| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.998691442030882, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010468463752944255, | |
| "grad_norm": 15.141688725114767, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": -2.770766496658325, | |
| "logits/rejected": -2.728191375732422, | |
| "logps/chosen": -1.027971625328064, | |
| "logps/rejected": -1.1730705499649048, | |
| "loss": 2.0284, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.027971625328064, | |
| "rewards/margins": 0.14509883522987366, | |
| "rewards/rejected": -1.1730705499649048, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 13.969517964651633, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.7511062622070312, | |
| "logits/rejected": -2.7253737449645996, | |
| "logps/chosen": -0.9954013824462891, | |
| "logps/rejected": -1.0719196796417236, | |
| "loss": 2.002, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.9954013824462891, | |
| "rewards/margins": 0.07651824504137039, | |
| "rewards/rejected": -1.0719196796417236, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 13.915856984059875, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": -2.7306742668151855, | |
| "logits/rejected": -2.697096347808838, | |
| "logps/chosen": -0.9895243644714355, | |
| "logps/rejected": -1.147679328918457, | |
| "loss": 1.9874, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.9895243644714355, | |
| "rewards/margins": 0.15815496444702148, | |
| "rewards/rejected": -1.147679328918457, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 13.588156017248505, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.8030834197998047, | |
| "logits/rejected": -2.709643602371216, | |
| "logps/chosen": -1.0228091478347778, | |
| "logps/rejected": -1.1238880157470703, | |
| "loss": 2.0265, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -1.0228091478347778, | |
| "rewards/margins": 0.10107902437448502, | |
| "rewards/rejected": -1.1238880157470703, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 14.042807702810773, | |
| "learning_rate": 2.604166666666667e-07, | |
| "logits/chosen": -2.7991013526916504, | |
| "logits/rejected": -2.7758243083953857, | |
| "logps/chosen": -1.0302048921585083, | |
| "logps/rejected": -1.086944818496704, | |
| "loss": 2.0091, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.0302048921585083, | |
| "rewards/margins": 0.05673985555768013, | |
| "rewards/rejected": -1.086944818496704, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 15.069602245174066, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.7733917236328125, | |
| "logits/rejected": -2.7317938804626465, | |
| "logps/chosen": -0.9855572581291199, | |
| "logps/rejected": -1.0908451080322266, | |
| "loss": 2.0118, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9855572581291199, | |
| "rewards/margins": 0.1052878275513649, | |
| "rewards/rejected": -1.0908451080322266, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 14.14588992747414, | |
| "learning_rate": 3.645833333333333e-07, | |
| "logits/chosen": -2.7773802280426025, | |
| "logits/rejected": -2.6910452842712402, | |
| "logps/chosen": -1.0310332775115967, | |
| "logps/rejected": -1.1711452007293701, | |
| "loss": 1.9903, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.0310332775115967, | |
| "rewards/margins": 0.1401120126247406, | |
| "rewards/rejected": -1.1711452007293701, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 20.106972236529245, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.7931580543518066, | |
| "logits/rejected": -2.6956787109375, | |
| "logps/chosen": -1.011854887008667, | |
| "logps/rejected": -1.2990570068359375, | |
| "loss": 1.9463, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.011854887008667, | |
| "rewards/margins": 0.28720206022262573, | |
| "rewards/rejected": -1.2990570068359375, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 15.19327904046142, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": -2.8017444610595703, | |
| "logits/rejected": -2.7550606727600098, | |
| "logps/chosen": -0.9639647603034973, | |
| "logps/rejected": -1.142878770828247, | |
| "loss": 1.9512, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.9639647603034973, | |
| "rewards/margins": 0.17891402542591095, | |
| "rewards/rejected": -1.142878770828247, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 20.91684390311358, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -2.7584781646728516, | |
| "logits/rejected": -2.6904306411743164, | |
| "logps/chosen": -1.0117145776748657, | |
| "logps/rejected": -1.30343496799469, | |
| "loss": 1.9653, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0117145776748657, | |
| "rewards/margins": 0.2917203903198242, | |
| "rewards/rejected": -1.30343496799469, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 16.247858117319236, | |
| "learning_rate": 4.996716052911017e-07, | |
| "logits/chosen": -2.754804849624634, | |
| "logits/rejected": -2.71933650970459, | |
| "logps/chosen": -1.0166981220245361, | |
| "logps/rejected": -1.268345832824707, | |
| "loss": 1.9611, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.0166981220245361, | |
| "rewards/margins": 0.25164780020713806, | |
| "rewards/rejected": -1.268345832824707, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 16.39781017532518, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -2.8165841102600098, | |
| "logits/rejected": -2.7710437774658203, | |
| "logps/chosen": -1.049938440322876, | |
| "logps/rejected": -1.2433346509933472, | |
| "loss": 1.9164, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.049938440322876, | |
| "rewards/margins": 0.19339624047279358, | |
| "rewards/rejected": -1.2433346509933472, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 15.239254500208718, | |
| "learning_rate": 4.980652179769217e-07, | |
| "logits/chosen": -2.8142337799072266, | |
| "logits/rejected": -2.6995322704315186, | |
| "logps/chosen": -1.0191227197647095, | |
| "logps/rejected": -1.544886827468872, | |
| "loss": 1.9025, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.0191227197647095, | |
| "rewards/margins": 0.5257640480995178, | |
| "rewards/rejected": -1.544886827468872, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 21.0566472158153, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -2.7174973487854004, | |
| "logits/rejected": -2.6667487621307373, | |
| "logps/chosen": -1.0492310523986816, | |
| "logps/rejected": -1.3482974767684937, | |
| "loss": 1.9091, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.0492310523986816, | |
| "rewards/margins": 0.2990665137767792, | |
| "rewards/rejected": -1.3482974767684937, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 30.761967865275885, | |
| "learning_rate": 4.951291206355559e-07, | |
| "logits/chosen": -2.773061752319336, | |
| "logits/rejected": -2.698946952819824, | |
| "logps/chosen": -0.9925507307052612, | |
| "logps/rejected": -1.3563807010650635, | |
| "loss": 1.8891, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.9925507307052612, | |
| "rewards/margins": 0.36382991075515747, | |
| "rewards/rejected": -1.3563807010650635, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 31.297744502118277, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -2.738257884979248, | |
| "logits/rejected": -2.6784729957580566, | |
| "logps/chosen": -1.0087690353393555, | |
| "logps/rejected": -1.4246900081634521, | |
| "loss": 1.9382, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.0087690353393555, | |
| "rewards/margins": 0.4159211218357086, | |
| "rewards/rejected": -1.4246900081634521, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 24.407477818447635, | |
| "learning_rate": 4.908790517010636e-07, | |
| "logits/chosen": -2.747776508331299, | |
| "logits/rejected": -2.7003962993621826, | |
| "logps/chosen": -0.9973554611206055, | |
| "logps/rejected": -1.3728947639465332, | |
| "loss": 1.9613, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.9973554611206055, | |
| "rewards/margins": 0.3755393624305725, | |
| "rewards/rejected": -1.3728947639465332, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 31.20847987328103, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -2.705894708633423, | |
| "logits/rejected": -2.6819891929626465, | |
| "logps/chosen": -1.0237383842468262, | |
| "logps/rejected": -1.404168725013733, | |
| "loss": 1.88, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0237383842468262, | |
| "rewards/margins": 0.3804302513599396, | |
| "rewards/rejected": -1.404168725013733, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 30.65914852598083, | |
| "learning_rate": 4.853377929214243e-07, | |
| "logits/chosen": -2.776717185974121, | |
| "logits/rejected": -2.6825289726257324, | |
| "logps/chosen": -1.038529872894287, | |
| "logps/rejected": -1.3572471141815186, | |
| "loss": 1.9062, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.038529872894287, | |
| "rewards/margins": 0.31871724128723145, | |
| "rewards/rejected": -1.3572471141815186, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 34.08856475096774, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -2.74165678024292, | |
| "logits/rejected": -2.6503825187683105, | |
| "logps/chosen": -1.051953673362732, | |
| "logps/rejected": -1.4617860317230225, | |
| "loss": 1.9147, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.051953673362732, | |
| "rewards/margins": 0.4098321497440338, | |
| "rewards/rejected": -1.4617860317230225, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 28.24345375738832, | |
| "learning_rate": 4.785350472409791e-07, | |
| "logits/chosen": -2.743473529815674, | |
| "logits/rejected": -2.7197020053863525, | |
| "logps/chosen": -0.931172251701355, | |
| "logps/rejected": -1.453914999961853, | |
| "loss": 1.8781, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.931172251701355, | |
| "rewards/margins": 0.522742748260498, | |
| "rewards/rejected": -1.453914999961853, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 35.845899664243596, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -2.817007064819336, | |
| "logits/rejected": -2.762336254119873, | |
| "logps/chosen": -1.0500761270523071, | |
| "logps/rejected": -1.4565006494522095, | |
| "loss": 1.88, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.0500761270523071, | |
| "rewards/margins": 0.40642452239990234, | |
| "rewards/rejected": -1.4565006494522095, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 23.382933706781618, | |
| "learning_rate": 4.70507279583015e-07, | |
| "logits/chosen": -2.7697396278381348, | |
| "logits/rejected": -2.710322856903076, | |
| "logps/chosen": -1.0177266597747803, | |
| "logps/rejected": -1.5666338205337524, | |
| "loss": 1.868, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.0177266597747803, | |
| "rewards/margins": 0.5489073991775513, | |
| "rewards/rejected": -1.5666338205337524, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 29.59771539393517, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -2.7786221504211426, | |
| "logits/rejected": -2.737684965133667, | |
| "logps/chosen": -1.0584137439727783, | |
| "logps/rejected": -1.4742605686187744, | |
| "loss": 1.9062, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.0584137439727783, | |
| "rewards/margins": 0.41584667563438416, | |
| "rewards/rejected": -1.4742605686187744, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 24.555574989451102, | |
| "learning_rate": 4.612975213859487e-07, | |
| "logits/chosen": -2.8200080394744873, | |
| "logits/rejected": -2.7997546195983887, | |
| "logps/chosen": -1.058563470840454, | |
| "logps/rejected": -1.3519017696380615, | |
| "loss": 1.9069, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.058563470840454, | |
| "rewards/margins": 0.29333844780921936, | |
| "rewards/rejected": -1.3519017696380615, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 26.634009671559898, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -2.788752794265747, | |
| "logits/rejected": -2.754631519317627, | |
| "logps/chosen": -1.0312286615371704, | |
| "logps/rejected": -1.2794463634490967, | |
| "loss": 1.8977, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.0312286615371704, | |
| "rewards/margins": 0.24821774661540985, | |
| "rewards/rejected": -1.2794463634490967, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 32.240888174368, | |
| "learning_rate": 4.5095513994085974e-07, | |
| "logits/chosen": -2.7877840995788574, | |
| "logits/rejected": -2.7001712322235107, | |
| "logps/chosen": -1.0880383253097534, | |
| "logps/rejected": -1.7845348119735718, | |
| "loss": 1.8912, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.0880383253097534, | |
| "rewards/margins": 0.696496307849884, | |
| "rewards/rejected": -1.7845348119735718, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 25.76204815720163, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -2.8313488960266113, | |
| "logits/rejected": -2.780134677886963, | |
| "logps/chosen": -0.9982128143310547, | |
| "logps/rejected": -1.4504693746566772, | |
| "loss": 1.8155, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.9982128143310547, | |
| "rewards/margins": 0.4522566795349121, | |
| "rewards/rejected": -1.4504693746566772, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 21.7794898263259, | |
| "learning_rate": 4.395355737667985e-07, | |
| "logits/chosen": -2.809255361557007, | |
| "logits/rejected": -2.761308193206787, | |
| "logps/chosen": -1.0206363201141357, | |
| "logps/rejected": -1.419768214225769, | |
| "loss": 1.8605, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0206363201141357, | |
| "rewards/margins": 0.39913198351860046, | |
| "rewards/rejected": -1.419768214225769, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 26.507909597761785, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -2.849714517593384, | |
| "logits/rejected": -2.8313772678375244, | |
| "logps/chosen": -0.9701302647590637, | |
| "logps/rejected": -1.3971799612045288, | |
| "loss": 1.8345, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.9701302647590637, | |
| "rewards/margins": 0.42704978585243225, | |
| "rewards/rejected": -1.3971799612045288, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 28.367719379820283, | |
| "learning_rate": 4.271000354423425e-07, | |
| "logits/chosen": -2.7890210151672363, | |
| "logits/rejected": -2.757753849029541, | |
| "logps/chosen": -1.0669995546340942, | |
| "logps/rejected": -1.4297670125961304, | |
| "loss": 1.894, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.0669995546340942, | |
| "rewards/margins": 0.36276736855506897, | |
| "rewards/rejected": -1.4297670125961304, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 32.029509535293904, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -2.7710726261138916, | |
| "logits/rejected": -2.7451701164245605, | |
| "logps/chosen": -1.0376036167144775, | |
| "logps/rejected": -1.4104793071746826, | |
| "loss": 1.8838, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.0376036167144775, | |
| "rewards/margins": 0.37287575006484985, | |
| "rewards/rejected": -1.4104793071746826, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 25.20687040858121, | |
| "learning_rate": 4.137151834863213e-07, | |
| "logits/chosen": -2.7923319339752197, | |
| "logits/rejected": -2.707219362258911, | |
| "logps/chosen": -0.9815086126327515, | |
| "logps/rejected": -1.4057986736297607, | |
| "loss": 1.8435, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.9815086126327515, | |
| "rewards/margins": 0.4242900311946869, | |
| "rewards/rejected": -1.4057986736297607, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 28.006785803959723, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -2.8175511360168457, | |
| "logits/rejected": -2.752450466156006, | |
| "logps/chosen": -1.0611494779586792, | |
| "logps/rejected": -1.5021220445632935, | |
| "loss": 1.8575, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0611494779586792, | |
| "rewards/margins": 0.4409725069999695, | |
| "rewards/rejected": -1.5021220445632935, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 25.596200434230976, | |
| "learning_rate": 3.994527650465352e-07, | |
| "logits/chosen": -2.7964577674865723, | |
| "logits/rejected": -2.709230661392212, | |
| "logps/chosen": -1.1404192447662354, | |
| "logps/rejected": -1.6063992977142334, | |
| "loss": 1.9076, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1404192447662354, | |
| "rewards/margins": 0.4659802317619324, | |
| "rewards/rejected": -1.6063992977142334, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 25.892094106029592, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -2.7817273139953613, | |
| "logits/rejected": -2.7617640495300293, | |
| "logps/chosen": -1.0391079187393188, | |
| "logps/rejected": -1.5876353979110718, | |
| "loss": 1.8544, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0391079187393188, | |
| "rewards/margins": 0.5485275387763977, | |
| "rewards/rejected": -1.5876353979110718, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 29.42463620034665, | |
| "learning_rate": 3.8438923131177237e-07, | |
| "logits/chosen": -2.7785096168518066, | |
| "logits/rejected": -2.6860692501068115, | |
| "logps/chosen": -1.0492454767227173, | |
| "logps/rejected": -1.6226046085357666, | |
| "loss": 1.8247, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0492454767227173, | |
| "rewards/margins": 0.573358952999115, | |
| "rewards/rejected": -1.6226046085357666, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 27.921360264789193, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": -2.7853875160217285, | |
| "logits/rejected": -2.7454705238342285, | |
| "logps/chosen": -1.0294712781906128, | |
| "logps/rejected": -1.4654343128204346, | |
| "loss": 1.9244, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0294712781906128, | |
| "rewards/margins": 0.4359629154205322, | |
| "rewards/rejected": -1.4654343128204346, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 29.87078620394206, | |
| "learning_rate": 3.6860532770864005e-07, | |
| "logits/chosen": -2.701105833053589, | |
| "logits/rejected": -2.6695313453674316, | |
| "logps/chosen": -1.098097324371338, | |
| "logps/rejected": -1.4831833839416504, | |
| "loss": 1.899, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.098097324371338, | |
| "rewards/margins": 0.3850858509540558, | |
| "rewards/rejected": -1.4831833839416504, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 28.048650935368237, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": -2.738492250442505, | |
| "logits/rejected": -2.7187023162841797, | |
| "logps/chosen": -1.0665818452835083, | |
| "logps/rejected": -1.46578049659729, | |
| "loss": 1.8737, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0665818452835083, | |
| "rewards/margins": 0.39919885993003845, | |
| "rewards/rejected": -1.46578049659729, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 27.850087928191183, | |
| "learning_rate": 3.5218566107988867e-07, | |
| "logits/chosen": -2.6582818031311035, | |
| "logits/rejected": -2.5809996128082275, | |
| "logps/chosen": -0.9890678524971008, | |
| "logps/rejected": -1.4678758382797241, | |
| "loss": 1.8403, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.9890678524971008, | |
| "rewards/margins": 0.47880783677101135, | |
| "rewards/rejected": -1.4678758382797241, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 29.109484926295075, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -2.6842644214630127, | |
| "logits/rejected": -2.627836227416992, | |
| "logps/chosen": -1.075455665588379, | |
| "logps/rejected": -1.4202773571014404, | |
| "loss": 1.8863, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.075455665588379, | |
| "rewards/margins": 0.34482160210609436, | |
| "rewards/rejected": -1.4202773571014404, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 23.99752736324705, | |
| "learning_rate": 3.3521824616429284e-07, | |
| "logits/chosen": -2.716024875640869, | |
| "logits/rejected": -2.6916396617889404, | |
| "logps/chosen": -1.0032670497894287, | |
| "logps/rejected": -1.5087072849273682, | |
| "loss": 1.8264, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -1.0032670497894287, | |
| "rewards/margins": 0.5054404139518738, | |
| "rewards/rejected": -1.5087072849273682, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 31.21916182230841, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -2.742882013320923, | |
| "logits/rejected": -2.658285617828369, | |
| "logps/chosen": -1.033542275428772, | |
| "logps/rejected": -1.5640771389007568, | |
| "loss": 1.8307, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.033542275428772, | |
| "rewards/margins": 0.5305347442626953, | |
| "rewards/rejected": -1.5640771389007568, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 26.98263419792695, | |
| "learning_rate": 3.1779403380910425e-07, | |
| "logits/chosen": -2.7341504096984863, | |
| "logits/rejected": -2.6276047229766846, | |
| "logps/chosen": -0.988808810710907, | |
| "logps/rejected": -1.5687158107757568, | |
| "loss": 1.8334, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.988808810710907, | |
| "rewards/margins": 0.5799070596694946, | |
| "rewards/rejected": -1.5687158107757568, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 29.311922113138767, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -2.74092698097229, | |
| "logits/rejected": -2.6369271278381348, | |
| "logps/chosen": -1.009143352508545, | |
| "logps/rejected": -1.665408730506897, | |
| "loss": 1.8283, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.009143352508545, | |
| "rewards/margins": 0.6562654376029968, | |
| "rewards/rejected": -1.665408730506897, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 32.62916846952035, | |
| "learning_rate": 3.000064234440111e-07, | |
| "logits/chosen": -2.729424476623535, | |
| "logits/rejected": -2.672799587249756, | |
| "logps/chosen": -1.0022896528244019, | |
| "logps/rejected": -1.5322816371917725, | |
| "loss": 1.86, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.0022896528244019, | |
| "rewards/margins": 0.5299919247627258, | |
| "rewards/rejected": -1.5322816371917725, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 25.60849652215575, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -2.666738986968994, | |
| "logits/rejected": -2.6147377490997314, | |
| "logps/chosen": -1.0380260944366455, | |
| "logps/rejected": -1.5499128103256226, | |
| "loss": 1.8806, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.0380260944366455, | |
| "rewards/margins": 0.5118868947029114, | |
| "rewards/rejected": -1.5499128103256226, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 23.307769710548996, | |
| "learning_rate": 2.8195076242990116e-07, | |
| "logits/chosen": -2.7083492279052734, | |
| "logits/rejected": -2.608227014541626, | |
| "logps/chosen": -1.1102759838104248, | |
| "logps/rejected": -1.5368731021881104, | |
| "loss": 1.8599, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.1102759838104248, | |
| "rewards/margins": 0.42659682035446167, | |
| "rewards/rejected": -1.5368731021881104, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 30.713855522791444, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -2.621891498565674, | |
| "logits/rejected": -2.572918653488159, | |
| "logps/chosen": -1.0056602954864502, | |
| "logps/rejected": -1.5289640426635742, | |
| "loss": 1.8611, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.0056602954864502, | |
| "rewards/margins": 0.5233036875724792, | |
| "rewards/rejected": -1.5289640426635742, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 21.027582284131125, | |
| "learning_rate": 2.6372383496608186e-07, | |
| "logits/chosen": -2.679741859436035, | |
| "logits/rejected": -2.588026762008667, | |
| "logps/chosen": -1.0767912864685059, | |
| "logps/rejected": -1.6802423000335693, | |
| "loss": 1.7962, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0767912864685059, | |
| "rewards/margins": 0.603451132774353, | |
| "rewards/rejected": -1.6802423000335693, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 21.083992579626138, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -2.639758825302124, | |
| "logits/rejected": -2.539482831954956, | |
| "logps/chosen": -0.9617997407913208, | |
| "logps/rejected": -1.643094778060913, | |
| "loss": 1.8469, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.9617997407913208, | |
| "rewards/margins": 0.6812952160835266, | |
| "rewards/rejected": -1.643094778060913, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 34.336932322095805, | |
| "learning_rate": 2.454233432955807e-07, | |
| "logits/chosen": -2.6934094429016113, | |
| "logits/rejected": -2.635524034500122, | |
| "logps/chosen": -1.0399844646453857, | |
| "logps/rejected": -1.3895108699798584, | |
| "loss": 1.8185, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0399844646453857, | |
| "rewards/margins": 0.3495265543460846, | |
| "rewards/rejected": -1.3895108699798584, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 35.56183499916552, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": -2.678424596786499, | |
| "logits/rejected": -2.6192078590393066, | |
| "logps/chosen": -1.1347005367279053, | |
| "logps/rejected": -1.5019694566726685, | |
| "loss": 1.8876, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.1347005367279053, | |
| "rewards/margins": 0.36726871132850647, | |
| "rewards/rejected": -1.5019694566726685, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 21.13133035084058, | |
| "learning_rate": 2.2714738398943308e-07, | |
| "logits/chosen": -2.749488592147827, | |
| "logits/rejected": -2.629812717437744, | |
| "logps/chosen": -1.0328668355941772, | |
| "logps/rejected": -1.6207387447357178, | |
| "loss": 1.833, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0328668355941772, | |
| "rewards/margins": 0.5878719091415405, | |
| "rewards/rejected": -1.6207387447357178, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 27.44411354972676, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": -2.7630438804626465, | |
| "logits/rejected": -2.692211866378784, | |
| "logps/chosen": -0.968722939491272, | |
| "logps/rejected": -1.3731743097305298, | |
| "loss": 1.8453, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.968722939491272, | |
| "rewards/margins": 0.40445131063461304, | |
| "rewards/rejected": -1.3731743097305298, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 32.499361452307625, | |
| "learning_rate": 2.089939221172446e-07, | |
| "logits/chosen": -2.682573080062866, | |
| "logits/rejected": -2.678396224975586, | |
| "logps/chosen": -1.0675628185272217, | |
| "logps/rejected": -1.5514373779296875, | |
| "loss": 1.8267, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.0675628185272217, | |
| "rewards/margins": 0.4838746190071106, | |
| "rewards/rejected": -1.5514373779296875, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 34.51002212540645, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": -2.7195992469787598, | |
| "logits/rejected": -2.6691336631774902, | |
| "logps/chosen": -1.0737675428390503, | |
| "logps/rejected": -1.6511163711547852, | |
| "loss": 1.8055, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.0737675428390503, | |
| "rewards/margins": 0.5773489475250244, | |
| "rewards/rejected": -1.6511163711547852, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 34.81875700322843, | |
| "learning_rate": 1.9106026612264315e-07, | |
| "logits/chosen": -2.7475931644439697, | |
| "logits/rejected": -2.661341667175293, | |
| "logps/chosen": -1.0815327167510986, | |
| "logps/rejected": -1.6415588855743408, | |
| "loss": 1.8548, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.0815327167510986, | |
| "rewards/margins": 0.5600259900093079, | |
| "rewards/rejected": -1.6415588855743408, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 39.656088577829394, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": -2.650444507598877, | |
| "logits/rejected": -2.601776123046875, | |
| "logps/chosen": -1.0726263523101807, | |
| "logps/rejected": -1.5858640670776367, | |
| "loss": 1.8352, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.0726263523101807, | |
| "rewards/margins": 0.5132376551628113, | |
| "rewards/rejected": -1.5858640670776367, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 34.266407421018876, | |
| "learning_rate": 1.7344254621846017e-07, | |
| "logits/chosen": -2.6358659267425537, | |
| "logits/rejected": -2.589916944503784, | |
| "logps/chosen": -1.0825355052947998, | |
| "logps/rejected": -1.695123314857483, | |
| "loss": 1.8122, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.0825355052947998, | |
| "rewards/margins": 0.6125878691673279, | |
| "rewards/rejected": -1.695123314857483, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 29.29392494715919, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": -2.6732287406921387, | |
| "logits/rejected": -2.6671674251556396, | |
| "logps/chosen": -0.9606215357780457, | |
| "logps/rejected": -1.507188081741333, | |
| "loss": 1.8154, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.9606215357780457, | |
| "rewards/margins": 0.5465664863586426, | |
| "rewards/rejected": -1.507188081741333, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 30.886215191002822, | |
| "learning_rate": 1.562351990976095e-07, | |
| "logits/chosen": -2.6945061683654785, | |
| "logits/rejected": -2.6447794437408447, | |
| "logps/chosen": -0.9400749206542969, | |
| "logps/rejected": -1.5918124914169312, | |
| "loss": 1.8296, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.9400749206542969, | |
| "rewards/margins": 0.651737630367279, | |
| "rewards/rejected": -1.5918124914169312, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 40.31455490523793, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": -2.6877899169921875, | |
| "logits/rejected": -2.6543657779693604, | |
| "logps/chosen": -1.0309157371520996, | |
| "logps/rejected": -1.5069587230682373, | |
| "loss": 1.7854, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0309157371520996, | |
| "rewards/margins": 0.47604283690452576, | |
| "rewards/rejected": -1.5069587230682373, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 33.29916453849858, | |
| "learning_rate": 1.3953046172178413e-07, | |
| "logits/chosen": -2.712068557739258, | |
| "logits/rejected": -2.6253581047058105, | |
| "logps/chosen": -1.0362852811813354, | |
| "logps/rejected": -1.6363003253936768, | |
| "loss": 1.7979, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.0362852811813354, | |
| "rewards/margins": 0.6000149846076965, | |
| "rewards/rejected": -1.6363003253936768, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 28.615351737244207, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": -2.672591209411621, | |
| "logits/rejected": -2.624178886413574, | |
| "logps/chosen": -1.0226625204086304, | |
| "logps/rejected": -1.4031275510787964, | |
| "loss": 1.8499, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0226625204086304, | |
| "rewards/margins": 0.38046500086784363, | |
| "rewards/rejected": -1.4031275510787964, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 31.872437043858163, | |
| "learning_rate": 1.2341787690142435e-07, | |
| "logits/chosen": -2.783500909805298, | |
| "logits/rejected": -2.7123568058013916, | |
| "logps/chosen": -0.9628676176071167, | |
| "logps/rejected": -1.5226846933364868, | |
| "loss": 1.8281, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.9628676176071167, | |
| "rewards/margins": 0.5598170161247253, | |
| "rewards/rejected": -1.5226846933364868, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 27.878376041354297, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": -2.7173564434051514, | |
| "logits/rejected": -2.6413800716400146, | |
| "logps/chosen": -1.1935253143310547, | |
| "logps/rejected": -1.5975086688995361, | |
| "loss": 1.8896, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.1935253143310547, | |
| "rewards/margins": 0.40398329496383667, | |
| "rewards/rejected": -1.5975086688995361, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 36.976534101742224, | |
| "learning_rate": 1.0798381331721107e-07, | |
| "logits/chosen": -2.7495174407958984, | |
| "logits/rejected": -2.700282573699951, | |
| "logps/chosen": -1.0389212369918823, | |
| "logps/rejected": -1.5958582162857056, | |
| "loss": 1.8433, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0389212369918823, | |
| "rewards/margins": 0.5569368600845337, | |
| "rewards/rejected": -1.5958582162857056, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 30.288124940758514, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": -2.7106757164001465, | |
| "logits/rejected": -2.687342882156372, | |
| "logps/chosen": -1.0641244649887085, | |
| "logps/rejected": -1.5430529117584229, | |
| "loss": 1.8735, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0641244649887085, | |
| "rewards/margins": 0.47892847657203674, | |
| "rewards/rejected": -1.5430529117584229, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 30.72874948597524, | |
| "learning_rate": 9.331100255592436e-08, | |
| "logits/chosen": -2.7169158458709717, | |
| "logits/rejected": -2.66363787651062, | |
| "logps/chosen": -0.9323965311050415, | |
| "logps/rejected": -1.5993990898132324, | |
| "loss": 1.7736, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.9323965311050415, | |
| "rewards/margins": 0.6670026779174805, | |
| "rewards/rejected": -1.5993990898132324, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 22.0079277817506, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -2.746021270751953, | |
| "logits/rejected": -2.6950621604919434, | |
| "logps/chosen": -0.9703041315078735, | |
| "logps/rejected": -1.4327679872512817, | |
| "loss": 1.783, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.9703041315078735, | |
| "rewards/margins": 0.4624637961387634, | |
| "rewards/rejected": -1.4327679872512817, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 33.139790420388316, | |
| "learning_rate": 7.947809564230445e-08, | |
| "logits/chosen": -2.6830432415008545, | |
| "logits/rejected": -2.5992660522460938, | |
| "logps/chosen": -1.0758744478225708, | |
| "logps/rejected": -1.6359386444091797, | |
| "loss": 1.8553, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.0758744478225708, | |
| "rewards/margins": 0.5600640773773193, | |
| "rewards/rejected": -1.6359386444091797, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 42.978648090018346, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": -2.7361717224121094, | |
| "logits/rejected": -2.6721107959747314, | |
| "logps/chosen": -1.0527479648590088, | |
| "logps/rejected": -1.7067893743515015, | |
| "loss": 1.8183, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.0527479648590088, | |
| "rewards/margins": 0.6540414094924927, | |
| "rewards/rejected": -1.7067893743515015, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 24.61410992203062, | |
| "learning_rate": 6.655924144404906e-08, | |
| "logits/chosen": -2.715893268585205, | |
| "logits/rejected": -2.7327394485473633, | |
| "logps/chosen": -1.0623342990875244, | |
| "logps/rejected": -1.5285212993621826, | |
| "loss": 1.8189, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0623342990875244, | |
| "rewards/margins": 0.4661870002746582, | |
| "rewards/rejected": -1.5285212993621826, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 29.084948650681554, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": -2.7230355739593506, | |
| "logits/rejected": -2.665907621383667, | |
| "logps/chosen": -1.0696966648101807, | |
| "logps/rejected": -1.5225203037261963, | |
| "loss": 1.7794, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.0696966648101807, | |
| "rewards/margins": 0.4528239369392395, | |
| "rewards/rejected": -1.5225203037261963, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 25.930022257563827, | |
| "learning_rate": 5.4623689209832484e-08, | |
| "logits/chosen": -2.7370214462280273, | |
| "logits/rejected": -2.6596484184265137, | |
| "logps/chosen": -1.0190281867980957, | |
| "logps/rejected": -1.4870203733444214, | |
| "loss": 1.8012, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0190281867980957, | |
| "rewards/margins": 0.46799221634864807, | |
| "rewards/rejected": -1.4870203733444214, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 30.16253644602296, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": -2.755403518676758, | |
| "logits/rejected": -2.6998536586761475, | |
| "logps/chosen": -1.0058339834213257, | |
| "logps/rejected": -1.6741714477539062, | |
| "loss": 1.7816, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.0058339834213257, | |
| "rewards/margins": 0.6683375239372253, | |
| "rewards/rejected": -1.6741714477539062, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 32.100503895364014, | |
| "learning_rate": 4.373541737087263e-08, | |
| "logits/chosen": -2.6867947578430176, | |
| "logits/rejected": -2.6116786003112793, | |
| "logps/chosen": -0.9958473443984985, | |
| "logps/rejected": -1.5259888172149658, | |
| "loss": 1.8098, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.9958473443984985, | |
| "rewards/margins": 0.5301414132118225, | |
| "rewards/rejected": -1.5259888172149658, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 39.50777886532678, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": -2.6500821113586426, | |
| "logits/rejected": -2.587491273880005, | |
| "logps/chosen": -0.9868978261947632, | |
| "logps/rejected": -1.5105135440826416, | |
| "loss": 1.81, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.9868978261947632, | |
| "rewards/margins": 0.523615837097168, | |
| "rewards/rejected": -1.5105135440826416, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 39.172472564486576, | |
| "learning_rate": 3.3952790595787986e-08, | |
| "logits/chosen": -2.7077910900115967, | |
| "logits/rejected": -2.671496868133545, | |
| "logps/chosen": -1.0807896852493286, | |
| "logps/rejected": -1.6041944026947021, | |
| "loss": 1.8475, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.0807896852493286, | |
| "rewards/margins": 0.5234047174453735, | |
| "rewards/rejected": -1.6041944026947021, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 34.16755927550673, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": -2.729513645172119, | |
| "logits/rejected": -2.68589448928833, | |
| "logps/chosen": -0.9558256268501282, | |
| "logps/rejected": -1.4403427839279175, | |
| "loss": 1.8083, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.9558256268501282, | |
| "rewards/margins": 0.4845171868801117, | |
| "rewards/rejected": -1.4403427839279175, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 27.600146859315487, | |
| "learning_rate": 2.5328246937043525e-08, | |
| "logits/chosen": -2.758931875228882, | |
| "logits/rejected": -2.693338394165039, | |
| "logps/chosen": -1.004372239112854, | |
| "logps/rejected": -1.6196565628051758, | |
| "loss": 1.7709, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.004372239112854, | |
| "rewards/margins": 0.615284264087677, | |
| "rewards/rejected": -1.6196565628051758, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 26.767731856804396, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": -2.7129735946655273, | |
| "logits/rejected": -2.6572775840759277, | |
| "logps/chosen": -0.9905831217765808, | |
| "logps/rejected": -1.5554085969924927, | |
| "loss": 1.8411, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.9905831217765808, | |
| "rewards/margins": 0.5648255944252014, | |
| "rewards/rejected": -1.5554085969924927, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 32.1190541072846, | |
| "learning_rate": 1.7908016745981856e-08, | |
| "logits/chosen": -2.656250476837158, | |
| "logits/rejected": -2.645900249481201, | |
| "logps/chosen": -1.1461454629898071, | |
| "logps/rejected": -1.604714035987854, | |
| "loss": 1.8218, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.1461454629898071, | |
| "rewards/margins": 0.45856842398643494, | |
| "rewards/rejected": -1.604714035987854, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 38.87548467523877, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": -2.661616325378418, | |
| "logits/rejected": -2.617453098297119, | |
| "logps/chosen": -1.1022528409957886, | |
| "logps/rejected": -1.628826379776001, | |
| "loss": 1.8088, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.1022528409957886, | |
| "rewards/margins": 0.5265735387802124, | |
| "rewards/rejected": -1.628826379776001, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 34.292720836263, | |
| "learning_rate": 1.1731874863145142e-08, | |
| "logits/chosen": -2.7260875701904297, | |
| "logits/rejected": -2.685364246368408, | |
| "logps/chosen": -1.0480232238769531, | |
| "logps/rejected": -1.6929914951324463, | |
| "loss": 1.8315, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0480232238769531, | |
| "rewards/margins": 0.6449683308601379, | |
| "rewards/rejected": -1.6929914951324463, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 36.60770449520838, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": -2.6789753437042236, | |
| "logits/rejected": -2.6536145210266113, | |
| "logps/chosen": -1.0216642618179321, | |
| "logps/rejected": -1.547012448310852, | |
| "loss": 1.8478, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0216642618179321, | |
| "rewards/margins": 0.5253482460975647, | |
| "rewards/rejected": -1.547012448310852, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 34.43665423362751, | |
| "learning_rate": 6.832927412229017e-09, | |
| "logits/chosen": -2.721832752227783, | |
| "logits/rejected": -2.660407781600952, | |
| "logps/chosen": -1.0646559000015259, | |
| "logps/rejected": -1.554300308227539, | |
| "loss": 1.8533, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0646559000015259, | |
| "rewards/margins": 0.4896443486213684, | |
| "rewards/rejected": -1.554300308227539, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 30.770535415455328, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": -2.6228880882263184, | |
| "logits/rejected": -2.5837976932525635, | |
| "logps/chosen": -1.0409653186798096, | |
| "logps/rejected": -1.537603735923767, | |
| "loss": 1.7905, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0409653186798096, | |
| "rewards/margins": 0.4966384768486023, | |
| "rewards/rejected": -1.537603735923767, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 27.915122014592697, | |
| "learning_rate": 3.2374343405217884e-09, | |
| "logits/chosen": -2.6800639629364014, | |
| "logits/rejected": -2.6128344535827637, | |
| "logps/chosen": -1.132541298866272, | |
| "logps/rejected": -1.5116033554077148, | |
| "loss": 1.825, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.132541298866272, | |
| "rewards/margins": 0.3790620267391205, | |
| "rewards/rejected": -1.5116033554077148, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 31.401214396279837, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": -2.6828560829162598, | |
| "logits/rejected": -2.6430163383483887, | |
| "logps/chosen": -0.9926983118057251, | |
| "logps/rejected": -1.5756382942199707, | |
| "loss": 1.7905, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.9926983118057251, | |
| "rewards/margins": 0.5829401612281799, | |
| "rewards/rejected": -1.5756382942199707, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 38.51782864695839, | |
| "learning_rate": 9.64668657069706e-10, | |
| "logits/chosen": -2.7383384704589844, | |
| "logits/rejected": -2.6524176597595215, | |
| "logps/chosen": -1.0901256799697876, | |
| "logps/rejected": -1.620060682296753, | |
| "loss": 1.8476, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.0901256799697876, | |
| "rewards/margins": 0.5299350023269653, | |
| "rewards/rejected": -1.620060682296753, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 27.923910450585048, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": -2.7122910022735596, | |
| "logits/rejected": -2.6457693576812744, | |
| "logps/chosen": -0.9990774393081665, | |
| "logps/rejected": -1.4476706981658936, | |
| "loss": 1.8714, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.9990774393081665, | |
| "rewards/margins": 0.4485931396484375, | |
| "rewards/rejected": -1.4476706981658936, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 33.978087151320835, | |
| "learning_rate": 2.6813123097352287e-11, | |
| "logits/chosen": -2.7624480724334717, | |
| "logits/rejected": -2.6619315147399902, | |
| "logps/chosen": -1.042180061340332, | |
| "logps/rejected": -1.5233901739120483, | |
| "loss": 1.8233, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.042180061340332, | |
| "rewards/margins": 0.4812101721763611, | |
| "rewards/rejected": -1.5233901739120483, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 1.8644854217705238, | |
| "train_runtime": 7378.4782, | |
| "train_samples_per_second": 8.286, | |
| "train_steps_per_second": 0.065 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |