| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9987943737441393, | |
| "eval_steps": 400, | |
| "global_step": 466, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010716677829872739, | |
| "grad_norm": 8.776682238147446, | |
| "learning_rate": 5.3191489361702123e-08, | |
| "logits/chosen": -2.8905415534973145, | |
| "logits/rejected": -2.8877615928649902, | |
| "logps/chosen": -0.9742914438247681, | |
| "logps/rejected": -0.9831219911575317, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.9742914438247681, | |
| "rewards/margins": 0.008830604143440723, | |
| "rewards/rejected": -0.9831219911575317, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.021433355659745478, | |
| "grad_norm": 9.810248541108757, | |
| "learning_rate": 1.0638297872340425e-07, | |
| "logits/chosen": -2.9070327281951904, | |
| "logits/rejected": -2.8929152488708496, | |
| "logps/chosen": -0.978168785572052, | |
| "logps/rejected": -1.001371145248413, | |
| "loss": -0.0, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.978168785572052, | |
| "rewards/margins": 0.023202406242489815, | |
| "rewards/rejected": -1.001371145248413, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.032150033489618215, | |
| "grad_norm": 7.0995253496741135, | |
| "learning_rate": 1.5957446808510638e-07, | |
| "logits/chosen": -2.895216703414917, | |
| "logits/rejected": -2.9180421829223633, | |
| "logps/chosen": -0.9596306085586548, | |
| "logps/rejected": -0.9847423434257507, | |
| "loss": -0.0001, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.9596306085586548, | |
| "rewards/margins": 0.02511175535619259, | |
| "rewards/rejected": -0.9847423434257507, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.042866711319490956, | |
| "grad_norm": 7.785343608601427, | |
| "learning_rate": 2.127659574468085e-07, | |
| "logits/chosen": -2.902580976486206, | |
| "logits/rejected": -2.9009933471679688, | |
| "logps/chosen": -0.9715825319290161, | |
| "logps/rejected": -0.9759886860847473, | |
| "loss": -0.0, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9715825319290161, | |
| "rewards/margins": 0.00440616998821497, | |
| "rewards/rejected": -0.9759886860847473, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0535833891493637, | |
| "grad_norm": 6.677386701450202, | |
| "learning_rate": 2.659574468085106e-07, | |
| "logits/chosen": -2.8314852714538574, | |
| "logits/rejected": -2.8363277912139893, | |
| "logps/chosen": -0.9519669413566589, | |
| "logps/rejected": -1.0002533197402954, | |
| "loss": -0.0003, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.9519669413566589, | |
| "rewards/margins": 0.048286281526088715, | |
| "rewards/rejected": -1.0002533197402954, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06430006697923643, | |
| "grad_norm": 8.4552870442767, | |
| "learning_rate": 3.1914893617021275e-07, | |
| "logits/chosen": -2.895077705383301, | |
| "logits/rejected": -2.9221832752227783, | |
| "logps/chosen": -0.983117938041687, | |
| "logps/rejected": -0.9829673767089844, | |
| "loss": -0.0002, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.983117938041687, | |
| "rewards/margins": -0.00015044640167616308, | |
| "rewards/rejected": -0.9829673767089844, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07501674480910918, | |
| "grad_norm": 7.60248558460589, | |
| "learning_rate": 3.7234042553191484e-07, | |
| "logits/chosen": -2.9559242725372314, | |
| "logits/rejected": -2.9249794483184814, | |
| "logps/chosen": -0.95380699634552, | |
| "logps/rejected": -0.9984177350997925, | |
| "loss": -0.0004, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.95380699634552, | |
| "rewards/margins": 0.04461073502898216, | |
| "rewards/rejected": -0.9984177350997925, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08573342263898191, | |
| "grad_norm": 8.131750300550957, | |
| "learning_rate": 4.25531914893617e-07, | |
| "logits/chosen": -2.8956754207611084, | |
| "logits/rejected": -2.8939049243927, | |
| "logps/chosen": -0.9987373352050781, | |
| "logps/rejected": -0.9930660128593445, | |
| "loss": -0.0015, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.9987373352050781, | |
| "rewards/margins": -0.005671312101185322, | |
| "rewards/rejected": -0.9930660128593445, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.09645010046885466, | |
| "grad_norm": 7.488747662199227, | |
| "learning_rate": 4.787234042553192e-07, | |
| "logits/chosen": -2.8674330711364746, | |
| "logits/rejected": -2.856853485107422, | |
| "logps/chosen": -0.9764472246170044, | |
| "logps/rejected": -0.9902997016906738, | |
| "loss": -0.001, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.9764472246170044, | |
| "rewards/margins": 0.013852563686668873, | |
| "rewards/rejected": -0.9902997016906738, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.1071667782987274, | |
| "grad_norm": 7.447612263721868, | |
| "learning_rate": 4.999367579541646e-07, | |
| "logits/chosen": -2.875795364379883, | |
| "logits/rejected": -2.8871512413024902, | |
| "logps/chosen": -0.9369581937789917, | |
| "logps/rejected": -0.9167054891586304, | |
| "loss": -0.0027, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.9369581937789917, | |
| "rewards/margins": -0.020252685993909836, | |
| "rewards/rejected": -0.9167054891586304, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11788345612860013, | |
| "grad_norm": 8.432358893020524, | |
| "learning_rate": 4.99550394651012e-07, | |
| "logits/chosen": -2.915107011795044, | |
| "logits/rejected": -2.925875186920166, | |
| "logps/chosen": -0.9737237095832825, | |
| "logps/rejected": -1.0020536184310913, | |
| "loss": -0.0022, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9737237095832825, | |
| "rewards/margins": 0.028329964727163315, | |
| "rewards/rejected": -1.0020536184310913, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12860013395847286, | |
| "grad_norm": 9.97321343058875, | |
| "learning_rate": 4.988133448023071e-07, | |
| "logits/chosen": -2.8224921226501465, | |
| "logits/rejected": -2.8447964191436768, | |
| "logps/chosen": -0.9149552583694458, | |
| "logps/rejected": -0.9350795745849609, | |
| "loss": -0.0041, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.9149552583694458, | |
| "rewards/margins": 0.02012430503964424, | |
| "rewards/rejected": -0.9350795745849609, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.13931681178834562, | |
| "grad_norm": 8.412802955896868, | |
| "learning_rate": 4.977266441646379e-07, | |
| "logits/chosen": -2.900336503982544, | |
| "logits/rejected": -2.9115452766418457, | |
| "logps/chosen": -0.8882263898849487, | |
| "logps/rejected": -0.9012020230293274, | |
| "loss": -0.0057, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.8882263898849487, | |
| "rewards/margins": 0.012975621037185192, | |
| "rewards/rejected": -0.9012020230293274, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.15003348961821836, | |
| "grad_norm": 8.21023142941744, | |
| "learning_rate": 4.962918198495654e-07, | |
| "logits/chosen": -2.816331386566162, | |
| "logits/rejected": -2.839690685272217, | |
| "logps/chosen": -0.9164741635322571, | |
| "logps/rejected": -0.9583846926689148, | |
| "loss": -0.007, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.9164741635322571, | |
| "rewards/margins": 0.04191048443317413, | |
| "rewards/rejected": -0.9583846926689148, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1607501674480911, | |
| "grad_norm": 8.354728288045035, | |
| "learning_rate": 4.945108881776135e-07, | |
| "logits/chosen": -2.8184382915496826, | |
| "logits/rejected": -2.81858229637146, | |
| "logps/chosen": -0.8914631009101868, | |
| "logps/rejected": -0.9294856190681458, | |
| "loss": -0.0066, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.8914631009101868, | |
| "rewards/margins": 0.038022562861442566, | |
| "rewards/rejected": -0.9294856190681458, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.17146684527796383, | |
| "grad_norm": 9.546162491226012, | |
| "learning_rate": 4.923863518447879e-07, | |
| "logits/chosen": -2.840766191482544, | |
| "logits/rejected": -2.8635830879211426, | |
| "logps/chosen": -0.887871265411377, | |
| "logps/rejected": -0.958912193775177, | |
| "loss": -0.006, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.887871265411377, | |
| "rewards/margins": 0.07104082405567169, | |
| "rewards/rejected": -0.958912193775177, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18218352310783656, | |
| "grad_norm": 9.354318155678731, | |
| "learning_rate": 4.899211964056035e-07, | |
| "logits/chosen": -2.854689359664917, | |
| "logits/rejected": -2.866008758544922, | |
| "logps/chosen": -0.9082525968551636, | |
| "logps/rejected": -0.9574602842330933, | |
| "loss": -0.0132, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9082525968551636, | |
| "rewards/margins": 0.04920773208141327, | |
| "rewards/rejected": -0.9574602842330933, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.19290020093770932, | |
| "grad_norm": 9.38237658589797, | |
| "learning_rate": 4.871188860775643e-07, | |
| "logits/chosen": -2.830294132232666, | |
| "logits/rejected": -2.8332040309906006, | |
| "logps/chosen": -0.9233630895614624, | |
| "logps/rejected": -0.9633795022964478, | |
| "loss": -0.0142, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.9233630895614624, | |
| "rewards/margins": 0.040016382932662964, | |
| "rewards/rejected": -0.9633795022964478, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20361687876758205, | |
| "grad_norm": 8.979685614931006, | |
| "learning_rate": 4.839833588729897e-07, | |
| "logits/chosen": -2.8387644290924072, | |
| "logits/rejected": -2.8495821952819824, | |
| "logps/chosen": -0.9509484171867371, | |
| "logps/rejected": -0.974162220954895, | |
| "loss": -0.0085, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.9509484171867371, | |
| "rewards/margins": 0.02321377769112587, | |
| "rewards/rejected": -0.974162220954895, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2143335565974548, | |
| "grad_norm": 9.427728619757332, | |
| "learning_rate": 4.80519021065031e-07, | |
| "logits/chosen": -2.860654354095459, | |
| "logits/rejected": -2.8680408000946045, | |
| "logps/chosen": -0.9869228601455688, | |
| "logps/rejected": -1.0385963916778564, | |
| "loss": -0.015, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.9869228601455688, | |
| "rewards/margins": 0.05167357996106148, | |
| "rewards/rejected": -1.0385963916778564, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.22505023442732752, | |
| "grad_norm": 8.191485445947501, | |
| "learning_rate": 4.767307409956528e-07, | |
| "logits/chosen": -2.859048843383789, | |
| "logits/rejected": -2.8625106811523438, | |
| "logps/chosen": -0.8909996151924133, | |
| "logps/rejected": -0.9282458424568176, | |
| "loss": -0.0154, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.8909996151924133, | |
| "rewards/margins": 0.03724620118737221, | |
| "rewards/rejected": -0.9282458424568176, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23576691225720026, | |
| "grad_norm": 9.657244741544229, | |
| "learning_rate": 4.726238422342805e-07, | |
| "logits/chosen": -2.8523192405700684, | |
| "logits/rejected": -2.84688138961792, | |
| "logps/chosen": -0.9256642460823059, | |
| "logps/rejected": -0.9944097399711609, | |
| "loss": -0.0136, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.9256642460823059, | |
| "rewards/margins": 0.06874561309814453, | |
| "rewards/rejected": -0.9944097399711609, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24648359008707302, | |
| "grad_norm": 11.45037121550009, | |
| "learning_rate": 4.6820409609673027e-07, | |
| "logits/chosen": -2.895726203918457, | |
| "logits/rejected": -2.900205612182617, | |
| "logps/chosen": -0.9234002828598022, | |
| "logps/rejected": -0.938769519329071, | |
| "loss": -0.0216, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.9234002828598022, | |
| "rewards/margins": 0.015369323082268238, | |
| "rewards/rejected": -0.938769519329071, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2572002679169457, | |
| "grad_norm": 9.77719736840442, | |
| "learning_rate": 4.6347771353493173e-07, | |
| "logits/chosen": -2.9136910438537598, | |
| "logits/rejected": -2.930593252182007, | |
| "logps/chosen": -0.9225813150405884, | |
| "logps/rejected": -0.917222797870636, | |
| "loss": -0.0168, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.9225813150405884, | |
| "rewards/margins": -0.005358507391065359, | |
| "rewards/rejected": -0.917222797870636, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.2679169457468185, | |
| "grad_norm": 10.417139695198527, | |
| "learning_rate": 4.5845133640884217e-07, | |
| "logits/chosen": -2.898656129837036, | |
| "logits/rejected": -2.8996481895446777, | |
| "logps/chosen": -0.9208585619926453, | |
| "logps/rejected": -0.9479470252990723, | |
| "loss": -0.0193, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.9208585619926453, | |
| "rewards/margins": 0.0270883496850729, | |
| "rewards/rejected": -0.9479470252990723, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.27863362357669125, | |
| "grad_norm": 11.697763554027677, | |
| "learning_rate": 4.531320281528169e-07, | |
| "logits/chosen": -2.9303739070892334, | |
| "logits/rejected": -2.9241251945495605, | |
| "logps/chosen": -0.9110285639762878, | |
| "logps/rejected": -0.9726667404174805, | |
| "loss": -0.0227, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.9110285639762878, | |
| "rewards/margins": 0.06163806468248367, | |
| "rewards/rejected": -0.9726667404174805, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.289350301406564, | |
| "grad_norm": 11.144902490144164, | |
| "learning_rate": 4.4752726384955293e-07, | |
| "logits/chosen": -2.9404830932617188, | |
| "logits/rejected": -2.947343111038208, | |
| "logps/chosen": -0.9363792538642883, | |
| "logps/rejected": -0.9744508862495422, | |
| "loss": -0.0195, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9363792538642883, | |
| "rewards/margins": 0.03807152435183525, | |
| "rewards/rejected": -0.9744508862495422, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3000669792364367, | |
| "grad_norm": 11.921950508886521, | |
| "learning_rate": 4.4164491972555294e-07, | |
| "logits/chosen": -2.8976681232452393, | |
| "logits/rejected": -2.898315906524658, | |
| "logps/chosen": -0.9219269752502441, | |
| "logps/rejected": -0.9640507698059082, | |
| "loss": -0.0259, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9219269752502441, | |
| "rewards/margins": 0.04212382063269615, | |
| "rewards/rejected": -0.9640507698059082, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31078365706630945, | |
| "grad_norm": 13.471098814591679, | |
| "learning_rate": 4.3549326208287393e-07, | |
| "logits/chosen": -2.9619617462158203, | |
| "logits/rejected": -2.9274888038635254, | |
| "logps/chosen": -0.9526675343513489, | |
| "logps/rejected": -1.007359266281128, | |
| "loss": -0.0355, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9526675343513489, | |
| "rewards/margins": 0.0546916201710701, | |
| "rewards/rejected": -1.007359266281128, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3215003348961822, | |
| "grad_norm": 12.766870563789471, | |
| "learning_rate": 4.2908093568271187e-07, | |
| "logits/chosen": -2.9606540203094482, | |
| "logits/rejected": -2.9726486206054688, | |
| "logps/chosen": -0.9194633364677429, | |
| "logps/rejected": -0.9709534645080566, | |
| "loss": -0.0371, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.9194633364677429, | |
| "rewards/margins": 0.05149019509553909, | |
| "rewards/rejected": -0.9709534645080566, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3322170127260549, | |
| "grad_norm": 12.28868938106036, | |
| "learning_rate": 4.2241695159714843e-07, | |
| "logits/chosen": -3.043233633041382, | |
| "logits/rejected": -3.0254199504852295, | |
| "logps/chosen": -0.9320791959762573, | |
| "logps/rejected": -0.9708281755447388, | |
| "loss": -0.0274, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9320791959762573, | |
| "rewards/margins": 0.038748957216739655, | |
| "rewards/rejected": -0.9708281755447388, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.34293369055592765, | |
| "grad_norm": 16.212616603609877, | |
| "learning_rate": 4.1551067454613074e-07, | |
| "logits/chosen": -2.943720579147339, | |
| "logits/rejected": -2.928537607192993, | |
| "logps/chosen": -0.921017050743103, | |
| "logps/rejected": -0.9882138967514038, | |
| "loss": -0.0359, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.921017050743103, | |
| "rewards/margins": 0.06719694286584854, | |
| "rewards/rejected": -0.9882138967514038, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.3536503683858004, | |
| "grad_norm": 15.792664568932544, | |
| "learning_rate": 4.0837180973747876e-07, | |
| "logits/chosen": -3.0298409461975098, | |
| "logits/rejected": -3.0069289207458496, | |
| "logps/chosen": -0.9174503087997437, | |
| "logps/rejected": -0.9726656079292297, | |
| "loss": -0.0342, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.9174503087997437, | |
| "rewards/margins": 0.05521535873413086, | |
| "rewards/rejected": -0.9726656079292297, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3643670462156731, | |
| "grad_norm": 10.787104513853434, | |
| "learning_rate": 4.0101038922841455e-07, | |
| "logits/chosen": -2.968522548675537, | |
| "logits/rejected": -2.9739623069763184, | |
| "logps/chosen": -0.9771224856376648, | |
| "logps/rejected": -1.0373847484588623, | |
| "loss": -0.0466, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9771224856376648, | |
| "rewards/margins": 0.06026231497526169, | |
| "rewards/rejected": -1.0373847484588623, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.3750837240455459, | |
| "grad_norm": 13.17830711446757, | |
| "learning_rate": 3.934367578277783e-07, | |
| "logits/chosen": -2.9837470054626465, | |
| "logits/rejected": -2.9880218505859375, | |
| "logps/chosen": -1.000985026359558, | |
| "logps/rejected": -1.0419981479644775, | |
| "loss": -0.0381, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.000985026359558, | |
| "rewards/margins": 0.041013117879629135, | |
| "rewards/rejected": -1.0419981479644775, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.38580040187541864, | |
| "grad_norm": 14.72612824066128, | |
| "learning_rate": 3.856615585587434e-07, | |
| "logits/chosen": -2.9933667182922363, | |
| "logits/rejected": -3.0047719478607178, | |
| "logps/chosen": -1.0653940439224243, | |
| "logps/rejected": -1.1087322235107422, | |
| "loss": -0.0319, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -1.0653940439224243, | |
| "rewards/margins": 0.043338216841220856, | |
| "rewards/rejected": -1.1087322235107422, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3965170797052914, | |
| "grad_norm": 13.397319728202142, | |
| "learning_rate": 3.776957177024581e-07, | |
| "logits/chosen": -2.9812378883361816, | |
| "logits/rejected": -2.9986259937286377, | |
| "logps/chosen": -1.0316323041915894, | |
| "logps/rejected": -1.0616919994354248, | |
| "loss": -0.033, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.0316323041915894, | |
| "rewards/margins": 0.030059661716222763, | |
| "rewards/rejected": -1.0616919994354248, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4072337575351641, | |
| "grad_norm": 14.059564629221503, | |
| "learning_rate": 3.69550429443633e-07, | |
| "logits/chosen": -2.8960227966308594, | |
| "logits/rejected": -2.913637161254883, | |
| "logps/chosen": -1.012019157409668, | |
| "logps/rejected": -1.215237021446228, | |
| "loss": -0.0447, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.012019157409668, | |
| "rewards/margins": 0.20321793854236603, | |
| "rewards/rejected": -1.215237021446228, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.41795043536503684, | |
| "grad_norm": 14.952805148520074, | |
| "learning_rate": 3.612371401396502e-07, | |
| "logits/chosen": -3.007540225982666, | |
| "logits/rejected": -2.9880564212799072, | |
| "logps/chosen": -1.0366661548614502, | |
| "logps/rejected": -1.0943790674209595, | |
| "loss": -0.0368, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.0366661548614502, | |
| "rewards/margins": 0.05771297961473465, | |
| "rewards/rejected": -1.0943790674209595, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4286671131949096, | |
| "grad_norm": 14.170450035952996, | |
| "learning_rate": 3.527675322353011e-07, | |
| "logits/chosen": -2.922602415084839, | |
| "logits/rejected": -2.9457390308380127, | |
| "logps/chosen": -0.970561146736145, | |
| "logps/rejected": -1.0033313035964966, | |
| "loss": -0.0473, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.970561146736145, | |
| "rewards/margins": 0.032770149409770966, | |
| "rewards/rejected": -1.0033313035964966, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.4393837910247823, | |
| "grad_norm": 15.488305590600266, | |
| "learning_rate": 3.4415350784575693e-07, | |
| "logits/chosen": -3.004404067993164, | |
| "logits/rejected": -2.9882898330688477, | |
| "logps/chosen": -0.9613070487976074, | |
| "logps/rejected": -1.0151447057724, | |
| "loss": -0.044, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.9613070487976074, | |
| "rewards/margins": 0.053837649524211884, | |
| "rewards/rejected": -1.0151447057724, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.45010046885465504, | |
| "grad_norm": 13.543647630514888, | |
| "learning_rate": 3.3540717203084224e-07, | |
| "logits/chosen": -2.956808090209961, | |
| "logits/rejected": -2.9622817039489746, | |
| "logps/chosen": -0.9170505404472351, | |
| "logps/rejected": -0.9884069561958313, | |
| "loss": -0.0514, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.9170505404472351, | |
| "rewards/margins": 0.07135634124279022, | |
| "rewards/rejected": -0.9884069561958313, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4608171466845278, | |
| "grad_norm": 14.183913951293327, | |
| "learning_rate": 3.265408157841153e-07, | |
| "logits/chosen": -2.9880080223083496, | |
| "logits/rejected": -2.986445903778076, | |
| "logps/chosen": -0.9918737411499023, | |
| "logps/rejected": -1.0286133289337158, | |
| "loss": -0.0446, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.9918737411499023, | |
| "rewards/margins": 0.036739740520715714, | |
| "rewards/rejected": -1.0286133289337158, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.4715338245144005, | |
| "grad_norm": 16.687524624276737, | |
| "learning_rate": 3.1756689876066187e-07, | |
| "logits/chosen": -2.940664291381836, | |
| "logits/rejected": -2.940556049346924, | |
| "logps/chosen": -0.9756967425346375, | |
| "logps/rejected": -1.0494619607925415, | |
| "loss": -0.0496, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9756967425346375, | |
| "rewards/margins": 0.07376515865325928, | |
| "rewards/rejected": -1.0494619607925415, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4822505023442733, | |
| "grad_norm": 15.106449736541158, | |
| "learning_rate": 3.0849803176787184e-07, | |
| "logits/chosen": -2.978764057159424, | |
| "logits/rejected": -3.02057147026062, | |
| "logps/chosen": -0.9675830602645874, | |
| "logps/rejected": -1.0782067775726318, | |
| "loss": -0.0528, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.9675830602645874, | |
| "rewards/margins": 0.11062371730804443, | |
| "rewards/rejected": -1.0782067775726318, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.49296718017414604, | |
| "grad_norm": 19.732642999044504, | |
| "learning_rate": 2.993469590438065e-07, | |
| "logits/chosen": -2.91894793510437, | |
| "logits/rejected": -2.92021107673645, | |
| "logps/chosen": -0.9777423143386841, | |
| "logps/rejected": -1.0478038787841797, | |
| "loss": -0.0487, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.9777423143386841, | |
| "rewards/margins": 0.07006163895130157, | |
| "rewards/rejected": -1.0478038787841797, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5036838580040187, | |
| "grad_norm": 15.470655040763342, | |
| "learning_rate": 2.901265403480597e-07, | |
| "logits/chosen": -2.868373394012451, | |
| "logits/rejected": -2.8822197914123535, | |
| "logps/chosen": -0.9746626615524292, | |
| "logps/rejected": -1.109392523765564, | |
| "loss": -0.0532, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.9746626615524292, | |
| "rewards/margins": 0.13472989201545715, | |
| "rewards/rejected": -1.109392523765564, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5144005358338914, | |
| "grad_norm": 18.117444556863273, | |
| "learning_rate": 2.808497328902782e-07, | |
| "logits/chosen": -2.829807758331299, | |
| "logits/rejected": -2.837395191192627, | |
| "logps/chosen": -1.0404951572418213, | |
| "logps/rejected": -1.1007030010223389, | |
| "loss": -0.0453, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0404951572418213, | |
| "rewards/margins": 0.06020789220929146, | |
| "rewards/rejected": -1.1007030010223389, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5251172136637642, | |
| "grad_norm": 18.628678925066545, | |
| "learning_rate": 2.715295731217396e-07, | |
| "logits/chosen": -2.81119966506958, | |
| "logits/rejected": -2.815497636795044, | |
| "logps/chosen": -1.0369569063186646, | |
| "logps/rejected": -1.112683892250061, | |
| "loss": -0.0596, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -1.0369569063186646, | |
| "rewards/margins": 0.07572712004184723, | |
| "rewards/rejected": -1.112683892250061, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.535833891493637, | |
| "grad_norm": 17.866763466807367, | |
| "learning_rate": 2.621791584155725e-07, | |
| "logits/chosen": -2.83900785446167, | |
| "logits/rejected": -2.859978199005127, | |
| "logps/chosen": -1.1023390293121338, | |
| "logps/rejected": -1.1515963077545166, | |
| "loss": -0.0617, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1023390293121338, | |
| "rewards/margins": 0.04925721138715744, | |
| "rewards/rejected": -1.1515963077545166, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.5465505693235098, | |
| "grad_norm": 17.478329638335467, | |
| "learning_rate": 2.52811628661367e-07, | |
| "logits/chosen": -2.8983101844787598, | |
| "logits/rejected": -2.908754348754883, | |
| "logps/chosen": -1.0288032293319702, | |
| "logps/rejected": -1.134108066558838, | |
| "loss": -0.0631, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -1.0288032293319702, | |
| "rewards/margins": 0.10530481487512589, | |
| "rewards/rejected": -1.134108066558838, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5572672471533825, | |
| "grad_norm": 15.049601923886751, | |
| "learning_rate": 2.434401478000353e-07, | |
| "logits/chosen": -2.8707144260406494, | |
| "logits/rejected": -2.876131057739258, | |
| "logps/chosen": -1.0516908168792725, | |
| "logps/rejected": -1.1409704685211182, | |
| "loss": -0.0405, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -1.0516908168792725, | |
| "rewards/margins": 0.08927960693836212, | |
| "rewards/rejected": -1.1409704685211182, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5679839249832552, | |
| "grad_norm": 14.714810625194913, | |
| "learning_rate": 2.340778853248759e-07, | |
| "logits/chosen": -2.86578369140625, | |
| "logits/rejected": -2.8567006587982178, | |
| "logps/chosen": -1.0564695596694946, | |
| "logps/rejected": -1.199512004852295, | |
| "loss": -0.0695, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.0564695596694946, | |
| "rewards/margins": 0.143042653799057, | |
| "rewards/rejected": -1.199512004852295, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.578700602813128, | |
| "grad_norm": 18.053080062016626, | |
| "learning_rate": 2.247379977748339e-07, | |
| "logits/chosen": -2.9022586345672607, | |
| "logits/rejected": -2.9201228618621826, | |
| "logps/chosen": -1.0403499603271484, | |
| "logps/rejected": -1.1265124082565308, | |
| "loss": -0.0673, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.0403499603271484, | |
| "rewards/margins": 0.08616244792938232, | |
| "rewards/rejected": -1.1265124082565308, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5894172806430007, | |
| "grad_norm": 14.15566071309189, | |
| "learning_rate": 2.1543361024596573e-07, | |
| "logits/chosen": -2.9802463054656982, | |
| "logits/rejected": -2.988959550857544, | |
| "logps/chosen": -0.9800962209701538, | |
| "logps/rejected": -1.0803120136260986, | |
| "loss": -0.0525, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9800962209701538, | |
| "rewards/margins": 0.1002158373594284, | |
| "rewards/rejected": -1.0803120136260986, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.6001339584728734, | |
| "grad_norm": 17.56629547418189, | |
| "learning_rate": 2.0617779794709082e-07, | |
| "logits/chosen": -2.8620638847351074, | |
| "logits/rejected": -2.8629186153411865, | |
| "logps/chosen": -1.103073000907898, | |
| "logps/rejected": -1.196608304977417, | |
| "loss": -0.0632, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.103073000907898, | |
| "rewards/margins": 0.09353531897068024, | |
| "rewards/rejected": -1.196608304977417, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6108506363027462, | |
| "grad_norm": 15.46761530330997, | |
| "learning_rate": 1.9698356782554687e-07, | |
| "logits/chosen": -2.9556548595428467, | |
| "logits/rejected": -2.9529995918273926, | |
| "logps/chosen": -1.109263300895691, | |
| "logps/rejected": -1.1731650829315186, | |
| "loss": -0.0564, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.109263300895691, | |
| "rewards/margins": 0.06390184909105301, | |
| "rewards/rejected": -1.1731650829315186, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6215673141326189, | |
| "grad_norm": 16.8037567850014, | |
| "learning_rate": 1.878638402888727e-07, | |
| "logits/chosen": -2.9940178394317627, | |
| "logits/rejected": -2.9970099925994873, | |
| "logps/chosen": -0.9795197248458862, | |
| "logps/rejected": -1.0418453216552734, | |
| "loss": -0.0668, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.9795197248458862, | |
| "rewards/margins": 0.06232555955648422, | |
| "rewards/rejected": -1.0418453216552734, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6322839919624916, | |
| "grad_norm": 16.60747046570722, | |
| "learning_rate": 1.7883143104810223e-07, | |
| "logits/chosen": -2.9415085315704346, | |
| "logits/rejected": -2.9544739723205566, | |
| "logps/chosen": -1.069349765777588, | |
| "logps/rejected": -1.2494323253631592, | |
| "loss": -0.0739, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.069349765777588, | |
| "rewards/margins": 0.1800827533006668, | |
| "rewards/rejected": -1.2494323253631592, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6430006697923644, | |
| "grad_norm": 15.422934564351497, | |
| "learning_rate": 1.6989903310818548e-07, | |
| "logits/chosen": -3.0317275524139404, | |
| "logits/rejected": -3.049278736114502, | |
| "logps/chosen": -1.0510863065719604, | |
| "logps/rejected": -1.1395753622055054, | |
| "loss": -0.0517, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.0510863065719604, | |
| "rewards/margins": 0.08848897367715836, | |
| "rewards/rejected": -1.1395753622055054, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6537173476222371, | |
| "grad_norm": 17.56071364411128, | |
| "learning_rate": 1.610791989308466e-07, | |
| "logits/chosen": -2.9625232219696045, | |
| "logits/rejected": -2.9480719566345215, | |
| "logps/chosen": -1.0810691118240356, | |
| "logps/rejected": -1.1474703550338745, | |
| "loss": -0.0711, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.0810691118240356, | |
| "rewards/margins": 0.06640125066041946, | |
| "rewards/rejected": -1.1474703550338745, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6644340254521098, | |
| "grad_norm": 19.400175535223255, | |
| "learning_rate": 1.523843227949418e-07, | |
| "logits/chosen": -3.0158638954162598, | |
| "logits/rejected": -3.014029026031494, | |
| "logps/chosen": -1.054434061050415, | |
| "logps/rejected": -1.1792142391204834, | |
| "loss": -0.0711, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.054434061050415, | |
| "rewards/margins": 0.12478007376194, | |
| "rewards/rejected": -1.1792142391204834, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6751507032819826, | |
| "grad_norm": 18.063066961906372, | |
| "learning_rate": 1.4382662337910867e-07, | |
| "logits/chosen": -2.9173102378845215, | |
| "logits/rejected": -2.9165103435516357, | |
| "logps/chosen": -1.1207153797149658, | |
| "logps/rejected": -1.2241191864013672, | |
| "loss": -0.0582, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.1207153797149658, | |
| "rewards/margins": 0.10340368747711182, | |
| "rewards/rejected": -1.2241191864013672, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6858673811118553, | |
| "grad_norm": 17.656998709658712, | |
| "learning_rate": 1.3541812659118104e-07, | |
| "logits/chosen": -2.924173593521118, | |
| "logits/rejected": -2.901674747467041, | |
| "logps/chosen": -1.1284511089324951, | |
| "logps/rejected": -1.2347378730773926, | |
| "loss": -0.0605, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1284511089324951, | |
| "rewards/margins": 0.10628668963909149, | |
| "rewards/rejected": -1.2347378730773926, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.696584058941728, | |
| "grad_norm": 17.674353062779463, | |
| "learning_rate": 1.2717064866850047e-07, | |
| "logits/chosen": -2.9069221019744873, | |
| "logits/rejected": -2.9266703128814697, | |
| "logps/chosen": -1.0635308027267456, | |
| "logps/rejected": -1.2443645000457764, | |
| "loss": -0.076, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0635308027267456, | |
| "rewards/margins": 0.1808338761329651, | |
| "rewards/rejected": -1.2443645000457764, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.7073007367716008, | |
| "grad_norm": 16.692003983984833, | |
| "learning_rate": 1.1909577957287117e-07, | |
| "logits/chosen": -2.9482550621032715, | |
| "logits/rejected": -2.9383444786071777, | |
| "logps/chosen": -1.1115998029708862, | |
| "logps/rejected": -1.269691824913025, | |
| "loss": -0.0875, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.1115998029708862, | |
| "rewards/margins": 0.15809203684329987, | |
| "rewards/rejected": -1.269691824913025, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7180174146014735, | |
| "grad_norm": 16.832132437998734, | |
| "learning_rate": 1.1120486670349441e-07, | |
| "logits/chosen": -3.0022974014282227, | |
| "logits/rejected": -3.0084285736083984, | |
| "logps/chosen": -1.1076433658599854, | |
| "logps/rejected": -1.2158708572387695, | |
| "loss": -0.0631, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1076433658599854, | |
| "rewards/margins": 0.10822737216949463, | |
| "rewards/rejected": -1.2158708572387695, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7287340924313462, | |
| "grad_norm": 16.746834458412387, | |
| "learning_rate": 1.0350899895076948e-07, | |
| "logits/chosen": -2.960242509841919, | |
| "logits/rejected": -2.9590137004852295, | |
| "logps/chosen": -1.1618969440460205, | |
| "logps/rejected": -1.3054715394973755, | |
| "loss": -0.063, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1618969440460205, | |
| "rewards/margins": 0.14357443153858185, | |
| "rewards/rejected": -1.3054715394973755, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.739450770261219, | |
| "grad_norm": 16.472011449839883, | |
| "learning_rate": 9.601899111337069e-08, | |
| "logits/chosen": -2.9834091663360596, | |
| "logits/rejected": -2.9975457191467285, | |
| "logps/chosen": -1.126527190208435, | |
| "logps/rejected": -1.3144327402114868, | |
| "loss": -0.0777, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -1.126527190208435, | |
| "rewards/margins": 0.18790560960769653, | |
| "rewards/rejected": -1.3144327402114868, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7501674480910918, | |
| "grad_norm": 17.215572295903478, | |
| "learning_rate": 8.874536870049765e-08, | |
| "logits/chosen": -2.969423294067383, | |
| "logits/rejected": -2.969230890274048, | |
| "logps/chosen": -1.127826452255249, | |
| "logps/rejected": -1.271470069885254, | |
| "loss": -0.0752, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.127826452255249, | |
| "rewards/margins": 0.1436435580253601, | |
| "rewards/rejected": -1.271470069885254, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7608841259209645, | |
| "grad_norm": 19.033752258043222, | |
| "learning_rate": 8.169835314065663e-08, | |
| "logits/chosen": -2.9632205963134766, | |
| "logits/rejected": -2.9548847675323486, | |
| "logps/chosen": -1.1820392608642578, | |
| "logps/rejected": -1.2698676586151123, | |
| "loss": -0.0668, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -1.1820392608642578, | |
| "rewards/margins": 0.08782847970724106, | |
| "rewards/rejected": -1.2698676586151123, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7716008037508373, | |
| "grad_norm": 17.064795493163807, | |
| "learning_rate": 7.488784741775816e-08, | |
| "logits/chosen": -2.976466178894043, | |
| "logits/rejected": -2.9713473320007324, | |
| "logps/chosen": -1.054131269454956, | |
| "logps/rejected": -1.1566834449768066, | |
| "loss": -0.0663, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.054131269454956, | |
| "rewards/margins": 0.10255217552185059, | |
| "rewards/rejected": -1.1566834449768066, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.78231748158071, | |
| "grad_norm": 17.868156711335107, | |
| "learning_rate": 6.832342215471715e-08, | |
| "logits/chosen": -2.98659086227417, | |
| "logits/rejected": -2.986912488937378, | |
| "logps/chosen": -1.1025737524032593, | |
| "logps/rejected": -1.242219090461731, | |
| "loss": -0.0823, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.1025737524032593, | |
| "rewards/margins": 0.13964524865150452, | |
| "rewards/rejected": -1.242219090461731, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7930341594105828, | |
| "grad_norm": 24.714287844140323, | |
| "learning_rate": 6.201430216410985e-08, | |
| "logits/chosen": -3.0034446716308594, | |
| "logits/rejected": -3.001354217529297, | |
| "logps/chosen": -1.1147871017456055, | |
| "logps/rejected": -1.200984239578247, | |
| "loss": -0.081, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.1147871017456055, | |
| "rewards/margins": 0.08619704097509384, | |
| "rewards/rejected": -1.200984239578247, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.8037508372404555, | |
| "grad_norm": 20.89166421346418, | |
| "learning_rate": 5.596935348479029e-08, | |
| "logits/chosen": -2.9388105869293213, | |
| "logits/rejected": -2.948521137237549, | |
| "logps/chosen": -1.2442572116851807, | |
| "logps/rejected": -1.393157958984375, | |
| "loss": -0.0861, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.2442572116851807, | |
| "rewards/margins": 0.1489005833864212, | |
| "rewards/rejected": -1.393157958984375, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.8144675150703282, | |
| "grad_norm": 22.406198534031674, | |
| "learning_rate": 5.01970709226808e-08, | |
| "logits/chosen": -3.028179883956909, | |
| "logits/rejected": -3.009589672088623, | |
| "logps/chosen": -1.0915172100067139, | |
| "logps/rejected": -1.243607997894287, | |
| "loss": -0.0837, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0915172100067139, | |
| "rewards/margins": 0.1520906686782837, | |
| "rewards/rejected": -1.243607997894287, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.825184192900201, | |
| "grad_norm": 24.898962036749563, | |
| "learning_rate": 4.4705566113246635e-08, | |
| "logits/chosen": -2.9715943336486816, | |
| "logits/rejected": -2.949638605117798, | |
| "logps/chosen": -1.228461503982544, | |
| "logps/rejected": -1.3865272998809814, | |
| "loss": -0.0897, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.228461503982544, | |
| "rewards/margins": 0.15806564688682556, | |
| "rewards/rejected": -1.3865272998809814, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.8359008707300737, | |
| "grad_norm": 19.449984589836117, | |
| "learning_rate": 3.950255612243042e-08, | |
| "logits/chosen": -2.951904058456421, | |
| "logits/rejected": -2.928208827972412, | |
| "logps/chosen": -1.2001017332077026, | |
| "logps/rejected": -1.3440535068511963, | |
| "loss": -0.0724, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.2001017332077026, | |
| "rewards/margins": 0.1439516693353653, | |
| "rewards/rejected": -1.3440535068511963, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8466175485599464, | |
| "grad_norm": 19.857222600639766, | |
| "learning_rate": 3.459535260206384e-08, | |
| "logits/chosen": -2.9977169036865234, | |
| "logits/rejected": -3.020946979522705, | |
| "logps/chosen": -1.096103549003601, | |
| "logps/rejected": -1.21526300907135, | |
| "loss": -0.0898, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.096103549003601, | |
| "rewards/margins": 0.11915931850671768, | |
| "rewards/rejected": -1.21526300907135, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8573342263898192, | |
| "grad_norm": 21.73532568075866, | |
| "learning_rate": 2.999085151499764e-08, | |
| "logits/chosen": -2.9534425735473633, | |
| "logits/rejected": -2.9359939098358154, | |
| "logps/chosen": -1.0908204317092896, | |
| "logps/rejected": -1.356081485748291, | |
| "loss": -0.0905, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.0908204317092896, | |
| "rewards/margins": 0.2652612626552582, | |
| "rewards/rejected": -1.356081485748291, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8573342263898192, | |
| "eval_logits/chosen": -2.8999884128570557, | |
| "eval_logits/rejected": -2.892754316329956, | |
| "eval_logps/chosen": -1.1346372365951538, | |
| "eval_logps/rejected": -1.2466005086898804, | |
| "eval_loss": -0.07080772519111633, | |
| "eval_rewards/accuracies": 0.5757978558540344, | |
| "eval_rewards/chosen": -1.1346372365951538, | |
| "eval_rewards/margins": 0.11196334660053253, | |
| "eval_rewards/rejected": -1.2466005086898804, | |
| "eval_runtime": 432.4273, | |
| "eval_samples_per_second": 6.924, | |
| "eval_steps_per_second": 0.435, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8680509042196919, | |
| "grad_norm": 19.440021908588918, | |
| "learning_rate": 2.5695523444387747e-08, | |
| "logits/chosen": -2.9388387203216553, | |
| "logits/rejected": -2.9628615379333496, | |
| "logps/chosen": -1.111882209777832, | |
| "logps/rejected": -1.2351758480072021, | |
| "loss": -0.09, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.111882209777832, | |
| "rewards/margins": 0.12329357862472534, | |
| "rewards/rejected": -1.2351758480072021, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8787675820495646, | |
| "grad_norm": 19.454852119565878, | |
| "learning_rate": 2.1715404500756875e-08, | |
| "logits/chosen": -2.98136305809021, | |
| "logits/rejected": -3.0059518814086914, | |
| "logps/chosen": -1.090047836303711, | |
| "logps/rejected": -1.203460693359375, | |
| "loss": -0.0823, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.090047836303711, | |
| "rewards/margins": 0.11341279745101929, | |
| "rewards/rejected": -1.203460693359375, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8894842598794374, | |
| "grad_norm": 27.119138813457198, | |
| "learning_rate": 1.8056087839608546e-08, | |
| "logits/chosen": -2.9424166679382324, | |
| "logits/rejected": -2.9254748821258545, | |
| "logps/chosen": -1.1260377168655396, | |
| "logps/rejected": -1.3184921741485596, | |
| "loss": -0.1086, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.1260377168655396, | |
| "rewards/margins": 0.19245454668998718, | |
| "rewards/rejected": -1.3184921741485596, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.9002009377093101, | |
| "grad_norm": 19.3060602023721, | |
| "learning_rate": 1.4722715801513935e-08, | |
| "logits/chosen": -2.9605276584625244, | |
| "logits/rejected": -2.967475414276123, | |
| "logps/chosen": -1.0344493389129639, | |
| "logps/rejected": -1.0916310548782349, | |
| "loss": -0.0802, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.0344493389129639, | |
| "rewards/margins": 0.057181768119335175, | |
| "rewards/rejected": -1.0916310548782349, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9109176155391828, | |
| "grad_norm": 20.048601077954377, | |
| "learning_rate": 1.1719972685717395e-08, | |
| "logits/chosen": -2.970942258834839, | |
| "logits/rejected": -2.9757351875305176, | |
| "logps/chosen": -1.0653189420700073, | |
| "logps/rejected": -1.3184258937835693, | |
| "loss": -0.1043, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0653189420700073, | |
| "rewards/margins": 0.2531070411205292, | |
| "rewards/rejected": -1.3184258937835693, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9216342933690556, | |
| "grad_norm": 21.66418470338373, | |
| "learning_rate": 9.05207816741485e-09, | |
| "logits/chosen": -2.939943313598633, | |
| "logits/rejected": -2.9273343086242676, | |
| "logps/chosen": -1.2503459453582764, | |
| "logps/rejected": -1.4009029865264893, | |
| "loss": -0.1013, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.2503459453582764, | |
| "rewards/margins": 0.15055687725543976, | |
| "rewards/rejected": -1.4009029865264893, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9323509711989283, | |
| "grad_norm": 21.774320932145653, | |
| "learning_rate": 6.722781367956481e-09, | |
| "logits/chosen": -3.0610954761505127, | |
| "logits/rejected": -3.0484843254089355, | |
| "logps/chosen": -1.0789310932159424, | |
| "logps/rejected": -1.2742266654968262, | |
| "loss": -0.1009, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.0789310932159424, | |
| "rewards/margins": 0.19529545307159424, | |
| "rewards/rejected": -1.2742266654968262, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.943067649028801, | |
| "grad_norm": 19.319233646112725, | |
| "learning_rate": 4.735355586305861e-09, | |
| "logits/chosen": -2.94116473197937, | |
| "logits/rejected": -2.949489116668701, | |
| "logps/chosen": -1.1311179399490356, | |
| "logps/rejected": -1.3173315525054932, | |
| "loss": -0.0993, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.1311179399490356, | |
| "rewards/margins": 0.1862136423587799, | |
| "rewards/rejected": -1.3173315525054932, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9537843268586738, | |
| "grad_norm": 19.556600822562356, | |
| "learning_rate": 3.0925936991598454e-09, | |
| "logits/chosen": -2.9415977001190186, | |
| "logits/rejected": -2.944352388381958, | |
| "logps/chosen": -1.1981102228164673, | |
| "logps/rejected": -1.329925775527954, | |
| "loss": -0.0746, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1981102228164673, | |
| "rewards/margins": 0.1318153738975525, | |
| "rewards/rejected": -1.329925775527954, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.9645010046885466, | |
| "grad_norm": 24.43813858171124, | |
| "learning_rate": 1.7968042361930225e-09, | |
| "logits/chosen": -2.958519697189331, | |
| "logits/rejected": -2.9468445777893066, | |
| "logps/chosen": -1.1451342105865479, | |
| "logps/rejected": -1.265005111694336, | |
| "loss": -0.0894, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.1451342105865479, | |
| "rewards/margins": 0.11987074464559555, | |
| "rewards/rejected": -1.265005111694336, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9752176825184193, | |
| "grad_norm": 23.495311684580507, | |
| "learning_rate": 8.498081359423759e-10, | |
| "logits/chosen": -2.9030869007110596, | |
| "logits/rejected": -2.932225465774536, | |
| "logps/chosen": -1.1186686754226685, | |
| "logps/rejected": -1.335756540298462, | |
| "loss": -0.1008, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.1186686754226685, | |
| "rewards/margins": 0.21708789467811584, | |
| "rewards/rejected": -1.335756540298462, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9859343603482921, | |
| "grad_norm": 20.358552883416746, | |
| "learning_rate": 2.5293618689056774e-10, | |
| "logits/chosen": -2.9370243549346924, | |
| "logits/rejected": -2.953622341156006, | |
| "logps/chosen": -1.1718332767486572, | |
| "logps/rejected": -1.4025189876556396, | |
| "loss": -0.0805, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.1718332767486572, | |
| "rewards/margins": 0.230685755610466, | |
| "rewards/rejected": -1.4025189876556396, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9966510381781648, | |
| "grad_norm": 27.657325299971138, | |
| "learning_rate": 7.02715734424153e-12, | |
| "logits/chosen": -2.9429261684417725, | |
| "logits/rejected": -2.951002359390259, | |
| "logps/chosen": -1.2534146308898926, | |
| "logps/rejected": -1.3702634572982788, | |
| "loss": -0.0688, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.2534146308898926, | |
| "rewards/margins": 0.11684894561767578, | |
| "rewards/rejected": -1.3702634572982788, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.9987943737441393, | |
| "step": 466, | |
| "total_flos": 0.0, | |
| "train_loss": 0.0, | |
| "train_runtime": 4.3106, | |
| "train_samples_per_second": 13854.197, | |
| "train_steps_per_second": 108.105 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 466, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |