| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 385, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.282051282051282e-07, | |
| "logits/chosen": -1.7278180122375488, | |
| "logits/rejected": -1.7377450466156006, | |
| "logps/chosen": -29.553977966308594, | |
| "logps/rejected": -42.813133239746094, | |
| "loss": 1.0, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.282051282051282e-06, | |
| "logits/chosen": -1.866829752922058, | |
| "logits/rejected": -1.8711390495300293, | |
| "logps/chosen": -36.97040557861328, | |
| "logps/rejected": -33.66280746459961, | |
| "loss": 0.9793, | |
| "rewards/accuracies": 0.5972222089767456, | |
| "rewards/chosen": 0.010857796296477318, | |
| "rewards/margins": 0.020676644518971443, | |
| "rewards/rejected": -0.009818848222494125, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.564102564102564e-06, | |
| "logits/chosen": -1.9976301193237305, | |
| "logits/rejected": -2.0002708435058594, | |
| "logps/chosen": -29.64252281188965, | |
| "logps/rejected": -29.05857276916504, | |
| "loss": 1.0016, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -9.895558468997478e-05, | |
| "rewards/margins": -0.0015598980244249105, | |
| "rewards/rejected": 0.0014609418576583266, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.846153846153847e-06, | |
| "logits/chosen": -1.9210872650146484, | |
| "logits/rejected": -1.918402910232544, | |
| "logps/chosen": -31.397838592529297, | |
| "logps/rejected": -33.22719192504883, | |
| "loss": 0.9919, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.005476226564496756, | |
| "rewards/margins": 0.00809534452855587, | |
| "rewards/rejected": -0.0026191179640591145, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999896948438434e-06, | |
| "logits/chosen": -2.017606258392334, | |
| "logits/rejected": -2.0088634490966797, | |
| "logps/chosen": -32.566612243652344, | |
| "logps/rejected": -32.52539825439453, | |
| "loss": 0.9931, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.0030602235347032547, | |
| "rewards/margins": 0.0068772160448133945, | |
| "rewards/rejected": -0.0038169927429407835, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.987541037542187e-06, | |
| "logits/chosen": -1.8625805377960205, | |
| "logits/rejected": -1.8518139123916626, | |
| "logps/chosen": -33.56818771362305, | |
| "logps/rejected": -35.4713020324707, | |
| "loss": 0.9975, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.0032131080515682697, | |
| "rewards/margins": 0.0024742281530052423, | |
| "rewards/rejected": -0.005687335971742868, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.954691471941119e-06, | |
| "logits/chosen": -1.9413166046142578, | |
| "logits/rejected": -1.9432499408721924, | |
| "logps/chosen": -32.55731964111328, | |
| "logps/rejected": -33.21186065673828, | |
| "loss": 0.975, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.012859411537647247, | |
| "rewards/margins": 0.02500327304005623, | |
| "rewards/rejected": -0.01214386522769928, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.901618883413549e-06, | |
| "logits/chosen": -2.0717270374298096, | |
| "logits/rejected": -2.0766983032226562, | |
| "logps/chosen": -34.008384704589844, | |
| "logps/rejected": -36.6549186706543, | |
| "loss": 0.9802, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.0059700957499444485, | |
| "rewards/margins": 0.01984369568526745, | |
| "rewards/rejected": -0.025813793763518333, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.828760511501322e-06, | |
| "logits/chosen": -1.932422399520874, | |
| "logits/rejected": -1.9355411529541016, | |
| "logps/chosen": -34.327762603759766, | |
| "logps/rejected": -34.66461944580078, | |
| "loss": 0.9483, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.03001437522470951, | |
| "rewards/margins": 0.05167943984270096, | |
| "rewards/rejected": -0.02166506089270115, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7367166013034295e-06, | |
| "logits/chosen": -1.9404243230819702, | |
| "logits/rejected": -1.944933533668518, | |
| "logps/chosen": -32.38780975341797, | |
| "logps/rejected": -32.35973358154297, | |
| "loss": 0.9759, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.02388269081711769, | |
| "rewards/margins": 0.024130593985319138, | |
| "rewards/rejected": -0.00024790168390609324, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.626245458345211e-06, | |
| "logits/chosen": -2.0375094413757324, | |
| "logits/rejected": -2.0355331897735596, | |
| "logps/chosen": -32.14101028442383, | |
| "logps/rejected": -31.31320571899414, | |
| "loss": 0.956, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": 0.029947593808174133, | |
| "rewards/margins": 0.04404589533805847, | |
| "rewards/rejected": -0.014098303392529488, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "eval_logits/chosen": -2.232485294342041, | |
| "eval_logits/rejected": -2.227635622024536, | |
| "eval_logps/chosen": -34.049468994140625, | |
| "eval_logps/rejected": -37.54916000366211, | |
| "eval_loss": 0.994844913482666, | |
| "eval_rewards/accuracies": 0.5631229281425476, | |
| "eval_rewards/chosen": -0.004474184010177851, | |
| "eval_rewards/margins": 0.0052884831093251705, | |
| "eval_rewards/rejected": -0.009762667119503021, | |
| "eval_runtime": 146.0056, | |
| "eval_samples_per_second": 2.349, | |
| "eval_steps_per_second": 0.295, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.498257201263691e-06, | |
| "logits/chosen": -1.9923181533813477, | |
| "logits/rejected": -1.989919662475586, | |
| "logps/chosen": -33.152462005615234, | |
| "logps/rejected": -34.039642333984375, | |
| "loss": 0.9695, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.027407139539718628, | |
| "rewards/margins": 0.03048800863325596, | |
| "rewards/rejected": -0.003080862807109952, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.353806263777678e-06, | |
| "logits/chosen": -2.0034618377685547, | |
| "logits/rejected": -1.9951326847076416, | |
| "logps/chosen": -32.344200134277344, | |
| "logps/rejected": -32.13230895996094, | |
| "loss": 0.9785, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": 0.03021082654595375, | |
| "rewards/margins": 0.02145235612988472, | |
| "rewards/rejected": 0.008758468553423882, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.1940827077152755e-06, | |
| "logits/chosen": -2.031149387359619, | |
| "logits/rejected": -2.0231785774230957, | |
| "logps/chosen": -30.327457427978516, | |
| "logps/rejected": -32.072906494140625, | |
| "loss": 0.9496, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.044478606432676315, | |
| "rewards/margins": 0.050448428839445114, | |
| "rewards/rejected": -0.005969815421849489, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.0204024186666215e-06, | |
| "logits/chosen": -1.961639642715454, | |
| "logits/rejected": -1.9718501567840576, | |
| "logps/chosen": -31.234424591064453, | |
| "logps/rejected": -32.57652282714844, | |
| "loss": 0.9319, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.05557093024253845, | |
| "rewards/margins": 0.06806287914514542, | |
| "rewards/rejected": -0.012491944245994091, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.834196265035119e-06, | |
| "logits/chosen": -1.8733562231063843, | |
| "logits/rejected": -1.8745276927947998, | |
| "logps/chosen": -33.9085578918457, | |
| "logps/rejected": -34.8173828125, | |
| "loss": 0.8973, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.08524172008037567, | |
| "rewards/margins": 0.10328052937984467, | |
| "rewards/rejected": -0.018038809299468994, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.636998309800573e-06, | |
| "logits/chosen": -1.9239585399627686, | |
| "logits/rejected": -1.9205586910247803, | |
| "logps/chosen": -35.99612045288086, | |
| "logps/rejected": -32.70136260986328, | |
| "loss": 0.9538, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.052666354924440384, | |
| "rewards/margins": 0.04624979570508003, | |
| "rewards/rejected": 0.006416561547666788, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4304331721118078e-06, | |
| "logits/chosen": -2.02461576461792, | |
| "logits/rejected": -2.017319440841675, | |
| "logps/chosen": -33.47701644897461, | |
| "logps/rejected": -31.457340240478516, | |
| "loss": 0.8721, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.09625236690044403, | |
| "rewards/margins": 0.12790000438690186, | |
| "rewards/rejected": -0.031647637486457825, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.2162026428305436e-06, | |
| "logits/chosen": -2.030789613723755, | |
| "logits/rejected": -2.0360310077667236, | |
| "logps/chosen": -32.23139572143555, | |
| "logps/rejected": -32.44371032714844, | |
| "loss": 0.9184, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.0940173864364624, | |
| "rewards/margins": 0.0815875232219696, | |
| "rewards/rejected": 0.012429863214492798, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.996071664294641e-06, | |
| "logits/chosen": -2.0321707725524902, | |
| "logits/rejected": -2.0294008255004883, | |
| "logps/chosen": -31.273319244384766, | |
| "logps/rejected": -31.368595123291016, | |
| "loss": 0.9202, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.06469077616930008, | |
| "rewards/margins": 0.0798054188489914, | |
| "rewards/rejected": -0.015114650130271912, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7718537898066833e-06, | |
| "logits/chosen": -1.9020545482635498, | |
| "logits/rejected": -1.9066784381866455, | |
| "logps/chosen": -31.315628051757812, | |
| "logps/rejected": -32.859519958496094, | |
| "loss": 0.8922, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": 0.0871955007314682, | |
| "rewards/margins": 0.1077716127038002, | |
| "rewards/rejected": -0.02057611010968685, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "eval_logits/chosen": -2.2279350757598877, | |
| "eval_logits/rejected": -2.2231225967407227, | |
| "eval_logps/chosen": -34.06590270996094, | |
| "eval_logps/rejected": -37.589378356933594, | |
| "eval_loss": 0.9875569939613342, | |
| "eval_rewards/accuracies": 0.530315637588501, | |
| "eval_rewards/chosen": -0.009404394775629044, | |
| "eval_rewards/margins": 0.012423668056726456, | |
| "eval_rewards/rejected": -0.0218280628323555, | |
| "eval_runtime": 145.8494, | |
| "eval_samples_per_second": 2.352, | |
| "eval_steps_per_second": 0.295, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.5453962426402006e-06, | |
| "logits/chosen": -2.0143890380859375, | |
| "logits/rejected": -2.0250439643859863, | |
| "logps/chosen": -31.765193939208984, | |
| "logps/rejected": -33.9805908203125, | |
| "loss": 0.9046, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.061833299696445465, | |
| "rewards/margins": 0.09720136225223541, | |
| "rewards/rejected": -0.03536807745695114, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3185646976551794e-06, | |
| "logits/chosen": -1.907122254371643, | |
| "logits/rejected": -1.9218772649765015, | |
| "logps/chosen": -29.810348510742188, | |
| "logps/rejected": -31.652551651000977, | |
| "loss": 0.8813, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.0903574600815773, | |
| "rewards/margins": 0.11866964399814606, | |
| "rewards/rejected": -0.028312185779213905, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0932279108998323e-06, | |
| "logits/chosen": -1.9638553857803345, | |
| "logits/rejected": -1.9678163528442383, | |
| "logps/chosen": -33.116092681884766, | |
| "logps/rejected": -31.662506103515625, | |
| "loss": 0.8826, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.09041241556406021, | |
| "rewards/margins": 0.1243731826543808, | |
| "rewards/rejected": -0.03396075591444969, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8712423238279358e-06, | |
| "logits/chosen": -1.9616014957427979, | |
| "logits/rejected": -1.9398181438446045, | |
| "logps/chosen": -33.844024658203125, | |
| "logps/rejected": -35.163543701171875, | |
| "loss": 0.8473, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": 0.09291732311248779, | |
| "rewards/margins": 0.15674278140068054, | |
| "rewards/rejected": -0.06382545083761215, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6544367689701824e-06, | |
| "logits/chosen": -2.002652883529663, | |
| "logits/rejected": -1.9993493556976318, | |
| "logps/chosen": -32.73945617675781, | |
| "logps/rejected": -36.27147674560547, | |
| "loss": 0.9305, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.055684663355350494, | |
| "rewards/margins": 0.0696500912308693, | |
| "rewards/rejected": -0.013965430669486523, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4445974030621963e-06, | |
| "logits/chosen": -1.869681715965271, | |
| "logits/rejected": -1.8672786951065063, | |
| "logps/chosen": -34.0054931640625, | |
| "logps/rejected": -35.55742645263672, | |
| "loss": 0.9293, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.05487124249339104, | |
| "rewards/margins": 0.07071295380592346, | |
| "rewards/rejected": -0.015841711312532425, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.243452991757889e-06, | |
| "logits/chosen": -1.8544126749038696, | |
| "logits/rejected": -1.852034330368042, | |
| "logps/chosen": -34.22079086303711, | |
| "logps/rejected": -31.864788055419922, | |
| "loss": 0.9205, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.0530419759452343, | |
| "rewards/margins": 0.08107996731996536, | |
| "rewards/rejected": -0.028037995100021362, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0526606671603523e-06, | |
| "logits/chosen": -1.957960844039917, | |
| "logits/rejected": -1.9474948644638062, | |
| "logps/chosen": -35.02719497680664, | |
| "logps/rejected": -31.89957046508789, | |
| "loss": 0.8836, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.09883318096399307, | |
| "rewards/margins": 0.11640648543834686, | |
| "rewards/rejected": -0.017573300749063492, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.737922755071455e-07, | |
| "logits/chosen": -2.0539374351501465, | |
| "logits/rejected": -2.0390708446502686, | |
| "logps/chosen": -30.736724853515625, | |
| "logps/rejected": -32.64698791503906, | |
| "loss": 0.9452, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.05622049421072006, | |
| "rewards/margins": 0.05476574972271919, | |
| "rewards/rejected": 0.001454742974601686, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.08321427484816e-07, | |
| "logits/chosen": -1.9243535995483398, | |
| "logits/rejected": -1.9218591451644897, | |
| "logps/chosen": -32.35340118408203, | |
| "logps/rejected": -30.938573837280273, | |
| "loss": 0.8033, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": 0.17211246490478516, | |
| "rewards/margins": 0.20864196121692657, | |
| "rewards/rejected": -0.036529459059238434, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "eval_logits/chosen": -2.224973440170288, | |
| "eval_logits/rejected": -2.2201507091522217, | |
| "eval_logps/chosen": -34.1029052734375, | |
| "eval_logps/rejected": -37.6210823059082, | |
| "eval_loss": 0.989064633846283, | |
| "eval_rewards/accuracies": 0.5182723999023438, | |
| "eval_rewards/chosen": -0.020507192239165306, | |
| "eval_rewards/margins": 0.010832280851900578, | |
| "eval_rewards/rejected": -0.03133947402238846, | |
| "eval_runtime": 145.7799, | |
| "eval_samples_per_second": 2.353, | |
| "eval_steps_per_second": 0.295, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.576113578589035e-07, | |
| "logits/chosen": -1.910521149635315, | |
| "logits/rejected": -1.9072707891464233, | |
| "logps/chosen": -31.341650009155273, | |
| "logps/rejected": -33.842002868652344, | |
| "loss": 0.9015, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.07851080596446991, | |
| "rewards/margins": 0.10846547037363052, | |
| "rewards/rejected": -0.02995465137064457, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.229036944380913e-07, | |
| "logits/chosen": -1.9592889547348022, | |
| "logits/rejected": -1.9470914602279663, | |
| "logps/chosen": -34.280879974365234, | |
| "logps/rejected": -33.69014358520508, | |
| "loss": 0.8601, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.09455753862857819, | |
| "rewards/margins": 0.1399478018283844, | |
| "rewards/rejected": -0.04539026692509651, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.053082288996112e-07, | |
| "logits/chosen": -1.9943225383758545, | |
| "logits/rejected": -1.992897629737854, | |
| "logps/chosen": -33.13710403442383, | |
| "logps/rejected": -32.56306838989258, | |
| "loss": 0.8752, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.10460783541202545, | |
| "rewards/margins": 0.1296529769897461, | |
| "rewards/rejected": -0.025045130401849747, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.0579377374915805e-07, | |
| "logits/chosen": -2.081005096435547, | |
| "logits/rejected": -2.0653510093688965, | |
| "logps/chosen": -33.79452896118164, | |
| "logps/rejected": -33.11925506591797, | |
| "loss": 0.8842, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.1154591292142868, | |
| "rewards/margins": 0.11582595109939575, | |
| "rewards/rejected": -0.00036681219353340566, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.2518018074041684e-07, | |
| "logits/chosen": -1.9534845352172852, | |
| "logits/rejected": -1.9526736736297607, | |
| "logps/chosen": -32.83572006225586, | |
| "logps/rejected": -32.52921676635742, | |
| "loss": 0.8567, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.1339126080274582, | |
| "rewards/margins": 0.15313318371772766, | |
| "rewards/rejected": -0.01922057382762432, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 6.41315865106129e-08, | |
| "logits/chosen": -1.9090850353240967, | |
| "logits/rejected": -1.919353723526001, | |
| "logps/chosen": -31.896703720092773, | |
| "logps/rejected": -35.348304748535156, | |
| "loss": 0.8898, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.09239637851715088, | |
| "rewards/margins": 0.11026018857955933, | |
| "rewards/rejected": -0.017863804474473, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.3150941078050325e-08, | |
| "logits/chosen": -2.0488381385803223, | |
| "logits/rejected": -2.0423855781555176, | |
| "logps/chosen": -33.353912353515625, | |
| "logps/rejected": -29.256216049194336, | |
| "loss": 0.9016, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.0894167423248291, | |
| "rewards/margins": 0.09835406392812729, | |
| "rewards/rejected": -0.008937308564782143, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 2.575864278703266e-09, | |
| "logits/chosen": -1.9087949991226196, | |
| "logits/rejected": -1.9110119342803955, | |
| "logps/chosen": -33.8967399597168, | |
| "logps/rejected": -31.023351669311523, | |
| "loss": 0.8526, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": 0.10269282758235931, | |
| "rewards/margins": 0.1479099541902542, | |
| "rewards/rejected": -0.0452171191573143, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 385, | |
| "total_flos": 0.0, | |
| "train_loss": 0.9217340717067967, | |
| "train_runtime": 3251.2729, | |
| "train_samples_per_second": 0.947, | |
| "train_steps_per_second": 0.118 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 385, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |