/margins": 0.06461935490369797, | |
| "rewards/rejected": -0.02814210020005703, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2.910360884749709e-09, | |
| "logits/chosen": -2.4887900352478027, | |
| "logits/rejected": -2.494776725769043, | |
| "logps/chosen": -259.7311706542969, | |
| "logps/rejected": -253.02346801757812, | |
| "loss": 0.6442, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": 0.02119762822985649, | |
| "rewards/margins": 0.04852701723575592, | |
| "rewards/rejected": -0.02732938900589943, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_logits/chosen": -2.4944536685943604, | |
| "eval_logits/rejected": -2.4963433742523193, | |
| "eval_logps/chosen": -269.25555419921875, | |
| "eval_logps/rejected": -253.21238708496094, | |
| "eval_loss": 0.6399702429771423, | |
| "eval_rewards/accuracies": 0.6370000243186951, | |
| "eval_rewards/chosen": 0.030110126361250877, | |
| "eval_rewards/margins": 0.05743245780467987, | |
| "eval_rewards/rejected": -0.027322327718138695, | |
| "eval_runtime": 806.7503, | |
| "eval_samples_per_second": 2.479, | |
| "eval_steps_per_second": 0.31, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 955, | |
| "total_flos": 0.0, | |
| "train_loss": 0.6598132096035942, | |
| "train_runtime": 45126.4521, | |
| "train_samples_per_second": 1.355, | |
| "train_steps_per_second": 0.021 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 955, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 0.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |