| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 478, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.0416666666666666e-08, | |
| "logits/chosen": -2.8099329471588135, | |
| "logits/rejected": -2.7572641372680664, | |
| "logps/chosen": -241.48843383789062, | |
| "logps/rejected": -197.4517822265625, | |
| "loss": 271.7943, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.83237361907959, | |
| "logits/rejected": -2.808957815170288, | |
| "logps/chosen": -292.6072692871094, | |
| "logps/rejected": -278.4604797363281, | |
| "loss": 286.0386, | |
| "rewards/accuracies": 0.4166666567325592, | |
| "rewards/chosen": 0.0008353570010513067, | |
| "rewards/margins": -0.0004216647648718208, | |
| "rewards/rejected": 0.0012570219114422798, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.81878662109375, | |
| "logits/rejected": -2.7905821800231934, | |
| "logps/chosen": -286.19378662109375, | |
| "logps/rejected": -286.7618103027344, | |
| "loss": 264.9378, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": 0.04205816239118576, | |
| "rewards/margins": 0.0025776384864002466, | |
| "rewards/rejected": 0.03948052600026131, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.7921807765960693, | |
| "logits/rejected": -2.7616498470306396, | |
| "logps/chosen": -232.4526824951172, | |
| "logps/rejected": -212.8272705078125, | |
| "loss": 266.1199, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": 0.13909421861171722, | |
| "rewards/margins": 0.004568194039165974, | |
| "rewards/rejected": 0.13452602922916412, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.8503639698028564, | |
| "logits/rejected": -2.819370985031128, | |
| "logps/chosen": -280.4808654785156, | |
| "logps/rejected": -243.86935424804688, | |
| "loss": 258.2878, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": 0.19040736556053162, | |
| "rewards/margins": 0.01646682806313038, | |
| "rewards/rejected": 0.1739405393600464, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999733114418725e-07, | |
| "logits/chosen": -2.764444589614868, | |
| "logits/rejected": -2.7427210807800293, | |
| "logps/chosen": -254.5093231201172, | |
| "logps/rejected": -240.6798858642578, | |
| "loss": 244.2991, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": 0.22921113669872284, | |
| "rewards/margins": -0.0009602505015209317, | |
| "rewards/rejected": 0.23017136752605438, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.990398100856366e-07, | |
| "logits/chosen": -2.7702980041503906, | |
| "logits/rejected": -2.742863893508911, | |
| "logps/chosen": -235.5380401611328, | |
| "logps/rejected": -211.11770629882812, | |
| "loss": 255.6714, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": 0.23538246750831604, | |
| "rewards/margins": 0.020409177988767624, | |
| "rewards/rejected": 0.2149733006954193, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.967775735898179e-07, | |
| "logits/chosen": -2.7509877681732178, | |
| "logits/rejected": -2.7182183265686035, | |
| "logps/chosen": -239.36514282226562, | |
| "logps/rejected": -210.6224822998047, | |
| "loss": 253.836, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": 0.29057446122169495, | |
| "rewards/margins": 0.04796246066689491, | |
| "rewards/rejected": 0.24261197447776794, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.931986719649298e-07, | |
| "logits/chosen": -2.699618339538574, | |
| "logits/rejected": -2.697359800338745, | |
| "logps/chosen": -241.5849609375, | |
| "logps/rejected": -241.43533325195312, | |
| "loss": 261.2479, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": 0.29146844148635864, | |
| "rewards/margins": 0.030574629083275795, | |
| "rewards/rejected": 0.2608937919139862, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.883222001996351e-07, | |
| "logits/chosen": -2.7679967880249023, | |
| "logits/rejected": -2.7326865196228027, | |
| "logps/chosen": -225.57156372070312, | |
| "logps/rejected": -208.76693725585938, | |
| "loss": 252.1535, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": 0.29392915964126587, | |
| "rewards/margins": 0.03465462103486061, | |
| "rewards/rejected": 0.25927454233169556, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.821741763807186e-07, | |
| "logits/chosen": -2.819113254547119, | |
| "logits/rejected": -2.783386707305908, | |
| "logps/chosen": -260.7179260253906, | |
| "logps/rejected": -238.54434204101562, | |
| "loss": 256.7688, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.28721123933792114, | |
| "rewards/margins": -0.005946027580648661, | |
| "rewards/rejected": 0.29315727949142456, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_logits/chosen": -2.7863247394561768, | |
| "eval_logits/rejected": -2.767011880874634, | |
| "eval_logps/chosen": -227.56491088867188, | |
| "eval_logps/rejected": -230.47303771972656, | |
| "eval_loss": 245.72764587402344, | |
| "eval_rewards/accuracies": 0.57421875, | |
| "eval_rewards/chosen": 0.29474756121635437, | |
| "eval_rewards/margins": 0.025945277884602547, | |
| "eval_rewards/rejected": 0.268802285194397, | |
| "eval_runtime": 53.5253, | |
| "eval_samples_per_second": 37.365, | |
| "eval_steps_per_second": 0.598, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.747874028753375e-07, | |
| "logits/chosen": -2.6783273220062256, | |
| "logits/rejected": -2.644078016281128, | |
| "logps/chosen": -237.9336395263672, | |
| "logps/rejected": -188.83895874023438, | |
| "loss": 240.9402, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.32163843512535095, | |
| "rewards/margins": -0.008066670037806034, | |
| "rewards/rejected": 0.3297051787376404, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.662012913161997e-07, | |
| "logits/chosen": -2.746994733810425, | |
| "logits/rejected": -2.724551200866699, | |
| "logps/chosen": -240.14242553710938, | |
| "logps/rejected": -237.6074981689453, | |
| "loss": 245.7042, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.2870863080024719, | |
| "rewards/margins": 0.03465163707733154, | |
| "rewards/rejected": 0.25243470072746277, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5646165232345103e-07, | |
| "logits/chosen": -2.7405877113342285, | |
| "logits/rejected": -2.7226414680480957, | |
| "logps/chosen": -242.31460571289062, | |
| "logps/rejected": -222.0249786376953, | |
| "loss": 239.1022, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.3218019902706146, | |
| "rewards/margins": 0.03552493453025818, | |
| "rewards/rejected": 0.28627708554267883, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.456204510851956e-07, | |
| "logits/chosen": -2.7363457679748535, | |
| "logits/rejected": -2.710092067718506, | |
| "logps/chosen": -248.31982421875, | |
| "logps/rejected": -242.44064331054688, | |
| "loss": 240.1779, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.32314616441726685, | |
| "rewards/margins": 0.03991778939962387, | |
| "rewards/rejected": 0.2832283675670624, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.337355301007335e-07, | |
| "logits/chosen": -2.6783087253570557, | |
| "logits/rejected": -2.707118511199951, | |
| "logps/chosen": -196.87429809570312, | |
| "logps/rejected": -205.2587890625, | |
| "loss": 236.8216, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": 0.2980460226535797, | |
| "rewards/margins": -0.012227327562868595, | |
| "rewards/rejected": 0.31027334928512573, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.2087030056579986e-07, | |
| "logits/chosen": -2.7062594890594482, | |
| "logits/rejected": -2.6792685985565186, | |
| "logps/chosen": -224.0284881591797, | |
| "logps/rejected": -215.8988494873047, | |
| "loss": 233.851, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": 0.31943365931510925, | |
| "rewards/margins": 0.01964866928756237, | |
| "rewards/rejected": 0.29978498816490173, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.070934040463998e-07, | |
| "logits/chosen": -2.6613574028015137, | |
| "logits/rejected": -2.6377127170562744, | |
| "logps/chosen": -249.140869140625, | |
| "logps/rejected": -229.2135009765625, | |
| "loss": 238.7709, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": 0.3302595019340515, | |
| "rewards/margins": 0.02071164920926094, | |
| "rewards/rejected": 0.30954790115356445, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.9247834624635404e-07, | |
| "logits/chosen": -2.697937488555908, | |
| "logits/rejected": -2.6982693672180176, | |
| "logps/chosen": -237.61752319335938, | |
| "logps/rejected": -222.94534301757812, | |
| "loss": 240.8094, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": 0.3156043589115143, | |
| "rewards/margins": -0.00877897534519434, | |
| "rewards/rejected": 0.32438334822654724, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7710310482256523e-07, | |
| "logits/chosen": -2.716660499572754, | |
| "logits/rejected": -2.6782338619232178, | |
| "logps/chosen": -256.08953857421875, | |
| "logps/rejected": -233.2887725830078, | |
| "loss": 241.1514, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.35306206345558167, | |
| "rewards/margins": 0.0451970100402832, | |
| "rewards/rejected": 0.3078650236129761, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.610497133404795e-07, | |
| "logits/chosen": -2.6975326538085938, | |
| "logits/rejected": -2.715744733810425, | |
| "logps/chosen": -260.04095458984375, | |
| "logps/rejected": -232.90744018554688, | |
| "loss": 239.4804, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": 0.3612653315067291, | |
| "rewards/margins": 0.047182150185108185, | |
| "rewards/rejected": 0.31408315896987915, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "eval_logits/chosen": -2.7289865016937256, | |
| "eval_logits/rejected": -2.7082109451293945, | |
| "eval_logps/chosen": -223.84622192382812, | |
| "eval_logps/rejected": -226.99331665039062, | |
| "eval_loss": 241.5078125, | |
| "eval_rewards/accuracies": 0.57421875, | |
| "eval_rewards/chosen": 0.33193421363830566, | |
| "eval_rewards/margins": 0.028334595263004303, | |
| "eval_rewards/rejected": 0.30359959602355957, | |
| "eval_runtime": 53.4363, | |
| "eval_samples_per_second": 37.428, | |
| "eval_steps_per_second": 0.599, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4440382358952115e-07, | |
| "logits/chosen": -2.653046131134033, | |
| "logits/rejected": -2.639833927154541, | |
| "logps/chosen": -236.7248992919922, | |
| "logps/rejected": -215.8306427001953, | |
| "loss": 232.2403, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": 0.3630830645561218, | |
| "rewards/margins": 0.002032138407230377, | |
| "rewards/rejected": 0.36105093359947205, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.272542485937368e-07, | |
| "logits/chosen": -2.6577677726745605, | |
| "logits/rejected": -2.610652446746826, | |
| "logps/chosen": -234.62515258789062, | |
| "logps/rejected": -224.5327911376953, | |
| "loss": 236.0736, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.403405100107193, | |
| "rewards/margins": 0.025665929540991783, | |
| "rewards/rejected": 0.3777391314506531, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.096924887558854e-07, | |
| "logits/chosen": -2.7138988971710205, | |
| "logits/rejected": -2.6695990562438965, | |
| "logps/chosen": -238.20166015625, | |
| "logps/rejected": -238.32510375976562, | |
| "loss": 243.5431, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": 0.36346226930618286, | |
| "rewards/margins": 0.04655776172876358, | |
| "rewards/rejected": 0.31690454483032227, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9181224366319943e-07, | |
| "logits/chosen": -2.6672616004943848, | |
| "logits/rejected": -2.6338298320770264, | |
| "logps/chosen": -245.7289581298828, | |
| "logps/rejected": -222.80068969726562, | |
| "loss": 241.8247, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.4044111371040344, | |
| "rewards/margins": 0.09945273399353027, | |
| "rewards/rejected": 0.30495840311050415, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7370891215954565e-07, | |
| "logits/chosen": -2.711887836456299, | |
| "logits/rejected": -2.660736083984375, | |
| "logps/chosen": -236.48410034179688, | |
| "logps/rejected": -220.5592803955078, | |
| "loss": 232.2589, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": 0.3508817255496979, | |
| "rewards/margins": 0.012243595905601978, | |
| "rewards/rejected": 0.3386381268501282, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.55479083351317e-07, | |
| "logits/chosen": -2.6873819828033447, | |
| "logits/rejected": -2.665743827819824, | |
| "logps/chosen": -226.0983428955078, | |
| "logps/rejected": -222.646484375, | |
| "loss": 249.1208, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": 0.33049115538597107, | |
| "rewards/margins": -0.025500113144516945, | |
| "rewards/rejected": 0.35599130392074585, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.3722002126275822e-07, | |
| "logits/chosen": -2.684255599975586, | |
| "logits/rejected": -2.6710708141326904, | |
| "logps/chosen": -250.91659545898438, | |
| "logps/rejected": -232.743896484375, | |
| "loss": 232.0498, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.3753630220890045, | |
| "rewards/margins": 0.05215846374630928, | |
| "rewards/rejected": 0.32320457696914673, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.19029145890313e-07, | |
| "logits/chosen": -2.642585515975952, | |
| "logits/rejected": -2.6435678005218506, | |
| "logps/chosen": -271.9496154785156, | |
| "logps/rejected": -232.7229461669922, | |
| "loss": 242.3832, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.3579340875148773, | |
| "rewards/margins": 0.029489517211914062, | |
| "rewards/rejected": 0.32844457030296326, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.0100351342479216e-07, | |
| "logits/chosen": -2.6854400634765625, | |
| "logits/rejected": -2.6508984565734863, | |
| "logps/chosen": -238.96533203125, | |
| "logps/rejected": -214.7186737060547, | |
| "loss": 237.8058, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": 0.38798388838768005, | |
| "rewards/margins": 0.002142349723726511, | |
| "rewards/rejected": 0.3858415484428406, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.8323929841460178e-07, | |
| "logits/chosen": -2.6979777812957764, | |
| "logits/rejected": -2.6885297298431396, | |
| "logps/chosen": -218.0729522705078, | |
| "logps/rejected": -206.6052703857422, | |
| "loss": 240.5041, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": 0.3353363871574402, | |
| "rewards/margins": -0.021952930837869644, | |
| "rewards/rejected": 0.35728925466537476, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_logits/chosen": -2.7041783332824707, | |
| "eval_logits/rejected": -2.680725574493408, | |
| "eval_logps/chosen": -222.30435180664062, | |
| "eval_logps/rejected": -225.71856689453125, | |
| "eval_loss": 239.8052978515625, | |
| "eval_rewards/accuracies": 0.5625, | |
| "eval_rewards/chosen": 0.3473527431488037, | |
| "eval_rewards/margins": 0.03100587986409664, | |
| "eval_rewards/rejected": 0.3163468837738037, | |
| "eval_runtime": 53.5009, | |
| "eval_samples_per_second": 37.383, | |
| "eval_steps_per_second": 0.598, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6583128063291573e-07, | |
| "logits/chosen": -2.681898832321167, | |
| "logits/rejected": -2.647827625274658, | |
| "logps/chosen": -249.84933471679688, | |
| "logps/rejected": -221.06466674804688, | |
| "loss": 244.5816, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.37856870889663696, | |
| "rewards/margins": 0.031483568251132965, | |
| "rewards/rejected": 0.347085177898407, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.488723393865766e-07, | |
| "logits/chosen": -2.6419777870178223, | |
| "logits/rejected": -2.6089978218078613, | |
| "logps/chosen": -247.5414581298828, | |
| "logps/rejected": -228.00442504882812, | |
| "loss": 227.1146, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.38319897651672363, | |
| "rewards/margins": 0.010034086182713509, | |
| "rewards/rejected": 0.3731648921966553, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3245295796480788e-07, | |
| "logits/chosen": -2.6546921730041504, | |
| "logits/rejected": -2.619576930999756, | |
| "logps/chosen": -235.1280059814453, | |
| "logps/rejected": -228.36550903320312, | |
| "loss": 235.3102, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.3890138268470764, | |
| "rewards/margins": 0.030328262597322464, | |
| "rewards/rejected": 0.35868555307388306, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1666074087171627e-07, | |
| "logits/chosen": -2.6908440589904785, | |
| "logits/rejected": -2.6842880249023438, | |
| "logps/chosen": -221.61752319335938, | |
| "logps/rejected": -221.1806182861328, | |
| "loss": 226.7886, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": 0.3297863006591797, | |
| "rewards/margins": 0.012335492298007011, | |
| "rewards/rejected": 0.3174508213996887, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0157994641835734e-07, | |
| "logits/chosen": -2.6660337448120117, | |
| "logits/rejected": -2.6440846920013428, | |
| "logps/chosen": -235.5915069580078, | |
| "logps/rejected": -227.4123077392578, | |
| "loss": 235.1443, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.4086208939552307, | |
| "rewards/margins": 0.052063293755054474, | |
| "rewards/rejected": 0.35655760765075684, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.729103716819111e-08, | |
| "logits/chosen": -2.631333112716675, | |
| "logits/rejected": -2.605966806411743, | |
| "logps/chosen": -223.34115600585938, | |
| "logps/rejected": -207.91500854492188, | |
| "loss": 246.5647, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": 0.35043418407440186, | |
| "rewards/margins": -0.0040539586916565895, | |
| "rewards/rejected": 0.35448816418647766, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.387025063449081e-08, | |
| "logits/chosen": -2.667217254638672, | |
| "logits/rejected": -2.6636977195739746, | |
| "logps/chosen": -240.44247436523438, | |
| "logps/rejected": -231.893798828125, | |
| "loss": 236.9863, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.37617605924606323, | |
| "rewards/margins": 0.05858853459358215, | |
| "rewards/rejected": 0.3175875246524811, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.138919252022435e-08, | |
| "logits/chosen": -2.6146702766418457, | |
| "logits/rejected": -2.611351728439331, | |
| "logps/chosen": -189.4939422607422, | |
| "logps/rejected": -206.30081176757812, | |
| "loss": 233.4487, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": 0.3334718644618988, | |
| "rewards/margins": -0.03314907103776932, | |
| "rewards/rejected": 0.3666209578514099, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.991445467064689e-08, | |
| "logits/chosen": -2.663173198699951, | |
| "logits/rejected": -2.634059429168701, | |
| "logps/chosen": -221.12240600585938, | |
| "logps/rejected": -204.8622589111328, | |
| "loss": 241.2531, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": 0.30437472462654114, | |
| "rewards/margins": -0.002047918038442731, | |
| "rewards/rejected": 0.3064226508140564, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.9507259776993954e-08, | |
| "logits/chosen": -2.646238088607788, | |
| "logits/rejected": -2.6048245429992676, | |
| "logps/chosen": -250.01693725585938, | |
| "logps/rejected": -215.4422607421875, | |
| "loss": 236.8453, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.35686779022216797, | |
| "rewards/margins": 0.03598792105913162, | |
| "rewards/rejected": 0.3208799362182617, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "eval_logits/chosen": -2.697472333908081, | |
| "eval_logits/rejected": -2.673011541366577, | |
| "eval_logps/chosen": -221.73487854003906, | |
| "eval_logps/rejected": -225.18995666503906, | |
| "eval_loss": 239.1992950439453, | |
| "eval_rewards/accuracies": 0.5625, | |
| "eval_rewards/chosen": 0.3530477285385132, | |
| "eval_rewards/margins": 0.031414665281772614, | |
| "eval_rewards/rejected": 0.3216330409049988, | |
| "eval_runtime": 53.4218, | |
| "eval_samples_per_second": 37.438, | |
| "eval_steps_per_second": 0.599, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.022313472693447e-08, | |
| "logits/chosen": -2.684819459915161, | |
| "logits/rejected": -2.622917652130127, | |
| "logps/chosen": -226.3255157470703, | |
| "logps/rejected": -246.3280487060547, | |
| "loss": 234.8678, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": 0.3726418912410736, | |
| "rewards/margins": 0.02128712832927704, | |
| "rewards/rejected": 0.3513546884059906, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.2111614344599684e-08, | |
| "logits/chosen": -2.6737873554229736, | |
| "logits/rejected": -2.638714551925659, | |
| "logps/chosen": -250.0021514892578, | |
| "logps/rejected": -255.8992156982422, | |
| "loss": 246.8406, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": 0.3681090474128723, | |
| "rewards/margins": 0.0030112355016171932, | |
| "rewards/rejected": 0.3650978207588196, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.521597710086439e-08, | |
| "logits/chosen": -2.6430375576019287, | |
| "logits/rejected": -2.6089277267456055, | |
| "logps/chosen": -255.19985961914062, | |
| "logps/rejected": -234.4241485595703, | |
| "loss": 234.6739, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": 0.3505329489707947, | |
| "rewards/margins": -0.049122948199510574, | |
| "rewards/rejected": 0.39965590834617615, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.57301420397924e-09, | |
| "logits/chosen": -2.665576934814453, | |
| "logits/rejected": -2.637112617492676, | |
| "logps/chosen": -278.3708190917969, | |
| "logps/rejected": -230.59423828125, | |
| "loss": 242.5247, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": 0.36680158972740173, | |
| "rewards/margins": 0.01681143045425415, | |
| "rewards/rejected": 0.34999018907546997, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.212833302556258e-09, | |
| "logits/chosen": -2.6472713947296143, | |
| "logits/rejected": -2.6246514320373535, | |
| "logps/chosen": -242.03775024414062, | |
| "logps/rejected": -210.9549560546875, | |
| "loss": 239.1569, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": 0.3881983160972595, | |
| "rewards/margins": -0.0041311681270599365, | |
| "rewards/rejected": 0.39232948422431946, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.158697848236607e-09, | |
| "logits/chosen": -2.6596500873565674, | |
| "logits/rejected": -2.6326732635498047, | |
| "logps/chosen": -259.3081359863281, | |
| "logps/rejected": -237.2527618408203, | |
| "loss": 246.6495, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": 0.4107862412929535, | |
| "rewards/margins": 0.07695204019546509, | |
| "rewards/rejected": 0.3338342308998108, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 4.269029751107489e-10, | |
| "logits/chosen": -2.6772663593292236, | |
| "logits/rejected": -2.6103363037109375, | |
| "logps/chosen": -256.0906677246094, | |
| "logps/rejected": -202.92539978027344, | |
| "loss": 248.9801, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": 0.42115721106529236, | |
| "rewards/margins": 0.12000129371881485, | |
| "rewards/rejected": 0.3011559247970581, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 478, | |
| "total_flos": 0.0, | |
| "train_loss": 243.0746703846185, | |
| "train_runtime": 4321.456, | |
| "train_samples_per_second": 14.147, | |
| "train_steps_per_second": 0.111 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 478, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |