| { |
| "best_global_step": 300, |
| "best_metric": 0.54357338, |
| "best_model_checkpoint": "/data2/kdd/crag/cjz/output/task1+3_dpo_absgt14_noidk__pari0.8_0608/v0-20250610-025054/checkpoint-300", |
| "epoch": 1.8009032743695896, |
| "eval_steps": 300, |
| "global_step": 300, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.0060218291305984195, |
| "grad_norm": 7.15871000289917, |
| "learning_rate": 1.1764705882352942e-05, |
| "logits/chosen": -0.32459306716918945, |
| "logits/rejected": -0.5000401735305786, |
| "logps/chosen": -3.3705074787139893, |
| "logps/rejected": -19.54801368713379, |
| "loss": 1.1142030954360962, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4210559129714966, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 1, |
| "train_speed(iter/s)": 0.005428 |
| }, |
| { |
| "epoch": 0.012043658261196839, |
| "grad_norm": 6.311911106109619, |
| "learning_rate": 2.3529411764705884e-05, |
| "logits/chosen": -0.41802966594696045, |
| "logits/rejected": -0.5096024870872498, |
| "logps/chosen": -8.828028678894043, |
| "logps/rejected": -13.557395935058594, |
| "loss": 1.4290869235992432, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.7359397411346436, |
| "rewards/accuracies": 0.0, |
| "rewards/chosen": 0.0, |
| "rewards/margins": 0.0, |
| "rewards/rejected": 0.0, |
| "step": 2, |
| "train_speed(iter/s)": 0.005545 |
| }, |
| { |
| "epoch": 0.018065487391795258, |
| "grad_norm": 6.693342685699463, |
| "learning_rate": 3.529411764705883e-05, |
| "logits/chosen": -0.3694347143173218, |
| "logits/rejected": -0.4798021912574768, |
| "logps/chosen": -6.542965412139893, |
| "logps/rejected": -17.62507438659668, |
| "loss": 1.2220425605773926, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5347560048103333, |
| "rewards/accuracies": 0.734375, |
| "rewards/chosen": 0.007337508723139763, |
| "rewards/margins": 0.011953208595514297, |
| "rewards/rejected": -0.004615699406713247, |
| "step": 3, |
| "train_speed(iter/s)": 0.005572 |
| }, |
| { |
| "epoch": 0.024087316522393678, |
| "grad_norm": 4.8458251953125, |
| "learning_rate": 4.705882352941177e-05, |
| "logits/chosen": -0.36237555742263794, |
| "logits/rejected": -0.5201014280319214, |
| "logps/chosen": -5.53680944442749, |
| "logps/rejected": -16.475929260253906, |
| "loss": 1.2521781921386719, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.579364538192749, |
| "rewards/accuracies": 0.734375, |
| "rewards/chosen": 0.01270484272390604, |
| "rewards/margins": 0.04416074603796005, |
| "rewards/rejected": -0.03145590424537659, |
| "step": 4, |
| "train_speed(iter/s)": 0.005594 |
| }, |
| { |
| "epoch": 0.030109145652992095, |
| "grad_norm": 3.684136152267456, |
| "learning_rate": 5.882352941176471e-05, |
| "logits/chosen": -0.3457816541194916, |
| "logits/rejected": -0.48865967988967896, |
| "logps/chosen": -4.437034606933594, |
| "logps/rejected": -20.260162353515625, |
| "loss": 0.9991621375083923, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.38671040534973145, |
| "rewards/accuracies": 0.734375, |
| "rewards/chosen": 0.002155877649784088, |
| "rewards/margins": 0.2208954095840454, |
| "rewards/rejected": -0.21873953938484192, |
| "step": 5, |
| "train_speed(iter/s)": 0.005606 |
| }, |
| { |
| "epoch": 0.036130974783590515, |
| "grad_norm": 6.8094706535339355, |
| "learning_rate": 7.058823529411765e-05, |
| "logits/chosen": -0.37496131658554077, |
| "logits/rejected": -0.499918133020401, |
| "logps/chosen": -5.312520980834961, |
| "logps/rejected": -18.596221923828125, |
| "loss": 1.1320774555206299, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5494582056999207, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": -0.11241161823272705, |
| "rewards/margins": 0.37036794424057007, |
| "rewards/rejected": -0.4827795624732971, |
| "step": 6, |
| "train_speed(iter/s)": 0.005617 |
| }, |
| { |
| "epoch": 0.042152803914188935, |
| "grad_norm": 6.563366889953613, |
| "learning_rate": 8.23529411764706e-05, |
| "logits/chosen": -0.3803170323371887, |
| "logits/rejected": -0.43647128343582153, |
| "logps/chosen": -12.155686378479004, |
| "logps/rejected": -19.252357482910156, |
| "loss": 1.275801420211792, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6294467449188232, |
| "rewards/accuracies": 0.609375, |
| "rewards/chosen": -0.2581723928451538, |
| "rewards/margins": 0.26009732484817505, |
| "rewards/rejected": -0.5182697176933289, |
| "step": 7, |
| "train_speed(iter/s)": 0.005615 |
| }, |
| { |
| "epoch": 0.048174633044787356, |
| "grad_norm": 5.275461673736572, |
| "learning_rate": 9.411764705882353e-05, |
| "logits/chosen": -0.30787912011146545, |
| "logits/rejected": -0.420376718044281, |
| "logps/chosen": -7.271793842315674, |
| "logps/rejected": -17.376087188720703, |
| "loss": 1.2506061792373657, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6039291024208069, |
| "rewards/accuracies": 0.609375, |
| "rewards/chosen": -0.07998734712600708, |
| "rewards/margins": 0.2555825710296631, |
| "rewards/rejected": -0.3355698883533478, |
| "step": 8, |
| "train_speed(iter/s)": 0.005615 |
| }, |
| { |
| "epoch": 0.054196462175385776, |
| "grad_norm": 3.0540287494659424, |
| "learning_rate": 0.00010588235294117647, |
| "logits/chosen": -0.30379173159599304, |
| "logits/rejected": -0.41216912865638733, |
| "logps/chosen": -6.820768356323242, |
| "logps/rejected": -16.3853816986084, |
| "loss": 1.1634117364883423, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5367689728736877, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": 0.04989926144480705, |
| "rewards/margins": 0.1963551640510559, |
| "rewards/rejected": -0.14645591378211975, |
| "step": 9, |
| "train_speed(iter/s)": 0.005615 |
| }, |
| { |
| "epoch": 0.06021829130598419, |
| "grad_norm": 2.3425395488739014, |
| "learning_rate": 0.00011764705882352942, |
| "logits/chosen": -0.3369804918766022, |
| "logits/rejected": -0.4523884356021881, |
| "logps/chosen": -7.222663402557373, |
| "logps/rejected": -17.98470687866211, |
| "loss": 1.0803574323654175, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4635474979877472, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.0936385914683342, |
| "rewards/margins": 0.19075465202331543, |
| "rewards/rejected": -0.09711606800556183, |
| "step": 10, |
| "train_speed(iter/s)": 0.005616 |
| }, |
| { |
| "epoch": 0.06624012043658262, |
| "grad_norm": 2.2927637100219727, |
| "learning_rate": 0.00012941176470588237, |
| "logits/chosen": -0.37337860465049744, |
| "logits/rejected": -0.4716382920742035, |
| "logps/chosen": -8.593006134033203, |
| "logps/rejected": -17.979745864868164, |
| "loss": 1.1164621114730835, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.49796244502067566, |
| "rewards/accuracies": 0.625, |
| "rewards/chosen": 0.06070924177765846, |
| "rewards/margins": 0.19886666536331177, |
| "rewards/rejected": -0.1381574273109436, |
| "step": 11, |
| "train_speed(iter/s)": 0.005619 |
| }, |
| { |
| "epoch": 0.07226194956718103, |
| "grad_norm": 2.380297899246216, |
| "learning_rate": 0.0001411764705882353, |
| "logits/chosen": -0.4056618809700012, |
| "logits/rejected": -0.47582757472991943, |
| "logps/chosen": -7.429853916168213, |
| "logps/rejected": -16.317821502685547, |
| "loss": 1.167059302330017, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5555555820465088, |
| "rewards/accuracies": 0.65625, |
| "rewards/chosen": -0.0010902846697717905, |
| "rewards/margins": 0.21717122197151184, |
| "rewards/rejected": -0.2182615101337433, |
| "step": 12, |
| "train_speed(iter/s)": 0.005622 |
| }, |
| { |
| "epoch": 0.07828377869777944, |
| "grad_norm": 1.7392877340316772, |
| "learning_rate": 0.00015294117647058822, |
| "logits/chosen": -0.4110736846923828, |
| "logits/rejected": -0.4770917296409607, |
| "logps/chosen": -8.422821044921875, |
| "logps/rejected": -16.356956481933594, |
| "loss": 1.1002569198608398, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5357962250709534, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.12992499768733978, |
| "rewards/margins": 0.3372637927532196, |
| "rewards/rejected": -0.20733879506587982, |
| "step": 13, |
| "train_speed(iter/s)": 0.005625 |
| }, |
| { |
| "epoch": 0.08430560782837787, |
| "grad_norm": 5.562183856964111, |
| "learning_rate": 0.0001647058823529412, |
| "logits/chosen": -0.32676243782043457, |
| "logits/rejected": -0.4272843599319458, |
| "logps/chosen": -8.393424034118652, |
| "logps/rejected": -19.974212646484375, |
| "loss": 1.156798005104065, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6415989995002747, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": -0.03989575803279877, |
| "rewards/margins": 0.5587274432182312, |
| "rewards/rejected": -0.5986231565475464, |
| "step": 14, |
| "train_speed(iter/s)": 0.005627 |
| }, |
| { |
| "epoch": 0.09032743695897628, |
| "grad_norm": 3.5578579902648926, |
| "learning_rate": 0.00017647058823529413, |
| "logits/chosen": -0.36884137988090515, |
| "logits/rejected": -0.47317999601364136, |
| "logps/chosen": -6.998829364776611, |
| "logps/rejected": -18.39483642578125, |
| "loss": 1.265438199043274, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.7606177926063538, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.002129599452018738, |
| "rewards/margins": 0.5497386455535889, |
| "rewards/rejected": -0.5476090908050537, |
| "step": 15, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.09634926608957471, |
| "grad_norm": 3.6414742469787598, |
| "learning_rate": 0.00018823529411764707, |
| "logits/chosen": -0.3789006471633911, |
| "logits/rejected": -0.4743453860282898, |
| "logps/chosen": -6.895637512207031, |
| "logps/rejected": -20.256675720214844, |
| "loss": 1.064023494720459, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5308047533035278, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.055957891047000885, |
| "rewards/margins": 0.40152865648269653, |
| "rewards/rejected": -0.34557074308395386, |
| "step": 16, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.10237109522017313, |
| "grad_norm": 3.263599157333374, |
| "learning_rate": 0.0002, |
| "logits/chosen": -0.28460752964019775, |
| "logits/rejected": -0.42314791679382324, |
| "logps/chosen": -5.295862197875977, |
| "logps/rejected": -21.666519165039062, |
| "loss": 0.9750788807868958, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5031211376190186, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.08322687447071075, |
| "rewards/margins": 0.5906113386154175, |
| "rewards/rejected": -0.5073844194412231, |
| "step": 17, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.10839292435077155, |
| "grad_norm": 2.0240275859832764, |
| "learning_rate": 0.00019999502669559432, |
| "logits/chosen": -0.3283552825450897, |
| "logits/rejected": -0.40596985816955566, |
| "logps/chosen": -6.424624919891357, |
| "logps/rejected": -17.71906852722168, |
| "loss": 0.9781989455223083, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.47217291593551636, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": -0.0207663644105196, |
| "rewards/margins": 0.5476341247558594, |
| "rewards/rejected": -0.568400502204895, |
| "step": 18, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.11441475348136997, |
| "grad_norm": 1.682889699935913, |
| "learning_rate": 0.00019998010727705236, |
| "logits/chosen": -0.2821420729160309, |
| "logits/rejected": -0.38918256759643555, |
| "logps/chosen": -5.882609844207764, |
| "logps/rejected": -20.25113296508789, |
| "loss": 0.9472936391830444, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4797293543815613, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.06502628326416016, |
| "rewards/margins": 0.6772950291633606, |
| "rewards/rejected": -0.6122686862945557, |
| "step": 19, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.12043658261196838, |
| "grad_norm": 2.9386143684387207, |
| "learning_rate": 0.00019995524322835034, |
| "logits/chosen": -0.2649421691894531, |
| "logits/rejected": -0.34010952711105347, |
| "logps/chosen": -10.95669174194336, |
| "logps/rejected": -24.044157028198242, |
| "loss": 1.1134816408157349, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.546178936958313, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": -0.18016880750656128, |
| "rewards/margins": 0.5494988560676575, |
| "rewards/rejected": -0.729667603969574, |
| "step": 20, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.1264584117425668, |
| "grad_norm": 1.6700323820114136, |
| "learning_rate": 0.00019992043702261793, |
| "logits/chosen": -0.2561277151107788, |
| "logits/rejected": -0.3126121163368225, |
| "logps/chosen": -6.485073566436768, |
| "logps/rejected": -19.772449493408203, |
| "loss": 0.9205772876739502, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.42214635014533997, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": -0.0028960183262825012, |
| "rewards/margins": 0.6638913750648499, |
| "rewards/rejected": -0.6667873859405518, |
| "step": 21, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.13248024087316523, |
| "grad_norm": 2.0114383697509766, |
| "learning_rate": 0.00019987569212189224, |
| "logits/chosen": -0.2601704001426697, |
| "logits/rejected": -0.3789623975753784, |
| "logps/chosen": -6.645152568817139, |
| "logps/rejected": -21.876020431518555, |
| "loss": 1.1017377376556396, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6103405356407166, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.04465393349528313, |
| "rewards/margins": 0.6134660840034485, |
| "rewards/rejected": -0.5688121318817139, |
| "step": 22, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.13850207000376363, |
| "grad_norm": 3.031888961791992, |
| "learning_rate": 0.0001998210129767735, |
| "logits/chosen": -0.26405903697013855, |
| "logits/rejected": -0.3669392466545105, |
| "logps/chosen": -6.921745300292969, |
| "logps/rejected": -21.70663070678711, |
| "loss": 0.9795225858688354, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4959058463573456, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.10062012821435928, |
| "rewards/margins": 0.6365569829940796, |
| "rewards/rejected": -0.5359368324279785, |
| "step": 23, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.14452389913436206, |
| "grad_norm": 2.15742564201355, |
| "learning_rate": 0.00019975640502598244, |
| "logits/chosen": -0.2354922890663147, |
| "logits/rejected": -0.36152032017707825, |
| "logps/chosen": -4.718395709991455, |
| "logps/rejected": -24.57033348083496, |
| "loss": 0.7706338763237, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3560357987880707, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.09812817722558975, |
| "rewards/margins": 0.9238024950027466, |
| "rewards/rejected": -0.825674295425415, |
| "step": 24, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.1505457282649605, |
| "grad_norm": 3.1512320041656494, |
| "learning_rate": 0.0001996818746958191, |
| "logits/chosen": -0.21456243097782135, |
| "logits/rejected": -0.34130796790122986, |
| "logps/chosen": -6.996722221374512, |
| "logps/rejected": -30.622554779052734, |
| "loss": 0.8399394750595093, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.49700063467025757, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.024456650018692017, |
| "rewards/margins": 1.4392318725585938, |
| "rewards/rejected": -1.4636887311935425, |
| "step": 25, |
| "train_speed(iter/s)": 0.005628 |
| }, |
| { |
| "epoch": 0.1565675573955589, |
| "grad_norm": 2.9886889457702637, |
| "learning_rate": 0.00019959742939952392, |
| "logits/chosen": -0.2622988224029541, |
| "logits/rejected": -0.3251183331012726, |
| "logps/chosen": -12.495084762573242, |
| "logps/rejected": -25.176090240478516, |
| "loss": 1.1550263166427612, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.7094165086746216, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.19754531979560852, |
| "rewards/margins": 1.1228585243225098, |
| "rewards/rejected": -1.3204039335250854, |
| "step": 26, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.16258938652615731, |
| "grad_norm": 2.7348685264587402, |
| "learning_rate": 0.00019950307753654017, |
| "logits/chosen": -0.23797279596328735, |
| "logits/rejected": -0.3161388337612152, |
| "logps/chosen": -8.345640182495117, |
| "logps/rejected": -24.480266571044922, |
| "loss": 1.0113354921340942, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5643945336341858, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.05142675340175629, |
| "rewards/margins": 1.0910050868988037, |
| "rewards/rejected": -1.0395784378051758, |
| "step": 27, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.16861121565675574, |
| "grad_norm": 3.4209959506988525, |
| "learning_rate": 0.00019939882849167852, |
| "logits/chosen": -0.2834317684173584, |
| "logits/rejected": -0.4013361930847168, |
| "logps/chosen": -6.27006196975708, |
| "logps/rejected": -23.9411678314209, |
| "loss": 0.9620530605316162, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5452341437339783, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.08509417623281479, |
| "rewards/margins": 1.0887510776519775, |
| "rewards/rejected": -1.0036569833755493, |
| "step": 28, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.17463304478735417, |
| "grad_norm": 3.021099805831909, |
| "learning_rate": 0.00019928469263418374, |
| "logits/chosen": -0.29641368985176086, |
| "logits/rejected": -0.3944772779941559, |
| "logps/chosen": -6.065124988555908, |
| "logps/rejected": -19.367380142211914, |
| "loss": 0.8545750379562378, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3696524500846863, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": -0.008128602989017963, |
| "rewards/margins": 0.7214774489402771, |
| "rewards/rejected": -0.729606032371521, |
| "step": 29, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.18065487391795257, |
| "grad_norm": 2.237170457839966, |
| "learning_rate": 0.00019916068131670302, |
| "logits/chosen": -0.2665432393550873, |
| "logits/rejected": -0.41999131441116333, |
| "logps/chosen": -3.6829960346221924, |
| "logps/rejected": -24.998981475830078, |
| "loss": 0.7719818353652954, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3786780834197998, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.15382343530654907, |
| "rewards/margins": 1.0922845602035522, |
| "rewards/rejected": -0.9384610652923584, |
| "step": 30, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.186676703048551, |
| "grad_norm": 2.299628496170044, |
| "learning_rate": 0.00019902680687415705, |
| "logits/chosen": -0.33755987882614136, |
| "logits/rejected": -0.37323832511901855, |
| "logps/chosen": -11.234481811523438, |
| "logps/rejected": -24.92943000793457, |
| "loss": 1.0785183906555176, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6123806834220886, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.03546600416302681, |
| "rewards/margins": 0.9260610342025757, |
| "rewards/rejected": -0.890595018863678, |
| "step": 31, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.19269853217914942, |
| "grad_norm": 2.0755157470703125, |
| "learning_rate": 0.00019888308262251285, |
| "logits/chosen": -0.2504952847957611, |
| "logits/rejected": -0.3852018713951111, |
| "logps/chosen": -5.893071174621582, |
| "logps/rejected": -33.010154724121094, |
| "loss": 0.7333768010139465, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4039708375930786, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": -0.07270973920822144, |
| "rewards/margins": 1.7266591787338257, |
| "rewards/rejected": -1.7993686199188232, |
| "step": 32, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.19872036130974782, |
| "grad_norm": 3.686795711517334, |
| "learning_rate": 0.00019872952285745959, |
| "logits/chosen": -0.30149099230766296, |
| "logits/rejected": -0.4157900810241699, |
| "logps/chosen": -7.993706703186035, |
| "logps/rejected": -27.117816925048828, |
| "loss": 1.0753562450408936, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6016828417778015, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": -0.10546419769525528, |
| "rewards/margins": 1.0370161533355713, |
| "rewards/rejected": -1.1424803733825684, |
| "step": 33, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.20474219044034625, |
| "grad_norm": 3.085048198699951, |
| "learning_rate": 0.0001985661428529863, |
| "logits/chosen": -0.31711238622665405, |
| "logits/rejected": -0.41008925437927246, |
| "logps/chosen": -6.8641557693481445, |
| "logps/rejected": -27.393281936645508, |
| "loss": 1.054256558418274, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6109166741371155, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": -0.06303416192531586, |
| "rewards/margins": 1.1662019491195679, |
| "rewards/rejected": -1.229236125946045, |
| "step": 34, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.21076401957094468, |
| "grad_norm": 1.7560195922851562, |
| "learning_rate": 0.00019839295885986296, |
| "logits/chosen": -0.337604820728302, |
| "logits/rejected": -0.39799487590789795, |
| "logps/chosen": -6.864170551300049, |
| "logps/rejected": -20.806344985961914, |
| "loss": 0.9619123339653015, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.48513123393058777, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.10183387249708176, |
| "rewards/margins": 0.7188172340393066, |
| "rewards/rejected": -0.6169832944869995, |
| "step": 35, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.2167858487015431, |
| "grad_norm": 2.6364471912384033, |
| "learning_rate": 0.0001982099881040239, |
| "logits/chosen": -0.36966580152511597, |
| "logits/rejected": -0.43631935119628906, |
| "logps/chosen": -6.647594451904297, |
| "logps/rejected": -21.291913986206055, |
| "loss": 0.9380239844322205, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4613775312900543, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.06358329951763153, |
| "rewards/margins": 0.7893975973129272, |
| "rewards/rejected": -0.7258143424987793, |
| "step": 36, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.2228076778321415, |
| "grad_norm": 2.354177474975586, |
| "learning_rate": 0.00019801724878485438, |
| "logits/chosen": -0.36403897404670715, |
| "logits/rejected": -0.45796939730644226, |
| "logps/chosen": -5.62083625793457, |
| "logps/rejected": -23.648834228515625, |
| "loss": 0.9243438243865967, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.46970799565315247, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.06422886252403259, |
| "rewards/margins": 0.8610265851020813, |
| "rewards/rejected": -0.7967977523803711, |
| "step": 37, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.22882950696273993, |
| "grad_norm": 2.2988836765289307, |
| "learning_rate": 0.00019781476007338058, |
| "logits/chosen": -0.3704385459423065, |
| "logits/rejected": -0.47973009943962097, |
| "logps/chosen": -4.231719017028809, |
| "logps/rejected": -20.56240463256836, |
| "loss": 0.916353702545166, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4209356904029846, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.06186368316411972, |
| "rewards/margins": 0.7635293006896973, |
| "rewards/rejected": -0.7016656398773193, |
| "step": 38, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.23485133609333836, |
| "grad_norm": 2.009106397628784, |
| "learning_rate": 0.00019760254211036244, |
| "logits/chosen": -0.3372359573841095, |
| "logits/rejected": -0.453107088804245, |
| "logps/chosen": -5.785426616668701, |
| "logps/rejected": -23.19124412536621, |
| "loss": 0.9124959707260132, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.44838690757751465, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.04832564294338226, |
| "rewards/margins": 0.937610387802124, |
| "rewards/rejected": -0.8892846703529358, |
| "step": 39, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.24087316522393676, |
| "grad_norm": 2.5412213802337646, |
| "learning_rate": 0.00019738061600429064, |
| "logits/chosen": -0.39427676796913147, |
| "logits/rejected": -0.48203372955322266, |
| "logps/chosen": -5.785412311553955, |
| "logps/rejected": -19.031665802001953, |
| "loss": 0.9683349132537842, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5110080242156982, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.022132933139801025, |
| "rewards/margins": 0.8074027299880981, |
| "rewards/rejected": -0.7852697372436523, |
| "step": 40, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.24689499435453519, |
| "grad_norm": 2.0049972534179688, |
| "learning_rate": 0.00019714900382928675, |
| "logits/chosen": -0.3044354021549225, |
| "logits/rejected": -0.41962650418281555, |
| "logps/chosen": -4.612288475036621, |
| "logps/rejected": -24.228839874267578, |
| "loss": 0.8179229497909546, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4278942048549652, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.07943032681941986, |
| "rewards/margins": 1.1561895608901978, |
| "rewards/rejected": -1.0767593383789062, |
| "step": 41, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.2529168234851336, |
| "grad_norm": 2.0874667167663574, |
| "learning_rate": 0.0001969077286229078, |
| "logits/chosen": -0.2736699879169464, |
| "logits/rejected": -0.4168338179588318, |
| "logps/chosen": -7.539179801940918, |
| "logps/rejected": -30.339994430541992, |
| "loss": 0.9962757229804993, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6084927320480347, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.046928200870752335, |
| "rewards/margins": 1.4784610271453857, |
| "rewards/rejected": -1.431532859802246, |
| "step": 42, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.258938652615732, |
| "grad_norm": 3.2748661041259766, |
| "learning_rate": 0.00019665681438385473, |
| "logits/chosen": -0.26249000430107117, |
| "logits/rejected": -0.3955192565917969, |
| "logps/chosen": -6.726841449737549, |
| "logps/rejected": -26.447267532348633, |
| "loss": 0.9923527836799622, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5636712312698364, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": -0.02616805024445057, |
| "rewards/margins": 1.2797741889953613, |
| "rewards/rejected": -1.3059422969818115, |
| "step": 43, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.26496048174633047, |
| "grad_norm": 4.499722957611084, |
| "learning_rate": 0.00019639628606958533, |
| "logits/chosen": -0.3380294442176819, |
| "logits/rejected": -0.4303723871707916, |
| "logps/chosen": -7.925315856933594, |
| "logps/rejected": -23.916906356811523, |
| "loss": 1.136991024017334, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6508135795593262, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": -0.1142488494515419, |
| "rewards/margins": 1.0733367204666138, |
| "rewards/rejected": -1.1875855922698975, |
| "step": 44, |
| "train_speed(iter/s)": 0.005638 |
| }, |
| { |
| "epoch": 0.27098231087692887, |
| "grad_norm": 3.695734739303589, |
| "learning_rate": 0.0001961261695938319, |
| "logits/chosen": -0.2465212196111679, |
| "logits/rejected": -0.39715391397476196, |
| "logps/chosen": -7.8324666023254395, |
| "logps/rejected": -24.97869110107422, |
| "loss": 0.9623777270317078, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5307109951972961, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.08475106954574585, |
| "rewards/margins": 0.9390297532081604, |
| "rewards/rejected": -0.8542786836624146, |
| "step": 45, |
| "train_speed(iter/s)": 0.005638 |
| }, |
| { |
| "epoch": 0.27700414000752727, |
| "grad_norm": 3.1841602325439453, |
| "learning_rate": 0.00019584649182402357, |
| "logits/chosen": -0.22775670886039734, |
| "logits/rejected": -0.3454330265522003, |
| "logps/chosen": -6.5436859130859375, |
| "logps/rejected": -24.402027130126953, |
| "loss": 1.0284608602523804, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5513083338737488, |
| "rewards/accuracies": 0.75, |
| "rewards/chosen": 0.05229588598012924, |
| "rewards/margins": 0.9507507085800171, |
| "rewards/rejected": -0.8984547853469849, |
| "step": 46, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.2830259691381257, |
| "grad_norm": 2.0325071811676025, |
| "learning_rate": 0.0001955572805786141, |
| "logits/chosen": -0.19194024801254272, |
| "logits/rejected": -0.32474127411842346, |
| "logps/chosen": -4.646670341491699, |
| "logps/rejected": -27.415727615356445, |
| "loss": 0.7930896878242493, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.38733866810798645, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.11078543961048126, |
| "rewards/margins": 1.1206859350204468, |
| "rewards/rejected": -1.009900450706482, |
| "step": 47, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.2890477982687241, |
| "grad_norm": 2.07770037651062, |
| "learning_rate": 0.0001952585646243146, |
| "logits/chosen": -0.25676336884498596, |
| "logits/rejected": -0.3494555950164795, |
| "logps/chosen": -7.388433933258057, |
| "logps/rejected": -23.070579528808594, |
| "loss": 0.8534475564956665, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.42261388897895813, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.006194199435412884, |
| "rewards/margins": 0.9558709859848022, |
| "rewards/rejected": -0.9496768712997437, |
| "step": 48, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.2950696273993225, |
| "grad_norm": 1.6862146854400635, |
| "learning_rate": 0.00019495037367323262, |
| "logits/chosen": -0.22853028774261475, |
| "logits/rejected": -0.3647540211677551, |
| "logps/chosen": -3.3972930908203125, |
| "logps/rejected": -25.391393661499023, |
| "loss": 0.7584111094474792, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.35165295004844666, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.12043334543704987, |
| "rewards/margins": 1.1430784463882446, |
| "rewards/rejected": -1.022645115852356, |
| "step": 49, |
| "train_speed(iter/s)": 0.005638 |
| }, |
| { |
| "epoch": 0.301091456529921, |
| "grad_norm": 3.027331829071045, |
| "learning_rate": 0.00019463273837991643, |
| "logits/chosen": -0.2110910564661026, |
| "logits/rejected": -0.38482916355133057, |
| "logps/chosen": -4.55904483795166, |
| "logps/rejected": -31.877506256103516, |
| "loss": 0.675049901008606, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.32799023389816284, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.05019056051969528, |
| "rewards/margins": 1.6790026426315308, |
| "rewards/rejected": -1.6288120746612549, |
| "step": 50, |
| "train_speed(iter/s)": 0.005638 |
| }, |
| { |
| "epoch": 0.3071132856605194, |
| "grad_norm": 2.950259208679199, |
| "learning_rate": 0.00019430569033830605, |
| "logits/chosen": -0.19638891518115997, |
| "logits/rejected": -0.3516000807285309, |
| "logps/chosen": -7.3598952293396, |
| "logps/rejected": -32.14203643798828, |
| "loss": 0.9957353472709656, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6411410570144653, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": -0.031813204288482666, |
| "rewards/margins": 1.6832952499389648, |
| "rewards/rejected": -1.7151083946228027, |
| "step": 51, |
| "train_speed(iter/s)": 0.005628 |
| }, |
| { |
| "epoch": 0.3131351147911178, |
| "grad_norm": 3.4817123413085938, |
| "learning_rate": 0.00019396926207859084, |
| "logits/chosen": -0.2925015687942505, |
| "logits/rejected": -0.4453135132789612, |
| "logps/chosen": -6.357159614562988, |
| "logps/rejected": -34.39080047607422, |
| "loss": 0.9899907112121582, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6184481382369995, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": -0.06884388625621796, |
| "rewards/margins": 1.8842136859893799, |
| "rewards/rejected": -1.9530572891235352, |
| "step": 52, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.31915694392171623, |
| "grad_norm": 3.4239518642425537, |
| "learning_rate": 0.00019362348706397373, |
| "logits/chosen": -0.22639772295951843, |
| "logits/rejected": -0.4221537709236145, |
| "logps/chosen": -5.600902080535889, |
| "logps/rejected": -36.40532684326172, |
| "loss": 0.8088786602020264, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4771510660648346, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.0569700188934803, |
| "rewards/margins": 1.8676366806030273, |
| "rewards/rejected": -1.810666561126709, |
| "step": 53, |
| "train_speed(iter/s)": 0.005628 |
| }, |
| { |
| "epoch": 0.32517877305231463, |
| "grad_norm": 3.8685495853424072, |
| "learning_rate": 0.00019326839968734279, |
| "logits/chosen": -0.2569487690925598, |
| "logits/rejected": -0.40343666076660156, |
| "logps/chosen": -5.900696277618408, |
| "logps/rejected": -36.28150177001953, |
| "loss": 0.7994194030761719, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.43986907601356506, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.030100831761956215, |
| "rewards/margins": 1.9989190101623535, |
| "rewards/rejected": -1.968818187713623, |
| "step": 54, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.3312006021829131, |
| "grad_norm": 3.0355138778686523, |
| "learning_rate": 0.00019290403526785025, |
| "logits/chosen": -0.25347286462783813, |
| "logits/rejected": -0.3757534623146057, |
| "logps/chosen": -7.442959308624268, |
| "logps/rejected": -32.45537567138672, |
| "loss": 0.903397798538208, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.531738817691803, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.0414905920624733, |
| "rewards/margins": 1.6564455032348633, |
| "rewards/rejected": -1.614954948425293, |
| "step": 55, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.3372224313135115, |
| "grad_norm": 3.739659070968628, |
| "learning_rate": 0.00019253043004739968, |
| "logits/chosen": -0.20877386629581451, |
| "logits/rejected": -0.37541496753692627, |
| "logps/chosen": -7.142071723937988, |
| "logps/rejected": -32.15766906738281, |
| "loss": 0.9619933366775513, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4930208921432495, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.08227778971195221, |
| "rewards/margins": 1.5348973274230957, |
| "rewards/rejected": -1.6171751022338867, |
| "step": 56, |
| "train_speed(iter/s)": 0.005628 |
| }, |
| { |
| "epoch": 0.3432442604441099, |
| "grad_norm": 2.4167847633361816, |
| "learning_rate": 0.00019214762118704076, |
| "logits/chosen": -0.32553619146347046, |
| "logits/rejected": -0.44049111008644104, |
| "logps/chosen": -6.96627140045166, |
| "logps/rejected": -24.777843475341797, |
| "loss": 0.8523041605949402, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.44423431158065796, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.08760333806276321, |
| "rewards/margins": 1.0987879037857056, |
| "rewards/rejected": -1.011184573173523, |
| "step": 57, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.34926608957470834, |
| "grad_norm": 1.849902629852295, |
| "learning_rate": 0.00019175564676327339, |
| "logits/chosen": -0.29927390813827515, |
| "logits/rejected": -0.4095182418823242, |
| "logps/chosen": -6.824729919433594, |
| "logps/rejected": -31.636598587036133, |
| "loss": 0.827016294002533, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.432888925075531, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.10196143388748169, |
| "rewards/margins": 1.6724345684051514, |
| "rewards/rejected": -1.5704729557037354, |
| "step": 58, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.35528791870530674, |
| "grad_norm": 2.3601560592651367, |
| "learning_rate": 0.0001913545457642601, |
| "logits/chosen": -0.31936806440353394, |
| "logits/rejected": -0.42600923776626587, |
| "logps/chosen": -6.396060943603516, |
| "logps/rejected": -26.946306228637695, |
| "loss": 0.9836419224739075, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5654138326644897, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.013314278796315193, |
| "rewards/margins": 1.2707021236419678, |
| "rewards/rejected": -1.2573878765106201, |
| "step": 59, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.36130974783590514, |
| "grad_norm": 2.7022287845611572, |
| "learning_rate": 0.00019094435808594823, |
| "logits/chosen": -0.3565162420272827, |
| "logits/rejected": -0.48129239678382874, |
| "logps/chosen": -4.523666858673096, |
| "logps/rejected": -30.854751586914062, |
| "loss": 0.7463247776031494, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.41281038522720337, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.06385684013366699, |
| "rewards/margins": 1.8442981243133545, |
| "rewards/rejected": -1.7804412841796875, |
| "step": 60, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.3673315769665036, |
| "grad_norm": 2.549546957015991, |
| "learning_rate": 0.0001905251245281015, |
| "logits/chosen": -0.2870449125766754, |
| "logits/rejected": -0.4699952006340027, |
| "logps/chosen": -3.242504596710205, |
| "logps/rejected": -35.58238220214844, |
| "loss": 0.6698994636535645, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3660232126712799, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.03870222344994545, |
| "rewards/margins": 1.8160982131958008, |
| "rewards/rejected": -1.7773959636688232, |
| "step": 61, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 0.373353406097102, |
| "grad_norm": 3.344719409942627, |
| "learning_rate": 0.0001900968867902419, |
| "logits/chosen": -0.26915398240089417, |
| "logits/rejected": -0.38059523701667786, |
| "logps/chosen": -6.996611595153809, |
| "logps/rejected": -36.200714111328125, |
| "loss": 0.8631665110588074, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5069284439086914, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": -0.0650690495967865, |
| "rewards/margins": 1.7960646152496338, |
| "rewards/rejected": -1.8611338138580322, |
| "step": 62, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.3793752352277004, |
| "grad_norm": 1.7888697385787964, |
| "learning_rate": 0.0001896596874675021, |
| "logits/chosen": -0.25557130575180054, |
| "logits/rejected": -0.4095707833766937, |
| "logps/chosen": -5.574047088623047, |
| "logps/rejected": -31.464292526245117, |
| "loss": 0.6977128386497498, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3594370484352112, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.08366499841213226, |
| "rewards/margins": 1.6534584760665894, |
| "rewards/rejected": -1.569793462753296, |
| "step": 63, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.38539706435829885, |
| "grad_norm": 2.8035733699798584, |
| "learning_rate": 0.00018921357004638835, |
| "logits/chosen": -0.22736577689647675, |
| "logits/rejected": -0.3719921410083771, |
| "logps/chosen": -6.317353248596191, |
| "logps/rejected": -30.95716094970703, |
| "loss": 0.8575116991996765, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5229077339172363, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.13013550639152527, |
| "rewards/margins": 1.687712550163269, |
| "rewards/rejected": -1.5575770139694214, |
| "step": 64, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.39141889348889725, |
| "grad_norm": 2.560202121734619, |
| "learning_rate": 0.00018875857890045543, |
| "logits/chosen": -0.24293088912963867, |
| "logits/rejected": -0.35710442066192627, |
| "logps/chosen": -7.503670692443848, |
| "logps/rejected": -32.99870300292969, |
| "loss": 0.8683996796607971, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.49205920100212097, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.03412435203790665, |
| "rewards/margins": 1.4795939922332764, |
| "rewards/rejected": -1.4454697370529175, |
| "step": 65, |
| "train_speed(iter/s)": 0.005629 |
| }, |
| { |
| "epoch": 0.39744072261949565, |
| "grad_norm": 1.993003487586975, |
| "learning_rate": 0.00018829475928589271, |
| "logits/chosen": -0.23541702330112457, |
| "logits/rejected": -0.3272259533405304, |
| "logps/chosen": -5.38151216506958, |
| "logps/rejected": -29.121570587158203, |
| "loss": 0.7320932149887085, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.36119604110717773, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.0803036242723465, |
| "rewards/margins": 1.6109750270843506, |
| "rewards/rejected": -1.53067147731781, |
| "step": 66, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.4034625517500941, |
| "grad_norm": 2.9964606761932373, |
| "learning_rate": 0.00018782215733702286, |
| "logits/chosen": -0.20383372902870178, |
| "logits/rejected": -0.34373044967651367, |
| "logps/chosen": -6.649240493774414, |
| "logps/rejected": -31.494646072387695, |
| "loss": 0.8884384036064148, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5308991074562073, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.10617589205503464, |
| "rewards/margins": 1.5828511714935303, |
| "rewards/rejected": -1.4766751527786255, |
| "step": 67, |
| "train_speed(iter/s)": 0.00563 |
| }, |
| { |
| "epoch": 0.4094843808806925, |
| "grad_norm": 2.774994134902954, |
| "learning_rate": 0.00018734082006171299, |
| "logits/chosen": -0.25100839138031006, |
| "logits/rejected": -0.4020751118659973, |
| "logps/chosen": -7.144831657409668, |
| "logps/rejected": -29.9566707611084, |
| "loss": 0.8242477178573608, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5139617919921875, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.13794061541557312, |
| "rewards/margins": 1.6364552974700928, |
| "rewards/rejected": -1.4985146522521973, |
| "step": 68, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 0.41550621001129096, |
| "grad_norm": 3.4358551502227783, |
| "learning_rate": 0.0001868507953366989, |
| "logits/chosen": -0.24417072534561157, |
| "logits/rejected": -0.36953434348106384, |
| "logps/chosen": -6.3338799476623535, |
| "logps/rejected": -32.21156311035156, |
| "loss": 0.7758980989456177, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.48221129179000854, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.14049482345581055, |
| "rewards/margins": 1.806667447090149, |
| "rewards/rejected": -1.666172742843628, |
| "step": 69, |
| "train_speed(iter/s)": 0.005631 |
| }, |
| { |
| "epoch": 0.42152803914188935, |
| "grad_norm": 2.821110248565674, |
| "learning_rate": 0.0001863521319028231, |
| "logits/chosen": -0.2586533725261688, |
| "logits/rejected": -0.36297786235809326, |
| "logps/chosen": -7.11786413192749, |
| "logps/rejected": -31.115772247314453, |
| "loss": 0.8807559013366699, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5770330429077148, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.07292432337999344, |
| "rewards/margins": 2.026425838470459, |
| "rewards/rejected": -1.9535014629364014, |
| "step": 70, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.42754986827248775, |
| "grad_norm": 2.6153159141540527, |
| "learning_rate": 0.00018584487936018661, |
| "logits/chosen": -0.305365651845932, |
| "logits/rejected": -0.4076971709728241, |
| "logps/chosen": -4.401376247406006, |
| "logps/rejected": -25.76744842529297, |
| "loss": 0.7574676275253296, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3759956955909729, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": -0.015199379995465279, |
| "rewards/margins": 1.5401445627212524, |
| "rewards/rejected": -1.5553438663482666, |
| "step": 71, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.4335716974030862, |
| "grad_norm": 4.708693027496338, |
| "learning_rate": 0.00018532908816321558, |
| "logits/chosen": -0.22257624566555023, |
| "logits/rejected": -0.36776047945022583, |
| "logps/chosen": -5.041079998016357, |
| "logps/rejected": -31.822460174560547, |
| "loss": 0.7604058980941772, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3996947705745697, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0631202757358551, |
| "rewards/margins": 1.8200846910476685, |
| "rewards/rejected": -1.7569644451141357, |
| "step": 72, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.4395935265336846, |
| "grad_norm": 4.16855001449585, |
| "learning_rate": 0.0001848048096156426, |
| "logits/chosen": -0.1833086460828781, |
| "logits/rejected": -0.40944039821624756, |
| "logps/chosen": -7.540615081787109, |
| "logps/rejected": -41.107208251953125, |
| "loss": 1.1337863206863403, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.7324954271316528, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": -0.13253508508205414, |
| "rewards/margins": 2.3366596698760986, |
| "rewards/rejected": -2.4691946506500244, |
| "step": 73, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.445615355664283, |
| "grad_norm": 2.806788444519043, |
| "learning_rate": 0.0001842720958654039, |
| "logits/chosen": -0.3181215524673462, |
| "logits/rejected": -0.39911144971847534, |
| "logps/chosen": -5.607314586639404, |
| "logps/rejected": -28.428564071655273, |
| "loss": 0.6960507035255432, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.40211787819862366, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.15752148628234863, |
| "rewards/margins": 1.9513351917266846, |
| "rewards/rejected": -1.7938138246536255, |
| "step": 74, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.45163718479488146, |
| "grad_norm": 3.420902967453003, |
| "learning_rate": 0.00018373099989945236, |
| "logits/chosen": -0.22940213978290558, |
| "logits/rejected": -0.4191284477710724, |
| "logps/chosen": -6.713252067565918, |
| "logps/rejected": -37.6962890625, |
| "loss": 0.920997679233551, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5668741464614868, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": -0.0326458215713501, |
| "rewards/margins": 2.1453044414520264, |
| "rewards/rejected": -2.177950382232666, |
| "step": 75, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.45765901392547986, |
| "grad_norm": 3.6164190769195557, |
| "learning_rate": 0.0001831815755384869, |
| "logits/chosen": -0.20445629954338074, |
| "logits/rejected": -0.3627314567565918, |
| "logps/chosen": -5.958117485046387, |
| "logps/rejected": -38.63805389404297, |
| "loss": 0.884059727191925, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5545799732208252, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": -0.049575597047805786, |
| "rewards/margins": 2.2317745685577393, |
| "rewards/rejected": -2.2813501358032227, |
| "step": 76, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.46368084305607826, |
| "grad_norm": 2.9762485027313232, |
| "learning_rate": 0.0001826238774315995, |
| "logits/chosen": -0.23990976810455322, |
| "logits/rejected": -0.38009804487228394, |
| "logps/chosen": -5.436191082000732, |
| "logps/rejected": -34.6729850769043, |
| "loss": 0.7682054042816162, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4302901029586792, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.1761879026889801, |
| "rewards/margins": 2.175489902496338, |
| "rewards/rejected": -1.9993020296096802, |
| "step": 77, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.4697026721866767, |
| "grad_norm": 2.6883339881896973, |
| "learning_rate": 0.00018205796105083915, |
| "logits/chosen": -0.27416834235191345, |
| "logits/rejected": -0.387048602104187, |
| "logps/chosen": -6.189967155456543, |
| "logps/rejected": -30.830598831176758, |
| "loss": 0.8499253988265991, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.46602901816368103, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.06820189207792282, |
| "rewards/margins": 1.8341140747070312, |
| "rewards/rejected": -1.7659121751785278, |
| "step": 78, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.4757245013172751, |
| "grad_norm": 2.6955106258392334, |
| "learning_rate": 0.00018148388268569453, |
| "logits/chosen": -0.22744283080101013, |
| "logits/rejected": -0.35057857632637024, |
| "logps/chosen": -4.485665798187256, |
| "logps/rejected": -29.74733543395996, |
| "loss": 0.7549667954444885, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4278162717819214, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.20732006430625916, |
| "rewards/margins": 2.0227808952331543, |
| "rewards/rejected": -1.8154606819152832, |
| "step": 79, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.4817463304478735, |
| "grad_norm": 2.907402515411377, |
| "learning_rate": 0.00018090169943749476, |
| "logits/chosen": -0.13104476034641266, |
| "logits/rejected": -0.36177706718444824, |
| "logps/chosen": -3.6902096271514893, |
| "logps/rejected": -34.731773376464844, |
| "loss": 0.6847038269042969, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.34349215030670166, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.11521060019731522, |
| "rewards/margins": 1.8666913509368896, |
| "rewards/rejected": -1.7514808177947998, |
| "step": 80, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.48776815957847197, |
| "grad_norm": 2.823246955871582, |
| "learning_rate": 0.00018031146921373018, |
| "logits/chosen": -0.22965751588344574, |
| "logits/rejected": -0.3291424512863159, |
| "logps/chosen": -5.761353969573975, |
| "logps/rejected": -36.051082611083984, |
| "loss": 0.7157370448112488, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4026964008808136, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.07981520146131516, |
| "rewards/margins": 2.387995958328247, |
| "rewards/rejected": -2.308180809020996, |
| "step": 81, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.49378998870907037, |
| "grad_norm": 2.73394513130188, |
| "learning_rate": 0.00017971325072229226, |
| "logits/chosen": -0.19317063689231873, |
| "logits/rejected": -0.40784093737602234, |
| "logps/chosen": -5.166616916656494, |
| "logps/rejected": -40.66303634643555, |
| "loss": 0.6980624198913574, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4219977855682373, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.030578728765249252, |
| "rewards/margins": 2.3669514656066895, |
| "rewards/rejected": -2.3363726139068604, |
| "step": 82, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.4998118178396688, |
| "grad_norm": 5.509173393249512, |
| "learning_rate": 0.00017910710346563416, |
| "logits/chosen": -0.1950196623802185, |
| "logits/rejected": -0.3517180383205414, |
| "logps/chosen": -8.197553634643555, |
| "logps/rejected": -42.4526252746582, |
| "loss": 0.9209753274917603, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.608447790145874, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.012901969254016876, |
| "rewards/margins": 2.6168861389160156, |
| "rewards/rejected": -2.6039838790893555, |
| "step": 83, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.5058336469702672, |
| "grad_norm": 3.090348482131958, |
| "learning_rate": 0.00017849308773485226, |
| "logits/chosen": -0.2090476155281067, |
| "logits/rejected": -0.3706829845905304, |
| "logps/chosen": -10.569226264953613, |
| "logps/rejected": -43.29594421386719, |
| "loss": 1.0815891027450562, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.7449980974197388, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": -0.03622851148247719, |
| "rewards/margins": 2.7713966369628906, |
| "rewards/rejected": -2.8076250553131104, |
| "step": 84, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.5118554761008657, |
| "grad_norm": 3.75370717048645, |
| "learning_rate": 0.0001778712646036894, |
| "logits/chosen": -0.2378547489643097, |
| "logits/rejected": -0.3873806297779083, |
| "logps/chosen": -8.168487548828125, |
| "logps/rejected": -37.17033386230469, |
| "loss": 1.0251762866973877, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6340765953063965, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": -0.11355286091566086, |
| "rewards/margins": 2.2919938564300537, |
| "rewards/rejected": -2.4055469036102295, |
| "step": 85, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.517877305231464, |
| "grad_norm": 3.9649507999420166, |
| "learning_rate": 0.00017724169592245995, |
| "logits/chosen": -0.20498239994049072, |
| "logits/rejected": -0.369476854801178, |
| "logps/chosen": -6.779036521911621, |
| "logps/rejected": -37.98490905761719, |
| "loss": 0.797798216342926, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5137072801589966, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.07833351194858551, |
| "rewards/margins": 2.144557476043701, |
| "rewards/rejected": -2.0662238597869873, |
| "step": 86, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.5238991343620625, |
| "grad_norm": 3.4646809101104736, |
| "learning_rate": 0.0001766044443118978, |
| "logits/chosen": -0.22029685974121094, |
| "logits/rejected": -0.33659541606903076, |
| "logps/chosen": -5.358316421508789, |
| "logps/rejected": -24.96926498413086, |
| "loss": 0.809937596321106, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3953189551830292, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.08655387908220291, |
| "rewards/margins": 1.2862589359283447, |
| "rewards/rejected": -1.1997051239013672, |
| "step": 87, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.5299209634926609, |
| "grad_norm": 2.3174376487731934, |
| "learning_rate": 0.00017595957315692782, |
| "logits/chosen": -0.23393434286117554, |
| "logits/rejected": -0.37657639384269714, |
| "logps/chosen": -5.694438934326172, |
| "logps/rejected": -26.54452896118164, |
| "loss": 0.8167567253112793, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.46906933188438416, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.12023281306028366, |
| "rewards/margins": 1.344594120979309, |
| "rewards/rejected": -1.2243614196777344, |
| "step": 88, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.5359427926232593, |
| "grad_norm": 2.29378342628479, |
| "learning_rate": 0.00017530714660036112, |
| "logits/chosen": -0.23060575127601624, |
| "logits/rejected": -0.34717997908592224, |
| "logps/chosen": -8.571242332458496, |
| "logps/rejected": -27.7784423828125, |
| "loss": 0.9238011837005615, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5501088500022888, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": 0.21566246449947357, |
| "rewards/margins": 1.4910237789154053, |
| "rewards/rejected": -1.2753612995147705, |
| "step": 89, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.5419646217538577, |
| "grad_norm": 4.33613395690918, |
| "learning_rate": 0.00017464722953651504, |
| "logits/chosen": -0.17152062058448792, |
| "logits/rejected": -0.3561325669288635, |
| "logps/chosen": -6.058459281921387, |
| "logps/rejected": -32.55611801147461, |
| "loss": 0.8752692341804504, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5144660472869873, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.0842672809958458, |
| "rewards/margins": 1.5552325248718262, |
| "rewards/rejected": -1.4709652662277222, |
| "step": 90, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.5479864508844562, |
| "grad_norm": 2.2437572479248047, |
| "learning_rate": 0.0001739798876047584, |
| "logits/chosen": -0.23511910438537598, |
| "logits/rejected": -0.3851989507675171, |
| "logps/chosen": -4.1521100997924805, |
| "logps/rejected": -28.886817932128906, |
| "loss": 0.7264631390571594, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.36743801832199097, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.10910709202289581, |
| "rewards/margins": 1.5735148191452026, |
| "rewards/rejected": -1.4644078016281128, |
| "step": 91, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.5540082800150545, |
| "grad_norm": 2.482116460800171, |
| "learning_rate": 0.00017330518718298264, |
| "logits/chosen": -0.19554060697555542, |
| "logits/rejected": -0.38464802503585815, |
| "logps/chosen": -5.105670928955078, |
| "logps/rejected": -42.76155471801758, |
| "loss": 0.6233892440795898, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.36078184843063354, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.06755763292312622, |
| "rewards/margins": 2.4900288581848145, |
| "rewards/rejected": -2.422471523284912, |
| "step": 92, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.560030109145653, |
| "grad_norm": 3.629793167114258, |
| "learning_rate": 0.0001726231953809993, |
| "logits/chosen": -0.23075315356254578, |
| "logits/rejected": -0.4017852246761322, |
| "logps/chosen": -7.799711227416992, |
| "logps/rejected": -38.58491516113281, |
| "loss": 0.7807697057723999, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5262271165847778, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.15829423069953918, |
| "rewards/margins": 2.4986987113952637, |
| "rewards/rejected": -2.340404510498047, |
| "step": 93, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.5660519382762514, |
| "grad_norm": 10.366227149963379, |
| "learning_rate": 0.0001719339800338651, |
| "logits/chosen": -0.29187849164009094, |
| "logits/rejected": -0.42478230595588684, |
| "logps/chosen": -7.801548004150391, |
| "logps/rejected": -34.29828643798828, |
| "loss": 1.1007416248321533, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.7008671164512634, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": -0.0017978232353925705, |
| "rewards/margins": 1.9699221849441528, |
| "rewards/rejected": -1.9717202186584473, |
| "step": 94, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.5720737674068498, |
| "grad_norm": 4.7859039306640625, |
| "learning_rate": 0.0001712376096951345, |
| "logits/chosen": -0.2881244421005249, |
| "logits/rejected": -0.39795249700546265, |
| "logps/chosen": -6.9338459968566895, |
| "logps/rejected": -34.682186126708984, |
| "loss": 0.8693129420280457, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5744621157646179, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.12829959392547607, |
| "rewards/margins": 2.2821671962738037, |
| "rewards/rejected": -2.153867244720459, |
| "step": 95, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.5780955965374482, |
| "grad_norm": 2.521221160888672, |
| "learning_rate": 0.0001705341536300409, |
| "logits/chosen": -0.2661621868610382, |
| "logits/rejected": -0.3923559784889221, |
| "logps/chosen": -4.780788898468018, |
| "logps/rejected": -28.810710906982422, |
| "loss": 0.7346460819244385, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.40230822563171387, |
| "rewards/accuracies": 0.796875, |
| "rewards/chosen": 0.10683214664459229, |
| "rewards/margins": 1.7658103704452515, |
| "rewards/rejected": -1.6589782238006592, |
| "step": 96, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.5841174256680467, |
| "grad_norm": 5.023887634277344, |
| "learning_rate": 0.00016982368180860728, |
| "logits/chosen": -0.25105586647987366, |
| "logits/rejected": -0.3866545855998993, |
| "logps/chosen": -7.320672512054443, |
| "logps/rejected": -34.95018005371094, |
| "loss": 0.9987191557884216, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6328321695327759, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.07585513591766357, |
| "rewards/margins": 2.0516769886016846, |
| "rewards/rejected": -2.127532482147217, |
| "step": 97, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.590139254798645, |
| "grad_norm": 1.8328591585159302, |
| "learning_rate": 0.00016910626489868649, |
| "logits/chosen": -0.2209641933441162, |
| "logits/rejected": -0.398262083530426, |
| "logps/chosen": -4.149725437164307, |
| "logps/rejected": -35.193931579589844, |
| "loss": 0.6401265263557434, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.31758564710617065, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.14480024576187134, |
| "rewards/margins": 2.074404716491699, |
| "rewards/rejected": -1.9296045303344727, |
| "step": 98, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.5961610839292435, |
| "grad_norm": 1.696344017982483, |
| "learning_rate": 0.00016838197425893202, |
| "logits/chosen": -0.2796581983566284, |
| "logits/rejected": -0.4003712832927704, |
| "logps/chosen": -5.847466468811035, |
| "logps/rejected": -30.844961166381836, |
| "loss": 0.7925845384597778, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4691714644432068, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.15563875436782837, |
| "rewards/margins": 1.9312589168548584, |
| "rewards/rejected": -1.7756202220916748, |
| "step": 99, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.602182913059842, |
| "grad_norm": 2.3960113525390625, |
| "learning_rate": 0.00016765088193170053, |
| "logits/chosen": -0.2749744653701782, |
| "logits/rejected": -0.39394593238830566, |
| "logps/chosen": -6.603066444396973, |
| "logps/rejected": -25.912214279174805, |
| "loss": 0.9519745707511902, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5554052591323853, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.11803217232227325, |
| "rewards/margins": 1.5442473888397217, |
| "rewards/rejected": -1.4262150526046753, |
| "step": 100, |
| "train_speed(iter/s)": 0.005637 |
| }, |
| { |
| "epoch": 0.6082047421904403, |
| "grad_norm": 2.1803297996520996, |
| "learning_rate": 0.00016691306063588583, |
| "logits/chosen": -0.14278045296669006, |
| "logits/rejected": -0.32641783356666565, |
| "logps/chosen": -6.367352485656738, |
| "logps/rejected": -36.09748077392578, |
| "loss": 0.7241327166557312, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4310902953147888, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.20738306641578674, |
| "rewards/margins": 1.9153938293457031, |
| "rewards/rejected": -1.7080106735229492, |
| "step": 101, |
| "train_speed(iter/s)": 0.005632 |
| }, |
| { |
| "epoch": 0.6142265713210388, |
| "grad_norm": 2.27510142326355, |
| "learning_rate": 0.00016616858375968595, |
| "logits/chosen": -0.20140963792800903, |
| "logits/rejected": -0.34892868995666504, |
| "logps/chosen": -4.8850579261779785, |
| "logps/rejected": -34.458396911621094, |
| "loss": 0.6944241523742676, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.39014381170272827, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.05396907776594162, |
| "rewards/margins": 2.1584324836730957, |
| "rewards/rejected": -2.1044633388519287, |
| "step": 102, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.6202484004516372, |
| "grad_norm": 3.8538124561309814, |
| "learning_rate": 0.00016541752535330345, |
| "logits/chosen": -0.15393608808517456, |
| "logits/rejected": -0.32232990860939026, |
| "logps/chosen": -7.490941047668457, |
| "logps/rejected": -38.019405364990234, |
| "loss": 0.899553656578064, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5592579245567322, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.04624926298856735, |
| "rewards/margins": 2.2129275798797607, |
| "rewards/rejected": -2.259176731109619, |
| "step": 103, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.6262702295822355, |
| "grad_norm": 3.2243363857269287, |
| "learning_rate": 0.00016465996012157995, |
| "logits/chosen": -0.1763302981853485, |
| "logits/rejected": -0.30002710223197937, |
| "logps/chosen": -7.91542387008667, |
| "logps/rejected": -35.163421630859375, |
| "loss": 0.8102483153343201, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5111361145973206, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.1345573365688324, |
| "rewards/margins": 2.1800684928894043, |
| "rewards/rejected": -2.045511245727539, |
| "step": 104, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.632292058712834, |
| "grad_norm": 2.4296176433563232, |
| "learning_rate": 0.0001638959634165656, |
| "logits/chosen": -0.2042100727558136, |
| "logits/rejected": -0.3498837947845459, |
| "logps/chosen": -5.70271635055542, |
| "logps/rejected": -33.96530532836914, |
| "loss": 0.7614078521728516, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.46085914969444275, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.10853511095046997, |
| "rewards/margins": 2.407238483428955, |
| "rewards/rejected": -2.298703670501709, |
| "step": 105, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.6383138878434325, |
| "grad_norm": 5.010378837585449, |
| "learning_rate": 0.0001631256112300239, |
| "logits/chosen": -0.17106319963932037, |
| "logits/rejected": -0.3643282651901245, |
| "logps/chosen": -5.012183666229248, |
| "logps/rejected": -39.59567642211914, |
| "loss": 0.7013410329818726, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4339035153388977, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.07484348863363266, |
| "rewards/margins": 2.594076156616211, |
| "rewards/rejected": -2.5192322731018066, |
| "step": 106, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.6443357169740309, |
| "grad_norm": 3.9870245456695557, |
| "learning_rate": 0.00016234898018587337, |
| "logits/chosen": -0.13750208914279938, |
| "logits/rejected": -0.3746242821216583, |
| "logps/chosen": -5.721056938171387, |
| "logps/rejected": -44.44032287597656, |
| "loss": 0.7298431396484375, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4898163080215454, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.036943770945072174, |
| "rewards/margins": 2.8432068824768066, |
| "rewards/rejected": -2.806262969970703, |
| "step": 107, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.6503575461046293, |
| "grad_norm": 2.7155423164367676, |
| "learning_rate": 0.0001615661475325658, |
| "logits/chosen": -0.1980631947517395, |
| "logits/rejected": -0.36777210235595703, |
| "logps/chosen": -5.860571384429932, |
| "logps/rejected": -45.095069885253906, |
| "loss": 0.6415181756019592, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.43772590160369873, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.029481690376996994, |
| "rewards/margins": 2.904158115386963, |
| "rewards/rejected": -2.874676465988159, |
| "step": 108, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.6563793752352277, |
| "grad_norm": 3.9366610050201416, |
| "learning_rate": 0.00016077719113540302, |
| "logits/chosen": -0.201637402176857, |
| "logits/rejected": -0.36231857538223267, |
| "logps/chosen": -8.20331859588623, |
| "logps/rejected": -47.63356018066406, |
| "loss": 0.8109980821609497, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.588591456413269, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": -0.11406072974205017, |
| "rewards/margins": 3.15336275100708, |
| "rewards/rejected": -3.267423629760742, |
| "step": 109, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.6624012043658262, |
| "grad_norm": 1.8952491283416748, |
| "learning_rate": 0.00015998218946879138, |
| "logits/chosen": -0.20580148696899414, |
| "logits/rejected": -0.37684956192970276, |
| "logps/chosen": -4.04914665222168, |
| "logps/rejected": -40.22568130493164, |
| "loss": 0.6388105154037476, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.39933112263679504, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.07669900357723236, |
| "rewards/margins": 2.701504945755005, |
| "rewards/rejected": -2.6248061656951904, |
| "step": 110, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.6684230334964245, |
| "grad_norm": 2.1146159172058105, |
| "learning_rate": 0.00015918122160843678, |
| "logits/chosen": -0.21242693066596985, |
| "logits/rejected": -0.3542603850364685, |
| "logps/chosen": -7.400607109069824, |
| "logps/rejected": -41.83246612548828, |
| "loss": 0.9009187817573547, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6562238931655884, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.07808566093444824, |
| "rewards/margins": 2.6408753395080566, |
| "rewards/rejected": -2.5627896785736084, |
| "step": 111, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.674444862627023, |
| "grad_norm": 3.891860008239746, |
| "learning_rate": 0.000158374367223479, |
| "logits/chosen": -0.2456948161125183, |
| "logits/rejected": -0.3476937413215637, |
| "logps/chosen": -6.91848087310791, |
| "logps/rejected": -35.34187698364258, |
| "loss": 0.9391785860061646, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.618550181388855, |
| "rewards/accuracies": 0.8125, |
| "rewards/chosen": -0.011886654421687126, |
| "rewards/margins": 2.1392247676849365, |
| "rewards/rejected": -2.151111602783203, |
| "step": 112, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.6804666917576214, |
| "grad_norm": 4.234856128692627, |
| "learning_rate": 0.00015756170656856737, |
| "logits/chosen": -0.11636840552091599, |
| "logits/rejected": -0.23480768501758575, |
| "logps/chosen": -7.3483710289001465, |
| "logps/rejected": -33.948822021484375, |
| "loss": 0.8160063624382019, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5110282301902771, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.08705638349056244, |
| "rewards/margins": 2.111095905303955, |
| "rewards/rejected": -2.0240395069122314, |
| "step": 113, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.6864885208882198, |
| "grad_norm": 1.5901049375534058, |
| "learning_rate": 0.0001567433204758782, |
| "logits/chosen": -0.15935438871383667, |
| "logits/rejected": -0.3056758642196655, |
| "logps/chosen": -5.887468338012695, |
| "logps/rejected": -39.561500549316406, |
| "loss": 0.7292969226837158, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4773525893688202, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.141588032245636, |
| "rewards/margins": 2.442899703979492, |
| "rewards/rejected": -2.301311731338501, |
| "step": 114, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.6925103500188182, |
| "grad_norm": 2.1147372722625732, |
| "learning_rate": 0.0001559192903470747, |
| "logits/chosen": -0.1537177711725235, |
| "logits/rejected": -0.2976364493370056, |
| "logps/chosen": -7.735870361328125, |
| "logps/rejected": -36.14551544189453, |
| "loss": 0.822877049446106, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5838936567306519, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.16609230637550354, |
| "rewards/margins": 2.184666633605957, |
| "rewards/rejected": -2.0185742378234863, |
| "step": 115, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.6985321791494167, |
| "grad_norm": 3.8271219730377197, |
| "learning_rate": 0.00015508969814521025, |
| "logits/chosen": -0.16476985812187195, |
| "logits/rejected": -0.2944994568824768, |
| "logps/chosen": -4.954969882965088, |
| "logps/rejected": -31.42076301574707, |
| "loss": 0.8119629621505737, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4658912420272827, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.09785570204257965, |
| "rewards/margins": 1.8094596862792969, |
| "rewards/rejected": -1.711604118347168, |
| "step": 116, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.704554008280015, |
| "grad_norm": 1.9744106531143188, |
| "learning_rate": 0.00015425462638657595, |
| "logits/chosen": -0.2048090547323227, |
| "logits/rejected": -0.2695329189300537, |
| "logps/chosen": -5.02174711227417, |
| "logps/rejected": -31.327930450439453, |
| "loss": 0.6611427068710327, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3884551525115967, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.18278352916240692, |
| "rewards/margins": 2.1040430068969727, |
| "rewards/rejected": -1.9212596416473389, |
| "step": 117, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.7105758374106135, |
| "grad_norm": 2.7671103477478027, |
| "learning_rate": 0.00015341415813249288, |
| "logits/chosen": -0.18153494596481323, |
| "logits/rejected": -0.31533682346343994, |
| "logps/chosen": -3.5208582878112793, |
| "logps/rejected": -28.974889755249023, |
| "loss": 0.6663014888763428, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.31734898686408997, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.08336202800273895, |
| "rewards/margins": 1.5573171377182007, |
| "rewards/rejected": -1.4739550352096558, |
| "step": 118, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.7165976665412119, |
| "grad_norm": 2.2519025802612305, |
| "learning_rate": 0.00015256837698105047, |
| "logits/chosen": -0.18909907341003418, |
| "logits/rejected": -0.32073426246643066, |
| "logps/chosen": -5.691015243530273, |
| "logps/rejected": -38.229949951171875, |
| "loss": 0.6945815086364746, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4534215033054352, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.016410622745752335, |
| "rewards/margins": 2.294341564178467, |
| "rewards/rejected": -2.3107523918151855, |
| "step": 119, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.7226194956718103, |
| "grad_norm": 2.6865475177764893, |
| "learning_rate": 0.00015171736705879126, |
| "logits/chosen": -0.11694711446762085, |
| "logits/rejected": -0.30947864055633545, |
| "logps/chosen": -4.004083156585693, |
| "logps/rejected": -42.21112060546875, |
| "loss": 0.5099095702171326, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2805330157279968, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.08305128663778305, |
| "rewards/margins": 2.5462100505828857, |
| "rewards/rejected": -2.46315860748291, |
| "step": 120, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.7286413248024087, |
| "grad_norm": 5.433821201324463, |
| "learning_rate": 0.00015086121301234316, |
| "logits/chosen": -0.15977299213409424, |
| "logits/rejected": -0.30839866399765015, |
| "logps/chosen": -7.34042501449585, |
| "logps/rejected": -38.30808639526367, |
| "loss": 0.9920263886451721, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6664140224456787, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": -0.023471834138035774, |
| "rewards/margins": 2.068791151046753, |
| "rewards/rejected": -2.0922627449035645, |
| "step": 121, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.7346631539330072, |
| "grad_norm": 1.8786602020263672, |
| "learning_rate": 0.00015000000000000001, |
| "logits/chosen": -0.12786100804805756, |
| "logits/rejected": -0.2967092990875244, |
| "logps/chosen": -4.349850654602051, |
| "logps/rejected": -32.289390563964844, |
| "loss": 0.6910618543624878, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.37633225321769714, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.09151757508516312, |
| "rewards/margins": 1.6974818706512451, |
| "rewards/rejected": -1.6059644222259521, |
| "step": 122, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.7406849830636055, |
| "grad_norm": 2.327937364578247, |
| "learning_rate": 0.00014913381368325115, |
| "logits/chosen": -0.12772715091705322, |
| "logits/rejected": -0.3204698860645294, |
| "logps/chosen": -3.4514901638031006, |
| "logps/rejected": -40.17311096191406, |
| "loss": 0.5442243814468384, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2678568661212921, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.06198100745677948, |
| "rewards/margins": 2.348788022994995, |
| "rewards/rejected": -2.286807060241699, |
| "step": 123, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.746706812194204, |
| "grad_norm": 3.345496892929077, |
| "learning_rate": 0.0001482627402182611, |
| "logits/chosen": -0.23605135083198547, |
| "logits/rejected": -0.3248905539512634, |
| "logps/chosen": -6.426857948303223, |
| "logps/rejected": -27.78033447265625, |
| "loss": 0.8108398914337158, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4387803077697754, |
| "rewards/accuracies": 0.828125, |
| "rewards/chosen": 0.12587979435920715, |
| "rewards/margins": 1.6669539213180542, |
| "rewards/rejected": -1.5410741567611694, |
| "step": 124, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.7527286413248024, |
| "grad_norm": 2.567275285720825, |
| "learning_rate": 0.00014738686624729986, |
| "logits/chosen": -0.19267147779464722, |
| "logits/rejected": -0.3316629230976105, |
| "logps/chosen": -5.075592994689941, |
| "logps/rejected": -33.312400817871094, |
| "loss": 0.7099701166152954, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3936714828014374, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.13089658319950104, |
| "rewards/margins": 1.870408535003662, |
| "rewards/rejected": -1.7395117282867432, |
| "step": 125, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.7587504704554008, |
| "grad_norm": 3.0261592864990234, |
| "learning_rate": 0.00014650627889012507, |
| "logits/chosen": -0.2155275195837021, |
| "logits/rejected": -0.37678274512290955, |
| "logps/chosen": -4.682234764099121, |
| "logps/rejected": -28.974559783935547, |
| "loss": 0.7498282790184021, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4256855249404907, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.15469788014888763, |
| "rewards/margins": 1.657031774520874, |
| "rewards/rejected": -1.502333641052246, |
| "step": 126, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.7647722995859992, |
| "grad_norm": 2.6022913455963135, |
| "learning_rate": 0.0001456210657353163, |
| "logits/chosen": -0.19964055716991425, |
| "logits/rejected": -0.35465607047080994, |
| "logps/chosen": -3.941540479660034, |
| "logps/rejected": -32.9716682434082, |
| "loss": 0.6142888069152832, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3339819610118866, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.15736861526966095, |
| "rewards/margins": 2.048983335494995, |
| "rewards/rejected": -1.8916147947311401, |
| "step": 127, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.7707941287165977, |
| "grad_norm": 2.2070999145507812, |
| "learning_rate": 0.00014473131483156327, |
| "logits/chosen": -0.24945035576820374, |
| "logits/rejected": -0.32592612504959106, |
| "logps/chosen": -5.566514015197754, |
| "logps/rejected": -27.138111114501953, |
| "loss": 0.7188521027565002, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4044472575187683, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.16954132914543152, |
| "rewards/margins": 1.7891260385513306, |
| "rewards/rejected": -1.6195846796035767, |
| "step": 128, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.776815957847196, |
| "grad_norm": 1.9873756170272827, |
| "learning_rate": 0.00014383711467890774, |
| "logits/chosen": -0.18453390896320343, |
| "logits/rejected": -0.3942897319793701, |
| "logps/chosen": -4.547736167907715, |
| "logps/rejected": -43.979976654052734, |
| "loss": 0.5406314730644226, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3523382544517517, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.08807562291622162, |
| "rewards/margins": 2.9467825889587402, |
| "rewards/rejected": -2.8587071895599365, |
| "step": 129, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.7828377869777945, |
| "grad_norm": 3.6719274520874023, |
| "learning_rate": 0.00014293855421994094, |
| "logits/chosen": -0.20188507437705994, |
| "logits/rejected": -0.34430867433547974, |
| "logps/chosen": -5.333648204803467, |
| "logps/rejected": -41.819515228271484, |
| "loss": 0.7074599862098694, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.44100311398506165, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.027486218139529228, |
| "rewards/margins": 2.8748302459716797, |
| "rewards/rejected": -2.847343921661377, |
| "step": 130, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.788859616108393, |
| "grad_norm": 3.5544822216033936, |
| "learning_rate": 0.00014203572283095657, |
| "logits/chosen": -0.1257411241531372, |
| "logits/rejected": -0.35228949785232544, |
| "logps/chosen": -5.203995704650879, |
| "logps/rejected": -53.26233673095703, |
| "loss": 0.763848602771759, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5348649024963379, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.0426684133708477, |
| "rewards/margins": 3.6239006519317627, |
| "rewards/rejected": -3.5812320709228516, |
| "step": 131, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.7948814452389913, |
| "grad_norm": 2.9171648025512695, |
| "learning_rate": 0.00014112871031306119, |
| "logits/chosen": -0.19912837445735931, |
| "logits/rejected": -0.3755126893520355, |
| "logps/chosen": -7.589679718017578, |
| "logps/rejected": -46.79793930053711, |
| "loss": 0.7475419640541077, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5176687240600586, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.19966475665569305, |
| "rewards/margins": 3.243659496307373, |
| "rewards/rejected": -3.043994426727295, |
| "step": 132, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.8009032743695897, |
| "grad_norm": 3.5576326847076416, |
| "learning_rate": 0.00014021760688324176, |
| "logits/chosen": -0.17760571837425232, |
| "logits/rejected": -0.34351277351379395, |
| "logps/chosen": -6.067855358123779, |
| "logps/rejected": -45.17057800292969, |
| "loss": 0.7542024850845337, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.46581241488456726, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": -0.06834451109170914, |
| "rewards/margins": 2.9498496055603027, |
| "rewards/rejected": -3.0181939601898193, |
| "step": 133, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.8069251035001882, |
| "grad_norm": 3.7524359226226807, |
| "learning_rate": 0.00013930250316539238, |
| "logits/chosen": -0.24070125818252563, |
| "logits/rejected": -0.36900123953819275, |
| "logps/chosen": -6.3118109703063965, |
| "logps/rejected": -41.5582389831543, |
| "loss": 0.8333712220191956, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5454068183898926, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.0363120511174202, |
| "rewards/margins": 2.832289695739746, |
| "rewards/rejected": -2.7959775924682617, |
| "step": 134, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.8129469326307867, |
| "grad_norm": 4.076293468475342, |
| "learning_rate": 0.00013838349018130007, |
| "logits/chosen": -0.22921916842460632, |
| "logits/rejected": -0.3980764150619507, |
| "logps/chosen": -6.020735263824463, |
| "logps/rejected": -36.49159622192383, |
| "loss": 0.7660678029060364, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5027486085891724, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.11397107690572739, |
| "rewards/margins": 2.2893691062927246, |
| "rewards/rejected": -2.175398349761963, |
| "step": 135, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.818968761761385, |
| "grad_norm": 3.8375930786132812, |
| "learning_rate": 0.00013746065934159123, |
| "logits/chosen": -0.21597175300121307, |
| "logits/rejected": -0.3748128414154053, |
| "logps/chosen": -5.917651176452637, |
| "logps/rejected": -34.7623176574707, |
| "loss": 0.7133091688156128, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.43554598093032837, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.14565804600715637, |
| "rewards/margins": 2.124354124069214, |
| "rewards/rejected": -1.9786962270736694, |
| "step": 136, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.8249905908919835, |
| "grad_norm": 3.4924089908599854, |
| "learning_rate": 0.00013653410243663952, |
| "logits/chosen": -0.18882140517234802, |
| "logits/rejected": -0.39621540904045105, |
| "logps/chosen": -3.5247278213500977, |
| "logps/rejected": -39.52235794067383, |
| "loss": 0.5699490308761597, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3287566304206848, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.16140440106391907, |
| "rewards/margins": 2.0828282833099365, |
| "rewards/rejected": -1.9214237928390503, |
| "step": 137, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.8310124200225819, |
| "grad_norm": 2.952988386154175, |
| "learning_rate": 0.00013560391162743569, |
| "logits/chosen": -0.1859571784734726, |
| "logits/rejected": -0.381632536649704, |
| "logps/chosen": -4.896010398864746, |
| "logps/rejected": -35.494384765625, |
| "loss": 0.7929538488388062, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4508986473083496, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.021028656512498856, |
| "rewards/margins": 1.990783452987671, |
| "rewards/rejected": -1.969754695892334, |
| "step": 138, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.8370342491531803, |
| "grad_norm": 2.0521535873413086, |
| "learning_rate": 0.00013467017943642073, |
| "logits/chosen": -0.24490389227867126, |
| "logits/rejected": -0.42894116044044495, |
| "logps/chosen": -4.2235565185546875, |
| "logps/rejected": -38.84959411621094, |
| "loss": 0.6944494247436523, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.43725934624671936, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.12278101593255997, |
| "rewards/margins": 2.2289106845855713, |
| "rewards/rejected": -2.1061296463012695, |
| "step": 139, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.8430560782837787, |
| "grad_norm": 3.664316415786743, |
| "learning_rate": 0.00013373299873828303, |
| "logits/chosen": -0.29347366094589233, |
| "logits/rejected": -0.41366317868232727, |
| "logps/chosen": -4.828797340393066, |
| "logps/rejected": -31.14942169189453, |
| "loss": 0.6968463659286499, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.41966789960861206, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.13118553161621094, |
| "rewards/margins": 1.9206351041793823, |
| "rewards/rejected": -1.7894494533538818, |
| "step": 140, |
| "train_speed(iter/s)": 0.005635 |
| }, |
| { |
| "epoch": 0.8490779074143772, |
| "grad_norm": 3.614262580871582, |
| "learning_rate": 0.00013279246275072046, |
| "logits/chosen": -0.31295719742774963, |
| "logits/rejected": -0.4447307288646698, |
| "logps/chosen": -8.66226577758789, |
| "logps/rejected": -31.42186737060547, |
| "loss": 1.0331419706344604, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6634517312049866, |
| "rewards/accuracies": 0.78125, |
| "rewards/chosen": 0.08492518961429596, |
| "rewards/margins": 1.7986928224563599, |
| "rewards/rejected": -1.713767647743225, |
| "step": 141, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.8550997365449755, |
| "grad_norm": 1.7650597095489502, |
| "learning_rate": 0.00013184866502516845, |
| "logits/chosen": -0.26669561862945557, |
| "logits/rejected": -0.4871166944503784, |
| "logps/chosen": -3.3100223541259766, |
| "logps/rejected": -39.79899215698242, |
| "loss": 0.5506975650787354, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.29588890075683594, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.10198746621608734, |
| "rewards/margins": 2.259812831878662, |
| "rewards/rejected": -2.157825231552124, |
| "step": 142, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.861121565675574, |
| "grad_norm": 3.4034411907196045, |
| "learning_rate": 0.00013090169943749476, |
| "logits/chosen": -0.3030458390712738, |
| "logits/rejected": -0.45216381549835205, |
| "logps/chosen": -5.469812393188477, |
| "logps/rejected": -35.740806579589844, |
| "loss": 0.6628592610359192, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.34965524077415466, |
| "rewards/accuracies": 0.84375, |
| "rewards/chosen": 0.07742348313331604, |
| "rewards/margins": 2.012543201446533, |
| "rewards/rejected": -1.935119867324829, |
| "step": 143, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.8671433948061724, |
| "grad_norm": 2.271817922592163, |
| "learning_rate": 0.00012995166017866193, |
| "logits/chosen": -0.35303038358688354, |
| "logits/rejected": -0.4534931778907776, |
| "logps/chosen": -7.304016590118408, |
| "logps/rejected": -30.363569259643555, |
| "loss": 0.8833200931549072, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5442834496498108, |
| "rewards/accuracies": 0.765625, |
| "rewards/chosen": 0.18977457284927368, |
| "rewards/margins": 2.04044508934021, |
| "rewards/rejected": -1.850670576095581, |
| "step": 144, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.8731652239367708, |
| "grad_norm": 1.6546629667282104, |
| "learning_rate": 0.00012899864174535864, |
| "logits/chosen": -0.30710387229919434, |
| "logits/rejected": -0.47071051597595215, |
| "logps/chosen": -4.921065330505371, |
| "logps/rejected": -37.83810043334961, |
| "loss": 0.611689567565918, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.36449986696243286, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.14039897918701172, |
| "rewards/margins": 2.364004373550415, |
| "rewards/rejected": -2.2236056327819824, |
| "step": 145, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.8791870530673692, |
| "grad_norm": 1.4133797883987427, |
| "learning_rate": 0.00012804273893060028, |
| "logits/chosen": -0.35278186202049255, |
| "logits/rejected": -0.4680350422859192, |
| "logps/chosen": -4.455732345581055, |
| "logps/rejected": -32.905887603759766, |
| "loss": 0.5695258975028992, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3188043534755707, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.18160954117774963, |
| "rewards/margins": 2.25382137298584, |
| "rewards/rejected": -2.072211742401123, |
| "step": 146, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.8852088821979677, |
| "grad_norm": 1.8506555557250977, |
| "learning_rate": 0.00012708404681430053, |
| "logits/chosen": -0.3219718635082245, |
| "logits/rejected": -0.4345773756504059, |
| "logps/chosen": -5.016592979431152, |
| "logps/rejected": -34.69657516479492, |
| "loss": 0.6405315399169922, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.35303816199302673, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.13450342416763306, |
| "rewards/margins": 2.0750272274017334, |
| "rewards/rejected": -1.9405235052108765, |
| "step": 147, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.891230711328566, |
| "grad_norm": 2.6582088470458984, |
| "learning_rate": 0.00012612266075381386, |
| "logits/chosen": -0.2835950255393982, |
| "logits/rejected": -0.4708768129348755, |
| "logps/chosen": -6.048505783081055, |
| "logps/rejected": -42.608131408691406, |
| "loss": 0.676105260848999, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4015389084815979, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.033125825226306915, |
| "rewards/margins": 2.4881484508514404, |
| "rewards/rejected": -2.4550228118896484, |
| "step": 148, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.8972525404591645, |
| "grad_norm": 3.3612794876098633, |
| "learning_rate": 0.00012515867637445086, |
| "logits/chosen": -0.3777806758880615, |
| "logits/rejected": -0.477069228887558, |
| "logps/chosen": -6.0037922859191895, |
| "logps/rejected": -32.48824691772461, |
| "loss": 0.825937032699585, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5271134972572327, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.07130320370197296, |
| "rewards/margins": 2.3660011291503906, |
| "rewards/rejected": -2.2946979999542236, |
| "step": 149, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.9032743695897629, |
| "grad_norm": 1.5847569704055786, |
| "learning_rate": 0.00012419218955996676, |
| "logits/chosen": -0.31665414571762085, |
| "logits/rejected": -0.4995279312133789, |
| "logps/chosen": -6.320664882659912, |
| "logps/rejected": -52.651451110839844, |
| "loss": 0.5624623894691467, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.368425577878952, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.14041581749916077, |
| "rewards/margins": 3.4388084411621094, |
| "rewards/rejected": -3.2983925342559814, |
| "step": 150, |
| "train_speed(iter/s)": 0.005636 |
| }, |
| { |
| "epoch": 0.9092961987203613, |
| "grad_norm": 1.733017921447754, |
| "learning_rate": 0.00012322329644302426, |
| "logits/chosen": -0.3288433849811554, |
| "logits/rejected": -0.5056887865066528, |
| "logps/chosen": -4.5702128410339355, |
| "logps/rejected": -42.928321838378906, |
| "loss": 0.534808337688446, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3458808958530426, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.14961260557174683, |
| "rewards/margins": 3.0105981826782227, |
| "rewards/rejected": -2.860985279083252, |
| "step": 151, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9153180278509597, |
| "grad_norm": 1.9891310930252075, |
| "learning_rate": 0.00012225209339563145, |
| "logits/chosen": -0.34356942772865295, |
| "logits/rejected": -0.46553438901901245, |
| "logps/chosen": -8.254244804382324, |
| "logps/rejected": -43.76982116699219, |
| "loss": 0.7122747302055359, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5048399567604065, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.2129802703857422, |
| "rewards/margins": 3.227698802947998, |
| "rewards/rejected": -3.0147182941436768, |
| "step": 152, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9213398569815582, |
| "grad_norm": 4.222365856170654, |
| "learning_rate": 0.00012127867701955622, |
| "logits/chosen": -0.3883526921272278, |
| "logits/rejected": -0.5774507522583008, |
| "logps/chosen": -6.422748565673828, |
| "logps/rejected": -53.63523483276367, |
| "loss": 0.6593223214149475, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.48106443881988525, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": -0.016244899481534958, |
| "rewards/margins": 4.012790203094482, |
| "rewards/rejected": -4.029035568237305, |
| "step": 153, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9273616861121565, |
| "grad_norm": 14.615867614746094, |
| "learning_rate": 0.00012030314413671762, |
| "logits/chosen": -0.39986497163772583, |
| "logits/rejected": -0.6344658136367798, |
| "logps/chosen": -4.43770170211792, |
| "logps/rejected": -55.2861213684082, |
| "loss": 0.5306422710418701, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3766982853412628, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.10156996548175812, |
| "rewards/margins": 3.934446334838867, |
| "rewards/rejected": -3.832875967025757, |
| "step": 154, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.933383515242755, |
| "grad_norm": 4.434772491455078, |
| "learning_rate": 0.00011932559177955533, |
| "logits/chosen": -0.4241827726364136, |
| "logits/rejected": -0.6216625571250916, |
| "logps/chosen": -5.785084247589111, |
| "logps/rejected": -56.10906982421875, |
| "loss": 0.7327705025672913, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5102080702781677, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.01768992468714714, |
| "rewards/margins": 4.036444187164307, |
| "rewards/rejected": -4.018754005432129, |
| "step": 155, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9394053443733534, |
| "grad_norm": 3.1120123863220215, |
| "learning_rate": 0.00011834611718137824, |
| "logits/chosen": -0.364039808511734, |
| "logits/rejected": -0.6007083654403687, |
| "logps/chosen": -5.2767415046691895, |
| "logps/rejected": -66.11264038085938, |
| "loss": 0.5869796276092529, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.47315409779548645, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.10729408264160156, |
| "rewards/margins": 4.741492748260498, |
| "rewards/rejected": -4.6341986656188965, |
| "step": 156, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9454271735039518, |
| "grad_norm": 9.935791015625, |
| "learning_rate": 0.00011736481776669306, |
| "logits/chosen": -0.43170925974845886, |
| "logits/rejected": -0.5715627670288086, |
| "logps/chosen": -9.369845390319824, |
| "logps/rejected": -57.28983688354492, |
| "loss": 1.0344016551971436, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.7142276763916016, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": -0.12030352652072906, |
| "rewards/margins": 3.983743190765381, |
| "rewards/rejected": -4.1040472984313965, |
| "step": 157, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9514490026345502, |
| "grad_norm": 3.4565649032592773, |
| "learning_rate": 0.00011638179114151377, |
| "logits/chosen": -0.4627074599266052, |
| "logits/rejected": -0.6269153952598572, |
| "logps/chosen": -6.801128387451172, |
| "logps/rejected": -50.93914031982422, |
| "loss": 0.7641481757164001, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.5224257111549377, |
| "rewards/accuracies": 0.859375, |
| "rewards/chosen": 0.020412985235452652, |
| "rewards/margins": 3.8560848236083984, |
| "rewards/rejected": -3.835671901702881, |
| "step": 158, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9574708317651487, |
| "grad_norm": 7.00652551651001, |
| "learning_rate": 0.00011539713508365335, |
| "logits/chosen": -0.4332171678543091, |
| "logits/rejected": -0.589859127998352, |
| "logps/chosen": -6.74440860748291, |
| "logps/rejected": -51.30781173706055, |
| "loss": 0.8142269253730774, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6082335710525513, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.04817220941185951, |
| "rewards/margins": 3.839707136154175, |
| "rewards/rejected": -3.791534662246704, |
| "step": 159, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.963492660895747, |
| "grad_norm": 4.027029514312744, |
| "learning_rate": 0.00011441094753299801, |
| "logits/chosen": -0.4457279443740845, |
| "logits/rejected": -0.5885544419288635, |
| "logps/chosen": -7.359399318695068, |
| "logps/rejected": -45.14234924316406, |
| "loss": 0.572091281414032, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.35912448167800903, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.04880950227379799, |
| "rewards/margins": 3.246192216873169, |
| "rewards/rejected": -3.1973824501037598, |
| "step": 160, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.9695144900263455, |
| "grad_norm": 3.832231044769287, |
| "learning_rate": 0.00011342332658176555, |
| "logits/chosen": -0.3761252760887146, |
| "logits/rejected": -0.5621005892753601, |
| "logps/chosen": -5.201396942138672, |
| "logps/rejected": -43.30262756347656, |
| "loss": 0.601041853427887, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.37146201729774475, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.1526133269071579, |
| "rewards/margins": 3.075632095336914, |
| "rewards/rejected": -2.92301869392395, |
| "step": 161, |
| "train_speed(iter/s)": 0.005634 |
| }, |
| { |
| "epoch": 0.9755363191569439, |
| "grad_norm": 3.592341184616089, |
| "learning_rate": 0.00011243437046474853, |
| "logits/chosen": -0.2978363037109375, |
| "logits/rejected": -0.5343031883239746, |
| "logps/chosen": -5.316001892089844, |
| "logps/rejected": -56.440513610839844, |
| "loss": 0.5654405951499939, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.40133601427078247, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.0053778961300849915, |
| "rewards/margins": 3.901097536087036, |
| "rewards/rejected": -3.895719289779663, |
| "step": 162, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9815581482875423, |
| "grad_norm": 3.0024077892303467, |
| "learning_rate": 0.0001114441775495432, |
| "logits/chosen": -0.3479576110839844, |
| "logits/rejected": -0.5726731419563293, |
| "logps/chosen": -5.025125026702881, |
| "logps/rejected": -43.54122543334961, |
| "loss": 0.7037834525108337, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.48029187321662903, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.18540027737617493, |
| "rewards/margins": 2.662545919418335, |
| "rewards/rejected": -2.4771456718444824, |
| "step": 163, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9875799774181407, |
| "grad_norm": 5.379988670349121, |
| "learning_rate": 0.00011045284632676536, |
| "logits/chosen": -0.3792319893836975, |
| "logits/rejected": -0.5281450152397156, |
| "logps/chosen": -8.132346153259277, |
| "logps/rejected": -42.75868606567383, |
| "loss": 0.7357421517372131, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4848954677581787, |
| "rewards/accuracies": 0.890625, |
| "rewards/chosen": 0.09983620047569275, |
| "rewards/margins": 2.911416530609131, |
| "rewards/rejected": -2.8115804195404053, |
| "step": 164, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9936018065487392, |
| "grad_norm": 7.149594306945801, |
| "learning_rate": 0.00010946047540025372, |
| "logits/chosen": -0.362922728061676, |
| "logits/rejected": -0.5330293774604797, |
| "logps/chosen": -5.657723903656006, |
| "logps/rejected": -44.190521240234375, |
| "loss": 0.777948796749115, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4741102457046509, |
| "rewards/accuracies": 0.875, |
| "rewards/chosen": 0.0894114077091217, |
| "rewards/margins": 2.840927839279175, |
| "rewards/rejected": -2.751516580581665, |
| "step": 165, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 0.9996236356793377, |
| "grad_norm": 2.048894166946411, |
| "learning_rate": 0.00010846716347726233, |
| "logits/chosen": -0.4384031891822815, |
| "logits/rejected": -0.5566267371177673, |
| "logps/chosen": -3.853196382522583, |
| "logps/rejected": -43.331207275390625, |
| "loss": 0.5384762287139893, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.346954882144928, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.1474352478981018, |
| "rewards/margins": 3.1648757457733154, |
| "rewards/rejected": -3.0174405574798584, |
| "step": 166, |
| "train_speed(iter/s)": 0.005633 |
| }, |
| { |
| "epoch": 1.0, |
| "grad_norm": 2.048894166946411, |
| "learning_rate": 0.00010747300935864243, |
| "logits/chosen": -0.37010034918785095, |
| "logits/rejected": -0.6208374500274658, |
| "logps/chosen": -8.17520809173584, |
| "logps/rejected": -61.29972839355469, |
| "loss": 0.12779855728149414, |
| "memory(GiB)": 46.13, |
| "nll_loss": 1.3625344038009644, |
| "rewards/accuracies": 0.5, |
| "rewards/chosen": -0.4605603814125061, |
| "rewards/margins": 4.346360683441162, |
| "rewards/rejected": -4.806921005249023, |
| "step": 167, |
| "train_speed(iter/s)": 0.005666 |
| }, |
| { |
| "epoch": 1.0060218291305985, |
| "grad_norm": 13.219114303588867, |
| "learning_rate": 0.00010647811192901518, |
| "logits/chosen": -0.34268608689308167, |
| "logits/rejected": -0.5246624946594238, |
| "logps/chosen": -4.731266975402832, |
| "logps/rejected": -46.866214752197266, |
| "loss": 0.5699129104614258, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.35907092690467834, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.17517316341400146, |
| "rewards/margins": 3.2273640632629395, |
| "rewards/rejected": -3.0521905422210693, |
| "step": 168, |
| "train_speed(iter/s)": 0.005665 |
| }, |
| { |
| "epoch": 1.012043658261197, |
| "grad_norm": 1.8148096799850464, |
| "learning_rate": 0.00010548257014693601, |
| "logits/chosen": -0.3512643575668335, |
| "logits/rejected": -0.5287365913391113, |
| "logps/chosen": -3.7313232421875, |
| "logps/rejected": -46.643409729003906, |
| "loss": 0.39852774143218994, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.22772511839866638, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.18213030695915222, |
| "rewards/margins": 3.319718837738037, |
| "rewards/rejected": -3.1375885009765625, |
| "step": 169, |
| "train_speed(iter/s)": 0.005665 |
| }, |
| { |
| "epoch": 1.0180654873917954, |
| "grad_norm": 3.1325247287750244, |
| "learning_rate": 0.00010448648303505151, |
| "logits/chosen": -0.3514709770679474, |
| "logits/rejected": -0.5060309171676636, |
| "logps/chosen": -2.770275592803955, |
| "logps/rejected": -42.078311920166016, |
| "loss": 0.4157463312149048, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2555791139602661, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.31045016646385193, |
| "rewards/margins": 3.08093523979187, |
| "rewards/rejected": -2.770484685897827, |
| "step": 170, |
| "train_speed(iter/s)": 0.005665 |
| }, |
| { |
| "epoch": 1.0240873165223936, |
| "grad_norm": 2.593531370162964, |
| "learning_rate": 0.00010348994967025012, |
| "logits/chosen": -0.39463332295417786, |
| "logits/rejected": -0.5413990616798401, |
| "logps/chosen": -5.247183322906494, |
| "logps/rejected": -46.65768051147461, |
| "loss": 0.5715495347976685, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3746676445007324, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.16656973958015442, |
| "rewards/margins": 3.0420989990234375, |
| "rewards/rejected": -2.8755292892456055, |
| "step": 171, |
| "train_speed(iter/s)": 0.005665 |
| }, |
| { |
| "epoch": 1.030109145652992, |
| "grad_norm": 1.7776541709899902, |
| "learning_rate": 0.0001024930691738073, |
| "logits/chosen": -0.3912968635559082, |
| "logits/rejected": -0.514131486415863, |
| "logps/chosen": -6.068740367889404, |
| "logps/rejected": -39.78652572631836, |
| "loss": 0.590811550617218, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.396287202835083, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2542215585708618, |
| "rewards/margins": 2.7374861240386963, |
| "rewards/rejected": -2.483264923095703, |
| "step": 172, |
| "train_speed(iter/s)": 0.005664 |
| }, |
| { |
| "epoch": 1.0361309747835905, |
| "grad_norm": 1.4022067785263062, |
| "learning_rate": 0.00010149594070152638, |
| "logits/chosen": -0.3674977123737335, |
| "logits/rejected": -0.5223217606544495, |
| "logps/chosen": -2.335606098175049, |
| "logps/rejected": -41.703453063964844, |
| "loss": 0.42932039499282837, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.24382930994033813, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.26927071809768677, |
| "rewards/margins": 2.965996503829956, |
| "rewards/rejected": -2.696725845336914, |
| "step": 173, |
| "train_speed(iter/s)": 0.005664 |
| }, |
| { |
| "epoch": 1.042152803914189, |
| "grad_norm": 2.1577670574188232, |
| "learning_rate": 0.00010049866343387581, |
| "logits/chosen": -0.33516278862953186, |
| "logits/rejected": -0.4972761273384094, |
| "logps/chosen": -3.2422990798950195, |
| "logps/rejected": -43.91548538208008, |
| "loss": 0.4766288697719574, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2899796962738037, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.25678345561027527, |
| "rewards/margins": 3.0347728729248047, |
| "rewards/rejected": -2.777989149093628, |
| "step": 174, |
| "train_speed(iter/s)": 0.005664 |
| }, |
| { |
| "epoch": 1.0481746330447874, |
| "grad_norm": 1.4100942611694336, |
| "learning_rate": 9.950133656612421e-05, |
| "logits/chosen": -0.3753768503665924, |
| "logits/rejected": -0.49912041425704956, |
| "logps/chosen": -3.9077322483062744, |
| "logps/rejected": -36.67354965209961, |
| "loss": 0.44091612100601196, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2640770673751831, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.37277600169181824, |
| "rewards/margins": 2.572587013244629, |
| "rewards/rejected": -2.1998109817504883, |
| "step": 175, |
| "train_speed(iter/s)": 0.005664 |
| }, |
| { |
| "epoch": 1.0541964621753859, |
| "grad_norm": 1.346986174583435, |
| "learning_rate": 9.850405929847366e-05, |
| "logits/chosen": -0.38442033529281616, |
| "logits/rejected": -0.5594974160194397, |
| "logps/chosen": -3.173485040664673, |
| "logps/rejected": -43.258094787597656, |
| "loss": 0.43492069840431213, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.28007155656814575, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.28305596113204956, |
| "rewards/margins": 3.2277960777282715, |
| "rewards/rejected": -2.9447402954101562, |
| "step": 176, |
| "train_speed(iter/s)": 0.005664 |
| }, |
| { |
| "epoch": 1.060218291305984, |
| "grad_norm": 1.9254361391067505, |
| "learning_rate": 9.750693082619273e-05, |
| "logits/chosen": -0.3788023889064789, |
| "logits/rejected": -0.5781509876251221, |
| "logps/chosen": -2.432941198348999, |
| "logps/rejected": -43.248146057128906, |
| "loss": 0.40009191632270813, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.26620903611183167, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.17993758618831635, |
| "rewards/margins": 3.0553667545318604, |
| "rewards/rejected": -2.8754286766052246, |
| "step": 177, |
| "train_speed(iter/s)": 0.005664 |
| }, |
| { |
| "epoch": 1.0662401204365826, |
| "grad_norm": 3.2801098823547363, |
| "learning_rate": 9.651005032974994e-05, |
| "logits/chosen": -0.33594202995300293, |
| "logits/rejected": -0.564365029335022, |
| "logps/chosen": -3.6123833656311035, |
| "logps/rejected": -49.69728088378906, |
| "loss": 0.4477875828742981, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2894076108932495, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.19575706124305725, |
| "rewards/margins": 3.35469388961792, |
| "rewards/rejected": -3.1589369773864746, |
| "step": 178, |
| "train_speed(iter/s)": 0.005664 |
| }, |
| { |
| "epoch": 1.072261949567181, |
| "grad_norm": 1.4043627977371216, |
| "learning_rate": 9.551351696494854e-05, |
| "logits/chosen": -0.37707191705703735, |
| "logits/rejected": -0.5827841758728027, |
| "logps/chosen": -2.930352210998535, |
| "logps/rejected": -50.44269943237305, |
| "loss": 0.3742366135120392, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2148890197277069, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.21785704791545868, |
| "rewards/margins": 3.5218472480773926, |
| "rewards/rejected": -3.303990364074707, |
| "step": 179, |
| "train_speed(iter/s)": 0.005664 |
| }, |
| { |
| "epoch": 1.0782837786977795, |
| "grad_norm": 1.5572144985198975, |
| "learning_rate": 9.451742985306398e-05, |
| "logits/chosen": -0.45742160081863403, |
| "logits/rejected": -0.6289136409759521, |
| "logps/chosen": -4.650609016418457, |
| "logps/rejected": -44.8750114440918, |
| "loss": 0.4705524444580078, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3358931541442871, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2931283414363861, |
| "rewards/margins": 3.50144624710083, |
| "rewards/rejected": -3.2083182334899902, |
| "step": 180, |
| "train_speed(iter/s)": 0.005664 |
| }, |
| { |
| "epoch": 1.084305607828378, |
| "grad_norm": 2.9546408653259277, |
| "learning_rate": 9.352188807098481e-05, |
| "logits/chosen": -0.4296163320541382, |
| "logits/rejected": -0.5809882879257202, |
| "logps/chosen": -4.599586009979248, |
| "logps/rejected": -42.98283386230469, |
| "loss": 0.5064187049865723, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.32983240485191345, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.33693230152130127, |
| "rewards/margins": 3.1911778450012207, |
| "rewards/rejected": -2.854245662689209, |
| "step": 181, |
| "train_speed(iter/s)": 0.005663 |
| }, |
| { |
| "epoch": 1.0903274369589764, |
| "grad_norm": 1.369982123374939, |
| "learning_rate": 9.252699064135758e-05, |
| "logits/chosen": -0.5194458961486816, |
| "logits/rejected": -0.7049198746681213, |
| "logps/chosen": -2.871542453765869, |
| "logps/rejected": -51.704444885253906, |
| "loss": 0.36053964495658875, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.25251564383506775, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.25039854645729065, |
| "rewards/margins": 4.2018303871154785, |
| "rewards/rejected": -3.9514317512512207, |
| "step": 182, |
| "train_speed(iter/s)": 0.005664 |
| }, |
| { |
| "epoch": 1.0963492660895746, |
| "grad_norm": 3.087183713912964, |
| "learning_rate": 9.153283652273768e-05, |
| "logits/chosen": -0.42322614789009094, |
| "logits/rejected": -0.687001645565033, |
| "logps/chosen": -5.198493957519531, |
| "logps/rejected": -59.92394256591797, |
| "loss": 0.49455440044403076, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.391512006521225, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.19199909269809723, |
| "rewards/margins": 4.593501091003418, |
| "rewards/rejected": -4.401501178741455, |
| "step": 183, |
| "train_speed(iter/s)": 0.005663 |
| }, |
| { |
| "epoch": 1.102371095220173, |
| "grad_norm": 2.106433391571045, |
| "learning_rate": 9.05395245997463e-05, |
| "logits/chosen": -0.44497790932655334, |
| "logits/rejected": -0.6745092272758484, |
| "logps/chosen": -4.499228477478027, |
| "logps/rejected": -65.7539291381836, |
| "loss": 0.4104112982749939, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3255768418312073, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.31253090500831604, |
| "rewards/margins": 5.569586753845215, |
| "rewards/rejected": -5.257055759429932, |
| "step": 184, |
| "train_speed(iter/s)": 0.005663 |
| }, |
| { |
| "epoch": 1.1083929243507715, |
| "grad_norm": 3.478254556655884, |
| "learning_rate": 8.954715367323468e-05, |
| "logits/chosen": -0.4755058288574219, |
| "logits/rejected": -0.7400538325309753, |
| "logps/chosen": -4.192931175231934, |
| "logps/rejected": -58.25489044189453, |
| "loss": 0.49360448122024536, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3856234550476074, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.23331968486309052, |
| "rewards/margins": 4.7285542488098145, |
| "rewards/rejected": -4.495234489440918, |
| "step": 185, |
| "train_speed(iter/s)": 0.005663 |
| }, |
| { |
| "epoch": 1.11441475348137, |
| "grad_norm": 3.7779037952423096, |
| "learning_rate": 8.855582245045683e-05, |
| "logits/chosen": -0.47070175409317017, |
| "logits/rejected": -0.6918365955352783, |
| "logps/chosen": -5.071094036102295, |
| "logps/rejected": -53.90576934814453, |
| "loss": 0.6311454772949219, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.44870883226394653, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.04279123246669769, |
| "rewards/margins": 4.136241912841797, |
| "rewards/rejected": -4.093451023101807, |
| "step": 186, |
| "train_speed(iter/s)": 0.005663 |
| }, |
| { |
| "epoch": 1.1204365826119684, |
| "grad_norm": 7.022064685821533, |
| "learning_rate": 8.756562953525152e-05, |
| "logits/chosen": -0.432133287191391, |
| "logits/rejected": -0.776434063911438, |
| "logps/chosen": -3.279095411300659, |
| "logps/rejected": -69.65541076660156, |
| "loss": 0.3344402611255646, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2780175805091858, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.25577378273010254, |
| "rewards/margins": 5.72814416885376, |
| "rewards/rejected": -5.472370624542236, |
| "step": 187, |
| "train_speed(iter/s)": 0.005663 |
| }, |
| { |
| "epoch": 1.126458411742567, |
| "grad_norm": 2.485914945602417, |
| "learning_rate": 8.657667341823448e-05, |
| "logits/chosen": -0.49354246258735657, |
| "logits/rejected": -0.7343668341636658, |
| "logps/chosen": -4.402050971984863, |
| "logps/rejected": -56.273170471191406, |
| "loss": 0.454555481672287, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3319138288497925, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.23367615044116974, |
| "rewards/margins": 4.513181209564209, |
| "rewards/rejected": -4.279505252838135, |
| "step": 188, |
| "train_speed(iter/s)": 0.005663 |
| }, |
| { |
| "epoch": 1.1324802408731651, |
| "grad_norm": 1.653633713722229, |
| "learning_rate": 8.558905246700201e-05, |
| "logits/chosen": -0.47012466192245483, |
| "logits/rejected": -0.7377257943153381, |
| "logps/chosen": -2.28665828704834, |
| "logps/rejected": -65.824462890625, |
| "loss": 0.31783220171928406, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.21150296926498413, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2025526762008667, |
| "rewards/margins": 5.142179012298584, |
| "rewards/rejected": -4.939626216888428, |
| "step": 189, |
| "train_speed(iter/s)": 0.005663 |
| }, |
| { |
| "epoch": 1.1385020700037636, |
| "grad_norm": 1.506331443786621, |
| "learning_rate": 8.460286491634663e-05, |
| "logits/chosen": -0.45535990595817566, |
| "logits/rejected": -0.6702480316162109, |
| "logps/chosen": -6.301774978637695, |
| "logps/rejected": -60.1036262512207, |
| "loss": 0.41861438751220703, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.35035622119903564, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.4653804302215576, |
| "rewards/margins": 4.745593547821045, |
| "rewards/rejected": -4.280213356018066, |
| "step": 190, |
| "train_speed(iter/s)": 0.005662 |
| }, |
| { |
| "epoch": 1.144523899134362, |
| "grad_norm": 7.771515369415283, |
| "learning_rate": 8.361820885848624e-05, |
| "logits/chosen": -0.559324324131012, |
| "logits/rejected": -0.7053715586662292, |
| "logps/chosen": -6.118768692016602, |
| "logps/rejected": -52.78406524658203, |
| "loss": 0.6141310930252075, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4540223479270935, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.3042090833187103, |
| "rewards/margins": 4.374350547790527, |
| "rewards/rejected": -4.070141792297363, |
| "step": 191, |
| "train_speed(iter/s)": 0.005662 |
| }, |
| { |
| "epoch": 1.1505457282649605, |
| "grad_norm": 2.293672800064087, |
| "learning_rate": 8.263518223330697e-05, |
| "logits/chosen": -0.49931055307388306, |
| "logits/rejected": -0.7118666768074036, |
| "logps/chosen": -4.773472785949707, |
| "logps/rejected": -52.72026824951172, |
| "loss": 0.43889304995536804, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.34974291920661926, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.32993441820144653, |
| "rewards/margins": 4.337234973907471, |
| "rewards/rejected": -4.00730037689209, |
| "step": 192, |
| "train_speed(iter/s)": 0.005662 |
| }, |
| { |
| "epoch": 1.156567557395559, |
| "grad_norm": 2.9582977294921875, |
| "learning_rate": 8.165388281862178e-05, |
| "logits/chosen": -0.5245535373687744, |
| "logits/rejected": -0.7488053441047668, |
| "logps/chosen": -4.319539546966553, |
| "logps/rejected": -49.7554817199707, |
| "loss": 0.4565528333187103, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.34413373470306396, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2616381347179413, |
| "rewards/margins": 3.892477512359619, |
| "rewards/rejected": -3.6308393478393555, |
| "step": 193, |
| "train_speed(iter/s)": 0.005662 |
| }, |
| { |
| "epoch": 1.1625893865261574, |
| "grad_norm": 2.9784886837005615, |
| "learning_rate": 8.067440822044469e-05, |
| "logits/chosen": -0.548271894454956, |
| "logits/rejected": -0.7532668113708496, |
| "logps/chosen": -6.1191020011901855, |
| "logps/rejected": -50.176265716552734, |
| "loss": 0.6044948101043701, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.49495670199394226, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.20567229390144348, |
| "rewards/margins": 4.019640922546387, |
| "rewards/rejected": -3.8139681816101074, |
| "step": 194, |
| "train_speed(iter/s)": 0.005662 |
| }, |
| { |
| "epoch": 1.1686112156567559, |
| "grad_norm": 1.896803617477417, |
| "learning_rate": 7.96968558632824e-05, |
| "logits/chosen": -0.4820174276828766, |
| "logits/rejected": -0.7427547574043274, |
| "logps/chosen": -6.175469875335693, |
| "logps/rejected": -60.446834564208984, |
| "loss": 0.5533992648124695, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4257928133010864, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.25794556736946106, |
| "rewards/margins": 4.815202236175537, |
| "rewards/rejected": -4.557256698608398, |
| "step": 195, |
| "train_speed(iter/s)": 0.005662 |
| }, |
| { |
| "epoch": 1.174633044787354, |
| "grad_norm": 2.890577554702759, |
| "learning_rate": 7.872132298044382e-05, |
| "logits/chosen": -0.47485801577568054, |
| "logits/rejected": -0.7618663311004639, |
| "logps/chosen": -2.814539909362793, |
| "logps/rejected": -54.596343994140625, |
| "loss": 0.3804952800273895, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.24290579557418823, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.3001152276992798, |
| "rewards/margins": 4.137256622314453, |
| "rewards/rejected": -3.8371410369873047, |
| "step": 196, |
| "train_speed(iter/s)": 0.005662 |
| }, |
| { |
| "epoch": 1.1806548739179525, |
| "grad_norm": 3.058959722518921, |
| "learning_rate": 7.774790660436858e-05, |
| "logits/chosen": -0.5394716262817383, |
| "logits/rejected": -0.7631605267524719, |
| "logps/chosen": -5.447549819946289, |
| "logps/rejected": -47.610145568847656, |
| "loss": 0.5637054443359375, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4707457423210144, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.3298027515411377, |
| "rewards/margins": 3.690687656402588, |
| "rewards/rejected": -3.36088490486145, |
| "step": 197, |
| "train_speed(iter/s)": 0.005662 |
| }, |
| { |
| "epoch": 1.186676703048551, |
| "grad_norm": 1.5582202672958374, |
| "learning_rate": 7.677670355697577e-05, |
| "logits/chosen": -0.3999323546886444, |
| "logits/rejected": -0.7442487478256226, |
| "logps/chosen": -2.5974080562591553, |
| "logps/rejected": -62.564483642578125, |
| "loss": 0.33909475803375244, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.25633975863456726, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.32342803478240967, |
| "rewards/margins": 4.781001091003418, |
| "rewards/rejected": -4.457572937011719, |
| "step": 198, |
| "train_speed(iter/s)": 0.005662 |
| }, |
| { |
| "epoch": 1.1926985321791495, |
| "grad_norm": 1.2910958528518677, |
| "learning_rate": 7.580781044003324e-05, |
| "logits/chosen": -0.4777570068836212, |
| "logits/rejected": -0.7280046343803406, |
| "logps/chosen": -3.076066493988037, |
| "logps/rejected": -55.281494140625, |
| "loss": 0.32068008184432983, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2149108499288559, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.284832626581192, |
| "rewards/margins": 4.372798442840576, |
| "rewards/rejected": -4.087965488433838, |
| "step": 199, |
| "train_speed(iter/s)": 0.005661 |
| }, |
| { |
| "epoch": 1.198720361309748, |
| "grad_norm": 2.6249725818634033, |
| "learning_rate": 7.484132362554915e-05, |
| "logits/chosen": -0.5193163752555847, |
| "logits/rejected": -0.7307865619659424, |
| "logps/chosen": -6.235736846923828, |
| "logps/rejected": -44.48322677612305, |
| "loss": 0.6363410353660583, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.44810640811920166, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.32792535424232483, |
| "rewards/margins": 3.486876964569092, |
| "rewards/rejected": -3.158951759338379, |
| "step": 200, |
| "train_speed(iter/s)": 0.005661 |
| }, |
| { |
| "epoch": 1.2047421904403461, |
| "grad_norm": 2.804173707962036, |
| "learning_rate": 7.387733924618617e-05, |
| "logits/chosen": -0.41156628727912903, |
| "logits/rejected": -0.6807573437690735, |
| "logps/chosen": -4.78164005279541, |
| "logps/rejected": -58.340606689453125, |
| "loss": 0.43140679597854614, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.33691924810409546, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.29949134588241577, |
| "rewards/margins": 4.446183681488037, |
| "rewards/rejected": -4.146692276000977, |
| "step": 201, |
| "train_speed(iter/s)": 0.005658 |
| }, |
| { |
| "epoch": 1.2107640195709446, |
| "grad_norm": 3.942891836166382, |
| "learning_rate": 7.291595318569951e-05, |
| "logits/chosen": -0.5371648669242859, |
| "logits/rejected": -0.7549322843551636, |
| "logps/chosen": -7.573888301849365, |
| "logps/rejected": -50.77197265625, |
| "loss": 0.8032181262969971, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.6219516396522522, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.18600612878799438, |
| "rewards/margins": 3.8017468452453613, |
| "rewards/rejected": -3.6157405376434326, |
| "step": 202, |
| "train_speed(iter/s)": 0.005658 |
| }, |
| { |
| "epoch": 1.216785848701543, |
| "grad_norm": 1.7948954105377197, |
| "learning_rate": 7.195726106939974e-05, |
| "logits/chosen": -0.49177029728889465, |
| "logits/rejected": -0.726800799369812, |
| "logps/chosen": -4.9699907302856445, |
| "logps/rejected": -60.005218505859375, |
| "loss": 0.4122886061668396, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.30634433031082153, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.36401644349098206, |
| "rewards/margins": 4.706804275512695, |
| "rewards/rejected": -4.342788219451904, |
| "step": 203, |
| "train_speed(iter/s)": 0.005658 |
| }, |
| { |
| "epoch": 1.2228076778321415, |
| "grad_norm": 4.069206237792969, |
| "learning_rate": 7.100135825464139e-05, |
| "logits/chosen": -0.45375001430511475, |
| "logits/rejected": -0.7056143283843994, |
| "logps/chosen": -6.165468692779541, |
| "logps/rejected": -48.77824783325195, |
| "loss": 0.5478510856628418, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4077027440071106, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.40980643033981323, |
| "rewards/margins": 3.759075403213501, |
| "rewards/rejected": -3.349268913269043, |
| "step": 204, |
| "train_speed(iter/s)": 0.005658 |
| }, |
| { |
| "epoch": 1.22882950696274, |
| "grad_norm": 3.2379000186920166, |
| "learning_rate": 7.004833982133808e-05, |
| "logits/chosen": -0.4625844955444336, |
| "logits/rejected": -0.7497645616531372, |
| "logps/chosen": -3.2215240001678467, |
| "logps/rejected": -54.89192199707031, |
| "loss": 0.4756178557872772, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3104082942008972, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.10858356952667236, |
| "rewards/margins": 4.205223083496094, |
| "rewards/rejected": -4.096639156341553, |
| "step": 205, |
| "train_speed(iter/s)": 0.005658 |
| }, |
| { |
| "epoch": 1.2348513360933384, |
| "grad_norm": 1.2772859334945679, |
| "learning_rate": 6.909830056250527e-05, |
| "logits/chosen": -0.46088141202926636, |
| "logits/rejected": -0.696212649345398, |
| "logps/chosen": -3.752298355102539, |
| "logps/rejected": -51.43436813354492, |
| "loss": 0.3476787805557251, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.24112334847450256, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3736254572868347, |
| "rewards/margins": 4.043981552124023, |
| "rewards/rejected": -3.670356512069702, |
| "step": 206, |
| "train_speed(iter/s)": 0.005658 |
| }, |
| { |
| "epoch": 1.2408731652239369, |
| "grad_norm": 3.96500301361084, |
| "learning_rate": 6.815133497483157e-05, |
| "logits/chosen": -0.4538869261741638, |
| "logits/rejected": -0.6587194204330444, |
| "logps/chosen": -3.7387635707855225, |
| "logps/rejected": -44.64501953125, |
| "loss": 0.4004911780357361, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2752317786216736, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.32666444778442383, |
| "rewards/margins": 3.603152275085449, |
| "rewards/rejected": -3.2764875888824463, |
| "step": 207, |
| "train_speed(iter/s)": 0.005658 |
| }, |
| { |
| "epoch": 1.246894994354535, |
| "grad_norm": 1.6064969301223755, |
| "learning_rate": 6.720753724927958e-05, |
| "logits/chosen": -0.49224531650543213, |
| "logits/rejected": -0.7212605476379395, |
| "logps/chosen": -8.330548286437988, |
| "logps/rejected": -52.39625930786133, |
| "loss": 0.5388084650039673, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.440222829580307, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.4413294792175293, |
| "rewards/margins": 4.095629692077637, |
| "rewards/rejected": -3.6542999744415283, |
| "step": 208, |
| "train_speed(iter/s)": 0.005657 |
| }, |
| { |
| "epoch": 1.2529168234851336, |
| "grad_norm": 1.888574481010437, |
| "learning_rate": 6.626700126171702e-05, |
| "logits/chosen": -0.46829870343208313, |
| "logits/rejected": -0.7399269342422485, |
| "logps/chosen": -3.1671807765960693, |
| "logps/rejected": -52.16673278808594, |
| "loss": 0.35412514209747314, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.24468238651752472, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.35541072487831116, |
| "rewards/margins": 4.117459774017334, |
| "rewards/rejected": -3.7620487213134766, |
| "step": 209, |
| "train_speed(iter/s)": 0.005657 |
| }, |
| { |
| "epoch": 1.258938652615732, |
| "grad_norm": 8.906493186950684, |
| "learning_rate": 6.532982056357928e-05, |
| "logits/chosen": -0.5025848150253296, |
| "logits/rejected": -0.647516131401062, |
| "logps/chosen": -4.769540309906006, |
| "logps/rejected": -46.10678482055664, |
| "loss": 0.487936407327652, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.35828882455825806, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3698485195636749, |
| "rewards/margins": 3.711308717727661, |
| "rewards/rejected": -3.3414597511291504, |
| "step": 210, |
| "train_speed(iter/s)": 0.005657 |
| }, |
| { |
| "epoch": 1.2649604817463305, |
| "grad_norm": 2.5342342853546143, |
| "learning_rate": 6.439608837256432e-05, |
| "logits/chosen": -0.43483102321624756, |
| "logits/rejected": -0.7881991863250732, |
| "logps/chosen": -2.7267870903015137, |
| "logps/rejected": -65.4534683227539, |
| "loss": 0.3906051218509674, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.29971927404403687, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2297130823135376, |
| "rewards/margins": 5.093780040740967, |
| "rewards/rejected": -4.864067077636719, |
| "step": 211, |
| "train_speed(iter/s)": 0.005657 |
| }, |
| { |
| "epoch": 1.270982310876929, |
| "grad_norm": 1.950996994972229, |
| "learning_rate": 6.34658975633605e-05, |
| "logits/chosen": -0.35362473130226135, |
| "logits/rejected": -0.6550623774528503, |
| "logps/chosen": -4.400423526763916, |
| "logps/rejected": -51.62921905517578, |
| "loss": 0.49274933338165283, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3038281798362732, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.2719678282737732, |
| "rewards/margins": 3.763463258743286, |
| "rewards/rejected": -3.4914956092834473, |
| "step": 212, |
| "train_speed(iter/s)": 0.005657 |
| }, |
| { |
| "epoch": 1.2770041400075272, |
| "grad_norm": 1.5276005268096924, |
| "learning_rate": 6.25393406584088e-05, |
| "logits/chosen": -0.40574443340301514, |
| "logits/rejected": -0.7509509921073914, |
| "logps/chosen": -3.0956404209136963, |
| "logps/rejected": -59.7081298828125, |
| "loss": 0.34677064418792725, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2606160342693329, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.23602759838104248, |
| "rewards/margins": 4.654949188232422, |
| "rewards/rejected": -4.418920993804932, |
| "step": 213, |
| "train_speed(iter/s)": 0.005657 |
| }, |
| { |
| "epoch": 1.2830259691381256, |
| "grad_norm": 1.9581797122955322, |
| "learning_rate": 6.161650981869998e-05, |
| "logits/chosen": -0.4477730691432953, |
| "logits/rejected": -0.708689272403717, |
| "logps/chosen": -4.111853122711182, |
| "logps/rejected": -65.05699157714844, |
| "loss": 0.45744559168815613, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.333187997341156, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.20490425825119019, |
| "rewards/margins": 4.838937759399414, |
| "rewards/rejected": -4.634034156799316, |
| "step": 214, |
| "train_speed(iter/s)": 0.005657 |
| }, |
| { |
| "epoch": 1.289047798268724, |
| "grad_norm": 0.9938692450523376, |
| "learning_rate": 6.069749683460765e-05, |
| "logits/chosen": -0.3912752568721771, |
| "logits/rejected": -0.7120423316955566, |
| "logps/chosen": -2.6046087741851807, |
| "logps/rejected": -72.46825408935547, |
| "loss": 0.28505194187164307, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.23823222517967224, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.31369316577911377, |
| "rewards/margins": 5.755457401275635, |
| "rewards/rejected": -5.441763877868652, |
| "step": 215, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.2950696273993225, |
| "grad_norm": 1.6541746854782104, |
| "learning_rate": 5.978239311675826e-05, |
| "logits/chosen": -0.4647775888442993, |
| "logits/rejected": -0.7389938831329346, |
| "logps/chosen": -4.554535865783691, |
| "logps/rejected": -61.292266845703125, |
| "loss": 0.44827941060066223, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3300400674343109, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.24124576151371002, |
| "rewards/margins": 4.951149940490723, |
| "rewards/rejected": -4.709904193878174, |
| "step": 216, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.301091456529921, |
| "grad_norm": 1.3506146669387817, |
| "learning_rate": 5.887128968693887e-05, |
| "logits/chosen": -0.5377171635627747, |
| "logits/rejected": -0.7490763664245605, |
| "logps/chosen": -4.447143077850342, |
| "logps/rejected": -55.9873161315918, |
| "loss": 0.44089001417160034, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3394245207309723, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.4647373557090759, |
| "rewards/margins": 4.664052963256836, |
| "rewards/rejected": -4.199316024780273, |
| "step": 217, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.3071132856605194, |
| "grad_norm": 3.494126796722412, |
| "learning_rate": 5.796427716904347e-05, |
| "logits/chosen": -0.5005620718002319, |
| "logits/rejected": -0.8207815885543823, |
| "logps/chosen": -2.703669309616089, |
| "logps/rejected": -65.87652587890625, |
| "loss": 0.33874765038490295, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2589980661869049, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.22972920536994934, |
| "rewards/margins": 5.275555610656738, |
| "rewards/rejected": -5.0458269119262695, |
| "step": 218, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.3131351147911179, |
| "grad_norm": 1.4259244203567505, |
| "learning_rate": 5.7061445780059074e-05, |
| "logits/chosen": -0.4546062648296356, |
| "logits/rejected": -0.7871499061584473, |
| "logps/chosen": -3.830498218536377, |
| "logps/rejected": -65.30023956298828, |
| "loss": 0.31216228008270264, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2483048439025879, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.3124459385871887, |
| "rewards/margins": 5.200329303741455, |
| "rewards/rejected": -4.88788366317749, |
| "step": 219, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.3191569439217163, |
| "grad_norm": 2.587780475616455, |
| "learning_rate": 5.616288532109225e-05, |
| "logits/chosen": -0.5172604322433472, |
| "logits/rejected": -0.7480805516242981, |
| "logps/chosen": -5.291163921356201, |
| "logps/rejected": -57.106868743896484, |
| "loss": 0.5239050388336182, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.406022310256958, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.26827341318130493, |
| "rewards/margins": 4.4456329345703125, |
| "rewards/rejected": -4.1773600578308105, |
| "step": 220, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.3251787730523146, |
| "grad_norm": 1.639807105064392, |
| "learning_rate": 5.526868516843673e-05, |
| "logits/chosen": -0.547332227230072, |
| "logits/rejected": -0.7309650182723999, |
| "logps/chosen": -5.008947372436523, |
| "logps/rejected": -56.65232849121094, |
| "loss": 0.4063935875892639, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.31995317339897156, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.2977209985256195, |
| "rewards/margins": 4.773816108703613, |
| "rewards/rejected": -4.476095676422119, |
| "step": 221, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.331200602182913, |
| "grad_norm": 4.514076232910156, |
| "learning_rate": 5.43789342646837e-05, |
| "logits/chosen": -0.4913971424102783, |
| "logits/rejected": -0.7436814308166504, |
| "logps/chosen": -4.23374605178833, |
| "logps/rejected": -58.28513717651367, |
| "loss": 0.4951242208480835, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3493507504463196, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2692214250564575, |
| "rewards/margins": 4.63377046585083, |
| "rewards/rejected": -4.364549160003662, |
| "step": 222, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.3372224313135115, |
| "grad_norm": 1.7904651165008545, |
| "learning_rate": 5.349372110987496e-05, |
| "logits/chosen": -0.525532603263855, |
| "logits/rejected": -0.7780842781066895, |
| "logps/chosen": -3.3643298149108887, |
| "logps/rejected": -65.86946868896484, |
| "loss": 0.3287546932697296, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.26226621866226196, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.34989917278289795, |
| "rewards/margins": 5.363834381103516, |
| "rewards/rejected": -5.013935089111328, |
| "step": 223, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.34324426044411, |
| "grad_norm": 1.8052747249603271, |
| "learning_rate": 5.261313375270014e-05, |
| "logits/chosen": -0.5206581950187683, |
| "logits/rejected": -0.7903106212615967, |
| "logps/chosen": -4.260349273681641, |
| "logps/rejected": -65.74819946289062, |
| "loss": 0.4206431806087494, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3237663805484772, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.32122018933296204, |
| "rewards/margins": 5.483022212982178, |
| "rewards/rejected": -5.161801815032959, |
| "step": 224, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.3492660895747084, |
| "grad_norm": 2.2413487434387207, |
| "learning_rate": 5.1737259781738936e-05, |
| "logits/chosen": -0.5118913650512695, |
| "logits/rejected": -0.7823872566223145, |
| "logps/chosen": -4.02402400970459, |
| "logps/rejected": -56.05274200439453, |
| "loss": 0.4401736855506897, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.34718388319015503, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.31864693760871887, |
| "rewards/margins": 4.531757354736328, |
| "rewards/rejected": -4.213110446929932, |
| "step": 225, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.3552879187053066, |
| "grad_norm": 2.4015703201293945, |
| "learning_rate": 5.086618631674888e-05, |
| "logits/chosen": -0.4304109215736389, |
| "logits/rejected": -0.8146275281906128, |
| "logps/chosen": -5.407003402709961, |
| "logps/rejected": -75.61699676513672, |
| "loss": 0.4439658224582672, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3463272750377655, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.24775578081607819, |
| "rewards/margins": 5.974477767944336, |
| "rewards/rejected": -5.726721286773682, |
| "step": 226, |
| "train_speed(iter/s)": 0.005656 |
| }, |
| { |
| "epoch": 1.361309747835905, |
| "grad_norm": 1.798647165298462, |
| "learning_rate": 5.000000000000002e-05, |
| "logits/chosen": -0.41718822717666626, |
| "logits/rejected": -0.8002769947052002, |
| "logps/chosen": -3.8272299766540527, |
| "logps/rejected": -80.6712646484375, |
| "loss": 0.41253572702407837, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.35548415780067444, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.15689942240715027, |
| "rewards/margins": 6.417231559753418, |
| "rewards/rejected": -6.260332107543945, |
| "step": 227, |
| "train_speed(iter/s)": 0.005655 |
| }, |
| { |
| "epoch": 1.3673315769665035, |
| "grad_norm": 1.539751648902893, |
| "learning_rate": 4.913878698765686e-05, |
| "logits/chosen": -0.5234534740447998, |
| "logits/rejected": -0.7054301500320435, |
| "logps/chosen": -7.054202079772949, |
| "logps/rejected": -59.50374221801758, |
| "loss": 0.46764418482780457, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.379008412361145, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3111666738986969, |
| "rewards/margins": 4.931512832641602, |
| "rewards/rejected": -4.620346546173096, |
| "step": 228, |
| "train_speed(iter/s)": 0.005655 |
| }, |
| { |
| "epoch": 1.373353406097102, |
| "grad_norm": 4.103943347930908, |
| "learning_rate": 4.8282632941208725e-05, |
| "logits/chosen": -0.40814512968063354, |
| "logits/rejected": -0.7456347942352295, |
| "logps/chosen": -4.591305255889893, |
| "logps/rejected": -72.21307373046875, |
| "loss": 0.372346967458725, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.33533719182014465, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.37759003043174744, |
| "rewards/margins": 5.916057586669922, |
| "rewards/rejected": -5.5384674072265625, |
| "step": 229, |
| "train_speed(iter/s)": 0.005655 |
| }, |
| { |
| "epoch": 1.3793752352277004, |
| "grad_norm": 2.3027875423431396, |
| "learning_rate": 4.743162301894952e-05, |
| "logits/chosen": -0.530301034450531, |
| "logits/rejected": -0.7990034222602844, |
| "logps/chosen": -4.5230021476745605, |
| "logps/rejected": -65.62507629394531, |
| "loss": 0.38261985778808594, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3187011778354645, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3396291434764862, |
| "rewards/margins": 5.285628318786621, |
| "rewards/rejected": -4.945999622344971, |
| "step": 230, |
| "train_speed(iter/s)": 0.005655 |
| }, |
| { |
| "epoch": 1.385397064358299, |
| "grad_norm": 3.5107035636901855, |
| "learning_rate": 4.658584186750713e-05, |
| "logits/chosen": -0.4405829906463623, |
| "logits/rejected": -0.7592389583587646, |
| "logps/chosen": -3.0335636138916016, |
| "logps/rejected": -64.45790100097656, |
| "loss": 0.3538115918636322, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.26427680253982544, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.22065187990665436, |
| "rewards/margins": 5.173749923706055, |
| "rewards/rejected": -4.953097343444824, |
| "step": 231, |
| "train_speed(iter/s)": 0.005655 |
| }, |
| { |
| "epoch": 1.3914188934888974, |
| "grad_norm": 3.794492244720459, |
| "learning_rate": 4.574537361342407e-05, |
| "logits/chosen": -0.5307365655899048, |
| "logits/rejected": -0.8200229406356812, |
| "logps/chosen": -2.5140573978424072, |
| "logps/rejected": -63.374088287353516, |
| "loss": 0.30093103647232056, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.23118244111537933, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.2798585295677185, |
| "rewards/margins": 5.135382175445557, |
| "rewards/rejected": -4.855523109436035, |
| "step": 232, |
| "train_speed(iter/s)": 0.005655 |
| }, |
| { |
| "epoch": 1.3974407226194956, |
| "grad_norm": 3.8283984661102295, |
| "learning_rate": 4.491030185478976e-05, |
| "logits/chosen": -0.5696873664855957, |
| "logits/rejected": -0.7703967690467834, |
| "logps/chosen": -4.258246421813965, |
| "logps/rejected": -60.22206497192383, |
| "loss": 0.4060544967651367, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3264697194099426, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.27311259508132935, |
| "rewards/margins": 5.165642261505127, |
| "rewards/rejected": -4.892529487609863, |
| "step": 233, |
| "train_speed(iter/s)": 0.005655 |
| }, |
| { |
| "epoch": 1.403462551750094, |
| "grad_norm": 3.2133429050445557, |
| "learning_rate": 4.4080709652925336e-05, |
| "logits/chosen": -0.45184794068336487, |
| "logits/rejected": -0.7718617916107178, |
| "logps/chosen": -5.11681604385376, |
| "logps/rejected": -76.45780181884766, |
| "loss": 0.43574121594429016, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3303045332431793, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2558760643005371, |
| "rewards/margins": 6.22172737121582, |
| "rewards/rejected": -5.965850830078125, |
| "step": 234, |
| "train_speed(iter/s)": 0.005654 |
| }, |
| { |
| "epoch": 1.4094843808806925, |
| "grad_norm": 9.291942596435547, |
| "learning_rate": 4.3256679524121834e-05, |
| "logits/chosen": -0.4269176423549652, |
| "logits/rejected": -0.7866188883781433, |
| "logps/chosen": -1.7651253938674927, |
| "logps/rejected": -71.36811065673828, |
| "loss": 0.26370862126350403, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.178541362285614, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.21358121931552887, |
| "rewards/margins": 5.760619163513184, |
| "rewards/rejected": -5.547037601470947, |
| "step": 235, |
| "train_speed(iter/s)": 0.005654 |
| }, |
| { |
| "epoch": 1.415506210011291, |
| "grad_norm": 1.960109829902649, |
| "learning_rate": 4.2438293431432665e-05, |
| "logits/chosen": -0.4432763159275055, |
| "logits/rejected": -0.7087511420249939, |
| "logps/chosen": -3.8448731899261475, |
| "logps/rejected": -70.58623504638672, |
| "loss": 0.3397082984447479, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2688691318035126, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3202817142009735, |
| "rewards/margins": 5.778705596923828, |
| "rewards/rejected": -5.458423614501953, |
| "step": 236, |
| "train_speed(iter/s)": 0.005654 |
| }, |
| { |
| "epoch": 1.4215280391418894, |
| "grad_norm": 1.571708083152771, |
| "learning_rate": 4.1625632776521037e-05, |
| "logits/chosen": -0.3686775863170624, |
| "logits/rejected": -0.7584090828895569, |
| "logps/chosen": -3.5422518253326416, |
| "logps/rejected": -74.61741638183594, |
| "loss": 0.36048567295074463, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3007555603981018, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.32932642102241516, |
| "rewards/margins": 5.9051923751831055, |
| "rewards/rejected": -5.575865745544434, |
| "step": 237, |
| "train_speed(iter/s)": 0.005654 |
| }, |
| { |
| "epoch": 1.4275498682724876, |
| "grad_norm": 1.7286449670791626, |
| "learning_rate": 4.081877839156325e-05, |
| "logits/chosen": -0.3519198000431061, |
| "logits/rejected": -0.7000710964202881, |
| "logps/chosen": -4.214243412017822, |
| "logps/rejected": -73.10647583007812, |
| "loss": 0.3545505106449127, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2885700464248657, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.24009957909584045, |
| "rewards/margins": 5.838498592376709, |
| "rewards/rejected": -5.5983991622924805, |
| "step": 238, |
| "train_speed(iter/s)": 0.005654 |
| }, |
| { |
| "epoch": 1.433571697403086, |
| "grad_norm": 4.094414234161377, |
| "learning_rate": 4.001781053120863e-05, |
| "logits/chosen": -0.5312204360961914, |
| "logits/rejected": -0.7494320869445801, |
| "logps/chosen": -3.4775454998016357, |
| "logps/rejected": -54.855403900146484, |
| "loss": 0.40386420488357544, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.24309346079826355, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.268097460269928, |
| "rewards/margins": 4.557480812072754, |
| "rewards/rejected": -4.289383411407471, |
| "step": 239, |
| "train_speed(iter/s)": 0.005653 |
| }, |
| { |
| "epoch": 1.4395935265336846, |
| "grad_norm": 1.7741433382034302, |
| "learning_rate": 3.9222808864597004e-05, |
| "logits/chosen": -0.5123165845870972, |
| "logits/rejected": -0.8100269436836243, |
| "logps/chosen": -3.479257106781006, |
| "logps/rejected": -70.73236083984375, |
| "loss": 0.29427477717399597, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2662776708602905, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.23656059801578522, |
| "rewards/margins": 5.880814075469971, |
| "rewards/rejected": -5.644253730773926, |
| "step": 240, |
| "train_speed(iter/s)": 0.005653 |
| }, |
| { |
| "epoch": 1.445615355664283, |
| "grad_norm": 1.358980417251587, |
| "learning_rate": 3.843385246743417e-05, |
| "logits/chosen": -0.43873101472854614, |
| "logits/rejected": -0.7396629452705383, |
| "logps/chosen": -2.895552396774292, |
| "logps/rejected": -58.965843200683594, |
| "loss": 0.27671289443969727, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.18713214993476868, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.253683865070343, |
| "rewards/margins": 4.665976047515869, |
| "rewards/rejected": -4.412292003631592, |
| "step": 241, |
| "train_speed(iter/s)": 0.005653 |
| }, |
| { |
| "epoch": 1.4516371847948815, |
| "grad_norm": 2.561663866043091, |
| "learning_rate": 3.7651019814126654e-05, |
| "logits/chosen": -0.5158201456069946, |
| "logits/rejected": -0.7556857466697693, |
| "logps/chosen": -6.659661769866943, |
| "logps/rejected": -61.460845947265625, |
| "loss": 0.540454626083374, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.43979448080062866, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.351629376411438, |
| "rewards/margins": 5.123873710632324, |
| "rewards/rejected": -4.772244453430176, |
| "step": 242, |
| "train_speed(iter/s)": 0.005653 |
| }, |
| { |
| "epoch": 1.45765901392548, |
| "grad_norm": 4.672154903411865, |
| "learning_rate": 3.687438876997612e-05, |
| "logits/chosen": -0.5328869819641113, |
| "logits/rejected": -0.78375244140625, |
| "logps/chosen": -4.969141960144043, |
| "logps/rejected": -57.542362213134766, |
| "loss": 0.5365437269210815, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3501831889152527, |
| "rewards/accuracies": 0.921875, |
| "rewards/chosen": 0.1367063820362091, |
| "rewards/margins": 4.617161273956299, |
| "rewards/rejected": -4.480454444885254, |
| "step": 243, |
| "train_speed(iter/s)": 0.005653 |
| }, |
| { |
| "epoch": 1.4636808430560784, |
| "grad_norm": 7.847040176391602, |
| "learning_rate": 3.610403658343443e-05, |
| "logits/chosen": -0.35487285256385803, |
| "logits/rejected": -0.7575075030326843, |
| "logps/chosen": -4.001574993133545, |
| "logps/rejected": -74.3595962524414, |
| "loss": 0.4401260018348694, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.35397177934646606, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.21161355078220367, |
| "rewards/margins": 5.741542816162109, |
| "rewards/rejected": -5.529928684234619, |
| "step": 244, |
| "train_speed(iter/s)": 0.005653 |
| }, |
| { |
| "epoch": 1.4697026721866768, |
| "grad_norm": 1.5019311904907227, |
| "learning_rate": 3.534003987842005e-05, |
| "logits/chosen": -0.5985352993011475, |
| "logits/rejected": -0.8159947395324707, |
| "logps/chosen": -3.0694146156311035, |
| "logps/rejected": -67.13231658935547, |
| "loss": 0.3305334746837616, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.26731085777282715, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.21900442242622375, |
| "rewards/margins": 5.598498821258545, |
| "rewards/rejected": -5.3794941902160645, |
| "step": 245, |
| "train_speed(iter/s)": 0.005653 |
| }, |
| { |
| "epoch": 1.475724501317275, |
| "grad_norm": 2.49977707862854, |
| "learning_rate": 3.458247464669657e-05, |
| "logits/chosen": -0.5261157155036926, |
| "logits/rejected": -0.7526097893714905, |
| "logps/chosen": -4.502616882324219, |
| "logps/rejected": -55.263763427734375, |
| "loss": 0.38907888531684875, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.29498404264450073, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2758098840713501, |
| "rewards/margins": 4.53220796585083, |
| "rewards/rejected": -4.2563982009887695, |
| "step": 246, |
| "train_speed(iter/s)": 0.005653 |
| }, |
| { |
| "epoch": 1.4817463304478735, |
| "grad_norm": 1.7574135065078735, |
| "learning_rate": 3.383141624031408e-05, |
| "logits/chosen": -0.5520765781402588, |
| "logits/rejected": -0.7475904822349548, |
| "logps/chosen": -6.128981590270996, |
| "logps/rejected": -56.70994186401367, |
| "loss": 0.43652814626693726, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3729217052459717, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.38195130228996277, |
| "rewards/margins": 4.705140113830566, |
| "rewards/rejected": -4.323188781738281, |
| "step": 247, |
| "train_speed(iter/s)": 0.005653 |
| }, |
| { |
| "epoch": 1.487768159578472, |
| "grad_norm": 1.5766643285751343, |
| "learning_rate": 3.308693936411421e-05, |
| "logits/chosen": -0.4617379307746887, |
| "logits/rejected": -0.6871187686920166, |
| "logps/chosen": -6.003100872039795, |
| "logps/rejected": -63.69612503051758, |
| "loss": 0.4494744539260864, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3877749443054199, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.3871355652809143, |
| "rewards/margins": 5.155317783355713, |
| "rewards/rejected": -4.768182277679443, |
| "step": 248, |
| "train_speed(iter/s)": 0.005652 |
| }, |
| { |
| "epoch": 1.4937899887090704, |
| "grad_norm": 2.092433214187622, |
| "learning_rate": 3.234911806829948e-05, |
| "logits/chosen": -0.4331311285495758, |
| "logits/rejected": -0.7093103528022766, |
| "logps/chosen": -4.4328694343566895, |
| "logps/rejected": -61.90970230102539, |
| "loss": 0.3958251476287842, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3243080973625183, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3177702724933624, |
| "rewards/margins": 5.146597385406494, |
| "rewards/rejected": -4.828826904296875, |
| "step": 249, |
| "train_speed(iter/s)": 0.005652 |
| }, |
| { |
| "epoch": 1.4998118178396689, |
| "grad_norm": 2.878459930419922, |
| "learning_rate": 3.161802574106799e-05, |
| "logits/chosen": -0.5477363467216492, |
| "logits/rejected": -0.73226398229599, |
| "logps/chosen": -4.833172798156738, |
| "logps/rejected": -54.378639221191406, |
| "loss": 0.42393556237220764, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3001037836074829, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.32397884130477905, |
| "rewards/margins": 4.320741653442383, |
| "rewards/rejected": -3.99676251411438, |
| "step": 250, |
| "train_speed(iter/s)": 0.005652 |
| }, |
| { |
| "epoch": 1.5058336469702671, |
| "grad_norm": 2.852609872817993, |
| "learning_rate": 3.089373510131354e-05, |
| "logits/chosen": -0.4120064973831177, |
| "logits/rejected": -0.7510567307472229, |
| "logps/chosen": -3.8237786293029785, |
| "logps/rejected": -68.50797271728516, |
| "loss": 0.38543784618377686, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.28033876419067383, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2964167594909668, |
| "rewards/margins": 5.442708969116211, |
| "rewards/rejected": -5.146292686462402, |
| "step": 251, |
| "train_speed(iter/s)": 0.00565 |
| }, |
| { |
| "epoch": 1.5118554761008656, |
| "grad_norm": 3.0803050994873047, |
| "learning_rate": 3.0176318191392726e-05, |
| "logits/chosen": -0.4008835256099701, |
| "logits/rejected": -0.7395371198654175, |
| "logps/chosen": -2.694051742553711, |
| "logps/rejected": -64.98765563964844, |
| "loss": 0.4233083128929138, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.28342971205711365, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.18508972227573395, |
| "rewards/margins": 5.180187702178955, |
| "rewards/rejected": -4.995098114013672, |
| "step": 252, |
| "train_speed(iter/s)": 0.00565 |
| }, |
| { |
| "epoch": 1.517877305231464, |
| "grad_norm": 3.8462982177734375, |
| "learning_rate": 2.9465846369959127e-05, |
| "logits/chosen": -0.44880083203315735, |
| "logits/rejected": -0.7400986552238464, |
| "logps/chosen": -4.19825553894043, |
| "logps/rejected": -68.70524597167969, |
| "loss": 0.3208198547363281, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.23766814172267914, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2140774428844452, |
| "rewards/margins": 5.5977983474731445, |
| "rewards/rejected": -5.383721351623535, |
| "step": 253, |
| "train_speed(iter/s)": 0.00565 |
| }, |
| { |
| "epoch": 1.5238991343620625, |
| "grad_norm": 2.542881488800049, |
| "learning_rate": 2.876239030486554e-05, |
| "logits/chosen": -0.5086544752120972, |
| "logits/rejected": -0.7899590730667114, |
| "logps/chosen": -3.355459451675415, |
| "logps/rejected": -63.492759704589844, |
| "loss": 0.31336328387260437, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.246277317404747, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.22459468245506287, |
| "rewards/margins": 5.192716121673584, |
| "rewards/rejected": -4.9681220054626465, |
| "step": 254, |
| "train_speed(iter/s)": 0.00565 |
| }, |
| { |
| "epoch": 1.529920963492661, |
| "grad_norm": 2.262983560562134, |
| "learning_rate": 2.8066019966134904e-05, |
| "logits/chosen": -0.4363003671169281, |
| "logits/rejected": -0.7306255102157593, |
| "logps/chosen": -2.433912515640259, |
| "logps/rejected": -67.02496337890625, |
| "loss": 0.31392139196395874, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.22945323586463928, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2390332818031311, |
| "rewards/margins": 5.437534332275391, |
| "rewards/rejected": -5.1985015869140625, |
| "step": 255, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 1.5359427926232594, |
| "grad_norm": 2.1863996982574463, |
| "learning_rate": 2.7376804619000707e-05, |
| "logits/chosen": -0.5122266411781311, |
| "logits/rejected": -0.7367725372314453, |
| "logps/chosen": -5.143320560455322, |
| "logps/rejected": -58.59758758544922, |
| "loss": 0.5367969274520874, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4321354627609253, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.3802338242530823, |
| "rewards/margins": 4.9757304191589355, |
| "rewards/rejected": -4.59549617767334, |
| "step": 256, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 1.5419646217538578, |
| "grad_norm": 2.382030725479126, |
| "learning_rate": 2.669481281701739e-05, |
| "logits/chosen": -0.4875583052635193, |
| "logits/rejected": -0.7425970435142517, |
| "logps/chosen": -3.6908037662506104, |
| "logps/rejected": -66.99980163574219, |
| "loss": 0.3160356283187866, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.24375540018081665, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3102016746997833, |
| "rewards/margins": 5.104434013366699, |
| "rewards/rejected": -4.7942328453063965, |
| "step": 257, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 1.5479864508844563, |
| "grad_norm": 1.3870384693145752, |
| "learning_rate": 2.6020112395241624e-05, |
| "logits/chosen": -0.5185623168945312, |
| "logits/rejected": -0.7977117300033569, |
| "logps/chosen": -3.2771565914154053, |
| "logps/rejected": -66.14156341552734, |
| "loss": 0.32793527841567993, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2557143270969391, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3175499141216278, |
| "rewards/margins": 5.339153289794922, |
| "rewards/rejected": -5.021603584289551, |
| "step": 258, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 1.5540082800150545, |
| "grad_norm": 1.6842197179794312, |
| "learning_rate": 2.5352770463484987e-05, |
| "logits/chosen": -0.5024842023849487, |
| "logits/rejected": -0.7918374538421631, |
| "logps/chosen": -1.9520148038864136, |
| "logps/rejected": -62.17351150512695, |
| "loss": 0.2610028386116028, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.18495017290115356, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.21339333057403564, |
| "rewards/margins": 5.084794998168945, |
| "rewards/rejected": -4.871401786804199, |
| "step": 259, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 1.560030109145653, |
| "grad_norm": 2.327899694442749, |
| "learning_rate": 2.4692853399638917e-05, |
| "logits/chosen": -0.45503148436546326, |
| "logits/rejected": -0.7623938918113708, |
| "logps/chosen": -3.5812909603118896, |
| "logps/rejected": -72.19535064697266, |
| "loss": 0.41872018575668335, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2947246730327606, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.15223976969718933, |
| "rewards/margins": 5.696834564208984, |
| "rewards/rejected": -5.544594764709473, |
| "step": 260, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 1.5660519382762514, |
| "grad_norm": 1.320288896560669, |
| "learning_rate": 2.4040426843072206e-05, |
| "logits/chosen": -0.47948870062828064, |
| "logits/rejected": -0.7575923204421997, |
| "logps/chosen": -2.7919044494628906, |
| "logps/rejected": -62.61322784423828, |
| "loss": 0.34768229722976685, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.23828013241291046, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.2685001492500305, |
| "rewards/margins": 5.049767971038818, |
| "rewards/rejected": -4.781268119812012, |
| "step": 261, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 1.5720737674068497, |
| "grad_norm": 2.9725844860076904, |
| "learning_rate": 2.339555568810221e-05, |
| "logits/chosen": -0.46958619356155396, |
| "logits/rejected": -0.7571808695793152, |
| "logps/chosen": -4.220214366912842, |
| "logps/rejected": -70.71320343017578, |
| "loss": 0.37105321884155273, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2905551791191101, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.11353044211864471, |
| "rewards/margins": 5.697925090789795, |
| "rewards/rejected": -5.5843939781188965, |
| "step": 262, |
| "train_speed(iter/s)": 0.005649 |
| }, |
| { |
| "epoch": 1.5780955965374481, |
| "grad_norm": 2.7388265132904053, |
| "learning_rate": 2.275830407754006e-05, |
| "logits/chosen": -0.43985384702682495, |
| "logits/rejected": -0.7053717374801636, |
| "logps/chosen": -3.578861951828003, |
| "logps/rejected": -65.03799438476562, |
| "loss": 0.34339943528175354, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2526337504386902, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.23234206438064575, |
| "rewards/margins": 5.231771469116211, |
| "rewards/rejected": -4.999429702758789, |
| "step": 263, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 1.5841174256680466, |
| "grad_norm": 2.7772486209869385, |
| "learning_rate": 2.212873539631061e-05, |
| "logits/chosen": -0.49956512451171875, |
| "logits/rejected": -0.7878684997558594, |
| "logps/chosen": -2.7876102924346924, |
| "logps/rejected": -65.78087615966797, |
| "loss": 0.3053550720214844, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2294946312904358, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.21735113859176636, |
| "rewards/margins": 5.3624958992004395, |
| "rewards/rejected": -5.145144462585449, |
| "step": 264, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 1.590139254798645, |
| "grad_norm": 1.685667634010315, |
| "learning_rate": 2.1506912265147772e-05, |
| "logits/chosen": -0.5206426382064819, |
| "logits/rejected": -0.7813724279403687, |
| "logps/chosen": -2.941920757293701, |
| "logps/rejected": -66.5394287109375, |
| "loss": 0.2972472906112671, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2362329214811325, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.33175477385520935, |
| "rewards/margins": 5.447272777557373, |
| "rewards/rejected": -5.115517616271973, |
| "step": 265, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 1.5961610839292435, |
| "grad_norm": 3.3678839206695557, |
| "learning_rate": 2.0892896534365904e-05, |
| "logits/chosen": -0.3419654369354248, |
| "logits/rejected": -0.7248679995536804, |
| "logps/chosen": -3.409761905670166, |
| "logps/rejected": -71.6166763305664, |
| "loss": 0.32658249139785767, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.26064980030059814, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.20473551750183105, |
| "rewards/margins": 5.657358646392822, |
| "rewards/rejected": -5.452622890472412, |
| "step": 266, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 1.602182913059842, |
| "grad_norm": 3.060711145401001, |
| "learning_rate": 2.0286749277707782e-05, |
| "logits/chosen": -0.4216618835926056, |
| "logits/rejected": -0.6746920347213745, |
| "logps/chosen": -7.840616703033447, |
| "logps/rejected": -59.754539489746094, |
| "loss": 0.5373175740242004, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.4301339387893677, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.4534755051136017, |
| "rewards/margins": 4.956415176391602, |
| "rewards/rejected": -4.502939701080322, |
| "step": 267, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 1.6082047421904404, |
| "grad_norm": 1.2512023448944092, |
| "learning_rate": 1.9688530786269855e-05, |
| "logits/chosen": -0.48612943291664124, |
| "logits/rejected": -0.78404700756073, |
| "logps/chosen": -3.4436018466949463, |
| "logps/rejected": -64.1688461303711, |
| "loss": 0.3891962468624115, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3172661066055298, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.27034294605255127, |
| "rewards/margins": 5.256752967834473, |
| "rewards/rejected": -4.986410140991211, |
| "step": 268, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 1.6142265713210389, |
| "grad_norm": 1.5459010601043701, |
| "learning_rate": 1.9098300562505266e-05, |
| "logits/chosen": -0.4943903386592865, |
| "logits/rejected": -0.8177123069763184, |
| "logps/chosen": -2.876307964324951, |
| "logps/rejected": -72.3165512084961, |
| "loss": 0.3240829110145569, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2640679180622101, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.26817601919174194, |
| "rewards/margins": 5.955280780792236, |
| "rewards/rejected": -5.687105178833008, |
| "step": 269, |
| "train_speed(iter/s)": 0.005648 |
| }, |
| { |
| "epoch": 1.6202484004516373, |
| "grad_norm": 2.2002980709075928, |
| "learning_rate": 1.8516117314305524e-05, |
| "logits/chosen": -0.4193212389945984, |
| "logits/rejected": -0.7498923540115356, |
| "logps/chosen": -2.122291088104248, |
| "logps/rejected": -78.85472106933594, |
| "loss": 0.29622161388397217, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.21589502692222595, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.19214794039726257, |
| "rewards/margins": 6.349010467529297, |
| "rewards/rejected": -6.156862735748291, |
| "step": 270, |
| "train_speed(iter/s)": 0.005647 |
| }, |
| { |
| "epoch": 1.6262702295822355, |
| "grad_norm": 2.6795237064361572, |
| "learning_rate": 1.7942038949160854e-05, |
| "logits/chosen": -0.44892024993896484, |
| "logits/rejected": -0.6841689348220825, |
| "logps/chosen": -6.089541435241699, |
| "logps/rejected": -60.286041259765625, |
| "loss": 0.5212526321411133, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.32065996527671814, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.24887697398662567, |
| "rewards/margins": 4.777753829956055, |
| "rewards/rejected": -4.528877258300781, |
| "step": 271, |
| "train_speed(iter/s)": 0.005647 |
| }, |
| { |
| "epoch": 1.632292058712834, |
| "grad_norm": 1.3338947296142578, |
| "learning_rate": 1.7376122568400532e-05, |
| "logits/chosen": -0.5868579149246216, |
| "logits/rejected": -0.9036458730697632, |
| "logps/chosen": -2.3813061714172363, |
| "logps/rejected": -66.72323608398438, |
| "loss": 0.29132747650146484, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.23172350227832794, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.2914634346961975, |
| "rewards/margins": 5.641142845153809, |
| "rewards/rejected": -5.349678993225098, |
| "step": 272, |
| "train_speed(iter/s)": 0.005647 |
| }, |
| { |
| "epoch": 1.6383138878434325, |
| "grad_norm": 1.2542729377746582, |
| "learning_rate": 1.681842446151313e-05, |
| "logits/chosen": -0.4844375252723694, |
| "logits/rejected": -0.7457984685897827, |
| "logps/chosen": -5.260637283325195, |
| "logps/rejected": -71.04159545898438, |
| "loss": 0.3770938515663147, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.33192208409309387, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.40421298146247864, |
| "rewards/margins": 5.93914270401001, |
| "rewards/rejected": -5.534929275512695, |
| "step": 273, |
| "train_speed(iter/s)": 0.005647 |
| }, |
| { |
| "epoch": 1.644335716974031, |
| "grad_norm": 2.0362296104431152, |
| "learning_rate": 1.6269000100547683e-05, |
| "logits/chosen": -0.4546166658401489, |
| "logits/rejected": -0.7672044634819031, |
| "logps/chosen": -3.0699520111083984, |
| "logps/rejected": -71.65580749511719, |
| "loss": 0.311799019575119, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.24750828742980957, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.31899750232696533, |
| "rewards/margins": 6.106335163116455, |
| "rewards/rejected": -5.787337303161621, |
| "step": 274, |
| "train_speed(iter/s)": 0.005647 |
| }, |
| { |
| "epoch": 1.6503575461046291, |
| "grad_norm": 3.0854225158691406, |
| "learning_rate": 1.5727904134596083e-05, |
| "logits/chosen": -0.5049324035644531, |
| "logits/rejected": -0.8067833185195923, |
| "logps/chosen": -3.4519662857055664, |
| "logps/rejected": -64.27713012695312, |
| "loss": 0.40550893545150757, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3140624165534973, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.23077721893787384, |
| "rewards/margins": 5.18168830871582, |
| "rewards/rejected": -4.950911045074463, |
| "step": 275, |
| "train_speed(iter/s)": 0.005647 |
| }, |
| { |
| "epoch": 1.6563793752352276, |
| "grad_norm": 2.3667151927948, |
| "learning_rate": 1.5195190384357404e-05, |
| "logits/chosen": -0.4429381787776947, |
| "logits/rejected": -0.7047520279884338, |
| "logps/chosen": -3.2484500408172607, |
| "logps/rejected": -63.5272216796875, |
| "loss": 0.36934536695480347, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.24999630451202393, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.2938525974750519, |
| "rewards/margins": 5.066824913024902, |
| "rewards/rejected": -4.772972583770752, |
| "step": 276, |
| "train_speed(iter/s)": 0.005646 |
| }, |
| { |
| "epoch": 1.662401204365826, |
| "grad_norm": 2.0824172496795654, |
| "learning_rate": 1.467091183678444e-05, |
| "logits/chosen": -0.4420183002948761, |
| "logits/rejected": -0.7609032988548279, |
| "logps/chosen": -4.130496025085449, |
| "logps/rejected": -72.41409301757812, |
| "loss": 0.3867420256137848, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.31039467453956604, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.35360175371170044, |
| "rewards/margins": 6.035587310791016, |
| "rewards/rejected": -5.681984901428223, |
| "step": 277, |
| "train_speed(iter/s)": 0.005646 |
| }, |
| { |
| "epoch": 1.6684230334964245, |
| "grad_norm": 1.2125921249389648, |
| "learning_rate": 1.415512063981339e-05, |
| "logits/chosen": -0.47004735469818115, |
| "logits/rejected": -0.7464879751205444, |
| "logps/chosen": -4.137353420257568, |
| "logps/rejected": -72.29967498779297, |
| "loss": 0.38727715611457825, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.32129383087158203, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.5302993059158325, |
| "rewards/margins": 6.306224822998047, |
| "rewards/rejected": -5.775925159454346, |
| "step": 278, |
| "train_speed(iter/s)": 0.005646 |
| }, |
| { |
| "epoch": 1.674444862627023, |
| "grad_norm": 1.5865390300750732, |
| "learning_rate": 1.364786809717692e-05, |
| "logits/chosen": -0.4276161789894104, |
| "logits/rejected": -0.7880982160568237, |
| "logps/chosen": -3.557345390319824, |
| "logps/rejected": -73.96890258789062, |
| "loss": 0.33818569779396057, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.28797072172164917, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.25843125581741333, |
| "rewards/margins": 6.005448818206787, |
| "rewards/rejected": -5.747016906738281, |
| "step": 279, |
| "train_speed(iter/s)": 0.005646 |
| }, |
| { |
| "epoch": 1.6804666917576214, |
| "grad_norm": 5.5078630447387695, |
| "learning_rate": 1.3149204663301118e-05, |
| "logits/chosen": -0.3974856734275818, |
| "logits/rejected": -0.7452144622802734, |
| "logps/chosen": -3.9654970169067383, |
| "logps/rejected": -71.51333618164062, |
| "loss": 0.35094186663627625, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2737097442150116, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.2904951870441437, |
| "rewards/margins": 5.682476043701172, |
| "rewards/rejected": -5.39198112487793, |
| "step": 280, |
| "train_speed(iter/s)": 0.005646 |
| }, |
| { |
| "epoch": 1.6864885208882199, |
| "grad_norm": 1.8228479623794556, |
| "learning_rate": 1.2659179938287035e-05, |
| "logits/chosen": -0.40838098526000977, |
| "logits/rejected": -0.7652671933174133, |
| "logps/chosen": -2.5664548873901367, |
| "logps/rejected": -64.62358856201172, |
| "loss": 0.24017123878002167, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.17215371131896973, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.20578134059906006, |
| "rewards/margins": 5.029583930969238, |
| "rewards/rejected": -4.823802471160889, |
| "step": 281, |
| "train_speed(iter/s)": 0.005646 |
| }, |
| { |
| "epoch": 1.6925103500188183, |
| "grad_norm": 1.196491003036499, |
| "learning_rate": 1.2177842662977135e-05, |
| "logits/chosen": -0.4218314588069916, |
| "logits/rejected": -0.7560523748397827, |
| "logps/chosen": -2.7297964096069336, |
| "logps/rejected": -65.94020080566406, |
| "loss": 0.2704247236251831, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.1988007128238678, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.31327253580093384, |
| "rewards/margins": 5.445861339569092, |
| "rewards/rejected": -5.132589340209961, |
| "step": 282, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.6985321791494168, |
| "grad_norm": 2.948547840118408, |
| "learning_rate": 1.1705240714107302e-05, |
| "logits/chosen": -0.3723593056201935, |
| "logits/rejected": -0.7366158366203308, |
| "logps/chosen": -3.405308485031128, |
| "logps/rejected": -75.39693450927734, |
| "loss": 0.40320032835006714, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2922399640083313, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.29622313380241394, |
| "rewards/margins": 6.236778736114502, |
| "rewards/rejected": -5.940556049346924, |
| "step": 283, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.704554008280015, |
| "grad_norm": 5.310917854309082, |
| "learning_rate": 1.124142109954459e-05, |
| "logits/chosen": -0.5253931283950806, |
| "logits/rejected": -0.8019742965698242, |
| "logps/chosen": -2.7380077838897705, |
| "logps/rejected": -66.83124542236328, |
| "loss": 0.42098596692085266, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3041377365589142, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.1617843061685562, |
| "rewards/margins": 5.515201568603516, |
| "rewards/rejected": -5.353416919708252, |
| "step": 284, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.7105758374106135, |
| "grad_norm": 1.8457385301589966, |
| "learning_rate": 1.0786429953611666e-05, |
| "logits/chosen": -0.4387832283973694, |
| "logits/rejected": -0.8179764747619629, |
| "logps/chosen": -2.1042332649230957, |
| "logps/rejected": -68.6592788696289, |
| "loss": 0.28756269812583923, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.21063612401485443, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.24719983339309692, |
| "rewards/margins": 5.616422176361084, |
| "rewards/rejected": -5.3692216873168945, |
| "step": 285, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.716597666541212, |
| "grad_norm": 1.7732220888137817, |
| "learning_rate": 1.034031253249792e-05, |
| "logits/chosen": -0.5289000272750854, |
| "logits/rejected": -0.8081372380256653, |
| "logps/chosen": -2.455111503601074, |
| "logps/rejected": -60.26368713378906, |
| "loss": 0.2956588566303253, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2198590785264969, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.19829076528549194, |
| "rewards/margins": 4.948812961578369, |
| "rewards/rejected": -4.750522136688232, |
| "step": 286, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.7226194956718102, |
| "grad_norm": 3.0396289825439453, |
| "learning_rate": 9.903113209758096e-06, |
| "logits/chosen": -0.5536311268806458, |
| "logits/rejected": -0.7566096186637878, |
| "logps/chosen": -4.647332191467285, |
| "logps/rejected": -53.316307067871094, |
| "loss": 0.4976504147052765, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.36106935143470764, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.3211871385574341, |
| "rewards/margins": 4.494439601898193, |
| "rewards/rejected": -4.173252582550049, |
| "step": 287, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.7286413248024086, |
| "grad_norm": 1.4801286458969116, |
| "learning_rate": 9.474875471898526e-06, |
| "logits/chosen": -0.4908003807067871, |
| "logits/rejected": -0.8340423703193665, |
| "logps/chosen": -2.6292285919189453, |
| "logps/rejected": -66.33879089355469, |
| "loss": 0.2964615523815155, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.23429438471794128, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.22690246999263763, |
| "rewards/margins": 5.54394006729126, |
| "rewards/rejected": -5.317037582397461, |
| "step": 288, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.734663153933007, |
| "grad_norm": 3.686876058578491, |
| "learning_rate": 9.055641914051782e-06, |
| "logits/chosen": -0.5189223885536194, |
| "logits/rejected": -0.8108649253845215, |
| "logps/chosen": -3.9785354137420654, |
| "logps/rejected": -68.66986083984375, |
| "loss": 0.4266186058521271, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3564129173755646, |
| "rewards/accuracies": 0.96875, |
| "rewards/chosen": 0.27698981761932373, |
| "rewards/margins": 5.682110786437988, |
| "rewards/rejected": -5.405120849609375, |
| "step": 289, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.7406849830636055, |
| "grad_norm": 4.0403594970703125, |
| "learning_rate": 8.645454235739903e-06, |
| "logits/chosen": -0.48102113604545593, |
| "logits/rejected": -0.8606957197189331, |
| "logps/chosen": -3.213351249694824, |
| "logps/rejected": -77.46415710449219, |
| "loss": 0.3690274953842163, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.31783145666122437, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.25699326395988464, |
| "rewards/margins": 6.534820556640625, |
| "rewards/rejected": -6.277827262878418, |
| "step": 290, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.746706812194204, |
| "grad_norm": 3.175973892211914, |
| "learning_rate": 8.24435323672661e-06, |
| "logits/chosen": -0.5072706937789917, |
| "logits/rejected": -0.8571786284446716, |
| "logps/chosen": -3.415510654449463, |
| "logps/rejected": -68.72850036621094, |
| "loss": 0.346034973859787, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.28102245926856995, |
| "rewards/accuracies": 1.0, |
| "rewards/chosen": 0.22457678616046906, |
| "rewards/margins": 5.616724491119385, |
| "rewards/rejected": -5.392147541046143, |
| "step": 291, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.7527286413248024, |
| "grad_norm": 3.0831406116485596, |
| "learning_rate": 7.852378812959227e-06, |
| "logits/chosen": -0.40548837184906006, |
| "logits/rejected": -0.818709135055542, |
| "logps/chosen": -2.9739181995391846, |
| "logps/rejected": -73.55524444580078, |
| "loss": 0.31632980704307556, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.22905109822750092, |
| "rewards/accuracies": 0.953125, |
| "rewards/chosen": 0.16724324226379395, |
| "rewards/margins": 5.85488224029541, |
| "rewards/rejected": -5.687638759613037, |
| "step": 292, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.758750470455401, |
| "grad_norm": 1.8791142702102661, |
| "learning_rate": 7.46956995260033e-06, |
| "logits/chosen": -0.4915006458759308, |
| "logits/rejected": -0.8246431946754456, |
| "logps/chosen": -6.461263179779053, |
| "logps/rejected": -72.80626678466797, |
| "loss": 0.43981942534446716, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3707093298435211, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.22866468131542206, |
| "rewards/margins": 5.882740020751953, |
| "rewards/rejected": -5.654075622558594, |
| "step": 293, |
| "train_speed(iter/s)": 0.005645 |
| }, |
| { |
| "epoch": 1.7647722995859993, |
| "grad_norm": 2.1687467098236084, |
| "learning_rate": 7.09596473214974e-06, |
| "logits/chosen": -0.4673860967159271, |
| "logits/rejected": -0.7422292232513428, |
| "logps/chosen": -5.505043983459473, |
| "logps/rejected": -82.50333404541016, |
| "loss": 0.41253846883773804, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3449009358882904, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.3176400065422058, |
| "rewards/margins": 6.987927436828613, |
| "rewards/rejected": -6.670287132263184, |
| "step": 294, |
| "train_speed(iter/s)": 0.005644 |
| }, |
| { |
| "epoch": 1.7707941287165978, |
| "grad_norm": 4.685088634490967, |
| "learning_rate": 6.731600312657238e-06, |
| "logits/chosen": -0.563675045967102, |
| "logits/rejected": -0.8160005807876587, |
| "logps/chosen": -3.448794364929199, |
| "logps/rejected": -61.5041389465332, |
| "loss": 0.45613980293273926, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3632497191429138, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.27636757493019104, |
| "rewards/margins": 5.065147399902344, |
| "rewards/rejected": -4.788780212402344, |
| "step": 295, |
| "train_speed(iter/s)": 0.005644 |
| }, |
| { |
| "epoch": 1.776815957847196, |
| "grad_norm": 4.02125883102417, |
| "learning_rate": 6.37651293602628e-06, |
| "logits/chosen": -0.5208415389060974, |
| "logits/rejected": -0.7839468121528625, |
| "logps/chosen": -5.326009750366211, |
| "logps/rejected": -63.45610809326172, |
| "loss": 0.501464307308197, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.3729536831378937, |
| "rewards/accuracies": 0.9375, |
| "rewards/chosen": 0.2640419602394104, |
| "rewards/margins": 5.044376850128174, |
| "rewards/rejected": -4.780335426330566, |
| "step": 296, |
| "train_speed(iter/s)": 0.005644 |
| }, |
| { |
| "epoch": 1.7828377869777945, |
| "grad_norm": 2.1763715744018555, |
| "learning_rate": 6.030737921409169e-06, |
| "logits/chosen": -0.48864227533340454, |
| "logits/rejected": -0.7334147095680237, |
| "logps/chosen": -5.976287841796875, |
| "logps/rejected": -67.99011993408203, |
| "loss": 0.46500977873802185, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.39140668511390686, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.31234222650527954, |
| "rewards/margins": 5.641515731811523, |
| "rewards/rejected": -5.329173564910889, |
| "step": 297, |
| "train_speed(iter/s)": 0.005644 |
| }, |
| { |
| "epoch": 1.788859616108393, |
| "grad_norm": 11.35409164428711, |
| "learning_rate": 5.694309661693942e-06, |
| "logits/chosen": -0.31258460879325867, |
| "logits/rejected": -0.6516164541244507, |
| "logps/chosen": -4.385894775390625, |
| "logps/rejected": -64.1596450805664, |
| "loss": 0.5977675914764404, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.399192750453949, |
| "rewards/accuracies": 0.90625, |
| "rewards/chosen": 0.13077813386917114, |
| "rewards/margins": 4.987821102142334, |
| "rewards/rejected": -4.85704231262207, |
| "step": 298, |
| "train_speed(iter/s)": 0.005644 |
| }, |
| { |
| "epoch": 1.7948814452389912, |
| "grad_norm": 1.2614425420761108, |
| "learning_rate": 5.367261620083575e-06, |
| "logits/chosen": -0.43785005807876587, |
| "logits/rejected": -0.7445405125617981, |
| "logps/chosen": -3.5625617504119873, |
| "logps/rejected": -69.88119506835938, |
| "loss": 0.37494975328445435, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.2672398090362549, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.20106351375579834, |
| "rewards/margins": 5.906167984008789, |
| "rewards/rejected": -5.705103397369385, |
| "step": 299, |
| "train_speed(iter/s)": 0.005644 |
| }, |
| { |
| "epoch": 1.8009032743695896, |
| "grad_norm": 2.9862029552459717, |
| "learning_rate": 5.049626326767365e-06, |
| "logits/chosen": -0.4020475745201111, |
| "logits/rejected": -0.722282350063324, |
| "logps/chosen": -2.223629951477051, |
| "logps/rejected": -74.56559753417969, |
| "loss": 0.2519979476928711, |
| "memory(GiB)": 46.13, |
| "nll_loss": 0.18500222265720367, |
| "rewards/accuracies": 0.984375, |
| "rewards/chosen": 0.25891339778900146, |
| "rewards/margins": 6.117902755737305, |
| "rewards/rejected": -5.858989238739014, |
| "step": 300, |
| "train_speed(iter/s)": 0.005644 |
| }, |
| { |
| "epoch": 1.8009032743695896, |
| "eval_logits/chosen": -0.5320332646369934, |
| "eval_logits/rejected": -0.7993212342262268, |
| "eval_logps/chosen": -5.09476375579834, |
| "eval_logps/rejected": -66.83241271972656, |
| "eval_loss": 0.5435733795166016, |
| "eval_nll_loss": 0.39369699358940125, |
| "eval_rewards/accuracies": 0.9444444179534912, |
| "eval_rewards/chosen": 0.06985757499933243, |
| "eval_rewards/margins": 5.177610397338867, |
| "eval_rewards/rejected": -5.107752323150635, |
| "eval_runtime": 117.9868, |
| "eval_samples_per_second": 0.907, |
| "eval_steps_per_second": 0.229, |
| "step": 300 |
| } |
| ], |
| "logging_steps": 1, |
| "max_steps": 332, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 2, |
| "save_steps": 50, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 3.444748519116636e+17, |
| "train_batch_size": 4, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
