| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.998691442030882, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010468463752944255, | |
| "grad_norm": 3.7513903074877035, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": -2.8171534538269043, | |
| "logits/rejected": -2.7752685546875, | |
| "logps/chosen": -1.073843240737915, | |
| "logps/rejected": -1.1337131261825562, | |
| "loss": -0.0245, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.073843240737915, | |
| "rewards/margins": 0.05986976623535156, | |
| "rewards/rejected": -1.1337131261825562, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 4.180855743914033, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.998121738433838, | |
| "logits/rejected": -2.951986074447632, | |
| "logps/chosen": -1.0525588989257812, | |
| "logps/rejected": -1.1949102878570557, | |
| "loss": -0.0274, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.0525588989257812, | |
| "rewards/margins": 0.14235138893127441, | |
| "rewards/rejected": -1.1949102878570557, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 4.394107831023165, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": -2.9039416313171387, | |
| "logits/rejected": -2.8798794746398926, | |
| "logps/chosen": -1.0741031169891357, | |
| "logps/rejected": -1.1714708805084229, | |
| "loss": -0.0311, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.0741031169891357, | |
| "rewards/margins": 0.09736768901348114, | |
| "rewards/rejected": -1.1714708805084229, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 4.084514515910866, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.9078304767608643, | |
| "logits/rejected": -2.8506627082824707, | |
| "logps/chosen": -0.9993938207626343, | |
| "logps/rejected": -1.060178518295288, | |
| "loss": -0.0327, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.9993938207626343, | |
| "rewards/margins": 0.060784727334976196, | |
| "rewards/rejected": -1.060178518295288, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 4.220975658371714, | |
| "learning_rate": 2.604166666666667e-07, | |
| "logits/chosen": -2.9366836547851562, | |
| "logits/rejected": -2.8801796436309814, | |
| "logps/chosen": -0.8963547945022583, | |
| "logps/rejected": -1.0184606313705444, | |
| "loss": -0.0413, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.8963547945022583, | |
| "rewards/margins": 0.12210593372583389, | |
| "rewards/rejected": -1.0184606313705444, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 5.657665328598262, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.926485538482666, | |
| "logits/rejected": -2.897592067718506, | |
| "logps/chosen": -0.8541727066040039, | |
| "logps/rejected": -0.9000848531723022, | |
| "loss": -0.0368, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.8541727066040039, | |
| "rewards/margins": 0.04591215401887894, | |
| "rewards/rejected": -0.9000848531723022, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 4.821432353040026, | |
| "learning_rate": 3.645833333333333e-07, | |
| "logits/chosen": -3.003875255584717, | |
| "logits/rejected": -2.975294589996338, | |
| "logps/chosen": -0.8470292091369629, | |
| "logps/rejected": -0.9920878410339355, | |
| "loss": -0.0474, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.8470292091369629, | |
| "rewards/margins": 0.14505870640277863, | |
| "rewards/rejected": -0.9920878410339355, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 4.5888718829315565, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.926931858062744, | |
| "logits/rejected": -2.911674976348877, | |
| "logps/chosen": -0.8758507966995239, | |
| "logps/rejected": -1.0194270610809326, | |
| "loss": -0.0483, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.8758507966995239, | |
| "rewards/margins": 0.14357635378837585, | |
| "rewards/rejected": -1.0194270610809326, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 4.790749815381475, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": -2.934635877609253, | |
| "logits/rejected": -2.9103052616119385, | |
| "logps/chosen": -0.8145984411239624, | |
| "logps/rejected": -0.9208289980888367, | |
| "loss": -0.0617, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.8145984411239624, | |
| "rewards/margins": 0.10623057931661606, | |
| "rewards/rejected": -0.9208289980888367, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 5.471597318206259, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -2.918330430984497, | |
| "logits/rejected": -2.956514835357666, | |
| "logps/chosen": -0.81634122133255, | |
| "logps/rejected": -1.0526411533355713, | |
| "loss": -0.0525, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.81634122133255, | |
| "rewards/margins": 0.23629975318908691, | |
| "rewards/rejected": -1.0526411533355713, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 7.76452501597249, | |
| "learning_rate": 4.996716052911017e-07, | |
| "logits/chosen": -2.9652035236358643, | |
| "logits/rejected": -2.9935660362243652, | |
| "logps/chosen": -0.9713274240493774, | |
| "logps/rejected": -1.0530242919921875, | |
| "loss": -0.0483, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.9713274240493774, | |
| "rewards/margins": 0.08169689774513245, | |
| "rewards/rejected": -1.0530242919921875, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 8.681841455192204, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -2.905766487121582, | |
| "logits/rejected": -2.9021992683410645, | |
| "logps/chosen": -0.9168001413345337, | |
| "logps/rejected": -1.0986061096191406, | |
| "loss": -0.0681, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9168001413345337, | |
| "rewards/margins": 0.18180601298809052, | |
| "rewards/rejected": -1.0986061096191406, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 6.6681874007549045, | |
| "learning_rate": 4.980652179769217e-07, | |
| "logits/chosen": -2.9656786918640137, | |
| "logits/rejected": -2.989439010620117, | |
| "logps/chosen": -0.9527915716171265, | |
| "logps/rejected": -1.0805647373199463, | |
| "loss": -0.0603, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9527915716171265, | |
| "rewards/margins": 0.12777289748191833, | |
| "rewards/rejected": -1.0805647373199463, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 6.9546346601931965, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -2.932742118835449, | |
| "logits/rejected": -2.920910358428955, | |
| "logps/chosen": -0.9588356018066406, | |
| "logps/rejected": -1.1663864850997925, | |
| "loss": -0.0804, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.9588356018066406, | |
| "rewards/margins": 0.2075507640838623, | |
| "rewards/rejected": -1.1663864850997925, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 8.658939501447122, | |
| "learning_rate": 4.951291206355559e-07, | |
| "logits/chosen": -2.9414222240448, | |
| "logits/rejected": -2.8595502376556396, | |
| "logps/chosen": -1.0185911655426025, | |
| "logps/rejected": -1.3531118631362915, | |
| "loss": -0.0632, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0185911655426025, | |
| "rewards/margins": 0.33452075719833374, | |
| "rewards/rejected": -1.3531118631362915, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 6.920392188242862, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -2.9239554405212402, | |
| "logits/rejected": -2.9125447273254395, | |
| "logps/chosen": -0.9926131367683411, | |
| "logps/rejected": -1.423416256904602, | |
| "loss": -0.0778, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.9926131367683411, | |
| "rewards/margins": 0.43080300092697144, | |
| "rewards/rejected": -1.423416256904602, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 9.131070177898815, | |
| "learning_rate": 4.908790517010636e-07, | |
| "logits/chosen": -2.950829029083252, | |
| "logits/rejected": -2.9169838428497314, | |
| "logps/chosen": -0.9384489059448242, | |
| "logps/rejected": -1.2703007459640503, | |
| "loss": -0.0827, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.9384489059448242, | |
| "rewards/margins": 0.33185189962387085, | |
| "rewards/rejected": -1.2703007459640503, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 6.205581176890279, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -2.9944369792938232, | |
| "logits/rejected": -2.9288582801818848, | |
| "logps/chosen": -1.0034880638122559, | |
| "logps/rejected": -1.5320912599563599, | |
| "loss": -0.0856, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0034880638122559, | |
| "rewards/margins": 0.528603196144104, | |
| "rewards/rejected": -1.5320912599563599, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 6.383852226299091, | |
| "learning_rate": 4.853377929214243e-07, | |
| "logits/chosen": -2.9386332035064697, | |
| "logits/rejected": -2.9376933574676514, | |
| "logps/chosen": -0.9994260668754578, | |
| "logps/rejected": -1.2795701026916504, | |
| "loss": -0.0757, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.9994260668754578, | |
| "rewards/margins": 0.2801441550254822, | |
| "rewards/rejected": -1.2795701026916504, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 11.112637282805201, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -3.124525785446167, | |
| "logits/rejected": -3.0877130031585693, | |
| "logps/chosen": -1.0537831783294678, | |
| "logps/rejected": -1.4097682237625122, | |
| "loss": -0.0768, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0537831783294678, | |
| "rewards/margins": 0.35598504543304443, | |
| "rewards/rejected": -1.4097682237625122, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 16.03176148360658, | |
| "learning_rate": 4.785350472409791e-07, | |
| "logits/chosen": -2.9793434143066406, | |
| "logits/rejected": -2.994006633758545, | |
| "logps/chosen": -1.1306521892547607, | |
| "logps/rejected": -1.6030582189559937, | |
| "loss": -0.1084, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.1306521892547607, | |
| "rewards/margins": 0.4724059998989105, | |
| "rewards/rejected": -1.6030582189559937, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 13.322882273152597, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -2.946302890777588, | |
| "logits/rejected": -2.9931533336639404, | |
| "logps/chosen": -1.016096830368042, | |
| "logps/rejected": -1.5478127002716064, | |
| "loss": -0.1078, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.016096830368042, | |
| "rewards/margins": 0.5317157506942749, | |
| "rewards/rejected": -1.5478127002716064, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 9.143168585235879, | |
| "learning_rate": 4.70507279583015e-07, | |
| "logits/chosen": -3.0106329917907715, | |
| "logits/rejected": -3.027531862258911, | |
| "logps/chosen": -1.142896294593811, | |
| "logps/rejected": -1.407957673072815, | |
| "loss": -0.0894, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.142896294593811, | |
| "rewards/margins": 0.26506131887435913, | |
| "rewards/rejected": -1.407957673072815, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 5.821978614016585, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -3.0557615756988525, | |
| "logits/rejected": -3.0477774143218994, | |
| "logps/chosen": -0.9973357319831848, | |
| "logps/rejected": -1.5232563018798828, | |
| "loss": -0.0749, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.9973357319831848, | |
| "rewards/margins": 0.525920569896698, | |
| "rewards/rejected": -1.5232563018798828, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 8.107257209423297, | |
| "learning_rate": 4.612975213859487e-07, | |
| "logits/chosen": -3.0169310569763184, | |
| "logits/rejected": -2.963167428970337, | |
| "logps/chosen": -1.074829339981079, | |
| "logps/rejected": -1.4796228408813477, | |
| "loss": -0.0827, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.074829339981079, | |
| "rewards/margins": 0.4047934412956238, | |
| "rewards/rejected": -1.4796228408813477, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 10.808245352281814, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -2.968205213546753, | |
| "logits/rejected": -2.987243175506592, | |
| "logps/chosen": -1.2360585927963257, | |
| "logps/rejected": -1.7557792663574219, | |
| "loss": -0.0975, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.2360585927963257, | |
| "rewards/margins": 0.5197205543518066, | |
| "rewards/rejected": -1.7557792663574219, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 6.440041105525622, | |
| "learning_rate": 4.5095513994085974e-07, | |
| "logits/chosen": -2.8724923133850098, | |
| "logits/rejected": -2.8528032302856445, | |
| "logps/chosen": -1.065212368965149, | |
| "logps/rejected": -1.674586296081543, | |
| "loss": -0.097, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.065212368965149, | |
| "rewards/margins": 0.609373927116394, | |
| "rewards/rejected": -1.674586296081543, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 10.975878079283369, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -2.9282472133636475, | |
| "logits/rejected": -2.9795360565185547, | |
| "logps/chosen": -1.4000217914581299, | |
| "logps/rejected": -1.768908143043518, | |
| "loss": -0.0838, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.4000217914581299, | |
| "rewards/margins": 0.3688863217830658, | |
| "rewards/rejected": -1.768908143043518, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 10.502089738944973, | |
| "learning_rate": 4.395355737667985e-07, | |
| "logits/chosen": -2.9197769165039062, | |
| "logits/rejected": -2.9231114387512207, | |
| "logps/chosen": -1.099702000617981, | |
| "logps/rejected": -1.8354318141937256, | |
| "loss": -0.1006, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.099702000617981, | |
| "rewards/margins": 0.735729992389679, | |
| "rewards/rejected": -1.8354318141937256, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 10.462991939664109, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -2.9763550758361816, | |
| "logits/rejected": -2.9400999546051025, | |
| "logps/chosen": -1.1915630102157593, | |
| "logps/rejected": -1.7106775045394897, | |
| "loss": -0.0907, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.1915630102157593, | |
| "rewards/margins": 0.5191145539283752, | |
| "rewards/rejected": -1.7106775045394897, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 19.362917244755998, | |
| "learning_rate": 4.271000354423425e-07, | |
| "logits/chosen": -2.88962984085083, | |
| "logits/rejected": -2.884782552719116, | |
| "logps/chosen": -1.2416050434112549, | |
| "logps/rejected": -1.970547080039978, | |
| "loss": -0.1124, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.2416050434112549, | |
| "rewards/margins": 0.7289418578147888, | |
| "rewards/rejected": -1.970547080039978, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 10.033852918157995, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -2.826140880584717, | |
| "logits/rejected": -2.8399477005004883, | |
| "logps/chosen": -1.1741280555725098, | |
| "logps/rejected": -1.8060003519058228, | |
| "loss": -0.109, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.1741280555725098, | |
| "rewards/margins": 0.6318725347518921, | |
| "rewards/rejected": -1.8060003519058228, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 7.788169857899764, | |
| "learning_rate": 4.137151834863213e-07, | |
| "logits/chosen": -2.919525384902954, | |
| "logits/rejected": -2.874574661254883, | |
| "logps/chosen": -1.2077335119247437, | |
| "logps/rejected": -1.4566949605941772, | |
| "loss": -0.1086, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.2077335119247437, | |
| "rewards/margins": 0.24896140396595, | |
| "rewards/rejected": -1.4566949605941772, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 8.399922056025794, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -2.8885338306427, | |
| "logits/rejected": -2.832759380340576, | |
| "logps/chosen": -1.1002306938171387, | |
| "logps/rejected": -1.5415006875991821, | |
| "loss": -0.1027, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.1002306938171387, | |
| "rewards/margins": 0.44127020239830017, | |
| "rewards/rejected": -1.5415006875991821, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 13.391590525406425, | |
| "learning_rate": 3.994527650465352e-07, | |
| "logits/chosen": -2.7812280654907227, | |
| "logits/rejected": -2.7597861289978027, | |
| "logps/chosen": -1.0584896802902222, | |
| "logps/rejected": -1.7376607656478882, | |
| "loss": -0.1093, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0584896802902222, | |
| "rewards/margins": 0.6791711449623108, | |
| "rewards/rejected": -1.7376607656478882, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 10.692790169158753, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -2.8191940784454346, | |
| "logits/rejected": -2.783812999725342, | |
| "logps/chosen": -1.0985336303710938, | |
| "logps/rejected": -1.7514450550079346, | |
| "loss": -0.1088, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.0985336303710938, | |
| "rewards/margins": 0.6529111862182617, | |
| "rewards/rejected": -1.7514450550079346, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 9.974751336008916, | |
| "learning_rate": 3.8438923131177237e-07, | |
| "logits/chosen": -2.7801730632781982, | |
| "logits/rejected": -2.789536237716675, | |
| "logps/chosen": -1.035691738128662, | |
| "logps/rejected": -2.0174129009246826, | |
| "loss": -0.1205, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.035691738128662, | |
| "rewards/margins": 0.9817211031913757, | |
| "rewards/rejected": -2.0174129009246826, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 7.918537129637632, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": -2.7641420364379883, | |
| "logits/rejected": -2.7668850421905518, | |
| "logps/chosen": -1.0616905689239502, | |
| "logps/rejected": -2.003706693649292, | |
| "loss": -0.1186, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0616905689239502, | |
| "rewards/margins": 0.9420160055160522, | |
| "rewards/rejected": -2.003706693649292, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 14.273863949897775, | |
| "learning_rate": 3.6860532770864005e-07, | |
| "logits/chosen": -2.7132833003997803, | |
| "logits/rejected": -2.7151715755462646, | |
| "logps/chosen": -1.184514045715332, | |
| "logps/rejected": -2.416534185409546, | |
| "loss": -0.1047, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.184514045715332, | |
| "rewards/margins": 1.2320199012756348, | |
| "rewards/rejected": -2.416534185409546, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 13.778909374638918, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": -2.5922656059265137, | |
| "logits/rejected": -2.541560649871826, | |
| "logps/chosen": -1.3440501689910889, | |
| "logps/rejected": -2.674525499343872, | |
| "loss": -0.1223, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.3440501689910889, | |
| "rewards/margins": 1.3304755687713623, | |
| "rewards/rejected": -2.674525499343872, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 19.066503947059758, | |
| "learning_rate": 3.5218566107988867e-07, | |
| "logits/chosen": -2.55405855178833, | |
| "logits/rejected": -2.4775567054748535, | |
| "logps/chosen": -1.4227993488311768, | |
| "logps/rejected": -2.808105707168579, | |
| "loss": -0.1218, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.4227993488311768, | |
| "rewards/margins": 1.3853065967559814, | |
| "rewards/rejected": -2.808105707168579, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 7.4083218354385405, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -2.4821341037750244, | |
| "logits/rejected": -2.4735965728759766, | |
| "logps/chosen": -1.382550835609436, | |
| "logps/rejected": -2.1365103721618652, | |
| "loss": -0.1053, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.382550835609436, | |
| "rewards/margins": 0.7539594173431396, | |
| "rewards/rejected": -2.1365103721618652, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 11.135941225237307, | |
| "learning_rate": 3.3521824616429284e-07, | |
| "logits/chosen": -2.494109630584717, | |
| "logits/rejected": -2.518179416656494, | |
| "logps/chosen": -1.4302706718444824, | |
| "logps/rejected": -2.0413661003112793, | |
| "loss": -0.1038, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.4302706718444824, | |
| "rewards/margins": 0.6110953688621521, | |
| "rewards/rejected": -2.0413661003112793, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 10.214876327489243, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -2.4926416873931885, | |
| "logits/rejected": -2.4311227798461914, | |
| "logps/chosen": -1.2597782611846924, | |
| "logps/rejected": -2.1729977130889893, | |
| "loss": -0.1168, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.2597782611846924, | |
| "rewards/margins": 0.9132192730903625, | |
| "rewards/rejected": -2.1729977130889893, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 14.878685792407815, | |
| "learning_rate": 3.1779403380910425e-07, | |
| "logits/chosen": -2.450490713119507, | |
| "logits/rejected": -2.3297905921936035, | |
| "logps/chosen": -1.5581631660461426, | |
| "logps/rejected": -2.340275287628174, | |
| "loss": -0.1145, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.5581631660461426, | |
| "rewards/margins": 0.7821120023727417, | |
| "rewards/rejected": -2.340275287628174, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 15.917709814115218, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -2.420201063156128, | |
| "logits/rejected": -2.347395181655884, | |
| "logps/chosen": -1.4262598752975464, | |
| "logps/rejected": -2.6726250648498535, | |
| "loss": -0.1278, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.4262598752975464, | |
| "rewards/margins": 1.246364951133728, | |
| "rewards/rejected": -2.6726250648498535, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 15.700390852097751, | |
| "learning_rate": 3.000064234440111e-07, | |
| "logits/chosen": -2.3450303077697754, | |
| "logits/rejected": -2.263624668121338, | |
| "logps/chosen": -1.4487098455429077, | |
| "logps/rejected": -2.293961763381958, | |
| "loss": -0.1174, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.4487098455429077, | |
| "rewards/margins": 0.845251739025116, | |
| "rewards/rejected": -2.293961763381958, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 11.565327009400553, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -2.2296805381774902, | |
| "logits/rejected": -2.2061212062835693, | |
| "logps/chosen": -1.7294538021087646, | |
| "logps/rejected": -2.555497884750366, | |
| "loss": -0.123, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.7294538021087646, | |
| "rewards/margins": 0.8260442018508911, | |
| "rewards/rejected": -2.555497884750366, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 12.147560315025176, | |
| "learning_rate": 2.8195076242990116e-07, | |
| "logits/chosen": -2.3435921669006348, | |
| "logits/rejected": -2.2495055198669434, | |
| "logps/chosen": -1.2541306018829346, | |
| "logps/rejected": -2.773146152496338, | |
| "loss": -0.1256, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.2541306018829346, | |
| "rewards/margins": 1.5190155506134033, | |
| "rewards/rejected": -2.773146152496338, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 13.46036439275393, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -2.356762647628784, | |
| "logits/rejected": -2.2052009105682373, | |
| "logps/chosen": -1.3296482563018799, | |
| "logps/rejected": -2.694124698638916, | |
| "loss": -0.1064, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.3296482563018799, | |
| "rewards/margins": 1.3644765615463257, | |
| "rewards/rejected": -2.694124698638916, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 11.303628784559335, | |
| "learning_rate": 2.6372383496608186e-07, | |
| "logits/chosen": -2.436122417449951, | |
| "logits/rejected": -2.3444714546203613, | |
| "logps/chosen": -1.372803807258606, | |
| "logps/rejected": -1.9248924255371094, | |
| "loss": -0.1079, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.372803807258606, | |
| "rewards/margins": 0.5520886182785034, | |
| "rewards/rejected": -1.9248924255371094, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 15.1501744128238, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -2.207700252532959, | |
| "logits/rejected": -2.203295946121216, | |
| "logps/chosen": -1.3273894786834717, | |
| "logps/rejected": -2.3319733142852783, | |
| "loss": -0.1254, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.3273894786834717, | |
| "rewards/margins": 1.004583716392517, | |
| "rewards/rejected": -2.3319733142852783, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 19.68796529060918, | |
| "learning_rate": 2.454233432955807e-07, | |
| "logits/chosen": -2.27662992477417, | |
| "logits/rejected": -2.262169361114502, | |
| "logps/chosen": -1.4423593282699585, | |
| "logps/rejected": -2.3584229946136475, | |
| "loss": -0.1259, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.4423593282699585, | |
| "rewards/margins": 0.916063666343689, | |
| "rewards/rejected": -2.3584229946136475, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 11.467800573919314, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": -2.202195644378662, | |
| "logits/rejected": -2.2492783069610596, | |
| "logps/chosen": -1.5429728031158447, | |
| "logps/rejected": -2.8726634979248047, | |
| "loss": -0.1223, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.5429728031158447, | |
| "rewards/margins": 1.3296908140182495, | |
| "rewards/rejected": -2.8726634979248047, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 16.984616647138346, | |
| "learning_rate": 2.2714738398943308e-07, | |
| "logits/chosen": -2.3055601119995117, | |
| "logits/rejected": -2.1988272666931152, | |
| "logps/chosen": -1.1838595867156982, | |
| "logps/rejected": -3.205533266067505, | |
| "loss": -0.1574, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.1838595867156982, | |
| "rewards/margins": 2.0216736793518066, | |
| "rewards/rejected": -3.205533266067505, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 67.43202579292445, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": -2.343538522720337, | |
| "logits/rejected": -2.2112088203430176, | |
| "logps/chosen": -1.3181947469711304, | |
| "logps/rejected": -2.941770076751709, | |
| "loss": -0.1317, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.3181947469711304, | |
| "rewards/margins": 1.6235754489898682, | |
| "rewards/rejected": -2.941770076751709, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 25.41054694704417, | |
| "learning_rate": 2.089939221172446e-07, | |
| "logits/chosen": -2.3341867923736572, | |
| "logits/rejected": -2.279958724975586, | |
| "logps/chosen": -1.3771889209747314, | |
| "logps/rejected": -2.7422072887420654, | |
| "loss": -0.1262, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.3771889209747314, | |
| "rewards/margins": 1.3650182485580444, | |
| "rewards/rejected": -2.7422072887420654, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 18.32126881217475, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": -2.319903612136841, | |
| "logits/rejected": -2.223397731781006, | |
| "logps/chosen": -1.745248794555664, | |
| "logps/rejected": -2.239058017730713, | |
| "loss": -0.1182, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.745248794555664, | |
| "rewards/margins": 0.49380961060523987, | |
| "rewards/rejected": -2.239058017730713, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 7.718830375673775, | |
| "learning_rate": 1.9106026612264315e-07, | |
| "logits/chosen": -2.2595322132110596, | |
| "logits/rejected": -2.1636531352996826, | |
| "logps/chosen": -1.4870641231536865, | |
| "logps/rejected": -1.8681867122650146, | |
| "loss": -0.1307, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.4870641231536865, | |
| "rewards/margins": 0.38112273812294006, | |
| "rewards/rejected": -1.8681867122650146, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 15.596899390350146, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": -2.1786293983459473, | |
| "logits/rejected": -2.0660665035247803, | |
| "logps/chosen": -1.4667835235595703, | |
| "logps/rejected": -2.425236701965332, | |
| "loss": -0.1163, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.4667835235595703, | |
| "rewards/margins": 0.9584531784057617, | |
| "rewards/rejected": -2.425236701965332, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 21.54797247836559, | |
| "learning_rate": 1.7344254621846017e-07, | |
| "logits/chosen": -2.256340503692627, | |
| "logits/rejected": -2.190368890762329, | |
| "logps/chosen": -1.1270886659622192, | |
| "logps/rejected": -1.6489307880401611, | |
| "loss": -0.1342, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.1270886659622192, | |
| "rewards/margins": 0.5218421220779419, | |
| "rewards/rejected": -1.6489307880401611, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 13.8459567378238, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": -2.215406656265259, | |
| "logits/rejected": -2.0842509269714355, | |
| "logps/chosen": -1.177166223526001, | |
| "logps/rejected": -3.131798505783081, | |
| "loss": -0.1307, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.177166223526001, | |
| "rewards/margins": 1.9546325206756592, | |
| "rewards/rejected": -3.131798505783081, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 18.48698198664654, | |
| "learning_rate": 1.562351990976095e-07, | |
| "logits/chosen": -2.16829514503479, | |
| "logits/rejected": -2.009229898452759, | |
| "logps/chosen": -1.3233706951141357, | |
| "logps/rejected": -2.3988609313964844, | |
| "loss": -0.1342, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.3233706951141357, | |
| "rewards/margins": 1.0754903554916382, | |
| "rewards/rejected": -2.3988609313964844, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 8.752357437921063, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": -2.0819640159606934, | |
| "logits/rejected": -1.9889743328094482, | |
| "logps/chosen": -1.5709710121154785, | |
| "logps/rejected": -2.7803707122802734, | |
| "loss": -0.1332, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.5709710121154785, | |
| "rewards/margins": 1.2093994617462158, | |
| "rewards/rejected": -2.7803707122802734, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 16.357463949606085, | |
| "learning_rate": 1.3953046172178413e-07, | |
| "logits/chosen": -2.047121524810791, | |
| "logits/rejected": -1.95102858543396, | |
| "logps/chosen": -1.2728681564331055, | |
| "logps/rejected": -2.9997382164001465, | |
| "loss": -0.1178, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.2728681564331055, | |
| "rewards/margins": 1.7268702983856201, | |
| "rewards/rejected": -2.9997382164001465, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 36.540239827092066, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": -2.0776524543762207, | |
| "logits/rejected": -2.049619674682617, | |
| "logps/chosen": -1.1679047346115112, | |
| "logps/rejected": -3.1095471382141113, | |
| "loss": -0.1188, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -1.1679047346115112, | |
| "rewards/margins": 1.941642165184021, | |
| "rewards/rejected": -3.1095471382141113, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 27.21209576066559, | |
| "learning_rate": 1.2341787690142435e-07, | |
| "logits/chosen": -2.070624589920044, | |
| "logits/rejected": -1.8952518701553345, | |
| "logps/chosen": -1.4785457849502563, | |
| "logps/rejected": -3.115140438079834, | |
| "loss": -0.1257, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.4785457849502563, | |
| "rewards/margins": 1.6365947723388672, | |
| "rewards/rejected": -3.115140438079834, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 23.86077500979542, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": -1.8781455755233765, | |
| "logits/rejected": -1.7856013774871826, | |
| "logps/chosen": -1.564313292503357, | |
| "logps/rejected": -3.2956881523132324, | |
| "loss": -0.1483, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.564313292503357, | |
| "rewards/margins": 1.7313750982284546, | |
| "rewards/rejected": -3.2956881523132324, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 32.743197556520705, | |
| "learning_rate": 1.0798381331721107e-07, | |
| "logits/chosen": -2.064187526702881, | |
| "logits/rejected": -1.9649896621704102, | |
| "logps/chosen": -1.2933770418167114, | |
| "logps/rejected": -2.960786819458008, | |
| "loss": -0.1287, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.2933770418167114, | |
| "rewards/margins": 1.667409896850586, | |
| "rewards/rejected": -2.960786819458008, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 23.239099915632792, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": -2.054849147796631, | |
| "logits/rejected": -1.7261388301849365, | |
| "logps/chosen": -1.6969807147979736, | |
| "logps/rejected": -3.3776707649230957, | |
| "loss": -0.1344, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -1.6969807147979736, | |
| "rewards/margins": 1.680689811706543, | |
| "rewards/rejected": -3.3776707649230957, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 22.775441361714698, | |
| "learning_rate": 9.331100255592436e-08, | |
| "logits/chosen": -1.925492525100708, | |
| "logits/rejected": -1.780735731124878, | |
| "logps/chosen": -1.6358855962753296, | |
| "logps/rejected": -3.2378578186035156, | |
| "loss": -0.1212, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.6358855962753296, | |
| "rewards/margins": 1.601971983909607, | |
| "rewards/rejected": -3.2378578186035156, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 26.750275729181787, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -1.9371429681777954, | |
| "logits/rejected": -1.8784688711166382, | |
| "logps/chosen": -1.6762475967407227, | |
| "logps/rejected": -3.2192864418029785, | |
| "loss": -0.1324, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -1.6762475967407227, | |
| "rewards/margins": 1.5430386066436768, | |
| "rewards/rejected": -3.2192864418029785, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 28.528016311020515, | |
| "learning_rate": 7.947809564230445e-08, | |
| "logits/chosen": -2.098036050796509, | |
| "logits/rejected": -1.8208805322647095, | |
| "logps/chosen": -1.6632845401763916, | |
| "logps/rejected": -3.5564017295837402, | |
| "loss": -0.1447, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.6632845401763916, | |
| "rewards/margins": 1.8931175470352173, | |
| "rewards/rejected": -3.5564017295837402, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 27.13113745463131, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": -2.0418508052825928, | |
| "logits/rejected": -1.8973302841186523, | |
| "logps/chosen": -1.3375906944274902, | |
| "logps/rejected": -2.4453277587890625, | |
| "loss": -0.1506, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -1.3375906944274902, | |
| "rewards/margins": 1.1077370643615723, | |
| "rewards/rejected": -2.4453277587890625, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 23.451359669795522, | |
| "learning_rate": 6.655924144404906e-08, | |
| "logits/chosen": -1.938241720199585, | |
| "logits/rejected": -1.8409382104873657, | |
| "logps/chosen": -1.1007254123687744, | |
| "logps/rejected": -3.7080078125, | |
| "loss": -0.1291, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.1007254123687744, | |
| "rewards/margins": 2.6072821617126465, | |
| "rewards/rejected": -3.7080078125, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 18.002790677072948, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": -2.0828349590301514, | |
| "logits/rejected": -2.072486400604248, | |
| "logps/chosen": -1.5133841037750244, | |
| "logps/rejected": -2.042178153991699, | |
| "loss": -0.1139, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.5133841037750244, | |
| "rewards/margins": 0.52879399061203, | |
| "rewards/rejected": -2.042178153991699, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 34.9962904822404, | |
| "learning_rate": 5.4623689209832484e-08, | |
| "logits/chosen": -2.016568660736084, | |
| "logits/rejected": -1.9771146774291992, | |
| "logps/chosen": -1.1945464611053467, | |
| "logps/rejected": -2.6259241104125977, | |
| "loss": -0.1238, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1945464611053467, | |
| "rewards/margins": 1.4313781261444092, | |
| "rewards/rejected": -2.6259241104125977, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 25.642946316101476, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": -2.136821746826172, | |
| "logits/rejected": -1.895983099937439, | |
| "logps/chosen": -1.6347423791885376, | |
| "logps/rejected": -2.6382884979248047, | |
| "loss": -0.1101, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.6347423791885376, | |
| "rewards/margins": 1.003546118736267, | |
| "rewards/rejected": -2.6382884979248047, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 20.450819747311563, | |
| "learning_rate": 4.373541737087263e-08, | |
| "logits/chosen": -2.059231996536255, | |
| "logits/rejected": -1.8570470809936523, | |
| "logps/chosen": -1.3500816822052002, | |
| "logps/rejected": -2.9107394218444824, | |
| "loss": -0.1211, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.3500816822052002, | |
| "rewards/margins": 1.5606579780578613, | |
| "rewards/rejected": -2.9107394218444824, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 19.246577992332785, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": -2.1224074363708496, | |
| "logits/rejected": -1.996233582496643, | |
| "logps/chosen": -1.3660318851470947, | |
| "logps/rejected": -2.114323139190674, | |
| "loss": -0.1138, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.3660318851470947, | |
| "rewards/margins": 0.7482913732528687, | |
| "rewards/rejected": -2.114323139190674, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 18.23142964128246, | |
| "learning_rate": 3.3952790595787986e-08, | |
| "logits/chosen": -2.071781635284424, | |
| "logits/rejected": -1.9027512073516846, | |
| "logps/chosen": -1.4462158679962158, | |
| "logps/rejected": -2.7868762016296387, | |
| "loss": -0.1347, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.4462158679962158, | |
| "rewards/margins": 1.3406604528427124, | |
| "rewards/rejected": -2.7868762016296387, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 35.59728011084227, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": -1.9791206121444702, | |
| "logits/rejected": -1.882555603981018, | |
| "logps/chosen": -1.6704654693603516, | |
| "logps/rejected": -3.460047483444214, | |
| "loss": -0.132, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.6704654693603516, | |
| "rewards/margins": 1.7895816564559937, | |
| "rewards/rejected": -3.460047483444214, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 32.113859967532896, | |
| "learning_rate": 2.5328246937043525e-08, | |
| "logits/chosen": -1.9948623180389404, | |
| "logits/rejected": -1.8517177104949951, | |
| "logps/chosen": -1.799766182899475, | |
| "logps/rejected": -2.450369358062744, | |
| "loss": -0.124, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.799766182899475, | |
| "rewards/margins": 0.6506029963493347, | |
| "rewards/rejected": -2.450369358062744, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 39.98225849203012, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": -2.0489373207092285, | |
| "logits/rejected": -1.8950214385986328, | |
| "logps/chosen": -1.3832013607025146, | |
| "logps/rejected": -2.896411418914795, | |
| "loss": -0.1384, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.3832013607025146, | |
| "rewards/margins": 1.5132099390029907, | |
| "rewards/rejected": -2.896411418914795, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 13.073405517814074, | |
| "learning_rate": 1.7908016745981856e-08, | |
| "logits/chosen": -2.124887228012085, | |
| "logits/rejected": -1.9995448589324951, | |
| "logps/chosen": -1.449738621711731, | |
| "logps/rejected": -3.0659384727478027, | |
| "loss": -0.1363, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.449738621711731, | |
| "rewards/margins": 1.6161997318267822, | |
| "rewards/rejected": -3.0659384727478027, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 37.25775553384719, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": -2.009740114212036, | |
| "logits/rejected": -1.8612778186798096, | |
| "logps/chosen": -1.4344267845153809, | |
| "logps/rejected": -3.088493585586548, | |
| "loss": -0.1351, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.4344267845153809, | |
| "rewards/margins": 1.6540664434432983, | |
| "rewards/rejected": -3.088493585586548, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 11.94801466556723, | |
| "learning_rate": 1.1731874863145142e-08, | |
| "logits/chosen": -1.9715747833251953, | |
| "logits/rejected": -2.0468873977661133, | |
| "logps/chosen": -1.4286668300628662, | |
| "logps/rejected": -2.368466854095459, | |
| "loss": -0.1292, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.4286668300628662, | |
| "rewards/margins": 0.9398000836372375, | |
| "rewards/rejected": -2.368466854095459, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 15.24664884840477, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": -2.1121363639831543, | |
| "logits/rejected": -2.0955002307891846, | |
| "logps/chosen": -1.60480535030365, | |
| "logps/rejected": -2.702631711959839, | |
| "loss": -0.129, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.60480535030365, | |
| "rewards/margins": 1.0978264808654785, | |
| "rewards/rejected": -2.702631711959839, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 18.40983828171781, | |
| "learning_rate": 6.832927412229017e-09, | |
| "logits/chosen": -2.048396587371826, | |
| "logits/rejected": -1.9267797470092773, | |
| "logps/chosen": -1.4430943727493286, | |
| "logps/rejected": -2.9393792152404785, | |
| "loss": -0.1373, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.4430943727493286, | |
| "rewards/margins": 1.496285080909729, | |
| "rewards/rejected": -2.9393792152404785, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 30.800432331307388, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": -2.1506505012512207, | |
| "logits/rejected": -1.9776548147201538, | |
| "logps/chosen": -1.1526744365692139, | |
| "logps/rejected": -2.0363593101501465, | |
| "loss": -0.1369, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1526744365692139, | |
| "rewards/margins": 0.8836847543716431, | |
| "rewards/rejected": -2.0363593101501465, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 18.63300124637018, | |
| "learning_rate": 3.2374343405217884e-09, | |
| "logits/chosen": -2.1101202964782715, | |
| "logits/rejected": -1.9711452722549438, | |
| "logps/chosen": -1.5566682815551758, | |
| "logps/rejected": -2.7286152839660645, | |
| "loss": -0.1161, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.5566682815551758, | |
| "rewards/margins": 1.1719467639923096, | |
| "rewards/rejected": -2.7286152839660645, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 17.7570417535466, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": -2.153376579284668, | |
| "logits/rejected": -1.8994888067245483, | |
| "logps/chosen": -1.110905647277832, | |
| "logps/rejected": -3.4143710136413574, | |
| "loss": -0.1345, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.110905647277832, | |
| "rewards/margins": 2.3034653663635254, | |
| "rewards/rejected": -3.4143710136413574, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 14.215442604935475, | |
| "learning_rate": 9.64668657069706e-10, | |
| "logits/chosen": -2.07485294342041, | |
| "logits/rejected": -1.8461986780166626, | |
| "logps/chosen": -1.4135997295379639, | |
| "logps/rejected": -2.7699055671691895, | |
| "loss": -0.148, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -1.4135997295379639, | |
| "rewards/margins": 1.356305718421936, | |
| "rewards/rejected": -2.7699055671691895, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 13.45220516924706, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": -2.0263218879699707, | |
| "logits/rejected": -1.7936031818389893, | |
| "logps/chosen": -1.155396580696106, | |
| "logps/rejected": -3.2296791076660156, | |
| "loss": -0.1373, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.155396580696106, | |
| "rewards/margins": 2.074282646179199, | |
| "rewards/rejected": -3.2296791076660156, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 14.04473344623843, | |
| "learning_rate": 2.6813123097352287e-11, | |
| "logits/chosen": -2.053156614303589, | |
| "logits/rejected": -1.9318702220916748, | |
| "logps/chosen": -1.1674525737762451, | |
| "logps/rejected": -3.196053981781006, | |
| "loss": -0.1442, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.1674525737762451, | |
| "rewards/margins": 2.028601884841919, | |
| "rewards/rejected": -3.196053981781006, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": -0.10723852930078967, | |
| "train_runtime": 12249.3424, | |
| "train_samples_per_second": 4.991, | |
| "train_steps_per_second": 0.039 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 239, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |