| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9987943737441393, | |
| "eval_steps": 400, | |
| "global_step": 466, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010716677829872739, | |
| "grad_norm": 8.743332175030668, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -2.890564441680908, | |
| "logits/rejected": -2.88779878616333, | |
| "logps/chosen": -0.9741678237915039, | |
| "logps/rejected": -0.9829432368278503, | |
| "loss": 1.9541, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -0.9741678237915039, | |
| "rewards/margins": 0.008775472640991211, | |
| "rewards/rejected": -0.9829432368278503, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021433355659745478, | |
| "grad_norm": 10.116815688406033, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -2.906921863555908, | |
| "logits/rejected": -2.892861843109131, | |
| "logps/chosen": -0.9780665636062622, | |
| "logps/rejected": -1.0012353658676147, | |
| "loss": 1.9659, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.9780665636062622, | |
| "rewards/margins": 0.02316875383257866, | |
| "rewards/rejected": -1.0012353658676147, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.032150033489618215, | |
| "grad_norm": 7.106110538153892, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -2.8951611518859863, | |
| "logits/rejected": -2.917839288711548, | |
| "logps/chosen": -0.9592474102973938, | |
| "logps/rejected": -0.9844042658805847, | |
| "loss": 1.9681, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.9592474102973938, | |
| "rewards/margins": 0.025157025083899498, | |
| "rewards/rejected": -0.9844042658805847, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042866711319490956, | |
| "grad_norm": 7.768435236696218, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -2.9024550914764404, | |
| "logits/rejected": -2.9008612632751465, | |
| "logps/chosen": -0.9704982042312622, | |
| "logps/rejected": -0.9749953150749207, | |
| "loss": 1.9515, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9704982042312622, | |
| "rewards/margins": 0.004497217480093241, | |
| "rewards/rejected": -0.9749953150749207, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0535833891493637, | |
| "grad_norm": 6.749888376224144, | |
| "learning_rate": 5.319148936170212e-07, | |
| "logits/chosen": -2.8302013874053955, | |
| "logits/rejected": -2.8350861072540283, | |
| "logps/chosen": -0.9493886828422546, | |
| "logps/rejected": -0.9983908534049988, | |
| "loss": 1.9589, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.9493886828422546, | |
| "rewards/margins": 0.04900217801332474, | |
| "rewards/rejected": -0.9983908534049988, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06430006697923643, | |
| "grad_norm": 8.485574549470025, | |
| "learning_rate": 6.382978723404255e-07, | |
| "logits/chosen": -2.8937394618988037, | |
| "logits/rejected": -2.9207406044006348, | |
| "logps/chosen": -0.9754034876823425, | |
| "logps/rejected": -0.9767441749572754, | |
| "loss": 1.9649, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.9754034876823425, | |
| "rewards/margins": 0.0013406850630417466, | |
| "rewards/rejected": -0.9767441749572754, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07501674480910918, | |
| "grad_norm": 7.870295088882763, | |
| "learning_rate": 7.446808510638297e-07, | |
| "logits/chosen": -2.9515388011932373, | |
| "logits/rejected": -2.9212729930877686, | |
| "logps/chosen": -0.943376362323761, | |
| "logps/rejected": -0.9886476397514343, | |
| "loss": 1.9685, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.943376362323761, | |
| "rewards/margins": 0.04527140408754349, | |
| "rewards/rejected": -0.9886476397514343, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08573342263898191, | |
| "grad_norm": 8.140974741892732, | |
| "learning_rate": 8.51063829787234e-07, | |
| "logits/chosen": -2.886676788330078, | |
| "logits/rejected": -2.884409189224243, | |
| "logps/chosen": -0.9962645769119263, | |
| "logps/rejected": -0.9912681579589844, | |
| "loss": 1.9721, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.9962645769119263, | |
| "rewards/margins": -0.004996386356651783, | |
| "rewards/rejected": -0.9912681579589844, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09645010046885466, | |
| "grad_norm": 7.451026538163931, | |
| "learning_rate": 9.574468085106384e-07, | |
| "logits/chosen": -2.845064878463745, | |
| "logits/rejected": -2.834542751312256, | |
| "logps/chosen": -0.9799020886421204, | |
| "logps/rejected": -1.0003235340118408, | |
| "loss": 1.9741, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.9799020886421204, | |
| "rewards/margins": 0.020421411842107773, | |
| "rewards/rejected": -1.0003235340118408, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1071667782987274, | |
| "grad_norm": 7.309795501443683, | |
| "learning_rate": 9.998735159083292e-07, | |
| "logits/chosen": -2.8403611183166504, | |
| "logits/rejected": -2.8530068397521973, | |
| "logps/chosen": -0.9515609741210938, | |
| "logps/rejected": -0.9347367286682129, | |
| "loss": 1.9306, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.9515609741210938, | |
| "rewards/margins": -0.01682429201900959, | |
| "rewards/rejected": -0.9347367286682129, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11788345612860013, | |
| "grad_norm": 8.763620887929697, | |
| "learning_rate": 9.99100789302024e-07, | |
| "logits/chosen": -2.861494541168213, | |
| "logits/rejected": -2.8731250762939453, | |
| "logps/chosen": -0.9942106008529663, | |
| "logps/rejected": -1.0270545482635498, | |
| "loss": 1.9564, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.9942106008529663, | |
| "rewards/margins": 0.032844070345163345, | |
| "rewards/rejected": -1.0270545482635498, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12860013395847286, | |
| "grad_norm": 10.67048824818839, | |
| "learning_rate": 9.976266896046142e-07, | |
| "logits/chosen": -2.745513439178467, | |
| "logits/rejected": -2.7677714824676514, | |
| "logps/chosen": -0.9298070669174194, | |
| "logps/rejected": -0.9584784507751465, | |
| "loss": 1.9575, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.9298070669174194, | |
| "rewards/margins": 0.028671354055404663, | |
| "rewards/rejected": -0.9584784507751465, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13931681178834562, | |
| "grad_norm": 8.587332020158266, | |
| "learning_rate": 9.954532883292758e-07, | |
| "logits/chosen": -2.8174448013305664, | |
| "logits/rejected": -2.827878713607788, | |
| "logps/chosen": -0.9096312522888184, | |
| "logps/rejected": -0.9340154528617859, | |
| "loss": 1.9445, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.9096312522888184, | |
| "rewards/margins": 0.02438436448574066, | |
| "rewards/rejected": -0.9340154528617859, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15003348961821836, | |
| "grad_norm": 8.28588297183277, | |
| "learning_rate": 9.925836396991307e-07, | |
| "logits/chosen": -2.7543578147888184, | |
| "logits/rejected": -2.7767698764801025, | |
| "logps/chosen": -0.9463087320327759, | |
| "logps/rejected": -1.000276803970337, | |
| "loss": 1.9328, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.9463087320327759, | |
| "rewards/margins": 0.05396796017885208, | |
| "rewards/rejected": -1.000276803970337, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1607501674480911, | |
| "grad_norm": 9.088492984557346, | |
| "learning_rate": 9.89021776355227e-07, | |
| "logits/chosen": -2.7640278339385986, | |
| "logits/rejected": -2.7647509574890137, | |
| "logps/chosen": -0.9192463755607605, | |
| "logps/rejected": -0.9625679850578308, | |
| "loss": 1.95, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9192463755607605, | |
| "rewards/margins": 0.043321557343006134, | |
| "rewards/rejected": -0.9625679850578308, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17146684527796383, | |
| "grad_norm": 9.431427301698536, | |
| "learning_rate": 9.847727036895757e-07, | |
| "logits/chosen": -2.8038690090179443, | |
| "logits/rejected": -2.82570481300354, | |
| "logps/chosen": -0.9232662320137024, | |
| "logps/rejected": -1.0234979391098022, | |
| "loss": 1.9471, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.9232662320137024, | |
| "rewards/margins": 0.10023164749145508, | |
| "rewards/rejected": -1.0234979391098022, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18218352310783656, | |
| "grad_norm": 10.888524371529982, | |
| "learning_rate": 9.79842392811207e-07, | |
| "logits/chosen": -2.8308658599853516, | |
| "logits/rejected": -2.840531349182129, | |
| "logps/chosen": -0.9489914774894714, | |
| "logps/rejected": -1.0153648853302002, | |
| "loss": 1.9556, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9489914774894714, | |
| "rewards/margins": 0.06637347489595413, | |
| "rewards/rejected": -1.0153648853302002, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19290020093770932, | |
| "grad_norm": 11.29650020138619, | |
| "learning_rate": 9.742377721551285e-07, | |
| "logits/chosen": -2.837176561355591, | |
| "logits/rejected": -2.8398804664611816, | |
| "logps/chosen": -0.9703192710876465, | |
| "logps/rejected": -1.0296133756637573, | |
| "loss": 1.9557, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.9703192710876465, | |
| "rewards/margins": 0.05929403379559517, | |
| "rewards/rejected": -1.0296133756637573, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20361687876758205, | |
| "grad_norm": 10.210746185136436, | |
| "learning_rate": 9.679667177459793e-07, | |
| "logits/chosen": -2.8701424598693848, | |
| "logits/rejected": -2.8827216625213623, | |
| "logps/chosen": -0.9922488927841187, | |
| "logps/rejected": -1.0364185571670532, | |
| "loss": 1.935, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9922488927841187, | |
| "rewards/margins": 0.04416964575648308, | |
| "rewards/rejected": -1.0364185571670532, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2143335565974548, | |
| "grad_norm": 12.483295768661268, | |
| "learning_rate": 9.61038042130062e-07, | |
| "logits/chosen": -2.8926491737365723, | |
| "logits/rejected": -2.901106357574463, | |
| "logps/chosen": -1.0361850261688232, | |
| "logps/rejected": -1.0975309610366821, | |
| "loss": 1.9596, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.0361850261688232, | |
| "rewards/margins": 0.06134594604372978, | |
| "rewards/rejected": -1.0975309610366821, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22505023442732752, | |
| "grad_norm": 10.251687371352437, | |
| "learning_rate": 9.534614819913056e-07, | |
| "logits/chosen": -2.8896324634552, | |
| "logits/rejected": -2.8919472694396973, | |
| "logps/chosen": -0.951892077922821, | |
| "logps/rejected": -1.0144917964935303, | |
| "loss": 1.9124, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.951892077922821, | |
| "rewards/margins": 0.06259982287883759, | |
| "rewards/rejected": -1.0144917964935303, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23576691225720026, | |
| "grad_norm": 12.050194786037592, | |
| "learning_rate": 9.45247684468561e-07, | |
| "logits/chosen": -2.884787082672119, | |
| "logits/rejected": -2.8778603076934814, | |
| "logps/chosen": -0.9767457842826843, | |
| "logps/rejected": -1.0685449838638306, | |
| "loss": 1.9478, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.9767457842826843, | |
| "rewards/margins": 0.09179918467998505, | |
| "rewards/rejected": -1.0685449838638306, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24648359008707302, | |
| "grad_norm": 16.70841820517718, | |
| "learning_rate": 9.364081921934605e-07, | |
| "logits/chosen": -2.938917636871338, | |
| "logits/rejected": -2.9450087547302246, | |
| "logps/chosen": -0.9636579751968384, | |
| "logps/rejected": -0.9759002923965454, | |
| "loss": 1.9752, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.9636579751968384, | |
| "rewards/margins": 0.01224228460341692, | |
| "rewards/rejected": -0.9759002923965454, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2572002679169457, | |
| "grad_norm": 11.293944567951943, | |
| "learning_rate": 9.269554270698635e-07, | |
| "logits/chosen": -2.945613384246826, | |
| "logits/rejected": -2.9629931449890137, | |
| "logps/chosen": -0.9591239094734192, | |
| "logps/rejected": -0.9764735102653503, | |
| "loss": 1.9573, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.9591239094734192, | |
| "rewards/margins": 0.01734955981373787, | |
| "rewards/rejected": -0.9764735102653503, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2679169457468185, | |
| "grad_norm": 11.980171410561908, | |
| "learning_rate": 9.169026728176843e-07, | |
| "logits/chosen": -2.933450222015381, | |
| "logits/rejected": -2.9331278800964355, | |
| "logps/chosen": -0.9632900357246399, | |
| "logps/rejected": -1.0157678127288818, | |
| "loss": 1.9492, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.9632900357246399, | |
| "rewards/margins": 0.05247765779495239, | |
| "rewards/rejected": -1.0157678127288818, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27863362357669125, | |
| "grad_norm": 13.043968582425784, | |
| "learning_rate": 9.062640563056338e-07, | |
| "logits/chosen": -2.991359233856201, | |
| "logits/rejected": -2.988865613937378, | |
| "logps/chosen": -0.9493595361709595, | |
| "logps/rejected": -1.034130334854126, | |
| "loss": 1.922, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9493595361709595, | |
| "rewards/margins": 0.08477075397968292, | |
| "rewards/rejected": -1.034130334854126, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.289350301406564, | |
| "grad_norm": 12.532276437925889, | |
| "learning_rate": 8.950545276991059e-07, | |
| "logits/chosen": -3.0021941661834717, | |
| "logits/rejected": -3.0065648555755615, | |
| "logps/chosen": -0.9963411092758179, | |
| "logps/rejected": -1.0470635890960693, | |
| "loss": 1.9424, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.9963411092758179, | |
| "rewards/margins": 0.05072233825922012, | |
| "rewards/rejected": -1.0470635890960693, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3000669792364367, | |
| "grad_norm": 12.705557690233402, | |
| "learning_rate": 8.832898394511059e-07, | |
| "logits/chosen": -2.9534518718719482, | |
| "logits/rejected": -2.955967903137207, | |
| "logps/chosen": -0.9860572814941406, | |
| "logps/rejected": -1.0404431819915771, | |
| "loss": 1.9178, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.9860572814941406, | |
| "rewards/margins": 0.054385989904403687, | |
| "rewards/rejected": -1.0404431819915771, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31078365706630945, | |
| "grad_norm": 13.072919082266676, | |
| "learning_rate": 8.709865241657479e-07, | |
| "logits/chosen": -3.0093483924865723, | |
| "logits/rejected": -2.979004383087158, | |
| "logps/chosen": -0.9930181503295898, | |
| "logps/rejected": -1.0835025310516357, | |
| "loss": 1.9157, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.9930181503295898, | |
| "rewards/margins": 0.09048454463481903, | |
| "rewards/rejected": -1.0835025310516357, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3215003348961822, | |
| "grad_norm": 12.426824715988731, | |
| "learning_rate": 8.581618713654237e-07, | |
| "logits/chosen": -3.013441324234009, | |
| "logits/rejected": -3.020658016204834, | |
| "logps/chosen": -0.9797670245170593, | |
| "logps/rejected": -1.0478479862213135, | |
| "loss": 1.9384, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.9797670245170593, | |
| "rewards/margins": 0.06808092445135117, | |
| "rewards/rejected": -1.0478479862213135, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3322170127260549, | |
| "grad_norm": 12.994046950866517, | |
| "learning_rate": 8.448339031942969e-07, | |
| "logits/chosen": -3.1380882263183594, | |
| "logits/rejected": -3.119088649749756, | |
| "logps/chosen": -1.0096795558929443, | |
| "logps/rejected": -1.0793818235397339, | |
| "loss": 1.9501, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0096795558929443, | |
| "rewards/margins": 0.0697021633386612, | |
| "rewards/rejected": -1.0793818235397339, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.34293369055592765, | |
| "grad_norm": 14.285933742292032, | |
| "learning_rate": 8.310213490922615e-07, | |
| "logits/chosen": -3.086894989013672, | |
| "logits/rejected": -3.0781033039093018, | |
| "logps/chosen": -1.01212477684021, | |
| "logps/rejected": -1.1041617393493652, | |
| "loss": 1.924, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.01212477684021, | |
| "rewards/margins": 0.09203706681728363, | |
| "rewards/rejected": -1.1041617393493652, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3536503683858004, | |
| "grad_norm": 12.902647964980785, | |
| "learning_rate": 8.167436194749575e-07, | |
| "logits/chosen": -3.154498815536499, | |
| "logits/rejected": -3.1380507946014404, | |
| "logps/chosen": -0.9696465730667114, | |
| "logps/rejected": -1.0397228002548218, | |
| "loss": 1.9234, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.9696465730667114, | |
| "rewards/margins": 0.07007621228694916, | |
| "rewards/rejected": -1.0397228002548218, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3643670462156731, | |
| "grad_norm": 10.461713458782954, | |
| "learning_rate": 8.020207784568291e-07, | |
| "logits/chosen": -3.0909907817840576, | |
| "logits/rejected": -3.0954232215881348, | |
| "logps/chosen": -0.9789296388626099, | |
| "logps/rejected": -1.0583703517913818, | |
| "loss": 1.8994, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9789296388626099, | |
| "rewards/margins": 0.07944078743457794, | |
| "rewards/rejected": -1.0583703517913818, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3750837240455459, | |
| "grad_norm": 14.39267394950559, | |
| "learning_rate": 7.868735156555566e-07, | |
| "logits/chosen": -3.113248348236084, | |
| "logits/rejected": -3.116093158721924, | |
| "logps/chosen": -0.9837858080863953, | |
| "logps/rejected": -1.0499708652496338, | |
| "loss": 1.9274, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.9837858080863953, | |
| "rewards/margins": 0.06618491560220718, | |
| "rewards/rejected": -1.0499708652496338, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.38580040187541864, | |
| "grad_norm": 15.99990919657707, | |
| "learning_rate": 7.713231171174868e-07, | |
| "logits/chosen": -3.117457866668701, | |
| "logits/rejected": -3.1266255378723145, | |
| "logps/chosen": -1.0582538843154907, | |
| "logps/rejected": -1.133825659751892, | |
| "loss": 1.9527, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.0582538843154907, | |
| "rewards/margins": 0.0755719467997551, | |
| "rewards/rejected": -1.133825659751892, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3965170797052914, | |
| "grad_norm": 14.766672490703176, | |
| "learning_rate": 7.553914354049162e-07, | |
| "logits/chosen": -3.0913758277893066, | |
| "logits/rejected": -3.103057384490967, | |
| "logps/chosen": -1.023158073425293, | |
| "logps/rejected": -1.06888747215271, | |
| "loss": 1.9093, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.023158073425293, | |
| "rewards/margins": 0.04572921246290207, | |
| "rewards/rejected": -1.06888747215271, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4072337575351641, | |
| "grad_norm": 15.970782695590072, | |
| "learning_rate": 7.39100858887266e-07, | |
| "logits/chosen": -3.0117366313934326, | |
| "logits/rejected": -3.0253918170928955, | |
| "logps/chosen": -0.9792415499687195, | |
| "logps/rejected": -1.192731499671936, | |
| "loss": 1.9159, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.9792415499687195, | |
| "rewards/margins": 0.2134898155927658, | |
| "rewards/rejected": -1.192731499671936, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.41795043536503684, | |
| "grad_norm": 14.772743027007902, | |
| "learning_rate": 7.224742802793004e-07, | |
| "logits/chosen": -3.117619514465332, | |
| "logits/rejected": -3.105710506439209, | |
| "logps/chosen": -0.9820570945739746, | |
| "logps/rejected": -1.0550075769424438, | |
| "loss": 1.9206, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.9820570945739746, | |
| "rewards/margins": 0.07295046001672745, | |
| "rewards/rejected": -1.0550075769424438, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4286671131949096, | |
| "grad_norm": 13.569195793364742, | |
| "learning_rate": 7.055350644706022e-07, | |
| "logits/chosen": -3.0321404933929443, | |
| "logits/rejected": -3.049098491668701, | |
| "logps/chosen": -0.9425796270370483, | |
| "logps/rejected": -1.015730857849121, | |
| "loss": 1.8967, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.9425796270370483, | |
| "rewards/margins": 0.07315118610858917, | |
| "rewards/rejected": -1.015730857849121, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4393837910247823, | |
| "grad_norm": 17.357578155356272, | |
| "learning_rate": 6.883070156915139e-07, | |
| "logits/chosen": -3.1063551902770996, | |
| "logits/rejected": -3.087735652923584, | |
| "logps/chosen": -0.9921137094497681, | |
| "logps/rejected": -1.0688540935516357, | |
| "loss": 1.9072, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.9921137094497681, | |
| "rewards/margins": 0.07674027979373932, | |
| "rewards/rejected": -1.0688540935516357, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.45010046885465504, | |
| "grad_norm": 18.638096035157822, | |
| "learning_rate": 6.708143440616845e-07, | |
| "logits/chosen": -3.037801504135132, | |
| "logits/rejected": -3.043381452560425, | |
| "logps/chosen": -0.9749434590339661, | |
| "logps/rejected": -1.093065619468689, | |
| "loss": 1.8807, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.9749434590339661, | |
| "rewards/margins": 0.11812222003936768, | |
| "rewards/rejected": -1.093065619468689, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4608171466845278, | |
| "grad_norm": 14.384225079310538, | |
| "learning_rate": 6.530816315682306e-07, | |
| "logits/chosen": -3.039210319519043, | |
| "logits/rejected": -3.036463499069214, | |
| "logps/chosen": -0.9878280758857727, | |
| "logps/rejected": -1.0471224784851074, | |
| "loss": 1.8936, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.9878280758857727, | |
| "rewards/margins": 0.0592944398522377, | |
| "rewards/rejected": -1.0471224784851074, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4715338245144005, | |
| "grad_norm": 15.635777348265833, | |
| "learning_rate": 6.351337975213237e-07, | |
| "logits/chosen": -2.952924966812134, | |
| "logits/rejected": -2.9542202949523926, | |
| "logps/chosen": -0.950057327747345, | |
| "logps/rejected": -1.0494937896728516, | |
| "loss": 1.9176, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.950057327747345, | |
| "rewards/margins": 0.0994364321231842, | |
| "rewards/rejected": -1.0494937896728516, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4822505023442733, | |
| "grad_norm": 15.744387463358793, | |
| "learning_rate": 6.169960635357437e-07, | |
| "logits/chosen": -2.9482624530792236, | |
| "logits/rejected": -2.989917278289795, | |
| "logps/chosen": -0.9638145565986633, | |
| "logps/rejected": -1.1090409755706787, | |
| "loss": 1.9033, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.9638145565986633, | |
| "rewards/margins": 0.14522647857666016, | |
| "rewards/rejected": -1.1090409755706787, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.49296718017414604, | |
| "grad_norm": 20.61441976992923, | |
| "learning_rate": 5.98693918087613e-07, | |
| "logits/chosen": -2.8699066638946533, | |
| "logits/rejected": -2.870300531387329, | |
| "logps/chosen": -0.9924839735031128, | |
| "logps/rejected": -1.0983588695526123, | |
| "loss": 1.8841, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.9924839735031128, | |
| "rewards/margins": 0.10587481409311295, | |
| "rewards/rejected": -1.0983588695526123, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5036838580040187, | |
| "grad_norm": 16.252130697815385, | |
| "learning_rate": 5.802530806961194e-07, | |
| "logits/chosen": -2.817322254180908, | |
| "logits/rejected": -2.829948663711548, | |
| "logps/chosen": -0.980228066444397, | |
| "logps/rejected": -1.1411757469177246, | |
| "loss": 1.8909, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.980228066444397, | |
| "rewards/margins": 0.16094763576984406, | |
| "rewards/rejected": -1.1411757469177246, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5144005358338914, | |
| "grad_norm": 18.85039464903285, | |
| "learning_rate": 5.616994657805565e-07, | |
| "logits/chosen": -2.8348546028137207, | |
| "logits/rejected": -2.8363900184631348, | |
| "logps/chosen": -1.0022538900375366, | |
| "logps/rejected": -1.091737985610962, | |
| "loss": 1.9144, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.0022538900375366, | |
| "rewards/margins": 0.08948404341936111, | |
| "rewards/rejected": -1.091737985610962, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5251172136637642, | |
| "grad_norm": 18.11195320561811, | |
| "learning_rate": 5.430591462434792e-07, | |
| "logits/chosen": -2.8302159309387207, | |
| "logits/rejected": -2.8293166160583496, | |
| "logps/chosen": -0.9994494318962097, | |
| "logps/rejected": -1.0923607349395752, | |
| "loss": 1.8816, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.9994494318962097, | |
| "rewards/margins": 0.09291131794452667, | |
| "rewards/rejected": -1.0923607349395752, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.535833891493637, | |
| "grad_norm": 17.05860095183427, | |
| "learning_rate": 5.24358316831145e-07, | |
| "logits/chosen": -2.8530397415161133, | |
| "logits/rejected": -2.8722071647644043, | |
| "logps/chosen": -1.039880394935608, | |
| "logps/rejected": -1.1407957077026367, | |
| "loss": 1.8947, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.039880394935608, | |
| "rewards/margins": 0.10091539472341537, | |
| "rewards/rejected": -1.1407957077026367, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5465505693235098, | |
| "grad_norm": 18.749033093794495, | |
| "learning_rate": 5.05623257322734e-07, | |
| "logits/chosen": -2.93331241607666, | |
| "logits/rejected": -2.944380283355713, | |
| "logps/chosen": -0.9659943580627441, | |
| "logps/rejected": -1.0837066173553467, | |
| "loss": 1.8979, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.9659943580627441, | |
| "rewards/margins": 0.11771215498447418, | |
| "rewards/rejected": -1.0837066173553467, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5572672471533825, | |
| "grad_norm": 16.94811716813612, | |
| "learning_rate": 4.868802956000706e-07, | |
| "logits/chosen": -2.931999683380127, | |
| "logits/rejected": -2.9354639053344727, | |
| "logps/chosen": -0.9535225033760071, | |
| "logps/rejected": -1.0598043203353882, | |
| "loss": 1.8988, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.9535225033760071, | |
| "rewards/margins": 0.10628175735473633, | |
| "rewards/rejected": -1.0598043203353882, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5679839249832552, | |
| "grad_norm": 16.773823816327198, | |
| "learning_rate": 4.681557706497518e-07, | |
| "logits/chosen": -2.9415526390075684, | |
| "logits/rejected": -2.928161144256592, | |
| "logps/chosen": -0.9740009307861328, | |
| "logps/rejected": -1.1567602157592773, | |
| "loss": 1.8795, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.9740009307861328, | |
| "rewards/margins": 0.1827593445777893, | |
| "rewards/rejected": -1.1567602157592773, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.578700602813128, | |
| "grad_norm": 18.918032339525077, | |
| "learning_rate": 4.494759955496678e-07, | |
| "logits/chosen": -2.977628231048584, | |
| "logits/rejected": -2.993778705596924, | |
| "logps/chosen": -0.9572176933288574, | |
| "logps/rejected": -1.0660667419433594, | |
| "loss": 1.8826, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.9572176933288574, | |
| "rewards/margins": 0.1088489443063736, | |
| "rewards/rejected": -1.0660667419433594, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5894172806430007, | |
| "grad_norm": 18.052342935862615, | |
| "learning_rate": 4.3086722049193145e-07, | |
| "logits/chosen": -3.0297939777374268, | |
| "logits/rejected": -3.0311572551727295, | |
| "logps/chosen": -0.9626353979110718, | |
| "logps/rejected": -1.1277216672897339, | |
| "loss": 1.8892, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.9626353979110718, | |
| "rewards/margins": 0.16508622467517853, | |
| "rewards/rejected": -1.1277216672897339, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6001339584728734, | |
| "grad_norm": 17.526484616917582, | |
| "learning_rate": 4.1235559589418164e-07, | |
| "logits/chosen": -2.9183692932128906, | |
| "logits/rejected": -2.9183740615844727, | |
| "logps/chosen": -1.0271321535110474, | |
| "logps/rejected": -1.1547820568084717, | |
| "loss": 1.8912, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.0271321535110474, | |
| "rewards/margins": 0.1276499330997467, | |
| "rewards/rejected": -1.1547820568084717, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6108506363027462, | |
| "grad_norm": 18.526748745941557, | |
| "learning_rate": 3.9396713565109375e-07, | |
| "logits/chosen": -2.999577045440674, | |
| "logits/rejected": -2.9965460300445557, | |
| "logps/chosen": -0.9945805668830872, | |
| "logps/rejected": -1.077164888381958, | |
| "loss": 1.9015, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.9945805668830872, | |
| "rewards/margins": 0.08258435875177383, | |
| "rewards/rejected": -1.077164888381958, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6215673141326189, | |
| "grad_norm": 20.379894797382978, | |
| "learning_rate": 3.757276805777454e-07, | |
| "logits/chosen": -3.0354697704315186, | |
| "logits/rejected": -3.0337672233581543, | |
| "logps/chosen": -0.9179645776748657, | |
| "logps/rejected": -1.0484569072723389, | |
| "loss": 1.8803, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.9179645776748657, | |
| "rewards/margins": 0.13049228489398956, | |
| "rewards/rejected": -1.0484569072723389, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6322839919624916, | |
| "grad_norm": 20.690228448794066, | |
| "learning_rate": 3.5766286209620446e-07, | |
| "logits/chosen": -3.001173973083496, | |
| "logits/rejected": -2.9999818801879883, | |
| "logps/chosen": -0.9978302717208862, | |
| "logps/rejected": -1.218730092048645, | |
| "loss": 1.8787, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.9978302717208862, | |
| "rewards/margins": 0.22089961171150208, | |
| "rewards/rejected": -1.218730092048645, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6430006697923644, | |
| "grad_norm": 21.088864907959156, | |
| "learning_rate": 3.3979806621637095e-07, | |
| "logits/chosen": -3.0734269618988037, | |
| "logits/rejected": -3.0850183963775635, | |
| "logps/chosen": -1.0158809423446655, | |
| "logps/rejected": -1.203379511833191, | |
| "loss": 1.8412, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0158809423446655, | |
| "rewards/margins": 0.1874985247850418, | |
| "rewards/rejected": -1.203379511833191, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6537173476222371, | |
| "grad_norm": 20.56239032846782, | |
| "learning_rate": 3.221583978616932e-07, | |
| "logits/chosen": -3.0289087295532227, | |
| "logits/rejected": -3.0159077644348145, | |
| "logps/chosen": -1.0385640859603882, | |
| "logps/rejected": -1.1604039669036865, | |
| "loss": 1.8697, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0385640859603882, | |
| "rewards/margins": 0.12183995544910431, | |
| "rewards/rejected": -1.1604039669036865, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6644340254521098, | |
| "grad_norm": 23.460225989452297, | |
| "learning_rate": 3.047686455898836e-07, | |
| "logits/chosen": -3.080087423324585, | |
| "logits/rejected": -3.0696513652801514, | |
| "logps/chosen": -1.0038082599639893, | |
| "logps/rejected": -1.1881027221679688, | |
| "loss": 1.8639, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0038082599639893, | |
| "rewards/margins": 0.18429455161094666, | |
| "rewards/rejected": -1.1881027221679688, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6751507032819826, | |
| "grad_norm": 20.943994771592788, | |
| "learning_rate": 2.8765324675821734e-07, | |
| "logits/chosen": -2.944671630859375, | |
| "logits/rejected": -2.9371066093444824, | |
| "logps/chosen": -1.0098899602890015, | |
| "logps/rejected": -1.1762586832046509, | |
| "loss": 1.8745, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0098899602890015, | |
| "rewards/margins": 0.16636863350868225, | |
| "rewards/rejected": -1.1762586832046509, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6858673811118553, | |
| "grad_norm": 23.376564204026522, | |
| "learning_rate": 2.708362531823621e-07, | |
| "logits/chosen": -2.9114675521850586, | |
| "logits/rejected": -2.8761847019195557, | |
| "logps/chosen": -1.0019965171813965, | |
| "logps/rejected": -1.203962802886963, | |
| "loss": 1.8847, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0019965171813965, | |
| "rewards/margins": 0.20196643471717834, | |
| "rewards/rejected": -1.203962802886963, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.696584058941728, | |
| "grad_norm": 21.72889429456692, | |
| "learning_rate": 2.5434129733700093e-07, | |
| "logits/chosen": -2.859740734100342, | |
| "logits/rejected": -2.850984573364258, | |
| "logps/chosen": -0.9476163983345032, | |
| "logps/rejected": -1.190605640411377, | |
| "loss": 1.8338, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.9476163983345032, | |
| "rewards/margins": 0.24298930168151855, | |
| "rewards/rejected": -1.190605640411377, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7073007367716008, | |
| "grad_norm": 23.976896941165123, | |
| "learning_rate": 2.3819155914574235e-07, | |
| "logits/chosen": -2.881459951400757, | |
| "logits/rejected": -2.860196113586426, | |
| "logps/chosen": -0.9749948382377625, | |
| "logps/rejected": -1.1787073612213135, | |
| "loss": 1.8473, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.9749948382377625, | |
| "rewards/margins": 0.20371253788471222, | |
| "rewards/rejected": -1.1787073612213135, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7180174146014735, | |
| "grad_norm": 28.808399946877696, | |
| "learning_rate": 2.2240973340698882e-07, | |
| "logits/chosen": -2.9146461486816406, | |
| "logits/rejected": -2.9025330543518066, | |
| "logps/chosen": -0.9778841137886047, | |
| "logps/rejected": -1.1417423486709595, | |
| "loss": 1.8579, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.9778841137886047, | |
| "rewards/margins": 0.16385838389396667, | |
| "rewards/rejected": -1.1417423486709595, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7287340924313462, | |
| "grad_norm": 25.882278976736618, | |
| "learning_rate": 2.0701799790153896e-07, | |
| "logits/chosen": -2.8679490089416504, | |
| "logits/rejected": -2.8503952026367188, | |
| "logps/chosen": -0.9974260330200195, | |
| "logps/rejected": -1.2173974514007568, | |
| "loss": 1.858, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.9974260330200195, | |
| "rewards/margins": 0.21997132897377014, | |
| "rewards/rejected": -1.2173974514007568, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.739450770261219, | |
| "grad_norm": 25.899929236131076, | |
| "learning_rate": 1.9203798222674138e-07, | |
| "logits/chosen": -2.881312608718872, | |
| "logits/rejected": -2.886810779571533, | |
| "logps/chosen": -0.9790380597114563, | |
| "logps/rejected": -1.2078144550323486, | |
| "loss": 1.8434, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.9790380597114563, | |
| "rewards/margins": 0.2287764996290207, | |
| "rewards/rejected": -1.2078144550323486, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7501674480910918, | |
| "grad_norm": 22.394258706384488, | |
| "learning_rate": 1.774907374009953e-07, | |
| "logits/chosen": -2.854478120803833, | |
| "logits/rejected": -2.8414759635925293, | |
| "logps/chosen": -0.9902400970458984, | |
| "logps/rejected": -1.1531636714935303, | |
| "loss": 1.8534, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.9902400970458984, | |
| "rewards/margins": 0.16292349994182587, | |
| "rewards/rejected": -1.1531636714935303, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7608841259209645, | |
| "grad_norm": 32.75022427627133, | |
| "learning_rate": 1.6339670628131326e-07, | |
| "logits/chosen": -2.8383326530456543, | |
| "logits/rejected": -2.819650888442993, | |
| "logps/chosen": -1.0634033679962158, | |
| "logps/rejected": -1.194883942604065, | |
| "loss": 1.876, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.0634033679962158, | |
| "rewards/margins": 0.13148045539855957, | |
| "rewards/rejected": -1.194883942604065, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7716008037508373, | |
| "grad_norm": 23.817047943255176, | |
| "learning_rate": 1.4977569483551632e-07, | |
| "logits/chosen": -2.8515446186065674, | |
| "logits/rejected": -2.8371310234069824, | |
| "logps/chosen": -0.9948540925979614, | |
| "logps/rejected": -1.1568361520767212, | |
| "loss": 1.8636, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.9948540925979614, | |
| "rewards/margins": 0.16198216378688812, | |
| "rewards/rejected": -1.1568361520767212, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.78231748158071, | |
| "grad_norm": 29.671187736811945, | |
| "learning_rate": 1.366468443094343e-07, | |
| "logits/chosen": -2.875363826751709, | |
| "logits/rejected": -2.8553128242492676, | |
| "logps/chosen": -1.039432168006897, | |
| "logps/rejected": -1.2274564504623413, | |
| "loss": 1.859, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.039432168006897, | |
| "rewards/margins": 0.18802431225776672, | |
| "rewards/rejected": -1.2274564504623413, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7930341594105828, | |
| "grad_norm": 33.954689650694874, | |
| "learning_rate": 1.240286043282197e-07, | |
| "logits/chosen": -2.8917102813720703, | |
| "logits/rejected": -2.8770086765289307, | |
| "logps/chosen": -1.0392128229141235, | |
| "logps/rejected": -1.1902668476104736, | |
| "loss": 1.8531, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0392128229141235, | |
| "rewards/margins": 0.15105411410331726, | |
| "rewards/rejected": -1.1902668476104736, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8037508372404555, | |
| "grad_norm": 29.25246301873259, | |
| "learning_rate": 1.1193870696958058e-07, | |
| "logits/chosen": -2.866624116897583, | |
| "logits/rejected": -2.865328073501587, | |
| "logps/chosen": -1.0088036060333252, | |
| "logps/rejected": -1.1752225160598755, | |
| "loss": 1.871, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0088036060333252, | |
| "rewards/margins": 0.16641899943351746, | |
| "rewards/rejected": -1.1752225160598755, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8144675150703282, | |
| "grad_norm": 23.976856879448345, | |
| "learning_rate": 1.003941418453616e-07, | |
| "logits/chosen": -2.9358296394348145, | |
| "logits/rejected": -2.9078102111816406, | |
| "logps/chosen": -0.9131346940994263, | |
| "logps/rejected": -1.0847430229187012, | |
| "loss": 1.8614, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.9131346940994263, | |
| "rewards/margins": 0.1716083586215973, | |
| "rewards/rejected": -1.0847430229187012, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.825184192900201, | |
| "grad_norm": 32.2123584818521, | |
| "learning_rate": 8.941113222649327e-08, | |
| "logits/chosen": -2.8775150775909424, | |
| "logits/rejected": -2.842041254043579, | |
| "logps/chosen": -1.0087697505950928, | |
| "logps/rejected": -1.1664918661117554, | |
| "loss": 1.8549, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.0087697505950928, | |
| "rewards/margins": 0.15772202610969543, | |
| "rewards/rejected": -1.1664918661117554, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8359008707300737, | |
| "grad_norm": 31.35752151311303, | |
| "learning_rate": 7.900511224486083e-08, | |
| "logits/chosen": -2.8315508365631104, | |
| "logits/rejected": -2.801546812057495, | |
| "logps/chosen": -1.1222208738327026, | |
| "logps/rejected": -1.3023992776870728, | |
| "loss": 1.8914, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.1222208738327026, | |
| "rewards/margins": 0.18017837405204773, | |
| "rewards/rejected": -1.3023992776870728, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8466175485599464, | |
| "grad_norm": 31.331425662268856, | |
| "learning_rate": 6.919070520412768e-08, | |
| "logits/chosen": -2.8807644844055176, | |
| "logits/rejected": -2.877798080444336, | |
| "logps/chosen": -1.002872347831726, | |
| "logps/rejected": -1.2327783107757568, | |
| "loss": 1.8638, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.002872347831726, | |
| "rewards/margins": 0.22990599274635315, | |
| "rewards/rejected": -1.2327783107757568, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8573342263898192, | |
| "grad_norm": 27.689197557194653, | |
| "learning_rate": 5.998170302999528e-08, | |
| "logits/chosen": -2.866481304168701, | |
| "logits/rejected": -2.851442575454712, | |
| "logps/chosen": -0.9658647775650024, | |
| "logps/rejected": -1.2617170810699463, | |
| "loss": 1.8319, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.9658647775650024, | |
| "rewards/margins": 0.2958524823188782, | |
| "rewards/rejected": -1.2617170810699463, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8573342263898192, | |
| "eval_logits/chosen": -2.7880678176879883, | |
| "eval_logits/rejected": -2.7752296924591064, | |
| "eval_logps/chosen": -0.9790560007095337, | |
| "eval_logps/rejected": -1.1183466911315918, | |
| "eval_loss": 1.8910752534866333, | |
| "eval_rewards/accuracies": 0.626329779624939, | |
| "eval_rewards/chosen": -0.9790560007095337, | |
| "eval_rewards/margins": 0.1392906755208969, | |
| "eval_rewards/rejected": -1.1183466911315918, | |
| "eval_runtime": 436.0547, | |
| "eval_samples_per_second": 6.866, | |
| "eval_steps_per_second": 0.431, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8680509042196919, | |
| "grad_norm": 32.62329228952369, | |
| "learning_rate": 5.1391046888775493e-08, | |
| "logits/chosen": -2.8493621349334717, | |
| "logits/rejected": -2.8683342933654785, | |
| "logps/chosen": -0.949614405632019, | |
| "logps/rejected": -1.1213003396987915, | |
| "loss": 1.8697, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.949614405632019, | |
| "rewards/margins": 0.1716858595609665, | |
| "rewards/rejected": -1.1213003396987915, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8787675820495646, | |
| "grad_norm": 25.8354383097588, | |
| "learning_rate": 4.343080900151375e-08, | |
| "logits/chosen": -2.881870985031128, | |
| "logits/rejected": -2.899685859680176, | |
| "logps/chosen": -0.9602924585342407, | |
| "logps/rejected": -1.1478478908538818, | |
| "loss": 1.8356, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.9602924585342407, | |
| "rewards/margins": 0.1875552237033844, | |
| "rewards/rejected": -1.1478478908538818, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8894842598794374, | |
| "grad_norm": 25.748109045396355, | |
| "learning_rate": 3.611217567921709e-08, | |
| "logits/chosen": -2.8336663246154785, | |
| "logits/rejected": -2.813194513320923, | |
| "logps/chosen": -1.000084400177002, | |
| "logps/rejected": -1.1840862035751343, | |
| "loss": 1.8298, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.000084400177002, | |
| "rewards/margins": 0.18400196731090546, | |
| "rewards/rejected": -1.1840862035751343, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9002009377093101, | |
| "grad_norm": 32.06643463040125, | |
| "learning_rate": 2.944543160302787e-08, | |
| "logits/chosen": -2.857529878616333, | |
| "logits/rejected": -2.8420491218566895, | |
| "logps/chosen": -0.9639240503311157, | |
| "logps/rejected": -1.1220653057098389, | |
| "loss": 1.8748, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.9639240503311157, | |
| "rewards/margins": 0.15814124047756195, | |
| "rewards/rejected": -1.1220653057098389, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9109176155391828, | |
| "grad_norm": 35.70189678978648, | |
| "learning_rate": 2.343994537143479e-08, | |
| "logits/chosen": -2.8647522926330566, | |
| "logits/rejected": -2.8519949913024902, | |
| "logps/chosen": -0.9895116686820984, | |
| "logps/rejected": -1.3003708124160767, | |
| "loss": 1.8221, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.9895116686820984, | |
| "rewards/margins": 0.31085896492004395, | |
| "rewards/rejected": -1.3003708124160767, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9216342933690556, | |
| "grad_norm": 34.791804330091374, | |
| "learning_rate": 1.81041563348297e-08, | |
| "logits/chosen": -2.837439775466919, | |
| "logits/rejected": -2.818183660507202, | |
| "logps/chosen": -1.0866570472717285, | |
| "logps/rejected": -1.2705637216567993, | |
| "loss": 1.8716, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.0866570472717285, | |
| "rewards/margins": 0.1839066594839096, | |
| "rewards/rejected": -1.2705637216567993, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9323509711989283, | |
| "grad_norm": 37.275898794915584, | |
| "learning_rate": 1.3445562735912962e-08, | |
| "logits/chosen": -2.9299168586730957, | |
| "logits/rejected": -2.9030609130859375, | |
| "logps/chosen": -0.9819711446762085, | |
| "logps/rejected": -1.2765450477600098, | |
| "loss": 1.8211, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.9819711446762085, | |
| "rewards/margins": 0.29457393288612366, | |
| "rewards/rejected": -1.2765450477600098, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.943067649028801, | |
| "grad_norm": 27.67934081660638, | |
| "learning_rate": 9.470711172611722e-09, | |
| "logits/chosen": -2.852858066558838, | |
| "logits/rejected": -2.8415844440460205, | |
| "logps/chosen": -0.9760105013847351, | |
| "logps/rejected": -1.2501945495605469, | |
| "loss": 1.8076, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.9760105013847351, | |
| "rewards/margins": 0.27418413758277893, | |
| "rewards/rejected": -1.2501945495605469, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9537843268586738, | |
| "grad_norm": 31.020987107038604, | |
| "learning_rate": 6.185187398319691e-09, | |
| "logits/chosen": -2.848702907562256, | |
| "logits/rejected": -2.8342273235321045, | |
| "logps/chosen": -1.029883623123169, | |
| "logps/rejected": -1.24784255027771, | |
| "loss": 1.8487, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.029883623123169, | |
| "rewards/margins": 0.2179589569568634, | |
| "rewards/rejected": -1.24784255027771, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9645010046885466, | |
| "grad_norm": 30.471294839063304, | |
| "learning_rate": 3.593608472386045e-09, | |
| "logits/chosen": -2.8422036170959473, | |
| "logits/rejected": -2.8164522647857666, | |
| "logps/chosen": -0.9777735471725464, | |
| "logps/rejected": -1.1690075397491455, | |
| "loss": 1.8265, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.9777735471725464, | |
| "rewards/margins": 0.19123412668704987, | |
| "rewards/rejected": -1.1690075397491455, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9752176825184193, | |
| "grad_norm": 30.36082211778746, | |
| "learning_rate": 1.6996162718847518e-09, | |
| "logits/chosen": -2.811438798904419, | |
| "logits/rejected": -2.8197503089904785, | |
| "logps/chosen": -0.9904121160507202, | |
| "logps/rejected": -1.257320523262024, | |
| "loss": 1.8079, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.9904121160507202, | |
| "rewards/margins": 0.2669083774089813, | |
| "rewards/rejected": -1.257320523262024, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9859343603482921, | |
| "grad_norm": 38.11034342674612, | |
| "learning_rate": 5.058723737811355e-10, | |
| "logits/chosen": -2.829425096511841, | |
| "logits/rejected": -2.828883647918701, | |
| "logps/chosen": -1.0095789432525635, | |
| "logps/rejected": -1.2887779474258423, | |
| "loss": 1.8606, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0095789432525635, | |
| "rewards/margins": 0.27919891476631165, | |
| "rewards/rejected": -1.2887779474258423, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9966510381781648, | |
| "grad_norm": 31.577486151371264, | |
| "learning_rate": 1.405431468848306e-11, | |
| "logits/chosen": -2.846008777618408, | |
| "logits/rejected": -2.832801580429077, | |
| "logps/chosen": -1.0791490077972412, | |
| "logps/rejected": -1.249805212020874, | |
| "loss": 1.8779, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0791490077972412, | |
| "rewards/margins": 0.170656219124794, | |
| "rewards/rejected": -1.249805212020874, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9987943737441393, | |
| "step": 466, | |
| "total_flos": 0.0, | |
| "train_loss": 1.899593046294773, | |
| "train_runtime": 19390.7723, | |
| "train_samples_per_second": 3.08, | |
| "train_steps_per_second": 0.024 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 466, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |