| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9982631930527722, | |
| "eval_steps": 400, | |
| "global_step": 467, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01068804275217101, | |
| "grad_norm": 7.8646928358072685, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "logits/chosen": 0.061868004500865936, | |
| "logits/rejected": 0.03774915263056755, | |
| "logps/chosen": -0.2700309157371521, | |
| "logps/rejected": -0.2683418393135071, | |
| "loss": 1.2747, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.2700309157371521, | |
| "rewards/margins": -0.0016891139093786478, | |
| "rewards/rejected": -0.2683418393135071, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02137608550434202, | |
| "grad_norm": 4.709071806627697, | |
| "learning_rate": 4.255319148936171e-06, | |
| "logits/chosen": -0.013552236370742321, | |
| "logits/rejected": -0.015203881077468395, | |
| "logps/chosen": -0.2748744487762451, | |
| "logps/rejected": -0.28091126680374146, | |
| "loss": 1.2665, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.2748744487762451, | |
| "rewards/margins": 0.006036859937012196, | |
| "rewards/rejected": -0.28091126680374146, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.03206412825651302, | |
| "grad_norm": 6.011039832471536, | |
| "learning_rate": 6.382978723404256e-06, | |
| "logits/chosen": 0.003249469446018338, | |
| "logits/rejected": 0.006007631309330463, | |
| "logps/chosen": -0.3281441628932953, | |
| "logps/rejected": -0.3555542826652527, | |
| "loss": 1.2824, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.3281441628932953, | |
| "rewards/margins": 0.027410129085183144, | |
| "rewards/rejected": -0.3555542826652527, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04275217100868404, | |
| "grad_norm": 6.004008006662261, | |
| "learning_rate": 8.510638297872341e-06, | |
| "logits/chosen": -0.01057378202676773, | |
| "logits/rejected": -0.014622092247009277, | |
| "logps/chosen": -0.3901517391204834, | |
| "logps/rejected": -0.46824750304222107, | |
| "loss": 1.2729, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.3901517391204834, | |
| "rewards/margins": 0.07809571921825409, | |
| "rewards/rejected": -0.46824750304222107, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.053440213760855046, | |
| "grad_norm": 3.977322200325917, | |
| "learning_rate": 1.0638297872340426e-05, | |
| "logits/chosen": -0.2300865650177002, | |
| "logits/rejected": -0.22464950382709503, | |
| "logps/chosen": -0.3714216351509094, | |
| "logps/rejected": -0.4496099352836609, | |
| "loss": 1.2469, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.3714216351509094, | |
| "rewards/margins": 0.07818831503391266, | |
| "rewards/rejected": -0.4496099352836609, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06412825651302605, | |
| "grad_norm": 3.540400129958154, | |
| "learning_rate": 1.2765957446808513e-05, | |
| "logits/chosen": -0.3409841060638428, | |
| "logits/rejected": -0.3320957124233246, | |
| "logps/chosen": -0.4569724202156067, | |
| "logps/rejected": -0.5337254405021667, | |
| "loss": 1.2712, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.4569724202156067, | |
| "rewards/margins": 0.07675307244062424, | |
| "rewards/rejected": -0.5337254405021667, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07481629926519706, | |
| "grad_norm": 3.7566095401283395, | |
| "learning_rate": 1.4893617021276596e-05, | |
| "logits/chosen": -0.179878830909729, | |
| "logits/rejected": -0.17120154201984406, | |
| "logps/chosen": -0.46021848917007446, | |
| "logps/rejected": -0.5379027128219604, | |
| "loss": 1.2803, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.46021848917007446, | |
| "rewards/margins": 0.0776842013001442, | |
| "rewards/rejected": -0.5379027128219604, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08550434201736808, | |
| "grad_norm": 3.1819948043330344, | |
| "learning_rate": 1.7021276595744682e-05, | |
| "logits/chosen": -0.007068359758704901, | |
| "logits/rejected": -0.01693672500550747, | |
| "logps/chosen": -0.4490579664707184, | |
| "logps/rejected": -0.4928591251373291, | |
| "loss": 1.2726, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.4490579664707184, | |
| "rewards/margins": 0.04380121827125549, | |
| "rewards/rejected": -0.4928591251373291, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09619238476953908, | |
| "grad_norm": 3.2270349702087966, | |
| "learning_rate": 1.914893617021277e-05, | |
| "logits/chosen": -0.08650992810726166, | |
| "logits/rejected": -0.054813969880342484, | |
| "logps/chosen": -0.5002081990242004, | |
| "logps/rejected": -0.5691278576850891, | |
| "loss": 1.2871, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.5002081990242004, | |
| "rewards/margins": 0.06891956925392151, | |
| "rewards/rejected": -0.5691278576850891, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10688042752171009, | |
| "grad_norm": 2.930228008373617, | |
| "learning_rate": 1.999748234942507e-05, | |
| "logits/chosen": 0.05123206228017807, | |
| "logits/rejected": 0.0639006644487381, | |
| "logps/chosen": -0.4998701512813568, | |
| "logps/rejected": -0.5959991812705994, | |
| "loss": 1.286, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.4998701512813568, | |
| "rewards/margins": 0.09612905234098434, | |
| "rewards/rejected": -0.5959991812705994, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11756847027388109, | |
| "grad_norm": 3.532397749748793, | |
| "learning_rate": 1.998210129767735e-05, | |
| "logits/chosen": 0.12292106449604034, | |
| "logits/rejected": 0.11930225789546967, | |
| "logps/chosen": -0.5049015879631042, | |
| "logps/rejected": -0.5526176691055298, | |
| "loss": 1.2864, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5049015879631042, | |
| "rewards/margins": 0.0477161705493927, | |
| "rewards/rejected": -0.5526176691055298, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1282565130260521, | |
| "grad_norm": 3.727525392306118, | |
| "learning_rate": 1.9952759374651266e-05, | |
| "logits/chosen": 0.13513417541980743, | |
| "logits/rejected": 0.11871670186519623, | |
| "logps/chosen": -0.5651683807373047, | |
| "logps/rejected": -0.629240095615387, | |
| "loss": 1.297, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.5651683807373047, | |
| "rewards/margins": 0.06407161056995392, | |
| "rewards/rejected": -0.629240095615387, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13894455577822312, | |
| "grad_norm": 4.935532042698089, | |
| "learning_rate": 1.990949761767935e-05, | |
| "logits/chosen": 0.2536969482898712, | |
| "logits/rejected": 0.29733771085739136, | |
| "logps/chosen": -0.5493020415306091, | |
| "logps/rejected": -0.6132219433784485, | |
| "loss": 1.2983, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.5493020415306091, | |
| "rewards/margins": 0.06391985714435577, | |
| "rewards/rejected": -0.6132219433784485, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14963259853039412, | |
| "grad_norm": 3.295200796877631, | |
| "learning_rate": 1.9852376532240594e-05, | |
| "logits/chosen": 0.4693821966648102, | |
| "logits/rejected": 0.4486643671989441, | |
| "logps/chosen": -0.5880988836288452, | |
| "logps/rejected": -0.6713588833808899, | |
| "loss": 1.2884, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.5880988836288452, | |
| "rewards/margins": 0.08326009660959244, | |
| "rewards/rejected": -0.6713588833808899, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.16032064128256512, | |
| "grad_norm": 3.1713468009980934, | |
| "learning_rate": 1.9781476007338058e-05, | |
| "logits/chosen": 0.4574071764945984, | |
| "logits/rejected": 0.5068231225013733, | |
| "logps/chosen": -0.5979748368263245, | |
| "logps/rejected": -0.7264337539672852, | |
| "loss": 1.2954, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5979748368263245, | |
| "rewards/margins": 0.1284589320421219, | |
| "rewards/rejected": -0.7264337539672852, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17100868403473615, | |
| "grad_norm": 3.662092336336613, | |
| "learning_rate": 1.969689520376687e-05, | |
| "logits/chosen": 0.4844590127468109, | |
| "logits/rejected": 0.48810499906539917, | |
| "logps/chosen": -0.5642470121383667, | |
| "logps/rejected": -0.6027047038078308, | |
| "loss": 1.3069, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.5642470121383667, | |
| "rewards/margins": 0.038457684218883514, | |
| "rewards/rejected": -0.6027047038078308, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18169672678690715, | |
| "grad_norm": 4.707016904902269, | |
| "learning_rate": 1.9598752415428893e-05, | |
| "logits/chosen": 0.6143449544906616, | |
| "logits/rejected": 0.6212804317474365, | |
| "logps/chosen": -0.5707159042358398, | |
| "logps/rejected": -0.6457289457321167, | |
| "loss": 1.3041, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.5707159042358398, | |
| "rewards/margins": 0.07501306384801865, | |
| "rewards/rejected": -0.6457289457321167, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19238476953907815, | |
| "grad_norm": 3.539384946251158, | |
| "learning_rate": 1.9487184903888e-05, | |
| "logits/chosen": 0.7592190504074097, | |
| "logits/rejected": 0.7624388933181763, | |
| "logps/chosen": -0.5937708616256714, | |
| "logps/rejected": -0.7089810967445374, | |
| "loss": 1.2995, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.5937708616256714, | |
| "rewards/margins": 0.11521027237176895, | |
| "rewards/rejected": -0.7089810967445374, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20307281229124916, | |
| "grad_norm": 3.622523084533489, | |
| "learning_rate": 1.9362348706397374e-05, | |
| "logits/chosen": 0.45539766550064087, | |
| "logits/rejected": 0.439523309469223, | |
| "logps/chosen": -0.6361597180366516, | |
| "logps/rejected": -0.660950243473053, | |
| "loss": 1.3105, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.6361597180366516, | |
| "rewards/margins": 0.024790499359369278, | |
| "rewards/rejected": -0.660950243473053, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.21376085504342018, | |
| "grad_norm": 3.7211660011686827, | |
| "learning_rate": 1.9224418417667295e-05, | |
| "logits/chosen": 0.4752308428287506, | |
| "logits/rejected": 0.4805218279361725, | |
| "logps/chosen": -0.6078037619590759, | |
| "logps/rejected": -0.7114612460136414, | |
| "loss": 1.3073, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.6078037619590759, | |
| "rewards/margins": 0.10365743935108185, | |
| "rewards/rejected": -0.7114612460136414, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22444889779559118, | |
| "grad_norm": 3.4955810534668945, | |
| "learning_rate": 1.907358694567865e-05, | |
| "logits/chosen": 0.4134964346885681, | |
| "logits/rejected": 0.39591750502586365, | |
| "logps/chosen": -0.5727143287658691, | |
| "logps/rejected": -0.6490830183029175, | |
| "loss": 1.2904, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.5727143287658691, | |
| "rewards/margins": 0.07636874169111252, | |
| "rewards/rejected": -0.6490830183029175, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23513694054776219, | |
| "grad_norm": 3.895186311710448, | |
| "learning_rate": 1.891006524188368e-05, | |
| "logits/chosen": 0.27901846170425415, | |
| "logits/rejected": 0.267717182636261, | |
| "logps/chosen": -0.5927436351776123, | |
| "logps/rejected": -0.6760529279708862, | |
| "loss": 1.2909, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.5927436351776123, | |
| "rewards/margins": 0.08330933004617691, | |
| "rewards/rejected": -0.6760529279708862, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2458249832999332, | |
| "grad_norm": 3.2529032172381034, | |
| "learning_rate": 1.87340820061713e-05, | |
| "logits/chosen": 0.4828861653804779, | |
| "logits/rejected": 0.4803077280521393, | |
| "logps/chosen": -0.6133522391319275, | |
| "logps/rejected": -0.6779380440711975, | |
| "loss": 1.3193, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.6133522391319275, | |
| "rewards/margins": 0.06458581984043121, | |
| "rewards/rejected": -0.6779380440711975, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2565130260521042, | |
| "grad_norm": 3.0442398494478504, | |
| "learning_rate": 1.854588336700962e-05, | |
| "logits/chosen": 0.2846220135688782, | |
| "logits/rejected": 0.2976106107234955, | |
| "logps/chosen": -0.6584545969963074, | |
| "logps/rejected": -0.7402211427688599, | |
| "loss": 1.2983, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.6584545969963074, | |
| "rewards/margins": 0.0817665308713913, | |
| "rewards/rejected": -0.7402211427688599, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26720106880427524, | |
| "grad_norm": 2.952766609523329, | |
| "learning_rate": 1.834573253721303e-05, | |
| "logits/chosen": 0.16905634105205536, | |
| "logits/rejected": 0.18389183282852173, | |
| "logps/chosen": -0.5878621339797974, | |
| "logps/rejected": -0.6856662034988403, | |
| "loss": 1.2725, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.5878621339797974, | |
| "rewards/margins": 0.097803995013237, | |
| "rewards/rejected": -0.6856662034988403, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27788911155644624, | |
| "grad_norm": 4.574556482577368, | |
| "learning_rate": 1.8133909445815277e-05, | |
| "logits/chosen": 0.2116955816745758, | |
| "logits/rejected": 0.23542623221874237, | |
| "logps/chosen": -0.6415301561355591, | |
| "logps/rejected": -0.711525559425354, | |
| "loss": 1.3254, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.6415301561355591, | |
| "rewards/margins": 0.06999534368515015, | |
| "rewards/rejected": -0.711525559425354, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.28857715430861725, | |
| "grad_norm": 3.3299570451518163, | |
| "learning_rate": 1.7910710346563417e-05, | |
| "logits/chosen": 0.21890752017498016, | |
| "logits/rejected": 0.2111731767654419, | |
| "logps/chosen": -0.6361859440803528, | |
| "logps/rejected": -0.7430610656738281, | |
| "loss": 1.3005, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.6361859440803528, | |
| "rewards/margins": 0.10687508434057236, | |
| "rewards/rejected": -0.7430610656738281, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29926519706078825, | |
| "grad_norm": 4.712724056108227, | |
| "learning_rate": 1.7676447403580114e-05, | |
| "logits/chosen": 0.17464594542980194, | |
| "logits/rejected": 0.15127554535865784, | |
| "logps/chosen": -0.6231650114059448, | |
| "logps/rejected": -0.7518295049667358, | |
| "loss": 1.3004, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.6231650114059448, | |
| "rewards/margins": 0.12866440415382385, | |
| "rewards/rejected": -0.7518295049667358, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.30995323981295925, | |
| "grad_norm": 3.8840639264672605, | |
| "learning_rate": 1.7431448254773943e-05, | |
| "logits/chosen": 0.09850597381591797, | |
| "logits/rejected": 0.06791912019252777, | |
| "logps/chosen": -0.6357108950614929, | |
| "logps/rejected": -0.7526232004165649, | |
| "loss": 1.2942, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.6357108950614929, | |
| "rewards/margins": 0.11691226810216904, | |
| "rewards/rejected": -0.7526232004165649, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.32064128256513025, | |
| "grad_norm": 3.8874041876031167, | |
| "learning_rate": 1.717605555360812e-05, | |
| "logits/chosen": -0.051757656037807465, | |
| "logits/rejected": -0.07511379569768906, | |
| "logps/chosen": -0.6289964914321899, | |
| "logps/rejected": -0.7442386746406555, | |
| "loss": 1.3017, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.6289964914321899, | |
| "rewards/margins": 0.115242138504982, | |
| "rewards/rejected": -0.7442386746406555, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33132932531730125, | |
| "grad_norm": 2.9263158094537807, | |
| "learning_rate": 1.691062648986865e-05, | |
| "logits/chosen": 0.0783982202410698, | |
| "logits/rejected": 0.06652794033288956, | |
| "logps/chosen": -0.6385728120803833, | |
| "logps/rejected": -0.7080265283584595, | |
| "loss": 1.2934, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.6385728120803833, | |
| "rewards/margins": 0.06945370137691498, | |
| "rewards/rejected": -0.7080265283584595, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3420173680694723, | |
| "grad_norm": 3.7316994379774573, | |
| "learning_rate": 1.6635532290102114e-05, | |
| "logits/chosen": 0.100054070353508, | |
| "logits/rejected": 0.10980580747127533, | |
| "logps/chosen": -0.6498998403549194, | |
| "logps/rejected": -0.7538723945617676, | |
| "loss": 1.3188, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.6498998403549194, | |
| "rewards/margins": 0.10397255420684814, | |
| "rewards/rejected": -0.7538723945617676, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3527054108216433, | |
| "grad_norm": 3.48431507554954, | |
| "learning_rate": 1.635115769842179e-05, | |
| "logits/chosen": -0.07162860035896301, | |
| "logits/rejected": -0.08312487602233887, | |
| "logps/chosen": -0.5983825325965881, | |
| "logps/rejected": -0.6531118750572205, | |
| "loss": 1.3115, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.5983825325965881, | |
| "rewards/margins": 0.0547294095158577, | |
| "rewards/rejected": -0.6531118750572205, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3633934535738143, | |
| "grad_norm": 3.1045296158878495, | |
| "learning_rate": 1.60579004384082e-05, | |
| "logits/chosen": -0.034972742199897766, | |
| "logits/rejected": -0.024472128599882126, | |
| "logps/chosen": -0.6385360956192017, | |
| "logps/rejected": -0.7788956761360168, | |
| "loss": 1.2984, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.6385360956192017, | |
| "rewards/margins": 0.14035962522029877, | |
| "rewards/rejected": -0.7788956761360168, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3740814963259853, | |
| "grad_norm": 2.8548400499038884, | |
| "learning_rate": 1.575617065685674e-05, | |
| "logits/chosen": -0.13102661073207855, | |
| "logits/rejected": -0.13152942061424255, | |
| "logps/chosen": -0.6120776534080505, | |
| "logps/rejected": -0.6730025410652161, | |
| "loss": 1.3073, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.6120776534080505, | |
| "rewards/margins": 0.06092479079961777, | |
| "rewards/rejected": -0.6730025410652161, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3847695390781563, | |
| "grad_norm": 5.270781840059861, | |
| "learning_rate": 1.5446390350150272e-05, | |
| "logits/chosen": -0.16384097933769226, | |
| "logits/rejected": -0.16462591290473938, | |
| "logps/chosen": -0.6268705725669861, | |
| "logps/rejected": -0.7125908136367798, | |
| "loss": 1.3028, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.6268705725669861, | |
| "rewards/margins": 0.08572021126747131, | |
| "rewards/rejected": -0.7125908136367798, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3954575818303273, | |
| "grad_norm": 2.9703550953722244, | |
| "learning_rate": 1.5128992774059063e-05, | |
| "logits/chosen": -0.2370837926864624, | |
| "logits/rejected": -0.2578278183937073, | |
| "logps/chosen": -0.6618776321411133, | |
| "logps/rejected": -0.6950840353965759, | |
| "loss": 1.3242, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.6618776321411133, | |
| "rewards/margins": 0.03320637717843056, | |
| "rewards/rejected": -0.6950840353965759, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4061456245824983, | |
| "grad_norm": 2.576969091118456, | |
| "learning_rate": 1.4804421837793379e-05, | |
| "logits/chosen": -0.31508585810661316, | |
| "logits/rejected": -0.31256750226020813, | |
| "logps/chosen": -0.6105505228042603, | |
| "logps/rejected": -0.6881141066551208, | |
| "loss": 1.2954, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.6105505228042603, | |
| "rewards/margins": 0.07756359130144119, | |
| "rewards/rejected": -0.6881141066551208, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4168336673346693, | |
| "grad_norm": 3.796929659092338, | |
| "learning_rate": 1.4473131483156326e-05, | |
| "logits/chosen": -0.25943300127983093, | |
| "logits/rejected": -0.26059621572494507, | |
| "logps/chosen": -0.6393094062805176, | |
| "logps/rejected": -0.7625299096107483, | |
| "loss": 1.2856, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.6393094062805176, | |
| "rewards/margins": 0.12322056293487549, | |
| "rewards/rejected": -0.7625299096107483, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.42752171008684037, | |
| "grad_norm": 3.4441112067059683, | |
| "learning_rate": 1.4135585049665207e-05, | |
| "logits/chosen": -0.24460437893867493, | |
| "logits/rejected": -0.2346183955669403, | |
| "logps/chosen": -0.6215813755989075, | |
| "logps/rejected": -0.7071408629417419, | |
| "loss": 1.2968, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.6215813755989075, | |
| "rewards/margins": 0.08555953204631805, | |
| "rewards/rejected": -0.7071408629417419, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43820975283901137, | |
| "grad_norm": 2.8503959168303465, | |
| "learning_rate": 1.3792254626529286e-05, | |
| "logits/chosen": -0.23945841193199158, | |
| "logits/rejected": -0.2520154118537903, | |
| "logps/chosen": -0.6485592722892761, | |
| "logps/rejected": -0.7163916826248169, | |
| "loss": 1.3058, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.6485592722892761, | |
| "rewards/margins": 0.06783244758844376, | |
| "rewards/rejected": -0.7163916826248169, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.44889779559118237, | |
| "grad_norm": 2.9623348839853323, | |
| "learning_rate": 1.3443620392390352e-05, | |
| "logits/chosen": -0.22061018645763397, | |
| "logits/rejected": -0.21657535433769226, | |
| "logps/chosen": -0.6273641586303711, | |
| "logps/rejected": -0.7573559880256653, | |
| "loss": 1.3037, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.6273641586303711, | |
| "rewards/margins": 0.1299917995929718, | |
| "rewards/rejected": -0.7573559880256653, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45958583834335337, | |
| "grad_norm": 3.2413099077478713, | |
| "learning_rate": 1.3090169943749475e-05, | |
| "logits/chosen": -0.11810100078582764, | |
| "logits/rejected": -0.14453695714473724, | |
| "logps/chosen": -0.6222162246704102, | |
| "logps/rejected": -0.7621700763702393, | |
| "loss": 1.2812, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.6222162246704102, | |
| "rewards/margins": 0.1399538516998291, | |
| "rewards/rejected": -0.7621700763702393, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.47027388109552437, | |
| "grad_norm": 3.1218400805304007, | |
| "learning_rate": 1.2732397613019203e-05, | |
| "logits/chosen": -0.26869744062423706, | |
| "logits/rejected": -0.277025043964386, | |
| "logps/chosen": -0.6378098726272583, | |
| "logps/rejected": -0.7766000628471375, | |
| "loss": 1.281, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.6378098726272583, | |
| "rewards/margins": 0.13879024982452393, | |
| "rewards/rejected": -0.7766000628471375, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48096192384769537, | |
| "grad_norm": 3.2992783572988698, | |
| "learning_rate": 1.2370803777154976e-05, | |
| "logits/chosen": -0.3226834833621979, | |
| "logits/rejected": -0.33692580461502075, | |
| "logps/chosen": -0.6211170554161072, | |
| "logps/rejected": -0.7601276636123657, | |
| "loss": 1.2728, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.6211170554161072, | |
| "rewards/margins": 0.13901060819625854, | |
| "rewards/rejected": -0.7601276636123657, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.4916499665998664, | |
| "grad_norm": 2.9698612659347132, | |
| "learning_rate": 1.200589415783273e-05, | |
| "logits/chosen": -0.2805679142475128, | |
| "logits/rejected": -0.3003981113433838, | |
| "logps/chosen": -0.6456371545791626, | |
| "logps/rejected": -0.7589985132217407, | |
| "loss": 1.2923, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.6456371545791626, | |
| "rewards/margins": 0.11336138099431992, | |
| "rewards/rejected": -0.7589985132217407, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5023380093520374, | |
| "grad_norm": 2.93442353979503, | |
| "learning_rate": 1.1638179114151378e-05, | |
| "logits/chosen": -0.2719253897666931, | |
| "logits/rejected": -0.2693996727466583, | |
| "logps/chosen": -0.5996414422988892, | |
| "logps/rejected": -0.7817131876945496, | |
| "loss": 1.2908, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.5996414422988892, | |
| "rewards/margins": 0.18207184970378876, | |
| "rewards/rejected": -0.7817131876945496, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5130260521042084, | |
| "grad_norm": 3.1656667859377405, | |
| "learning_rate": 1.1268172928849486e-05, | |
| "logits/chosen": -0.36759647727012634, | |
| "logits/rejected": -0.36635178327560425, | |
| "logps/chosen": -0.5924197435379028, | |
| "logps/rejected": -0.7910494208335876, | |
| "loss": 1.286, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.5924197435379028, | |
| "rewards/margins": 0.1986297070980072, | |
| "rewards/rejected": -0.7910494208335876, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5237140948563794, | |
| "grad_norm": 2.891600548101634, | |
| "learning_rate": 1.0896393089034336e-05, | |
| "logits/chosen": -0.3432336151599884, | |
| "logits/rejected": -0.33940672874450684, | |
| "logps/chosen": -0.6054046154022217, | |
| "logps/rejected": -0.6738919615745544, | |
| "loss": 1.2829, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.6054046154022217, | |
| "rewards/margins": 0.06848734617233276, | |
| "rewards/rejected": -0.6738919615745544, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5344021376085505, | |
| "grad_norm": 5.01628908139822, | |
| "learning_rate": 1.0523359562429441e-05, | |
| "logits/chosen": -0.383030503988266, | |
| "logits/rejected": -0.399156779050827, | |
| "logps/chosen": -0.6492289900779724, | |
| "logps/rejected": -0.769420862197876, | |
| "loss": 1.2834, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.6492289900779724, | |
| "rewards/margins": 0.12019187211990356, | |
| "rewards/rejected": -0.769420862197876, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5450901803607214, | |
| "grad_norm": 2.9846407948611695, | |
| "learning_rate": 1.0149594070152638e-05, | |
| "logits/chosen": -0.41133952140808105, | |
| "logits/rejected": -0.41932740807533264, | |
| "logps/chosen": -0.6495887041091919, | |
| "logps/rejected": -0.7451075315475464, | |
| "loss": 1.3063, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.6495887041091919, | |
| "rewards/margins": 0.09551887214183807, | |
| "rewards/rejected": -0.7451075315475464, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5557782231128925, | |
| "grad_norm": 3.025145774637567, | |
| "learning_rate": 9.775619357041952e-06, | |
| "logits/chosen": -0.37427011132240295, | |
| "logits/rejected": -0.38933342695236206, | |
| "logps/chosen": -0.5939801931381226, | |
| "logps/rejected": -0.7346431016921997, | |
| "loss": 1.2816, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.5939801931381226, | |
| "rewards/margins": 0.14066289365291595, | |
| "rewards/rejected": -0.7346431016921997, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5664662658650634, | |
| "grad_norm": 3.7873867621914634, | |
| "learning_rate": 9.401958460549658e-06, | |
| "logits/chosen": -0.3902246356010437, | |
| "logits/rejected": -0.39489641785621643, | |
| "logps/chosen": -0.5874339938163757, | |
| "logps/rejected": -0.6752017736434937, | |
| "loss": 1.2809, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.5874339938163757, | |
| "rewards/margins": 0.08776768296957016, | |
| "rewards/rejected": -0.6752017736434937, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5771543086172345, | |
| "grad_norm": 4.336316474366534, | |
| "learning_rate": 9.02913397922712e-06, | |
| "logits/chosen": -0.4125131666660309, | |
| "logits/rejected": -0.4208359122276306, | |
| "logps/chosen": -0.607820987701416, | |
| "logps/rejected": -0.905206024646759, | |
| "loss": 1.2747, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.607820987701416, | |
| "rewards/margins": 0.29738494753837585, | |
| "rewards/rejected": -0.905206024646759, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5878423513694054, | |
| "grad_norm": 2.9455182934401654, | |
| "learning_rate": 8.657667341823449e-06, | |
| "logits/chosen": -0.3412848114967346, | |
| "logits/rejected": -0.3707515597343445, | |
| "logps/chosen": -0.6197370290756226, | |
| "logps/rejected": -0.7178043127059937, | |
| "loss": 1.2785, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.6197370290756226, | |
| "rewards/margins": 0.09806738793849945, | |
| "rewards/rejected": -0.7178043127059937, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5985303941215765, | |
| "grad_norm": 3.0985019376115464, | |
| "learning_rate": 8.28807807802025e-06, | |
| "logits/chosen": -0.32350999116897583, | |
| "logits/rejected": -0.3327878415584564, | |
| "logps/chosen": -0.5803725123405457, | |
| "logps/rejected": -0.6662936210632324, | |
| "loss": 1.2821, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.5803725123405457, | |
| "rewards/margins": 0.08592123538255692, | |
| "rewards/rejected": -0.6662936210632324, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6092184368737475, | |
| "grad_norm": 3.037932780161429, | |
| "learning_rate": 7.92088309182241e-06, | |
| "logits/chosen": -0.3814396262168884, | |
| "logits/rejected": -0.38753852248191833, | |
| "logps/chosen": -0.5815099477767944, | |
| "logps/rejected": -0.6666026711463928, | |
| "loss": 1.294, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.5815099477767944, | |
| "rewards/margins": 0.0850927084684372, | |
| "rewards/rejected": -0.6666026711463928, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6199064796259185, | |
| "grad_norm": 3.0961871877321485, | |
| "learning_rate": 7.556595938621058e-06, | |
| "logits/chosen": -0.46853989362716675, | |
| "logits/rejected": -0.49206361174583435, | |
| "logps/chosen": -0.6127103567123413, | |
| "logps/rejected": -0.7595942616462708, | |
| "loss": 1.2796, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.6127103567123413, | |
| "rewards/margins": 0.14688387513160706, | |
| "rewards/rejected": -0.7595942616462708, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6305945223780896, | |
| "grad_norm": 2.5558798414818837, | |
| "learning_rate": 7.1957261069399745e-06, | |
| "logits/chosen": -0.41997450590133667, | |
| "logits/rejected": -0.41455715894699097, | |
| "logps/chosen": -0.5393789410591125, | |
| "logps/rejected": -0.6705390214920044, | |
| "loss": 1.2859, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.5393789410591125, | |
| "rewards/margins": 0.13116006553173065, | |
| "rewards/rejected": -0.6705390214920044, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6412825651302605, | |
| "grad_norm": 4.199327343754458, | |
| "learning_rate": 6.83877830586976e-06, | |
| "logits/chosen": -0.4186905324459076, | |
| "logits/rejected": -0.400879442691803, | |
| "logps/chosen": -0.5945369601249695, | |
| "logps/rejected": -0.8209554553031921, | |
| "loss": 1.2546, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.5945369601249695, | |
| "rewards/margins": 0.22641853988170624, | |
| "rewards/rejected": -0.8209554553031921, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6519706078824316, | |
| "grad_norm": 3.5426555178859074, | |
| "learning_rate": 6.486251759186573e-06, | |
| "logits/chosen": -0.42441505193710327, | |
| "logits/rejected": -0.4325898587703705, | |
| "logps/chosen": -0.6070866584777832, | |
| "logps/rejected": -0.7520343661308289, | |
| "loss": 1.2695, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.6070866584777832, | |
| "rewards/margins": 0.14494772255420685, | |
| "rewards/rejected": -0.7520343661308289, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6626586506346025, | |
| "grad_norm": 3.399076738544932, | |
| "learning_rate": 6.138639507142539e-06, | |
| "logits/chosen": -0.40451377630233765, | |
| "logits/rejected": -0.41942930221557617, | |
| "logps/chosen": -0.5666729211807251, | |
| "logps/rejected": -0.685453474521637, | |
| "loss": 1.2546, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.5666729211807251, | |
| "rewards/margins": 0.11878061294555664, | |
| "rewards/rejected": -0.685453474521637, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6733466933867736, | |
| "grad_norm": 3.5429592529416407, | |
| "learning_rate": 5.796427716904347e-06, | |
| "logits/chosen": -0.3595660328865051, | |
| "logits/rejected": -0.3509676158428192, | |
| "logps/chosen": -0.6435049772262573, | |
| "logps/rejected": -0.7910181879997253, | |
| "loss": 1.2803, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.6435049772262573, | |
| "rewards/margins": 0.14751319587230682, | |
| "rewards/rejected": -0.7910181879997253, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6840347361389446, | |
| "grad_norm": 3.6719145043569146, | |
| "learning_rate": 5.460095002604533e-06, | |
| "logits/chosen": -0.3422035574913025, | |
| "logits/rejected": -0.34498724341392517, | |
| "logps/chosen": -0.5467932224273682, | |
| "logps/rejected": -0.6095358729362488, | |
| "loss": 1.2877, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.5467932224273682, | |
| "rewards/margins": 0.06274259090423584, | |
| "rewards/rejected": -0.6095358729362488, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6947227788911156, | |
| "grad_norm": 3.978020135849161, | |
| "learning_rate": 5.130111755956327e-06, | |
| "logits/chosen": -0.3481397032737732, | |
| "logits/rejected": -0.34341806173324585, | |
| "logps/chosen": -0.5532098412513733, | |
| "logps/rejected": -0.7315355539321899, | |
| "loss": 1.2661, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.5532098412513733, | |
| "rewards/margins": 0.17832568287849426, | |
| "rewards/rejected": -0.7315355539321899, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7054108216432866, | |
| "grad_norm": 3.442968094749812, | |
| "learning_rate": 4.806939488368308e-06, | |
| "logits/chosen": -0.4118443429470062, | |
| "logits/rejected": -0.4033087193965912, | |
| "logps/chosen": -0.5545328855514526, | |
| "logps/rejected": -0.7638643383979797, | |
| "loss": 1.2502, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.5545328855514526, | |
| "rewards/margins": 0.2093314677476883, | |
| "rewards/rejected": -0.7638643383979797, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7160988643954576, | |
| "grad_norm": 3.693118587496843, | |
| "learning_rate": 4.491030185478976e-06, | |
| "logits/chosen": -0.3698023855686188, | |
| "logits/rejected": -0.3537530303001404, | |
| "logps/chosen": -0.5946871042251587, | |
| "logps/rejected": -0.9203634262084961, | |
| "loss": 1.2694, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.5946871042251587, | |
| "rewards/margins": 0.32567623257637024, | |
| "rewards/rejected": -0.9203634262084961, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7267869071476286, | |
| "grad_norm": 3.528628571085271, | |
| "learning_rate": 4.182825675013945e-06, | |
| "logits/chosen": -0.4216648042201996, | |
| "logits/rejected": -0.4200260639190674, | |
| "logps/chosen": -0.5487005710601807, | |
| "logps/rejected": -0.6899628639221191, | |
| "loss": 1.2749, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.5487005710601807, | |
| "rewards/margins": 0.14126229286193848, | |
| "rewards/rejected": -0.6899628639221191, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7374749498997996, | |
| "grad_norm": 3.8114111942749536, | |
| "learning_rate": 3.882757008849936e-06, | |
| "logits/chosen": -0.41083043813705444, | |
| "logits/rejected": -0.4161206781864166, | |
| "logps/chosen": -0.575046181678772, | |
| "logps/rejected": -0.7536992430686951, | |
| "loss": 1.2744, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.575046181678772, | |
| "rewards/margins": 0.17865313589572906, | |
| "rewards/rejected": -0.7536992430686951, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7481629926519706, | |
| "grad_norm": 4.0438053008699315, | |
| "learning_rate": 3.591243860149759e-06, | |
| "logits/chosen": -0.3628717064857483, | |
| "logits/rejected": -0.3853951096534729, | |
| "logps/chosen": -0.5914100408554077, | |
| "logps/rejected": -0.7692973613739014, | |
| "loss": 1.2695, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.5914100408554077, | |
| "rewards/margins": 0.17788729071617126, | |
| "rewards/rejected": -0.7692973613739014, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7588510354041417, | |
| "grad_norm": 3.0837058616543813, | |
| "learning_rate": 3.308693936411421e-06, | |
| "logits/chosen": -0.41618281602859497, | |
| "logits/rejected": -0.41979169845581055, | |
| "logps/chosen": -0.6286286115646362, | |
| "logps/rejected": -0.8485868573188782, | |
| "loss": 1.2669, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.6286286115646362, | |
| "rewards/margins": 0.2199581414461136, | |
| "rewards/rejected": -0.8485868573188782, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7695390781563126, | |
| "grad_norm": 3.528809280951808, | |
| "learning_rate": 3.0355024092523334e-06, | |
| "logits/chosen": -0.38231906294822693, | |
| "logits/rejected": -0.40169796347618103, | |
| "logps/chosen": -0.5484600067138672, | |
| "logps/rejected": -0.6940116882324219, | |
| "loss": 1.2726, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.5484600067138672, | |
| "rewards/margins": 0.14555169641971588, | |
| "rewards/rejected": -0.6940116882324219, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7802271209084837, | |
| "grad_norm": 3.5483575752162406, | |
| "learning_rate": 2.7720513617260857e-06, | |
| "logits/chosen": -0.44837936758995056, | |
| "logits/rejected": -0.4736010432243347, | |
| "logps/chosen": -0.6057429313659668, | |
| "logps/rejected": -0.7651041746139526, | |
| "loss": 1.2659, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.6057429313659668, | |
| "rewards/margins": 0.15936115384101868, | |
| "rewards/rejected": -0.7651041746139526, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7909151636606546, | |
| "grad_norm": 3.7870821722349985, | |
| "learning_rate": 2.51870925394473e-06, | |
| "logits/chosen": -0.43106943368911743, | |
| "logits/rejected": -0.44449687004089355, | |
| "logps/chosen": -0.5907775163650513, | |
| "logps/rejected": -0.7539359331130981, | |
| "loss": 1.2464, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.5907775163650513, | |
| "rewards/margins": 0.1631583422422409, | |
| "rewards/rejected": -0.7539359331130981, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8016032064128257, | |
| "grad_norm": 3.7264373621641713, | |
| "learning_rate": 2.275830407754006e-06, | |
| "logits/chosen": -0.4176546633243561, | |
| "logits/rejected": -0.4209866523742676, | |
| "logps/chosen": -0.6014735102653503, | |
| "logps/rejected": -0.7754963040351868, | |
| "loss": 1.2615, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.6014735102653503, | |
| "rewards/margins": 0.1740228235721588, | |
| "rewards/rejected": -0.7754963040351868, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8122912491649966, | |
| "grad_norm": 3.211374199470538, | |
| "learning_rate": 2.043754511182191e-06, | |
| "logits/chosen": -0.42597073316574097, | |
| "logits/rejected": -0.4239738881587982, | |
| "logps/chosen": -0.5504560470581055, | |
| "logps/rejected": -0.7494019269943237, | |
| "loss": 1.2622, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.5504560470581055, | |
| "rewards/margins": 0.19894585013389587, | |
| "rewards/rejected": -0.7494019269943237, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8229792919171677, | |
| "grad_norm": 3.301850545380218, | |
| "learning_rate": 1.8228061433556866e-06, | |
| "logits/chosen": -0.408210426568985, | |
| "logits/rejected": -0.41355451941490173, | |
| "logps/chosen": -0.5664467811584473, | |
| "logps/rejected": -0.7051645517349243, | |
| "loss": 1.2613, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.5664467811584473, | |
| "rewards/margins": 0.13871778547763824, | |
| "rewards/rejected": -0.7051645517349243, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8336673346693386, | |
| "grad_norm": 3.141220152295375, | |
| "learning_rate": 1.6132943205457607e-06, | |
| "logits/chosen": -0.41336697340011597, | |
| "logits/rejected": -0.4139551520347595, | |
| "logps/chosen": -0.556975245475769, | |
| "logps/rejected": -0.6668010950088501, | |
| "loss": 1.257, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.556975245475769, | |
| "rewards/margins": 0.10982589423656464, | |
| "rewards/rejected": -0.6668010950088501, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8443553774215097, | |
| "grad_norm": 4.402129582330891, | |
| "learning_rate": 1.4155120639813392e-06, | |
| "logits/chosen": -0.4578641355037689, | |
| "logits/rejected": -0.4619896411895752, | |
| "logps/chosen": -0.5758166909217834, | |
| "logps/rejected": -0.6729660630226135, | |
| "loss": 1.271, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.5758166909217834, | |
| "rewards/margins": 0.09714937210083008, | |
| "rewards/rejected": -0.6729660630226135, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "grad_norm": 3.440496458892442, | |
| "learning_rate": 1.2297359900323346e-06, | |
| "logits/chosen": -0.48626774549484253, | |
| "logits/rejected": -0.48332077264785767, | |
| "logps/chosen": -0.6083348989486694, | |
| "logps/rejected": -0.7751791477203369, | |
| "loss": 1.2571, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.6083348989486694, | |
| "rewards/margins": 0.1668442189693451, | |
| "rewards/rejected": -0.7751791477203369, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8550434201736807, | |
| "eval_logits/chosen": -0.3830186724662781, | |
| "eval_logits/rejected": -0.3938583731651306, | |
| "eval_logps/chosen": -0.5791661143302917, | |
| "eval_logps/rejected": -0.7269699573516846, | |
| "eval_loss": 1.2552496194839478, | |
| "eval_rewards/accuracies": 0.6626016497612, | |
| "eval_rewards/chosen": -0.5791661143302917, | |
| "eval_rewards/margins": 0.14780384302139282, | |
| "eval_rewards/rejected": -0.7269699573516846, | |
| "eval_runtime": 426.1833, | |
| "eval_samples_per_second": 4.601, | |
| "eval_steps_per_second": 0.289, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8657314629258517, | |
| "grad_norm": 3.9745302359804944, | |
| "learning_rate": 1.0562259233366334e-06, | |
| "logits/chosen": -0.48229265213012695, | |
| "logits/rejected": -0.5193549394607544, | |
| "logps/chosen": -0.6197930574417114, | |
| "logps/rejected": -0.764507532119751, | |
| "loss": 1.2391, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.6197930574417114, | |
| "rewards/margins": 0.14471451938152313, | |
| "rewards/rejected": -0.764507532119751, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8764195056780227, | |
| "grad_norm": 3.419466411083352, | |
| "learning_rate": 8.952245334118415e-07, | |
| "logits/chosen": -0.4593311846256256, | |
| "logits/rejected": -0.4776662290096283, | |
| "logps/chosen": -0.5977698564529419, | |
| "logps/rejected": -0.7118576765060425, | |
| "loss": 1.2719, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.5977698564529419, | |
| "rewards/margins": 0.11408782005310059, | |
| "rewards/rejected": -0.7118576765060425, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8871075484301937, | |
| "grad_norm": 3.431721917530868, | |
| "learning_rate": 7.46956995260033e-07, | |
| "logits/chosen": -0.4408165514469147, | |
| "logits/rejected": -0.45343226194381714, | |
| "logps/chosen": -0.5183443427085876, | |
| "logps/rejected": -0.6987067461013794, | |
| "loss": 1.251, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.5183443427085876, | |
| "rewards/margins": 0.1803622990846634, | |
| "rewards/rejected": -0.6987067461013794, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8977955911823647, | |
| "grad_norm": 3.3661467211282274, | |
| "learning_rate": 6.116306744401391e-07, | |
| "logits/chosen": -0.4190942645072937, | |
| "logits/rejected": -0.4262022376060486, | |
| "logps/chosen": -0.5752037763595581, | |
| "logps/rejected": -0.7899097204208374, | |
| "loss": 1.2457, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.5752037763595581, | |
| "rewards/margins": 0.2147059142589569, | |
| "rewards/rejected": -0.7899097204208374, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9084836339345357, | |
| "grad_norm": 3.4491228706828267, | |
| "learning_rate": 4.894348370484648e-07, | |
| "logits/chosen": -0.3854565918445587, | |
| "logits/rejected": -0.41155076026916504, | |
| "logps/chosen": -0.5380185842514038, | |
| "logps/rejected": -0.7508363127708435, | |
| "loss": 1.2377, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.5380185842514038, | |
| "rewards/margins": 0.21281783282756805, | |
| "rewards/rejected": -0.7508363127708435, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9191716766867067, | |
| "grad_norm": 3.846770405186625, | |
| "learning_rate": 3.8054038501294077e-07, | |
| "logits/chosen": -0.42405056953430176, | |
| "logits/rejected": -0.4208356738090515, | |
| "logps/chosen": -0.5479136109352112, | |
| "logps/rejected": -0.7237703204154968, | |
| "loss": 1.2312, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.5479136109352112, | |
| "rewards/margins": 0.17585670948028564, | |
| "rewards/rejected": -0.7237703204154968, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9298597194388778, | |
| "grad_norm": 3.8565624225807036, | |
| "learning_rate": 2.8509961707132496e-07, | |
| "logits/chosen": -0.4456121325492859, | |
| "logits/rejected": -0.4567331373691559, | |
| "logps/chosen": -0.6145879030227661, | |
| "logps/rejected": -0.7536846995353699, | |
| "loss": 1.2534, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.6145879030227661, | |
| "rewards/margins": 0.13909675180912018, | |
| "rewards/rejected": -0.7536846995353699, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9405477621910487, | |
| "grad_norm": 3.992184128956239, | |
| "learning_rate": 2.0324601576764525e-07, | |
| "logits/chosen": -0.42855358123779297, | |
| "logits/rejected": -0.44918686151504517, | |
| "logps/chosen": -0.6049205660820007, | |
| "logps/rejected": -0.7878492474555969, | |
| "loss": 1.2545, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.6049205660820007, | |
| "rewards/margins": 0.1829286515712738, | |
| "rewards/rejected": -0.7878492474555969, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9512358049432198, | |
| "grad_norm": 3.818733547378664, | |
| "learning_rate": 1.350940607647866e-07, | |
| "logits/chosen": -0.3801869750022888, | |
| "logits/rejected": -0.38896074891090393, | |
| "logps/chosen": -0.5800012350082397, | |
| "logps/rejected": -0.7279914617538452, | |
| "loss": 1.2581, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.5800012350082397, | |
| "rewards/margins": 0.14799019694328308, | |
| "rewards/rejected": -0.7279914617538452, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9619238476953907, | |
| "grad_norm": 3.601185509771085, | |
| "learning_rate": 8.07390687343379e-08, | |
| "logits/chosen": -0.4334098696708679, | |
| "logits/rejected": -0.4463852047920227, | |
| "logps/chosen": -0.589901864528656, | |
| "logps/rejected": -0.768067479133606, | |
| "loss": 1.2422, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.589901864528656, | |
| "rewards/margins": 0.1781655251979828, | |
| "rewards/rejected": -0.768067479133606, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9726118904475618, | |
| "grad_norm": 3.7947367879452023, | |
| "learning_rate": 4.025706004760932e-08, | |
| "logits/chosen": -0.39869770407676697, | |
| "logits/rejected": -0.39332014322280884, | |
| "logps/chosen": -0.6198989152908325, | |
| "logps/rejected": -0.8089291453361511, | |
| "loss": 1.2643, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.6198989152908325, | |
| "rewards/margins": 0.18903018534183502, | |
| "rewards/rejected": -0.8089291453361511, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9832999331997327, | |
| "grad_norm": 3.5592218644016613, | |
| "learning_rate": 1.370465245426167e-08, | |
| "logits/chosen": -0.39606115221977234, | |
| "logits/rejected": -0.4214437007904053, | |
| "logps/chosen": -0.5898042917251587, | |
| "logps/rejected": -0.7647892832756042, | |
| "loss": 1.2623, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.5898042917251587, | |
| "rewards/margins": 0.17498508095741272, | |
| "rewards/rejected": -0.7647892832756042, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9939879759519038, | |
| "grad_norm": 3.587979644510153, | |
| "learning_rate": 1.1189818972656697e-09, | |
| "logits/chosen": -0.3767338693141937, | |
| "logits/rejected": -0.40465354919433594, | |
| "logps/chosen": -0.5555558204650879, | |
| "logps/rejected": -0.7796151041984558, | |
| "loss": 1.2483, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.5555558204650879, | |
| "rewards/margins": 0.2240593433380127, | |
| "rewards/rejected": -0.7796151041984558, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9982631930527722, | |
| "step": 467, | |
| "total_flos": 0.0, | |
| "train_loss": 1.2796013329728513, | |
| "train_runtime": 21025.6961, | |
| "train_samples_per_second": 2.848, | |
| "train_steps_per_second": 0.022 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 467, | |
| "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 | |
| } | |