| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.992, | |
| "eval_steps": 1000, | |
| "global_step": 561, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.005333333333333333, | |
| "grad_norm": 0.7276772136468688, | |
| "learning_rate": 1.7543859649122808e-07, | |
| "logits/chosen": -1.480194330215454, | |
| "logits/rejected": -1.445469856262207, | |
| "logps/chosen": -200.37672424316406, | |
| "logps/rejected": -252.16812133789062, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.05333333333333334, | |
| "grad_norm": 0.7041154730069451, | |
| "learning_rate": 1.7543859649122807e-06, | |
| "logits/chosen": -1.4611287117004395, | |
| "logits/rejected": -1.460557222366333, | |
| "logps/chosen": -251.83726501464844, | |
| "logps/rejected": -254.89569091796875, | |
| "loss": 0.6926, | |
| "rewards/accuracies": 0.5208333134651184, | |
| "rewards/chosen": 0.0008822197560220957, | |
| "rewards/margins": 0.0008583312737755477, | |
| "rewards/rejected": 2.3888502255431376e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10666666666666667, | |
| "grad_norm": 0.6867796553002457, | |
| "learning_rate": 3.5087719298245615e-06, | |
| "logits/chosen": -1.4618045091629028, | |
| "logits/rejected": -1.460260272026062, | |
| "logps/chosen": -249.17172241210938, | |
| "logps/rejected": -243.20840454101562, | |
| "loss": 0.6825, | |
| "rewards/accuracies": 0.875, | |
| "rewards/chosen": 0.011592308059334755, | |
| "rewards/margins": 0.01997370831668377, | |
| "rewards/rejected": -0.00838139932602644, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 0.9201550701932262, | |
| "learning_rate": 5.263157894736842e-06, | |
| "logits/chosen": -1.449467420578003, | |
| "logits/rejected": -1.4497499465942383, | |
| "logps/chosen": -244.9186248779297, | |
| "logps/rejected": -258.53594970703125, | |
| "loss": 0.6393, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": 0.0312686562538147, | |
| "rewards/margins": 0.10987410694360733, | |
| "rewards/rejected": -0.07860545814037323, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.21333333333333335, | |
| "grad_norm": 2.166122053069534, | |
| "learning_rate": 7.017543859649123e-06, | |
| "logits/chosen": -1.3264390230178833, | |
| "logits/rejected": -1.2739282846450806, | |
| "logps/chosen": -248.67361450195312, | |
| "logps/rejected": -315.01190185546875, | |
| "loss": 0.493, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": 0.05745680257678032, | |
| "rewards/margins": 0.5744289755821228, | |
| "rewards/rejected": -0.5169721841812134, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.26666666666666666, | |
| "grad_norm": 0.5760301626216394, | |
| "learning_rate": 8.771929824561405e-06, | |
| "logits/chosen": -1.0679562091827393, | |
| "logits/rejected": -0.7663384675979614, | |
| "logps/chosen": -266.1653747558594, | |
| "logps/rejected": -710.7122802734375, | |
| "loss": 0.227, | |
| "rewards/accuracies": 0.9937499761581421, | |
| "rewards/chosen": -0.03745045140385628, | |
| "rewards/margins": 4.618867874145508, | |
| "rewards/rejected": -4.656317710876465, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.1657312187471681, | |
| "learning_rate": 9.999125804563732e-06, | |
| "logits/chosen": -0.7736077308654785, | |
| "logits/rejected": -0.20556649565696716, | |
| "logps/chosen": -297.69696044921875, | |
| "logps/rejected": -1320.8775634765625, | |
| "loss": 0.0582, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.49763980507850647, | |
| "rewards/margins": 10.216059684753418, | |
| "rewards/rejected": -10.713699340820312, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.37333333333333335, | |
| "grad_norm": 0.0045546225580099286, | |
| "learning_rate": 9.983593054785776e-06, | |
| "logits/chosen": -0.42266106605529785, | |
| "logits/rejected": 0.3064606785774231, | |
| "logps/chosen": -336.424072265625, | |
| "logps/rejected": -2042.226318359375, | |
| "loss": 0.0033, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.825412392616272, | |
| "rewards/margins": 16.916912078857422, | |
| "rewards/rejected": -17.74232292175293, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4266666666666667, | |
| "grad_norm": 0.009511386498326472, | |
| "learning_rate": 9.948703190391131e-06, | |
| "logits/chosen": -0.2592306137084961, | |
| "logits/rejected": 0.5354924201965332, | |
| "logps/chosen": -330.1050109863281, | |
| "logps/rejected": -2244.28857421875, | |
| "loss": 0.0007, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9192193150520325, | |
| "rewards/margins": 18.9534912109375, | |
| "rewards/rejected": -19.872713088989258, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.04227567141303192, | |
| "learning_rate": 9.894591729488243e-06, | |
| "logits/chosen": -0.03826441615819931, | |
| "logits/rejected": 0.9021320343017578, | |
| "logps/chosen": -337.2659912109375, | |
| "logps/rejected": -2535.1923828125, | |
| "loss": 0.0004, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.057660460472107, | |
| "rewards/margins": 21.727779388427734, | |
| "rewards/rejected": -22.785442352294922, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5333333333333333, | |
| "grad_norm": 0.018772815508711668, | |
| "learning_rate": 9.82146885010314e-06, | |
| "logits/chosen": -0.08644755929708481, | |
| "logits/rejected": 0.9283016324043274, | |
| "logps/chosen": -351.6574401855469, | |
| "logps/rejected": -2719.936767578125, | |
| "loss": 0.0003, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9892961382865906, | |
| "rewards/margins": 23.653213500976562, | |
| "rewards/rejected": -24.642505645751953, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5866666666666667, | |
| "grad_norm": 0.00020416355795206898, | |
| "learning_rate": 9.72961857381258e-06, | |
| "logits/chosen": 0.040968358516693115, | |
| "logits/rejected": 1.012926697731018, | |
| "logps/chosen": -324.0478210449219, | |
| "logps/rejected": -2775.6318359375, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9146736264228821, | |
| "rewards/margins": 24.339845657348633, | |
| "rewards/rejected": -25.254520416259766, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.00020565435473269487, | |
| "learning_rate": 9.619397662556434e-06, | |
| "logits/chosen": -0.06887398660182953, | |
| "logits/rejected": 0.8867729902267456, | |
| "logps/chosen": -389.4439697265625, | |
| "logps/rejected": -2713.948974609375, | |
| "loss": 0.0011, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1910871267318726, | |
| "rewards/margins": 23.515239715576172, | |
| "rewards/rejected": -24.706327438354492, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6933333333333334, | |
| "grad_norm": 0.07831943931959111, | |
| "learning_rate": 9.49123423291422e-06, | |
| "logits/chosen": -0.011257803067564964, | |
| "logits/rejected": 1.0701903104782104, | |
| "logps/chosen": -373.259033203125, | |
| "logps/rejected": -2927.7978515625, | |
| "loss": 0.0005, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.190225601196289, | |
| "rewards/margins": 25.528148651123047, | |
| "rewards/rejected": -26.718372344970703, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7466666666666667, | |
| "grad_norm": 0.0004181495770625809, | |
| "learning_rate": 9.345626093228233e-06, | |
| "logits/chosen": 0.1232583150267601, | |
| "logits/rejected": 1.1058192253112793, | |
| "logps/chosen": -368.08062744140625, | |
| "logps/rejected": -2857.86767578125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0861897468566895, | |
| "rewards/margins": 25.07080841064453, | |
| "rewards/rejected": -26.156997680664062, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.0003206578966702994, | |
| "learning_rate": 9.1831388100321e-06, | |
| "logits/chosen": 0.22049780189990997, | |
| "logits/rejected": 1.3007761240005493, | |
| "logps/chosen": -348.95294189453125, | |
| "logps/rejected": -2912.65673828125, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9624420404434204, | |
| "rewards/margins": 25.764638900756836, | |
| "rewards/rejected": -26.72708511352539, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8533333333333334, | |
| "grad_norm": 0.0002525473318545097, | |
| "learning_rate": 9.004403511295141e-06, | |
| "logits/chosen": 0.25433215498924255, | |
| "logits/rejected": 1.5667898654937744, | |
| "logps/chosen": -354.92938232421875, | |
| "logps/rejected": -3299.885986328125, | |
| "loss": 0.0002, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9957138895988464, | |
| "rewards/margins": 29.404260635375977, | |
| "rewards/rejected": -30.39996910095215, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.9066666666666666, | |
| "grad_norm": 0.00013079558655504706, | |
| "learning_rate": 8.810114435015055e-06, | |
| "logits/chosen": 0.4517189562320709, | |
| "logits/rejected": 1.8216791152954102, | |
| "logps/chosen": -349.7914733886719, | |
| "logps/rejected": -3455.74755859375, | |
| "loss": 0.0005, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1400539875030518, | |
| "rewards/margins": 30.764062881469727, | |
| "rewards/rejected": -31.90411376953125, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.0003519489046110966, | |
| "learning_rate": 8.601026232680634e-06, | |
| "logits/chosen": 0.08518455922603607, | |
| "logits/rejected": 1.142094373703003, | |
| "logps/chosen": -370.423828125, | |
| "logps/rejected": -3112.953125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.124570608139038, | |
| "rewards/margins": 27.508525848388672, | |
| "rewards/rejected": -28.633098602294922, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.0133333333333334, | |
| "grad_norm": 9.418447063320835e-05, | |
| "learning_rate": 8.377951038078303e-06, | |
| "logits/chosen": 0.09172124415636063, | |
| "logits/rejected": 1.2133045196533203, | |
| "logps/chosen": -377.74609375, | |
| "logps/rejected": -3092.125, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.2323256731033325, | |
| "rewards/margins": 27.186447143554688, | |
| "rewards/rejected": -28.418771743774414, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0666666666666667, | |
| "grad_norm": 5.617178391629701e-05, | |
| "learning_rate": 8.141755312827737e-06, | |
| "logits/chosen": 0.26229405403137207, | |
| "logits/rejected": 1.4855743646621704, | |
| "logps/chosen": -363.44500732421875, | |
| "logps/rejected": -3202.00830078125, | |
| "loss": 0.0003, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1322368383407593, | |
| "rewards/margins": 28.364980697631836, | |
| "rewards/rejected": -29.49721908569336, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.0006911681869226501, | |
| "learning_rate": 7.89335648089903e-06, | |
| "logits/chosen": 0.27199310064315796, | |
| "logits/rejected": 1.5671254396438599, | |
| "logps/chosen": -383.5852966308594, | |
| "logps/rejected": -3335.39453125, | |
| "loss": 0.0002, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.2137231826782227, | |
| "rewards/margins": 29.575260162353516, | |
| "rewards/rejected": -30.78898048400879, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.1733333333333333, | |
| "grad_norm": 0.0001949526485184765, | |
| "learning_rate": 7.633719365183505e-06, | |
| "logits/chosen": 0.48210054636001587, | |
| "logits/rejected": 2.022475481033325, | |
| "logps/chosen": -336.2737121582031, | |
| "logps/rejected": -3361.983642578125, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.048081636428833, | |
| "rewards/margins": 30.004507064819336, | |
| "rewards/rejected": -31.052587509155273, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.2266666666666666, | |
| "grad_norm": 0.0003148592569737297, | |
| "learning_rate": 7.363852439959136e-06, | |
| "logits/chosen": 0.2750876247882843, | |
| "logits/rejected": 1.6405880451202393, | |
| "logps/chosen": -348.73944091796875, | |
| "logps/rejected": -3349.078857421875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9442297220230103, | |
| "rewards/margins": 30.076862335205078, | |
| "rewards/rejected": -31.02109146118164, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.0001452254973725037, | |
| "learning_rate": 7.084803913806642e-06, | |
| "logits/chosen": 0.2403218001127243, | |
| "logits/rejected": 1.4615638256072998, | |
| "logps/chosen": -350.90313720703125, | |
| "logps/rejected": -3641.379638671875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9335042238235474, | |
| "rewards/margins": 32.821163177490234, | |
| "rewards/rejected": -33.75466537475586, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 0.0017729334665338538, | |
| "learning_rate": 6.797657658190838e-06, | |
| "logits/chosen": 0.20432284474372864, | |
| "logits/rejected": 1.5277740955352783, | |
| "logps/chosen": -366.70648193359375, | |
| "logps/rejected": -3363.55615234375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9871833920478821, | |
| "rewards/margins": 30.192123413085938, | |
| "rewards/rejected": -31.179306030273438, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.3866666666666667, | |
| "grad_norm": 4.839956002070519e-05, | |
| "learning_rate": 6.503528997521365e-06, | |
| "logits/chosen": 0.35634833574295044, | |
| "logits/rejected": 1.6659228801727295, | |
| "logps/chosen": -353.92230224609375, | |
| "logps/rejected": -3349.120361328125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9974803924560547, | |
| "rewards/margins": 30.04400062561035, | |
| "rewards/rejected": -31.04148292541504, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.00010640905181507166, | |
| "learning_rate": 6.2035603770448664e-06, | |
| "logits/chosen": 0.20499210059642792, | |
| "logits/rejected": 1.4853886365890503, | |
| "logps/chosen": -349.24420166015625, | |
| "logps/rejected": -3330.802001953125, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9278079271316528, | |
| "rewards/margins": 29.96002197265625, | |
| "rewards/rejected": -30.887828826904297, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.4933333333333334, | |
| "grad_norm": 0.0008385515993980591, | |
| "learning_rate": 5.8989169253952635e-06, | |
| "logits/chosen": 0.48648738861083984, | |
| "logits/rejected": 1.7824747562408447, | |
| "logps/chosen": -365.4317321777344, | |
| "logps/rejected": -3538.17529296875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0642529726028442, | |
| "rewards/margins": 31.774890899658203, | |
| "rewards/rejected": -32.83914566040039, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.5466666666666666, | |
| "grad_norm": 0.006853385916627181, | |
| "learning_rate": 5.590781929037965e-06, | |
| "logits/chosen": 0.32097750902175903, | |
| "logits/rejected": 1.6314767599105835, | |
| "logps/chosen": -339.53387451171875, | |
| "logps/rejected": -3474.714111328125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9878216981887817, | |
| "rewards/margins": 31.292749404907227, | |
| "rewards/rejected": -32.28057098388672, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 5.799629098534532e-05, | |
| "learning_rate": 5.2803522361859596e-06, | |
| "logits/chosen": 0.33614903688430786, | |
| "logits/rejected": 1.679677963256836, | |
| "logps/chosen": -359.19586181640625, | |
| "logps/rejected": -3670.31787109375, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0481712818145752, | |
| "rewards/margins": 33.0955924987793, | |
| "rewards/rejected": -34.143768310546875, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.6533333333333333, | |
| "grad_norm": 4.24276661954427e-05, | |
| "learning_rate": 4.968833608039832e-06, | |
| "logits/chosen": 0.534166693687439, | |
| "logits/rejected": 1.9130985736846924, | |
| "logps/chosen": -358.578125, | |
| "logps/rejected": -3584.983154296875, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0471018552780151, | |
| "rewards/margins": 32.28336715698242, | |
| "rewards/rejected": -33.33047103881836, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.7066666666666666, | |
| "grad_norm": 0.00023914971694466748, | |
| "learning_rate": 4.657436035408217e-06, | |
| "logits/chosen": 0.5404218435287476, | |
| "logits/rejected": 2.118741273880005, | |
| "logps/chosen": -337.90423583984375, | |
| "logps/rejected": -3778.611328125, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9940769076347351, | |
| "rewards/margins": 34.16416549682617, | |
| "rewards/rejected": -35.15824508666992, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 5.3234427116302415e-05, | |
| "learning_rate": 4.347369038899744e-06, | |
| "logits/chosen": 0.2553441822528839, | |
| "logits/rejected": 1.6488205194473267, | |
| "logps/chosen": -381.4933166503906, | |
| "logps/rejected": -3569.164794921875, | |
| "loss": 0.0007, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.153830885887146, | |
| "rewards/margins": 32.051170349121094, | |
| "rewards/rejected": -33.20500183105469, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.8133333333333335, | |
| "grad_norm": 3.284575847964367e-05, | |
| "learning_rate": 4.0398369709413195e-06, | |
| "logits/chosen": 0.4040505886077881, | |
| "logits/rejected": 1.7738959789276123, | |
| "logps/chosen": -354.72760009765625, | |
| "logps/rejected": -3770.258544921875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0133658647537231, | |
| "rewards/margins": 34.131263732910156, | |
| "rewards/rejected": -35.14462661743164, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.8666666666666667, | |
| "grad_norm": 7.061849396581742e-05, | |
| "learning_rate": 3.736034337870512e-06, | |
| "logits/chosen": 0.33974844217300415, | |
| "logits/rejected": 1.8519665002822876, | |
| "logps/chosen": -352.6244201660156, | |
| "logps/rejected": -3790.14453125, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0042585134506226, | |
| "rewards/margins": 34.339210510253906, | |
| "rewards/rejected": -35.34347152709961, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.0037822509705480955, | |
| "learning_rate": 3.4371411602717785e-06, | |
| "logits/chosen": 0.2804495394229889, | |
| "logits/rejected": 1.6284122467041016, | |
| "logps/chosen": -372.8134460449219, | |
| "logps/rejected": -3452.75537109375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0843387842178345, | |
| "rewards/margins": 31.060293197631836, | |
| "rewards/rejected": -32.144630432128906, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.9733333333333334, | |
| "grad_norm": 0.0032698987512048447, | |
| "learning_rate": 3.1443183895778105e-06, | |
| "logits/chosen": 0.39163050055503845, | |
| "logits/rejected": 1.7898658514022827, | |
| "logps/chosen": -356.69293212890625, | |
| "logps/rejected": -3789.28662109375, | |
| "loss": 0.0002, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0389353036880493, | |
| "rewards/margins": 34.344696044921875, | |
| "rewards/rejected": -35.383628845214844, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 2.026666666666667, | |
| "grad_norm": 7.942034446736206e-05, | |
| "learning_rate": 2.8587033987386857e-06, | |
| "logits/chosen": 0.646866500377655, | |
| "logits/rejected": 2.1818718910217285, | |
| "logps/chosen": -347.6737365722656, | |
| "logps/rejected": -3894.88525390625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -0.9586876630783081, | |
| "rewards/margins": 35.39875030517578, | |
| "rewards/rejected": -36.35744094848633, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.007394158789638307, | |
| "learning_rate": 2.5814055644738013e-06, | |
| "logits/chosen": 0.6307215094566345, | |
| "logits/rejected": 2.338592529296875, | |
| "logps/chosen": -339.7705993652344, | |
| "logps/rejected": -4155.8330078125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.053788661956787, | |
| "rewards/margins": 37.81401824951172, | |
| "rewards/rejected": -38.8678092956543, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.1333333333333333, | |
| "grad_norm": 0.00019979686681868912, | |
| "learning_rate": 2.3135019582658803e-06, | |
| "logits/chosen": 0.3243342936038971, | |
| "logits/rejected": 1.6741666793823242, | |
| "logps/chosen": -371.49658203125, | |
| "logps/rejected": -3645.52880859375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1034276485443115, | |
| "rewards/margins": 32.93607711791992, | |
| "rewards/rejected": -34.03950881958008, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.1866666666666665, | |
| "grad_norm": 0.00033866067681241304, | |
| "learning_rate": 2.056033162833977e-06, | |
| "logits/chosen": 0.493365615606308, | |
| "logits/rejected": 2.103060007095337, | |
| "logps/chosen": -371.1989440917969, | |
| "logps/rejected": -3984.021484375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1410273313522339, | |
| "rewards/margins": 36.14289093017578, | |
| "rewards/rejected": -37.283912658691406, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.00014521714619967945, | |
| "learning_rate": 1.809999230334958e-06, | |
| "logits/chosen": 0.731221079826355, | |
| "logits/rejected": 2.3306963443756104, | |
| "logps/chosen": -349.55670166015625, | |
| "logps/rejected": -3872.18896484375, | |
| "loss": 0.0002, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0938142538070679, | |
| "rewards/margins": 35.13497543334961, | |
| "rewards/rejected": -36.22879409790039, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.2933333333333334, | |
| "grad_norm": 0.00020768738588162756, | |
| "learning_rate": 1.5763557979925326e-06, | |
| "logits/chosen": 0.5012251138687134, | |
| "logits/rejected": 1.9018030166625977, | |
| "logps/chosen": -372.3318176269531, | |
| "logps/rejected": -3688.88134765625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1943247318267822, | |
| "rewards/margins": 33.225135803222656, | |
| "rewards/rejected": -34.419456481933594, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.3466666666666667, | |
| "grad_norm": 7.310800602605835e-05, | |
| "learning_rate": 1.3560103762413584e-06, | |
| "logits/chosen": 0.5643904805183411, | |
| "logits/rejected": 2.1081748008728027, | |
| "logps/chosen": -370.6814880371094, | |
| "logps/rejected": -4141.08740234375, | |
| "loss": 0.0004, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1442639827728271, | |
| "rewards/margins": 37.65648651123047, | |
| "rewards/rejected": -38.800743103027344, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.00024154584051485787, | |
| "learning_rate": 1.1498188238036862e-06, | |
| "logits/chosen": 0.644447922706604, | |
| "logits/rejected": 2.188000440597534, | |
| "logps/chosen": -363.6133728027344, | |
| "logps/rejected": -3879.87939453125, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.153423547744751, | |
| "rewards/margins": 35.144954681396484, | |
| "rewards/rejected": -36.298377990722656, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.453333333333333, | |
| "grad_norm": 0.00016005089303980743, | |
| "learning_rate": 9.585820233899739e-07, | |
| "logits/chosen": 0.5130532383918762, | |
| "logits/rejected": 2.036015033721924, | |
| "logps/chosen": -363.82366943359375, | |
| "logps/rejected": -3935.55712890625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1580547094345093, | |
| "rewards/margins": 35.62702560424805, | |
| "rewards/rejected": -36.78507995605469, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.506666666666667, | |
| "grad_norm": 0.00036818670972699614, | |
| "learning_rate": 7.830427709355726e-07, | |
| "logits/chosen": 0.5470496416091919, | |
| "logits/rejected": 2.2380988597869873, | |
| "logps/chosen": -364.50726318359375, | |
| "logps/rejected": -4161.2060546875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1360366344451904, | |
| "rewards/margins": 37.85028839111328, | |
| "rewards/rejected": -38.986324310302734, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.00012922481748816765, | |
| "learning_rate": 6.238828904562316e-07, | |
| "logits/chosen": 0.6629059314727783, | |
| "logits/rejected": 2.3191587924957275, | |
| "logps/chosen": -363.80462646484375, | |
| "logps/rejected": -4275.02099609375, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0997378826141357, | |
| "rewards/margins": 38.973384857177734, | |
| "rewards/rejected": -40.073123931884766, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.6133333333333333, | |
| "grad_norm": 5.441705463376079e-05, | |
| "learning_rate": 4.817205857288176e-07, | |
| "logits/chosen": 0.45334744453430176, | |
| "logits/rejected": 1.8800970315933228, | |
| "logps/chosen": -363.90594482421875, | |
| "logps/rejected": -3663.01806640625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1525468826293945, | |
| "rewards/margins": 33.03852081298828, | |
| "rewards/rejected": -34.19106674194336, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 5.433719318270229e-05, | |
| "learning_rate": 3.5710803908373226e-07, | |
| "logits/chosen": 0.6295723915100098, | |
| "logits/rejected": 2.0429253578186035, | |
| "logps/chosen": -356.60894775390625, | |
| "logps/rejected": -3701.161376953125, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.0938208103179932, | |
| "rewards/margins": 33.477996826171875, | |
| "rewards/rejected": -34.571815490722656, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 0.0008541567247667036, | |
| "learning_rate": 2.5052926663577006e-07, | |
| "logits/chosen": 0.5647148489952087, | |
| "logits/rejected": 2.2748687267303467, | |
| "logps/chosen": -367.0702209472656, | |
| "logps/rejected": -4109.466796875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1226731538772583, | |
| "rewards/margins": 37.33594512939453, | |
| "rewards/rejected": -38.458614349365234, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.7733333333333334, | |
| "grad_norm": 9.149130407162097e-05, | |
| "learning_rate": 1.6239823828401945e-07, | |
| "logits/chosen": 0.8614528775215149, | |
| "logits/rejected": 2.60591721534729, | |
| "logps/chosen": -346.8439636230469, | |
| "logps/rejected": -4330.7294921875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1519685983657837, | |
| "rewards/margins": 39.40824890136719, | |
| "rewards/rejected": -40.56021499633789, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.8266666666666667, | |
| "grad_norm": 6.199788721668637e-05, | |
| "learning_rate": 9.305726978306173e-08, | |
| "logits/chosen": 0.4548477530479431, | |
| "logits/rejected": 1.8193861246109009, | |
| "logps/chosen": -368.3951721191406, | |
| "logps/rejected": -3689.09765625, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1117929220199585, | |
| "rewards/margins": 33.348350524902344, | |
| "rewards/rejected": -34.460140228271484, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.00048754024723981533, | |
| "learning_rate": 4.2775693130948094e-08, | |
| "logits/chosen": 0.601469874382019, | |
| "logits/rejected": 1.8848406076431274, | |
| "logps/chosen": -371.9225158691406, | |
| "logps/rejected": -3655.969482421875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.2053592205047607, | |
| "rewards/margins": 32.930843353271484, | |
| "rewards/rejected": -34.13620376586914, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.9333333333333336, | |
| "grad_norm": 3.647961634896807e-05, | |
| "learning_rate": 1.174881043835563e-08, | |
| "logits/chosen": 0.16144999861717224, | |
| "logits/rejected": 1.4864904880523682, | |
| "logps/chosen": -379.0067138671875, | |
| "logps/rejected": -3737.619873046875, | |
| "loss": 0.0, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.1336318254470825, | |
| "rewards/margins": 33.75095748901367, | |
| "rewards/rejected": -34.88459014892578, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.986666666666667, | |
| "grad_norm": 0.008316077390120028, | |
| "learning_rate": 9.713534230904043e-11, | |
| "logits/chosen": 0.6818611025810242, | |
| "logits/rejected": 2.075132131576538, | |
| "logps/chosen": -358.7864685058594, | |
| "logps/rejected": -3695.71484375, | |
| "loss": 0.0001, | |
| "rewards/accuracies": 1.0, | |
| "rewards/chosen": -1.12270188331604, | |
| "rewards/margins": 33.42768859863281, | |
| "rewards/rejected": -34.550392150878906, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.992, | |
| "step": 561, | |
| "total_flos": 0.0, | |
| "train_loss": 0.049965105635520604, | |
| "train_runtime": 8901.9362, | |
| "train_samples_per_second": 4.044, | |
| "train_steps_per_second": 0.063 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 561, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |