| { | |
| "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": 37.4633897932904, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "logits/chosen": -2.397902250289917, | |
| "logits/rejected": -2.391808271408081, | |
| "logps/chosen": -0.5661157369613647, | |
| "logps/rejected": -0.5547963380813599, | |
| "loss": 1.5468, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -0.5661157369613647, | |
| "rewards/margins": -0.011319375596940517, | |
| "rewards/rejected": -0.5547963380813599, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021433355659745478, | |
| "grad_norm": 18.848503699490514, | |
| "learning_rate": 4.255319148936171e-06, | |
| "logits/chosen": -2.374025344848633, | |
| "logits/rejected": -2.345165729522705, | |
| "logps/chosen": -0.5381986498832703, | |
| "logps/rejected": -0.5637714266777039, | |
| "loss": 1.5538, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.5381986498832703, | |
| "rewards/margins": 0.025572726503014565, | |
| "rewards/rejected": -0.5637714266777039, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.032150033489618215, | |
| "grad_norm": 18.865749434279863, | |
| "learning_rate": 6.382978723404256e-06, | |
| "logits/chosen": -2.228337526321411, | |
| "logits/rejected": -2.2436676025390625, | |
| "logps/chosen": -0.6078538298606873, | |
| "logps/rejected": -0.6489665508270264, | |
| "loss": 1.5486, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.6078538298606873, | |
| "rewards/margins": 0.04111265391111374, | |
| "rewards/rejected": -0.6489665508270264, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042866711319490956, | |
| "grad_norm": 11.695416903513719, | |
| "learning_rate": 8.510638297872341e-06, | |
| "logits/chosen": -2.3822503089904785, | |
| "logits/rejected": -2.399451732635498, | |
| "logps/chosen": -0.6396132111549377, | |
| "logps/rejected": -0.6679492592811584, | |
| "loss": 1.5283, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.6396132111549377, | |
| "rewards/margins": 0.028336137533187866, | |
| "rewards/rejected": -0.6679492592811584, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0535833891493637, | |
| "grad_norm": 8.657694670203071, | |
| "learning_rate": 1.0638297872340426e-05, | |
| "logits/chosen": -2.6283469200134277, | |
| "logits/rejected": -2.6289353370666504, | |
| "logps/chosen": -0.6763823628425598, | |
| "logps/rejected": -0.7394057512283325, | |
| "loss": 1.5522, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.6763823628425598, | |
| "rewards/margins": 0.0630234032869339, | |
| "rewards/rejected": -0.7394057512283325, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06430006697923643, | |
| "grad_norm": 11.701639197469074, | |
| "learning_rate": 1.2765957446808513e-05, | |
| "logits/chosen": -2.7873454093933105, | |
| "logits/rejected": -2.815898895263672, | |
| "logps/chosen": -0.8007046580314636, | |
| "logps/rejected": -0.8376566767692566, | |
| "loss": 1.5826, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.8007046580314636, | |
| "rewards/margins": 0.03695206344127655, | |
| "rewards/rejected": -0.8376566767692566, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07501674480910918, | |
| "grad_norm": 85.64689375367887, | |
| "learning_rate": 1.4893617021276596e-05, | |
| "logits/chosen": -2.5493829250335693, | |
| "logits/rejected": -2.5297470092773438, | |
| "logps/chosen": -0.9185592532157898, | |
| "logps/rejected": -1.0066051483154297, | |
| "loss": 1.6494, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.9185592532157898, | |
| "rewards/margins": 0.08804582059383392, | |
| "rewards/rejected": -1.0066051483154297, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08573342263898191, | |
| "grad_norm": 12.618515943070657, | |
| "learning_rate": 1.7021276595744682e-05, | |
| "logits/chosen": -2.7925362586975098, | |
| "logits/rejected": -2.7926783561706543, | |
| "logps/chosen": -0.95704585313797, | |
| "logps/rejected": -0.9239652752876282, | |
| "loss": 1.6286, | |
| "rewards/accuracies": 0.4437499940395355, | |
| "rewards/chosen": -0.95704585313797, | |
| "rewards/margins": -0.033080630004405975, | |
| "rewards/rejected": -0.9239652752876282, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09645010046885466, | |
| "grad_norm": 7.437533737509709, | |
| "learning_rate": 1.914893617021277e-05, | |
| "logits/chosen": -2.668349027633667, | |
| "logits/rejected": -2.6576356887817383, | |
| "logps/chosen": -0.9402163624763489, | |
| "logps/rejected": -0.9749400019645691, | |
| "loss": 1.5995, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.9402163624763489, | |
| "rewards/margins": 0.034723617136478424, | |
| "rewards/rejected": -0.9749400019645691, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1071667782987274, | |
| "grad_norm": 8.099072281553982, | |
| "learning_rate": 1.999747031816659e-05, | |
| "logits/chosen": -2.796558380126953, | |
| "logits/rejected": -2.7897918224334717, | |
| "logps/chosen": -0.8397698402404785, | |
| "logps/rejected": -0.8385842442512512, | |
| "loss": 1.566, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.8397698402404785, | |
| "rewards/margins": -0.0011855497723445296, | |
| "rewards/rejected": -0.8385842442512512, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11788345612860013, | |
| "grad_norm": 13.021189847208735, | |
| "learning_rate": 1.9982015786040484e-05, | |
| "logits/chosen": -2.577963352203369, | |
| "logits/rejected": -2.5813965797424316, | |
| "logps/chosen": -0.9185325503349304, | |
| "logps/rejected": -0.9634258151054382, | |
| "loss": 1.587, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9185325503349304, | |
| "rewards/margins": 0.044893164187669754, | |
| "rewards/rejected": -0.9634258151054382, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12860013395847286, | |
| "grad_norm": 29.275115325498533, | |
| "learning_rate": 1.9952533792092286e-05, | |
| "logits/chosen": -2.7369465827941895, | |
| "logits/rejected": -2.746910810470581, | |
| "logps/chosen": -0.8204312324523926, | |
| "logps/rejected": -0.871277928352356, | |
| "loss": 1.6101, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.8204312324523926, | |
| "rewards/margins": 0.05084659531712532, | |
| "rewards/rejected": -0.871277928352356, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13931681178834562, | |
| "grad_norm": 9.013295050879966, | |
| "learning_rate": 1.9909065766585522e-05, | |
| "logits/chosen": -2.476846694946289, | |
| "logits/rejected": -2.4859910011291504, | |
| "logps/chosen": -0.8948932886123657, | |
| "logps/rejected": -0.9379276037216187, | |
| "loss": 1.6324, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.8948932886123657, | |
| "rewards/margins": 0.043034326285123825, | |
| "rewards/rejected": -0.9379276037216187, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15003348961821836, | |
| "grad_norm": 8.308683516584036, | |
| "learning_rate": 1.9851672793982618e-05, | |
| "logits/chosen": -2.443450450897217, | |
| "logits/rejected": -2.465125799179077, | |
| "logps/chosen": -0.957764744758606, | |
| "logps/rejected": -0.9942213296890259, | |
| "loss": 1.6029, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.957764744758606, | |
| "rewards/margins": 0.03645653277635574, | |
| "rewards/rejected": -0.9942213296890259, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1607501674480911, | |
| "grad_norm": 7.915844945947729, | |
| "learning_rate": 1.978043552710454e-05, | |
| "logits/chosen": -2.4849705696105957, | |
| "logits/rejected": -2.4826462268829346, | |
| "logps/chosen": -0.9050397872924805, | |
| "logps/rejected": -0.9400200843811035, | |
| "loss": 1.6052, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.9050397872924805, | |
| "rewards/margins": 0.03498036414384842, | |
| "rewards/rejected": -0.9400200843811035, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17146684527796383, | |
| "grad_norm": 304.0629881690128, | |
| "learning_rate": 1.9695454073791517e-05, | |
| "logits/chosen": -2.500096559524536, | |
| "logits/rejected": -2.5168652534484863, | |
| "logps/chosen": -0.906775951385498, | |
| "logps/rejected": -1.02707040309906, | |
| "loss": 1.82, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.906775951385498, | |
| "rewards/margins": 0.12029434740543365, | |
| "rewards/rejected": -1.02707040309906, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18218352310783656, | |
| "grad_norm": 10.960148765645817, | |
| "learning_rate": 1.9596847856224143e-05, | |
| "logits/chosen": -2.257906436920166, | |
| "logits/rejected": -2.265631675720215, | |
| "logps/chosen": -1.1177845001220703, | |
| "logps/rejected": -1.1891874074935913, | |
| "loss": 1.6996, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.1177845001220703, | |
| "rewards/margins": 0.07140282541513443, | |
| "rewards/rejected": -1.1891874074935913, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19290020093770932, | |
| "grad_norm": 10.383486663685009, | |
| "learning_rate": 1.9484755443102573e-05, | |
| "logits/chosen": -2.215013027191162, | |
| "logits/rejected": -2.2123208045959473, | |
| "logps/chosen": -1.035982370376587, | |
| "logps/rejected": -1.0804064273834229, | |
| "loss": 1.6576, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -1.035982370376587, | |
| "rewards/margins": 0.044424109160900116, | |
| "rewards/rejected": -1.0804064273834229, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20361687876758205, | |
| "grad_norm": 9.444434646488055, | |
| "learning_rate": 1.9359334354919587e-05, | |
| "logits/chosen": -2.5854439735412598, | |
| "logits/rejected": -2.604400157928467, | |
| "logps/chosen": -1.0645908117294312, | |
| "logps/rejected": -1.1030633449554443, | |
| "loss": 1.6255, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.0645908117294312, | |
| "rewards/margins": 0.03847253695130348, | |
| "rewards/rejected": -1.1030633449554443, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2143335565974548, | |
| "grad_norm": 10.008429713827578, | |
| "learning_rate": 1.9220760842601245e-05, | |
| "logits/chosen": -2.41093373298645, | |
| "logits/rejected": -2.416090488433838, | |
| "logps/chosen": -1.0662959814071655, | |
| "logps/rejected": -1.0942766666412354, | |
| "loss": 1.6366, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -1.0662959814071655, | |
| "rewards/margins": 0.027980733662843704, | |
| "rewards/rejected": -1.0942766666412354, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22505023442732752, | |
| "grad_norm": 13.793372325395232, | |
| "learning_rate": 1.9069229639826113e-05, | |
| "logits/chosen": -2.593897581100464, | |
| "logits/rejected": -2.592832326889038, | |
| "logps/chosen": -1.0412575006484985, | |
| "logps/rejected": -1.1209934949874878, | |
| "loss": 1.6663, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.0412575006484985, | |
| "rewards/margins": 0.0797361508011818, | |
| "rewards/rejected": -1.1209934949874878, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23576691225720026, | |
| "grad_norm": 25.707371820916816, | |
| "learning_rate": 1.8904953689371223e-05, | |
| "logits/chosen": -2.1585533618927, | |
| "logits/rejected": -2.143022060394287, | |
| "logps/chosen": -1.0903128385543823, | |
| "logps/rejected": -1.1769222021102905, | |
| "loss": 1.6708, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0903128385543823, | |
| "rewards/margins": 0.08660950511693954, | |
| "rewards/rejected": -1.1769222021102905, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24648359008707302, | |
| "grad_norm": 15.08396540818952, | |
| "learning_rate": 1.8728163843869215e-05, | |
| "logits/chosen": -2.3684897422790527, | |
| "logits/rejected": -2.3717474937438965, | |
| "logps/chosen": -1.1836369037628174, | |
| "logps/rejected": -1.1988976001739502, | |
| "loss": 1.8024, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -1.1836369037628174, | |
| "rewards/margins": 0.015260684303939342, | |
| "rewards/rejected": -1.1988976001739502, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2572002679169457, | |
| "grad_norm": 10.565734009587494, | |
| "learning_rate": 1.853910854139727e-05, | |
| "logits/chosen": -2.122260570526123, | |
| "logits/rejected": -2.138091802597046, | |
| "logps/chosen": -1.133177399635315, | |
| "logps/rejected": -1.1549627780914307, | |
| "loss": 1.7056, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -1.133177399635315, | |
| "rewards/margins": 0.02178535796701908, | |
| "rewards/rejected": -1.1549627780914307, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2679169457468185, | |
| "grad_norm": 7.929268982931603, | |
| "learning_rate": 1.833805345635369e-05, | |
| "logits/chosen": -2.467066526412964, | |
| "logits/rejected": -2.470454692840576, | |
| "logps/chosen": -1.0294818878173828, | |
| "logps/rejected": -1.0639444589614868, | |
| "loss": 1.6829, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.0294818878173828, | |
| "rewards/margins": 0.03446248173713684, | |
| "rewards/rejected": -1.0639444589614868, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27863362357669125, | |
| "grad_norm": 9.40947063841382, | |
| "learning_rate": 1.8125281126112678e-05, | |
| "logits/chosen": -2.5180554389953613, | |
| "logits/rejected": -2.5261032581329346, | |
| "logps/chosen": -1.044611930847168, | |
| "logps/rejected": -1.1444398164749146, | |
| "loss": 1.6283, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.044611930847168, | |
| "rewards/margins": 0.09982797503471375, | |
| "rewards/rejected": -1.1444398164749146, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.289350301406564, | |
| "grad_norm": 7.797174649236775, | |
| "learning_rate": 1.790109055398212e-05, | |
| "logits/chosen": -2.5548009872436523, | |
| "logits/rejected": -2.5629782676696777, | |
| "logps/chosen": -0.9959312677383423, | |
| "logps/rejected": -1.0395903587341309, | |
| "loss": 1.6327, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.9959312677383423, | |
| "rewards/margins": 0.04365905374288559, | |
| "rewards/rejected": -1.0395903587341309, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3000669792364367, | |
| "grad_norm": 8.65910654888172, | |
| "learning_rate": 1.766579678902212e-05, | |
| "logits/chosen": -2.452951431274414, | |
| "logits/rejected": -2.459916114807129, | |
| "logps/chosen": -0.9893513917922974, | |
| "logps/rejected": -1.0323150157928467, | |
| "loss": 1.6058, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9893513917922974, | |
| "rewards/margins": 0.042963750660419464, | |
| "rewards/rejected": -1.0323150157928467, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31078365706630945, | |
| "grad_norm": 8.475342441948424, | |
| "learning_rate": 1.741973048331496e-05, | |
| "logits/chosen": -2.608076810836792, | |
| "logits/rejected": -2.592602491378784, | |
| "logps/chosen": -0.959572434425354, | |
| "logps/rejected": -1.034380316734314, | |
| "loss": 1.5962, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.959572434425354, | |
| "rewards/margins": 0.07480800151824951, | |
| "rewards/rejected": -1.034380316734314, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3215003348961822, | |
| "grad_norm": 13.2413080829705, | |
| "learning_rate": 1.7163237427308477e-05, | |
| "logits/chosen": -2.6177635192871094, | |
| "logits/rejected": -2.6168532371520996, | |
| "logps/chosen": -1.004525899887085, | |
| "logps/rejected": -1.0910793542861938, | |
| "loss": 1.6239, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.004525899887085, | |
| "rewards/margins": 0.08655332773923874, | |
| "rewards/rejected": -1.0910793542861938, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3322170127260549, | |
| "grad_norm": 8.151298227822702, | |
| "learning_rate": 1.689667806388594e-05, | |
| "logits/chosen": -2.7449164390563965, | |
| "logits/rejected": -2.7318592071533203, | |
| "logps/chosen": -0.9720630645751953, | |
| "logps/rejected": -1.047321081161499, | |
| "loss": 1.5985, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.9720630645751953, | |
| "rewards/margins": 0.0752580463886261, | |
| "rewards/rejected": -1.047321081161499, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.34293369055592765, | |
| "grad_norm": 7.660153117545113, | |
| "learning_rate": 1.6620426981845233e-05, | |
| "logits/chosen": -2.5920422077178955, | |
| "logits/rejected": -2.5721192359924316, | |
| "logps/chosen": -0.9702916145324707, | |
| "logps/rejected": -1.0599411725997925, | |
| "loss": 1.6107, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.9702916145324707, | |
| "rewards/margins": 0.08964961767196655, | |
| "rewards/rejected": -1.0599411725997925, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3536503683858004, | |
| "grad_norm": 469.42247386416926, | |
| "learning_rate": 1.633487238949915e-05, | |
| "logits/chosen": -2.528369426727295, | |
| "logits/rejected": -2.5209622383117676, | |
| "logps/chosen": -0.9299852252006531, | |
| "logps/rejected": -0.9966603517532349, | |
| "loss": 1.6799, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.9299852252006531, | |
| "rewards/margins": 0.06667519360780716, | |
| "rewards/rejected": -0.9966603517532349, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3643670462156731, | |
| "grad_norm": 13.123351141683822, | |
| "learning_rate": 1.6040415569136586e-05, | |
| "logits/chosen": -2.3123598098754883, | |
| "logits/rejected": -2.3209071159362793, | |
| "logps/chosen": -1.2778890132904053, | |
| "logps/rejected": -1.3249436616897583, | |
| "loss": 1.7418, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.2778890132904053, | |
| "rewards/margins": 0.04705450311303139, | |
| "rewards/rejected": -1.3249436616897583, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3750837240455459, | |
| "grad_norm": 7.9321661768835865, | |
| "learning_rate": 1.5737470313111134e-05, | |
| "logits/chosen": -2.1733322143554688, | |
| "logits/rejected": -2.188865900039673, | |
| "logps/chosen": -1.1108664274215698, | |
| "logps/rejected": -1.1735206842422485, | |
| "loss": 1.6869, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.1108664274215698, | |
| "rewards/margins": 0.06265412271022797, | |
| "rewards/rejected": -1.1735206842422485, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.38580040187541864, | |
| "grad_norm": 8.557996815044758, | |
| "learning_rate": 1.5426462342349736e-05, | |
| "logits/chosen": -2.3044323921203613, | |
| "logits/rejected": -2.3116490840911865, | |
| "logps/chosen": -1.1034280061721802, | |
| "logps/rejected": -1.1603246927261353, | |
| "loss": 1.6567, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -1.1034280061721802, | |
| "rewards/margins": 0.05689678341150284, | |
| "rewards/rejected": -1.1603246927261353, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3965170797052914, | |
| "grad_norm": 10.818770194479542, | |
| "learning_rate": 1.5107828708098327e-05, | |
| "logits/chosen": -2.3942196369171143, | |
| "logits/rejected": -2.4129745960235596, | |
| "logps/chosen": -1.0460340976715088, | |
| "logps/rejected": -1.105322241783142, | |
| "loss": 1.6177, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0460340976715088, | |
| "rewards/margins": 0.059288132935762405, | |
| "rewards/rejected": -1.105322241783142, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4072337575351641, | |
| "grad_norm": 15.512656922129596, | |
| "learning_rate": 1.4782017177745322e-05, | |
| "logits/chosen": -2.4357337951660156, | |
| "logits/rejected": -2.458134889602661, | |
| "logps/chosen": -1.0308730602264404, | |
| "logps/rejected": -1.2367624044418335, | |
| "loss": 1.6917, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0308730602264404, | |
| "rewards/margins": 0.20588929951190948, | |
| "rewards/rejected": -1.2367624044418335, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.41795043536503684, | |
| "grad_norm": 10.838461128892558, | |
| "learning_rate": 1.444948560558601e-05, | |
| "logits/chosen": -2.4340341091156006, | |
| "logits/rejected": -2.4456231594085693, | |
| "logps/chosen": -1.2891473770141602, | |
| "logps/rejected": -1.3041666746139526, | |
| "loss": 1.7583, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -1.2891473770141602, | |
| "rewards/margins": 0.015019327402114868, | |
| "rewards/rejected": -1.3041666746139526, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4286671131949096, | |
| "grad_norm": 8.28412245267914, | |
| "learning_rate": 1.4110701289412045e-05, | |
| "logits/chosen": -2.515929937362671, | |
| "logits/rejected": -2.5307044982910156, | |
| "logps/chosen": -1.1040905714035034, | |
| "logps/rejected": -1.1781259775161743, | |
| "loss": 1.6734, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.1040905714035034, | |
| "rewards/margins": 0.07403533160686493, | |
| "rewards/rejected": -1.1781259775161743, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4393837910247823, | |
| "grad_norm": 9.96354155003577, | |
| "learning_rate": 1.3766140313830279e-05, | |
| "logits/chosen": -2.531589984893799, | |
| "logits/rejected": -2.520904302597046, | |
| "logps/chosen": -1.0145483016967773, | |
| "logps/rejected": -1.0664831399917603, | |
| "loss": 1.6419, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.0145483016967773, | |
| "rewards/margins": 0.05193489044904709, | |
| "rewards/rejected": -1.0664831399917603, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.45010046885465504, | |
| "grad_norm": 7.434219003968733, | |
| "learning_rate": 1.341628688123369e-05, | |
| "logits/chosen": -2.517782211303711, | |
| "logits/rejected": -2.5262608528137207, | |
| "logps/chosen": -1.0025173425674438, | |
| "logps/rejected": -1.0794353485107422, | |
| "loss": 1.606, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.0025173425674438, | |
| "rewards/margins": 0.07691796869039536, | |
| "rewards/rejected": -1.0794353485107422, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4608171466845278, | |
| "grad_norm": 7.337174471827815, | |
| "learning_rate": 1.3061632631364613e-05, | |
| "logits/chosen": -2.5459511280059814, | |
| "logits/rejected": -2.546247959136963, | |
| "logps/chosen": -0.9931095242500305, | |
| "logps/rejected": -1.094226598739624, | |
| "loss": 1.594, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.9931095242500305, | |
| "rewards/margins": 0.10111699998378754, | |
| "rewards/rejected": -1.094226598739624, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4715338245144005, | |
| "grad_norm": 8.701123471362035, | |
| "learning_rate": 1.2702675950426477e-05, | |
| "logits/chosen": -2.7032644748687744, | |
| "logits/rejected": -2.685612440109253, | |
| "logps/chosen": -0.9561999440193176, | |
| "logps/rejected": -1.0159614086151123, | |
| "loss": 1.63, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.9561999440193176, | |
| "rewards/margins": 0.05976147577166557, | |
| "rewards/rejected": -1.0159614086151123, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4822505023442733, | |
| "grad_norm": 10.855107466518529, | |
| "learning_rate": 1.2339921270714873e-05, | |
| "logits/chosen": -2.7168991565704346, | |
| "logits/rejected": -2.742966651916504, | |
| "logps/chosen": -0.9448912739753723, | |
| "logps/rejected": -1.0768978595733643, | |
| "loss": 1.5893, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9448912739753723, | |
| "rewards/margins": 0.1320066601037979, | |
| "rewards/rejected": -1.0768978595733643, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.49296718017414604, | |
| "grad_norm": 7.921384206654596, | |
| "learning_rate": 1.1973878361752263e-05, | |
| "logits/chosen": -2.7181243896484375, | |
| "logits/rejected": -2.7184972763061523, | |
| "logps/chosen": -0.9301810264587402, | |
| "logps/rejected": -1.000950813293457, | |
| "loss": 1.5781, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9301810264587402, | |
| "rewards/margins": 0.07076982408761978, | |
| "rewards/rejected": -1.000950813293457, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5036838580040187, | |
| "grad_norm": 8.000512303019988, | |
| "learning_rate": 1.160506161392239e-05, | |
| "logits/chosen": -2.7740559577941895, | |
| "logits/rejected": -2.767812967300415, | |
| "logps/chosen": -0.906406044960022, | |
| "logps/rejected": -1.006753921508789, | |
| "loss": 1.5802, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.906406044960022, | |
| "rewards/margins": 0.10034787654876709, | |
| "rewards/rejected": -1.006753921508789, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5144005358338914, | |
| "grad_norm": 9.398458173467265, | |
| "learning_rate": 1.1233989315611132e-05, | |
| "logits/chosen": -2.792450428009033, | |
| "logits/rejected": -2.7874622344970703, | |
| "logps/chosen": -0.9394813776016235, | |
| "logps/rejected": -1.022139549255371, | |
| "loss": 1.587, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.9394813776016235, | |
| "rewards/margins": 0.08265805244445801, | |
| "rewards/rejected": -1.022139549255371, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5251172136637642, | |
| "grad_norm": 10.738263842024553, | |
| "learning_rate": 1.0861182924869584e-05, | |
| "logits/chosen": -2.721609115600586, | |
| "logits/rejected": -2.712308406829834, | |
| "logps/chosen": -0.9511491656303406, | |
| "logps/rejected": -1.0187407732009888, | |
| "loss": 1.5773, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9511491656303406, | |
| "rewards/margins": 0.06759153306484222, | |
| "rewards/rejected": -1.0187407732009888, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.535833891493637, | |
| "grad_norm": 8.609658733354994, | |
| "learning_rate": 1.0487166336622903e-05, | |
| "logits/chosen": -2.714517116546631, | |
| "logits/rejected": -2.7152135372161865, | |
| "logps/chosen": -0.9705566167831421, | |
| "logps/rejected": -1.0084294080734253, | |
| "loss": 1.5971, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.9705566167831421, | |
| "rewards/margins": 0.03787278011441231, | |
| "rewards/rejected": -1.0084294080734253, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5465505693235098, | |
| "grad_norm": 8.407135051661365, | |
| "learning_rate": 1.0112465146454681e-05, | |
| "logits/chosen": -2.790414333343506, | |
| "logits/rejected": -2.799837112426758, | |
| "logps/chosen": -0.8985916376113892, | |
| "logps/rejected": -0.9925104379653931, | |
| "loss": 1.5696, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.8985916376113892, | |
| "rewards/margins": 0.09391861408948898, | |
| "rewards/rejected": -0.9925104379653931, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5572672471533825, | |
| "grad_norm": 8.166210902327252, | |
| "learning_rate": 9.737605912001413e-06, | |
| "logits/chosen": -2.7387256622314453, | |
| "logits/rejected": -2.7400269508361816, | |
| "logps/chosen": -0.9064778089523315, | |
| "logps/rejected": -1.0061310529708862, | |
| "loss": 1.5824, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.9064778089523315, | |
| "rewards/margins": 0.09965313225984573, | |
| "rewards/rejected": -1.0061310529708862, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5679839249832552, | |
| "grad_norm": 7.963553030396463, | |
| "learning_rate": 9.363115412995038e-06, | |
| "logits/chosen": -2.733085870742798, | |
| "logits/rejected": -2.7209997177124023, | |
| "logps/chosen": -0.9366081357002258, | |
| "logps/rejected": -1.074143648147583, | |
| "loss": 1.5601, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.9366081357002258, | |
| "rewards/margins": 0.13753557205200195, | |
| "rewards/rejected": -1.074143648147583, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.578700602813128, | |
| "grad_norm": 8.394463914962087, | |
| "learning_rate": 8.989519910993357e-06, | |
| "logits/chosen": -2.741532802581787, | |
| "logits/rejected": -2.7389678955078125, | |
| "logps/chosen": -0.8945878744125366, | |
| "logps/rejected": -0.9697369337081909, | |
| "loss": 1.5652, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.8945878744125366, | |
| "rewards/margins": 0.07514907419681549, | |
| "rewards/rejected": -0.9697369337081909, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5894172806430007, | |
| "grad_norm": 7.2839753266908875, | |
| "learning_rate": 8.617344409838631e-06, | |
| "logits/chosen": -2.837791919708252, | |
| "logits/rejected": -2.8426520824432373, | |
| "logps/chosen": -0.8537940979003906, | |
| "logps/rejected": -0.9409109354019165, | |
| "loss": 1.5685, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.8537940979003906, | |
| "rewards/margins": 0.08711685240268707, | |
| "rewards/rejected": -0.9409109354019165, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6001339584728734, | |
| "grad_norm": 7.994187957031587, | |
| "learning_rate": 8.247111917883634e-06, | |
| "logits/chosen": -2.6431148052215576, | |
| "logits/rejected": -2.644808530807495, | |
| "logps/chosen": -0.8992069959640503, | |
| "logps/rejected": -1.0336244106292725, | |
| "loss": 1.5658, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.8992069959640503, | |
| "rewards/margins": 0.1344173550605774, | |
| "rewards/rejected": -1.0336244106292725, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6108506363027462, | |
| "grad_norm": 10.3254576486645, | |
| "learning_rate": 7.879342713021875e-06, | |
| "logits/chosen": -2.745375156402588, | |
| "logits/rejected": -2.744263172149658, | |
| "logps/chosen": -0.9114044308662415, | |
| "logps/rejected": -0.9392461776733398, | |
| "loss": 1.5753, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9114044308662415, | |
| "rewards/margins": 0.027841681614518166, | |
| "rewards/rejected": -0.9392461776733398, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6215673141326189, | |
| "grad_norm": 14.256488881900221, | |
| "learning_rate": 7.514553611554909e-06, | |
| "logits/chosen": -2.9054572582244873, | |
| "logits/rejected": -2.915325880050659, | |
| "logps/chosen": -0.8290088772773743, | |
| "logps/rejected": -0.9459854364395142, | |
| "loss": 1.5614, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.8290088772773743, | |
| "rewards/margins": 0.11697645485401154, | |
| "rewards/rejected": -0.9459854364395142, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6322839919624916, | |
| "grad_norm": 8.049072856223585, | |
| "learning_rate": 7.15325724192409e-06, | |
| "logits/chosen": -2.861290454864502, | |
| "logits/rejected": -2.8629579544067383, | |
| "logps/chosen": -0.8957312703132629, | |
| "logps/rejected": -1.0799039602279663, | |
| "loss": 1.5436, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.8957312703132629, | |
| "rewards/margins": 0.1841726452112198, | |
| "rewards/rejected": -1.0799039602279663, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6430006697923644, | |
| "grad_norm": 8.353215812366958, | |
| "learning_rate": 6.79596132432742e-06, | |
| "logits/chosen": -2.8521814346313477, | |
| "logits/rejected": -2.8686671257019043, | |
| "logps/chosen": -0.8559802770614624, | |
| "logps/rejected": -0.9652966260910034, | |
| "loss": 1.5273, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.8559802770614624, | |
| "rewards/margins": 0.10931621491909027, | |
| "rewards/rejected": -0.9652966260910034, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6537173476222371, | |
| "grad_norm": 8.558746762469774, | |
| "learning_rate": 6.443167957233864e-06, | |
| "logits/chosen": -2.8060519695281982, | |
| "logits/rejected": -2.793569326400757, | |
| "logps/chosen": -0.8676208257675171, | |
| "logps/rejected": -0.9558470845222473, | |
| "loss": 1.5333, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.8676208257675171, | |
| "rewards/margins": 0.08822615444660187, | |
| "rewards/rejected": -0.9558470845222473, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6644340254521098, | |
| "grad_norm": 8.944155497280653, | |
| "learning_rate": 6.095372911797673e-06, | |
| "logits/chosen": -2.8632490634918213, | |
| "logits/rejected": -2.859429359436035, | |
| "logps/chosen": -0.8489475250244141, | |
| "logps/rejected": -0.9311763644218445, | |
| "loss": 1.5397, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.8489475250244141, | |
| "rewards/margins": 0.08222876489162445, | |
| "rewards/rejected": -0.9311763644218445, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6751507032819826, | |
| "grad_norm": 7.822389173887281, | |
| "learning_rate": 5.753064935164348e-06, | |
| "logits/chosen": -2.7866902351379395, | |
| "logits/rejected": -2.792357921600342, | |
| "logps/chosen": -0.84296715259552, | |
| "logps/rejected": -0.959184467792511, | |
| "loss": 1.5381, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.84296715259552, | |
| "rewards/margins": 0.11621735244989395, | |
| "rewards/rejected": -0.959184467792511, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6858673811118553, | |
| "grad_norm": 9.32294397022148, | |
| "learning_rate": 5.4167250636472426e-06, | |
| "logits/chosen": -2.7923617362976074, | |
| "logits/rejected": -2.7717807292938232, | |
| "logps/chosen": -0.8463284373283386, | |
| "logps/rejected": -0.9291221499443054, | |
| "loss": 1.5458, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.8463284373283386, | |
| "rewards/margins": 0.08279372751712799, | |
| "rewards/rejected": -0.9291221499443054, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.696584058941728, | |
| "grad_norm": 8.808118982464391, | |
| "learning_rate": 5.08682594674002e-06, | |
| "logits/chosen": -2.7373509407043457, | |
| "logits/rejected": -2.7527077198028564, | |
| "logps/chosen": -0.8056324124336243, | |
| "logps/rejected": -0.957785964012146, | |
| "loss": 1.5196, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.8056324124336243, | |
| "rewards/margins": 0.15215365588665009, | |
| "rewards/rejected": -0.957785964012146, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7073007367716008, | |
| "grad_norm": 8.169267761293227, | |
| "learning_rate": 4.763831182914848e-06, | |
| "logits/chosen": -2.7620842456817627, | |
| "logits/rejected": -2.77044415473938, | |
| "logps/chosen": -0.8720673322677612, | |
| "logps/rejected": -0.9474166631698608, | |
| "loss": 1.5373, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.8720673322677612, | |
| "rewards/margins": 0.07534926384687424, | |
| "rewards/rejected": -0.9474166631698608, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7180174146014735, | |
| "grad_norm": 8.7345666789259, | |
| "learning_rate": 4.448194668139777e-06, | |
| "logits/chosen": -2.794130325317383, | |
| "logits/rejected": -2.7945706844329834, | |
| "logps/chosen": -0.8377755284309387, | |
| "logps/rejected": -0.9672821760177612, | |
| "loss": 1.5233, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.8377755284309387, | |
| "rewards/margins": 0.12950672209262848, | |
| "rewards/rejected": -0.9672821760177612, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7287340924313462, | |
| "grad_norm": 8.389766148627231, | |
| "learning_rate": 4.1403599580307795e-06, | |
| "logits/chosen": -2.714247226715088, | |
| "logits/rejected": -2.7236764430999756, | |
| "logps/chosen": -0.8354495167732239, | |
| "logps/rejected": -0.9634197354316711, | |
| "loss": 1.5335, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.8354495167732239, | |
| "rewards/margins": 0.1279701441526413, | |
| "rewards/rejected": -0.9634197354316711, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.739450770261219, | |
| "grad_norm": 9.147040621292957, | |
| "learning_rate": 3.840759644534828e-06, | |
| "logits/chosen": -2.790487766265869, | |
| "logits/rejected": -2.8023035526275635, | |
| "logps/chosen": -0.8659473657608032, | |
| "logps/rejected": -0.9856260418891907, | |
| "loss": 1.5348, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.8659473657608032, | |
| "rewards/margins": 0.11967878043651581, | |
| "rewards/rejected": -0.9856260418891907, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7501674480910918, | |
| "grad_norm": 8.599069402922064, | |
| "learning_rate": 3.549814748019906e-06, | |
| "logits/chosen": -2.815507411956787, | |
| "logits/rejected": -2.824169158935547, | |
| "logps/chosen": -0.808633029460907, | |
| "logps/rejected": -0.9542997479438782, | |
| "loss": 1.529, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.808633029460907, | |
| "rewards/margins": 0.14566662907600403, | |
| "rewards/rejected": -0.9542997479438782, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7608841259209645, | |
| "grad_norm": 7.982319756803084, | |
| "learning_rate": 3.2679341256262653e-06, | |
| "logits/chosen": -2.8472940921783447, | |
| "logits/rejected": -2.8238396644592285, | |
| "logps/chosen": -0.9044294357299805, | |
| "logps/rejected": -0.9391825795173645, | |
| "loss": 1.5418, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.9044294357299805, | |
| "rewards/margins": 0.03475295007228851, | |
| "rewards/rejected": -0.9391825795173645, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7716008037508373, | |
| "grad_norm": 11.32319755235924, | |
| "learning_rate": 2.995513896710327e-06, | |
| "logits/chosen": -2.8410215377807617, | |
| "logits/rejected": -2.83413028717041, | |
| "logps/chosen": -0.7960796356201172, | |
| "logps/rejected": -0.9083698391914368, | |
| "loss": 1.5306, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.7960796356201172, | |
| "rewards/margins": 0.1122901439666748, | |
| "rewards/rejected": -0.9083698391914368, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.78231748158071, | |
| "grad_norm": 7.8205965789155325, | |
| "learning_rate": 2.7329368861886862e-06, | |
| "logits/chosen": -2.8469526767730713, | |
| "logits/rejected": -2.8535444736480713, | |
| "logps/chosen": -0.8426831960678101, | |
| "logps/rejected": -0.9201619029045105, | |
| "loss": 1.5523, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.8426831960678101, | |
| "rewards/margins": 0.07747860997915268, | |
| "rewards/rejected": -0.9201619029045105, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7930341594105828, | |
| "grad_norm": 11.791063151570222, | |
| "learning_rate": 2.4805720865643943e-06, | |
| "logits/chosen": -2.8243472576141357, | |
| "logits/rejected": -2.8304176330566406, | |
| "logps/chosen": -0.8739852905273438, | |
| "logps/rejected": -0.9622358083724976, | |
| "loss": 1.5445, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.8739852905273438, | |
| "rewards/margins": 0.08825065940618515, | |
| "rewards/rejected": -0.9622358083724976, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8037508372404555, | |
| "grad_norm": 10.592008554970972, | |
| "learning_rate": 2.2387741393916117e-06, | |
| "logits/chosen": -2.7737154960632324, | |
| "logits/rejected": -2.7636284828186035, | |
| "logps/chosen": -0.9277487993240356, | |
| "logps/rejected": -1.0097343921661377, | |
| "loss": 1.5368, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9277487993240356, | |
| "rewards/margins": 0.08198549598455429, | |
| "rewards/rejected": -1.0097343921661377, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8144675150703282, | |
| "grad_norm": 8.911271573494423, | |
| "learning_rate": 2.0078828369072323e-06, | |
| "logits/chosen": -2.7896554470062256, | |
| "logits/rejected": -2.770059823989868, | |
| "logps/chosen": -0.8297151327133179, | |
| "logps/rejected": -0.9309379458427429, | |
| "loss": 1.531, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.8297151327133179, | |
| "rewards/margins": 0.10122283548116684, | |
| "rewards/rejected": -0.9309379458427429, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.825184192900201, | |
| "grad_norm": 10.364577364972547, | |
| "learning_rate": 1.7882226445298656e-06, | |
| "logits/chosen": -2.740999460220337, | |
| "logits/rejected": -2.733097791671753, | |
| "logps/chosen": -0.8539925813674927, | |
| "logps/rejected": -0.9543533325195312, | |
| "loss": 1.5174, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.8539925813674927, | |
| "rewards/margins": 0.10036077350378036, | |
| "rewards/rejected": -0.9543533325195312, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8359008707300737, | |
| "grad_norm": 10.862441952775951, | |
| "learning_rate": 1.580102244897217e-06, | |
| "logits/chosen": -2.7668397426605225, | |
| "logits/rejected": -2.7470200061798096, | |
| "logps/chosen": -0.8713840246200562, | |
| "logps/rejected": -0.940260112285614, | |
| "loss": 1.5632, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.8713840246200562, | |
| "rewards/margins": 0.0688760057091713, | |
| "rewards/rejected": -0.940260112285614, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8466175485599464, | |
| "grad_norm": 7.998296949983092, | |
| "learning_rate": 1.3838141040825536e-06, | |
| "logits/chosen": -2.85998272895813, | |
| "logits/rejected": -2.8610522747039795, | |
| "logps/chosen": -0.8187233805656433, | |
| "logps/rejected": -0.8954393267631531, | |
| "loss": 1.5253, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.8187233805656433, | |
| "rewards/margins": 0.0767158791422844, | |
| "rewards/rejected": -0.8954393267631531, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8573342263898192, | |
| "grad_norm": 10.009174586154066, | |
| "learning_rate": 1.1996340605999058e-06, | |
| "logits/chosen": -2.842839479446411, | |
| "logits/rejected": -2.819075107574463, | |
| "logps/chosen": -0.7899017333984375, | |
| "logps/rejected": -0.9768079519271851, | |
| "loss": 1.5218, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.7899017333984375, | |
| "rewards/margins": 0.18690629303455353, | |
| "rewards/rejected": -0.9768079519271851, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8573342263898192, | |
| "eval_logits/chosen": -2.840418815612793, | |
| "eval_logits/rejected": -2.8387703895568848, | |
| "eval_logps/chosen": -0.8695030808448792, | |
| "eval_logps/rejected": -0.9116575121879578, | |
| "eval_loss": 1.5846024751663208, | |
| "eval_rewards/accuracies": 0.5558510422706604, | |
| "eval_rewards/chosen": -0.8695030808448792, | |
| "eval_rewards/margins": 0.0421544648706913, | |
| "eval_rewards/rejected": -0.9116575121879578, | |
| "eval_runtime": 432.5597, | |
| "eval_samples_per_second": 6.922, | |
| "eval_steps_per_second": 0.435, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8680509042196919, | |
| "grad_norm": 8.778213042419416, | |
| "learning_rate": 1.0278209377755099e-06, | |
| "logits/chosen": -2.829361915588379, | |
| "logits/rejected": -2.8452582359313965, | |
| "logps/chosen": -0.8199126124382019, | |
| "logps/rejected": -0.9467391967773438, | |
| "loss": 1.5156, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.8199126124382019, | |
| "rewards/margins": 0.12682652473449707, | |
| "rewards/rejected": -0.9467391967773438, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8787675820495646, | |
| "grad_norm": 8.779487119016105, | |
| "learning_rate": 8.68616180030275e-07, | |
| "logits/chosen": -2.865104913711548, | |
| "logits/rejected": -2.867384672164917, | |
| "logps/chosen": -0.8341816067695618, | |
| "logps/rejected": -0.8998939394950867, | |
| "loss": 1.5114, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.8341816067695618, | |
| "rewards/margins": 0.06571228057146072, | |
| "rewards/rejected": -0.8998939394950867, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8894842598794374, | |
| "grad_norm": 8.92068989780619, | |
| "learning_rate": 7.22243513584342e-07, | |
| "logits/chosen": -2.77596378326416, | |
| "logits/rejected": -2.7669265270233154, | |
| "logps/chosen": -0.8306072354316711, | |
| "logps/rejected": -0.9651764035224915, | |
| "loss": 1.5034, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.8306072354316711, | |
| "rewards/margins": 0.13456913828849792, | |
| "rewards/rejected": -0.9651764035224915, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9002009377093101, | |
| "grad_norm": 9.219885917035732, | |
| "learning_rate": 5.889086320605575e-07, | |
| "logits/chosen": -2.7921142578125, | |
| "logits/rejected": -2.7921335697174072, | |
| "logps/chosen": -0.7846695184707642, | |
| "logps/rejected": -0.8718813061714172, | |
| "loss": 1.5351, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.7846695184707642, | |
| "rewards/margins": 0.08721177279949188, | |
| "rewards/rejected": -0.8718813061714172, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9109176155391828, | |
| "grad_norm": 9.448964270757202, | |
| "learning_rate": 4.6879890742869584e-07, | |
| "logits/chosen": -2.7913196086883545, | |
| "logits/rejected": -2.789302110671997, | |
| "logps/chosen": -0.7865716218948364, | |
| "logps/rejected": -0.9528288841247559, | |
| "loss": 1.5108, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.7865716218948364, | |
| "rewards/margins": 0.16625720262527466, | |
| "rewards/rejected": -0.9528288841247559, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9216342933690556, | |
| "grad_norm": 10.369844004542797, | |
| "learning_rate": 3.6208312669659405e-07, | |
| "logits/chosen": -2.817171812057495, | |
| "logits/rejected": -2.8241400718688965, | |
| "logps/chosen": -0.8919852375984192, | |
| "logps/rejected": -1.037209153175354, | |
| "loss": 1.5259, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.8919852375984192, | |
| "rewards/margins": 0.14522376656532288, | |
| "rewards/rejected": -1.037209153175354, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9323509711989283, | |
| "grad_norm": 12.055758939525129, | |
| "learning_rate": 2.689112547182593e-07, | |
| "logits/chosen": -2.890130043029785, | |
| "logits/rejected": -2.882509708404541, | |
| "logps/chosen": -0.7903082370758057, | |
| "logps/rejected": -0.9467129707336426, | |
| "loss": 1.5203, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.7903082370758057, | |
| "rewards/margins": 0.15640470385551453, | |
| "rewards/rejected": -0.9467129707336426, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.943067649028801, | |
| "grad_norm": 9.50477372459136, | |
| "learning_rate": 1.8941422345223448e-07, | |
| "logits/chosen": -2.7733447551727295, | |
| "logits/rejected": -2.77803635597229, | |
| "logps/chosen": -0.7930269241333008, | |
| "logps/rejected": -0.9639067649841309, | |
| "loss": 1.4896, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.7930269241333008, | |
| "rewards/margins": 0.17087996006011963, | |
| "rewards/rejected": -0.9639067649841309, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9537843268586738, | |
| "grad_norm": 10.00023586076567, | |
| "learning_rate": 1.2370374796639383e-07, | |
| "logits/chosen": -2.794691324234009, | |
| "logits/rejected": -2.799956798553467, | |
| "logps/chosen": -0.8402734994888306, | |
| "logps/rejected": -0.9617452621459961, | |
| "loss": 1.5232, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.8402734994888306, | |
| "rewards/margins": 0.12147176265716553, | |
| "rewards/rejected": -0.9617452621459961, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9645010046885466, | |
| "grad_norm": 9.634395288787493, | |
| "learning_rate": 7.18721694477209e-08, | |
| "logits/chosen": -2.8351221084594727, | |
| "logits/rejected": -2.833106756210327, | |
| "logps/chosen": -0.8035479784011841, | |
| "logps/rejected": -0.9219916462898254, | |
| "loss": 1.5027, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.8035479784011841, | |
| "rewards/margins": 0.11844360828399658, | |
| "rewards/rejected": -0.9219916462898254, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9752176825184193, | |
| "grad_norm": 9.629343705982949, | |
| "learning_rate": 3.399232543769504e-08, | |
| "logits/chosen": -2.807852268218994, | |
| "logits/rejected": -2.815563678741455, | |
| "logps/chosen": -0.8062378168106079, | |
| "logps/rejected": -0.9637410044670105, | |
| "loss": 1.5073, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.8062378168106079, | |
| "rewards/margins": 0.15750321745872498, | |
| "rewards/rejected": -0.9637410044670105, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9859343603482921, | |
| "grad_norm": 10.887896034397038, | |
| "learning_rate": 1.011744747562271e-08, | |
| "logits/chosen": -2.826111316680908, | |
| "logits/rejected": -2.830981731414795, | |
| "logps/chosen": -0.8098823428153992, | |
| "logps/rejected": -0.9950210452079773, | |
| "loss": 1.5206, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.8098823428153992, | |
| "rewards/margins": 0.1851385533809662, | |
| "rewards/rejected": -0.9950210452079773, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9966510381781648, | |
| "grad_norm": 10.914112703731178, | |
| "learning_rate": 2.810862937696612e-10, | |
| "logits/chosen": -2.813309907913208, | |
| "logits/rejected": -2.8205251693725586, | |
| "logps/chosen": -0.8651328086853027, | |
| "logps/rejected": -0.9807795286178589, | |
| "loss": 1.5128, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.8651328086853027, | |
| "rewards/margins": 0.11564669758081436, | |
| "rewards/rejected": -0.9807795286178589, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9987943737441393, | |
| "step": 466, | |
| "total_flos": 0.0, | |
| "train_loss": 1.5861326811651304, | |
| "train_runtime": 19371.3426, | |
| "train_samples_per_second": 3.083, | |
| "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 | |
| } | |