| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9998423955870764, | |
| "eval_steps": 500, | |
| "global_step": 1586, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 3.1446540880503146e-08, | |
| "logits/chosen": 0.20935016870498657, | |
| "logits/rejected": 0.28666022419929504, | |
| "logps/chosen": -338.6566467285156, | |
| "logps/rejected": -308.9756164550781, | |
| "loss": 0.1265, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 3.144654088050315e-07, | |
| "logits/chosen": 0.21327850222587585, | |
| "logits/rejected": 0.2737487554550171, | |
| "logps/chosen": -325.64306640625, | |
| "logps/rejected": -319.2493896484375, | |
| "loss": 0.1202, | |
| "rewards/accuracies": 0.4027777910232544, | |
| "rewards/chosen": 4.140691817156039e-05, | |
| "rewards/margins": -2.3560824047308415e-05, | |
| "rewards/rejected": 6.496775313280523e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 6.28930817610063e-07, | |
| "logits/chosen": 0.260469526052475, | |
| "logits/rejected": 0.3009797930717468, | |
| "logps/chosen": -294.43804931640625, | |
| "logps/rejected": -289.858154296875, | |
| "loss": 0.1078, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": 3.346197263454087e-05, | |
| "rewards/margins": 5.3948046115692705e-05, | |
| "rewards/rejected": -2.0486073481151834e-05, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 9.433962264150944e-07, | |
| "logits/chosen": 0.19516317546367645, | |
| "logits/rejected": 0.22890889644622803, | |
| "logps/chosen": -330.41143798828125, | |
| "logps/rejected": -330.6087341308594, | |
| "loss": 0.1203, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -5.2339473768370226e-05, | |
| "rewards/margins": 0.000176291519892402, | |
| "rewards/rejected": -0.0002286309900227934, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.257861635220126e-06, | |
| "logits/chosen": 0.22363004088401794, | |
| "logits/rejected": 0.2845616936683655, | |
| "logps/chosen": -351.8126525878906, | |
| "logps/rejected": -298.5321044921875, | |
| "loss": 0.1223, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.0003051063104066998, | |
| "rewards/margins": 0.0003100560570601374, | |
| "rewards/rejected": -0.0006151623092591763, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.5723270440251573e-06, | |
| "logits/chosen": 0.2019948959350586, | |
| "logits/rejected": 0.25790125131607056, | |
| "logps/chosen": -321.6785888671875, | |
| "logps/rejected": -317.89813232421875, | |
| "loss": 0.1163, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.0004650927730835974, | |
| "rewards/margins": 0.0002983523881994188, | |
| "rewards/rejected": -0.0007634450448676944, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1.8867924528301889e-06, | |
| "logits/chosen": 0.23230671882629395, | |
| "logits/rejected": 0.24606795608997345, | |
| "logps/chosen": -320.841552734375, | |
| "logps/rejected": -297.08306884765625, | |
| "loss": 0.1075, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.0008995781536214054, | |
| "rewards/margins": 0.0004379089514259249, | |
| "rewards/rejected": -0.0013374872505664825, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.2012578616352204e-06, | |
| "logits/chosen": 0.18308493494987488, | |
| "logits/rejected": 0.2341008186340332, | |
| "logps/chosen": -288.20172119140625, | |
| "logps/rejected": -298.20050048828125, | |
| "loss": 0.1169, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.0013604220002889633, | |
| "rewards/margins": 0.0006924796034581959, | |
| "rewards/rejected": -0.0020529015455394983, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.515723270440252e-06, | |
| "logits/chosen": 0.24107146263122559, | |
| "logits/rejected": 0.2765248417854309, | |
| "logps/chosen": -327.1878356933594, | |
| "logps/rejected": -324.80706787109375, | |
| "loss": 0.1138, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -0.0024016976822167635, | |
| "rewards/margins": 0.0004348896036390215, | |
| "rewards/rejected": -0.0028365873731672764, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 2.830188679245283e-06, | |
| "logits/chosen": 0.23600873351097107, | |
| "logits/rejected": 0.2604605257511139, | |
| "logps/chosen": -322.81610107421875, | |
| "logps/rejected": -312.8374328613281, | |
| "loss": 0.1158, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.0036561130546033382, | |
| "rewards/margins": 0.0011059448588639498, | |
| "rewards/rejected": -0.004762057680636644, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.1446540880503146e-06, | |
| "logits/chosen": 0.17113199830055237, | |
| "logits/rejected": 0.19848307967185974, | |
| "logps/chosen": -339.1964416503906, | |
| "logps/rejected": -318.428955078125, | |
| "loss": 0.1188, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.0050820643082261086, | |
| "rewards/margins": 0.002656476106494665, | |
| "rewards/rejected": -0.007738540414720774, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.4591194968553458e-06, | |
| "logits/chosen": 0.16253027319908142, | |
| "logits/rejected": 0.20139364898204803, | |
| "logps/chosen": -339.23040771484375, | |
| "logps/rejected": -336.61444091796875, | |
| "loss": 0.1205, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.008081575855612755, | |
| "rewards/margins": 0.0035987249575555325, | |
| "rewards/rejected": -0.011680299416184425, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.7735849056603777e-06, | |
| "logits/chosen": 0.17907488346099854, | |
| "logits/rejected": 0.1907232403755188, | |
| "logps/chosen": -347.43646240234375, | |
| "logps/rejected": -350.59881591796875, | |
| "loss": 0.1156, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.012615310959517956, | |
| "rewards/margins": 0.0076774852350354195, | |
| "rewards/rejected": -0.020292794331908226, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.08805031446541e-06, | |
| "logits/chosen": 0.14514799416065216, | |
| "logits/rejected": 0.1459697037935257, | |
| "logps/chosen": -318.0201416015625, | |
| "logps/rejected": -335.25506591796875, | |
| "loss": 0.1111, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.01840413734316826, | |
| "rewards/margins": 0.015484845265746117, | |
| "rewards/rejected": -0.033888980746269226, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.402515723270441e-06, | |
| "logits/chosen": 0.10525090992450714, | |
| "logits/rejected": 0.12337298691272736, | |
| "logps/chosen": -353.93353271484375, | |
| "logps/rejected": -395.66522216796875, | |
| "loss": 0.117, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.026070892810821533, | |
| "rewards/margins": 0.020400088280439377, | |
| "rewards/rejected": -0.04647097736597061, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.716981132075472e-06, | |
| "logits/chosen": 0.104182168841362, | |
| "logits/rejected": 0.09100546687841415, | |
| "logps/chosen": -375.31732177734375, | |
| "logps/rejected": -394.6401062011719, | |
| "loss": 0.1104, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.04359036684036255, | |
| "rewards/margins": 0.019685441628098488, | |
| "rewards/rejected": -0.06327580660581589, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999993941548484e-06, | |
| "logits/chosen": 0.06796596199274063, | |
| "logits/rejected": 0.10237328708171844, | |
| "logps/chosen": -380.8462829589844, | |
| "logps/rejected": -407.0494079589844, | |
| "loss": 0.1092, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.05505872890353203, | |
| "rewards/margins": 0.028139878064393997, | |
| "rewards/rejected": -0.08319860696792603, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.999266962896092e-06, | |
| "logits/chosen": 0.08151620626449585, | |
| "logits/rejected": 0.07926828414201736, | |
| "logps/chosen": -364.17974853515625, | |
| "logps/rejected": -370.903564453125, | |
| "loss": 0.1007, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.053988486528396606, | |
| "rewards/margins": 0.021424416452646255, | |
| "rewards/rejected": -0.07541291415691376, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.997328697661372e-06, | |
| "logits/chosen": 0.0645218938589096, | |
| "logits/rejected": 0.09341205656528473, | |
| "logps/chosen": -343.18609619140625, | |
| "logps/rejected": -353.36810302734375, | |
| "loss": 0.1208, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.049203019589185715, | |
| "rewards/margins": 0.016046565026044846, | |
| "rewards/rejected": -0.06524958461523056, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.994180085237635e-06, | |
| "logits/chosen": 0.13387538492679596, | |
| "logits/rejected": 0.13471734523773193, | |
| "logps/chosen": -333.0649108886719, | |
| "logps/rejected": -387.0137939453125, | |
| "loss": 0.1179, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.05196870118379593, | |
| "rewards/margins": 0.02883673831820488, | |
| "rewards/rejected": -0.08080543577671051, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.98982265162112e-06, | |
| "logits/chosen": 0.09480293095111847, | |
| "logits/rejected": 0.08700738847255707, | |
| "logps/chosen": -373.3707580566406, | |
| "logps/rejected": -419.16131591796875, | |
| "loss": 0.0986, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.058679305016994476, | |
| "rewards/margins": 0.025583749637007713, | |
| "rewards/rejected": -0.08426305651664734, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.984258508671404e-06, | |
| "logits/chosen": 0.12381438910961151, | |
| "logits/rejected": 0.11012091487646103, | |
| "logps/chosen": -403.3990478515625, | |
| "logps/rejected": -435.538818359375, | |
| "loss": 0.0994, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.051610808819532394, | |
| "rewards/margins": 0.041461896151304245, | |
| "rewards/rejected": -0.09307270497083664, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.977490353087878e-06, | |
| "logits/chosen": 0.13996745645999908, | |
| "logits/rejected": 0.15085197985172272, | |
| "logps/chosen": -365.3772277832031, | |
| "logps/rejected": -394.3701477050781, | |
| "loss": 0.0994, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.046448659151792526, | |
| "rewards/margins": 0.0269100870937109, | |
| "rewards/rejected": -0.07335875183343887, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.969521465102775e-06, | |
| "logits/chosen": 0.10541512072086334, | |
| "logits/rejected": 0.13170066475868225, | |
| "logps/chosen": -372.6682434082031, | |
| "logps/rejected": -432.30670166015625, | |
| "loss": 0.1162, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.055505841970443726, | |
| "rewards/margins": 0.03007945977151394, | |
| "rewards/rejected": -0.08558530360460281, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.960355706891375e-06, | |
| "logits/chosen": 0.08097681403160095, | |
| "logits/rejected": 0.08701089024543762, | |
| "logps/chosen": -377.55023193359375, | |
| "logps/rejected": -453.13702392578125, | |
| "loss": 0.1107, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.061971623450517654, | |
| "rewards/margins": 0.04781196638941765, | |
| "rewards/rejected": -0.1097835898399353, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.94999752070019e-06, | |
| "logits/chosen": 0.10144323110580444, | |
| "logits/rejected": 0.13508056104183197, | |
| "logps/chosen": -362.32012939453125, | |
| "logps/rejected": -365.26947021484375, | |
| "loss": 0.1084, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.06062735244631767, | |
| "rewards/margins": 0.024900587275624275, | |
| "rewards/rejected": -0.0855279415845871, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.938451926693985e-06, | |
| "logits/chosen": 0.11075524240732193, | |
| "logits/rejected": 0.1454174965620041, | |
| "logps/chosen": -375.95538330078125, | |
| "logps/rejected": -407.412109375, | |
| "loss": 0.1061, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.07085081189870834, | |
| "rewards/margins": 0.03195417672395706, | |
| "rewards/rejected": -0.10280498117208481, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.9257245205227296e-06, | |
| "logits/chosen": 0.14314480125904083, | |
| "logits/rejected": 0.122921422123909, | |
| "logps/chosen": -381.03472900390625, | |
| "logps/rejected": -460.407470703125, | |
| "loss": 0.0954, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.06635353714227676, | |
| "rewards/margins": 0.05833283066749573, | |
| "rewards/rejected": -0.12468637526035309, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.911821470609627e-06, | |
| "logits/chosen": 0.11854016780853271, | |
| "logits/rejected": 0.0927438884973526, | |
| "logps/chosen": -393.5327453613281, | |
| "logps/rejected": -386.33795166015625, | |
| "loss": 0.1045, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.08982465416193008, | |
| "rewards/margins": 0.011397195979952812, | |
| "rewards/rejected": -0.10122185945510864, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.896749515161544e-06, | |
| "logits/chosen": 0.16842451691627502, | |
| "logits/rejected": 0.13145343959331512, | |
| "logps/chosen": -371.6466064453125, | |
| "logps/rejected": -430.08587646484375, | |
| "loss": 0.0924, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.06959854066371918, | |
| "rewards/margins": 0.037486035376787186, | |
| "rewards/rejected": -0.10708458721637726, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.880515958903296e-06, | |
| "logits/chosen": 0.14044585824012756, | |
| "logits/rejected": 0.14415714144706726, | |
| "logps/chosen": -387.0291442871094, | |
| "logps/rejected": -432.3477478027344, | |
| "loss": 0.0995, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.07424234598875046, | |
| "rewards/margins": 0.029624024406075478, | |
| "rewards/rejected": -0.10386637598276138, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.8631286695373626e-06, | |
| "logits/chosen": 0.12493617832660675, | |
| "logits/rejected": 0.16041526198387146, | |
| "logps/chosen": -432.5501403808594, | |
| "logps/rejected": -449.15167236328125, | |
| "loss": 0.0916, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.08585993945598602, | |
| "rewards/margins": 0.024735430255532265, | |
| "rewards/rejected": -0.11059536784887314, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.844596073930749e-06, | |
| "logits/chosen": 0.11006021499633789, | |
| "logits/rejected": 0.15848229825496674, | |
| "logps/chosen": -401.59051513671875, | |
| "logps/rejected": -428.4137268066406, | |
| "loss": 0.1124, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.08783275634050369, | |
| "rewards/margins": 0.04196416586637497, | |
| "rewards/rejected": -0.12979692220687866, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.824927154030856e-06, | |
| "logits/chosen": 0.14780564606189728, | |
| "logits/rejected": 0.1997770518064499, | |
| "logps/chosen": -412.12664794921875, | |
| "logps/rejected": -425.85711669921875, | |
| "loss": 0.1016, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.08760993927717209, | |
| "rewards/margins": 0.03202151134610176, | |
| "rewards/rejected": -0.11963143199682236, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.804131442512309e-06, | |
| "logits/chosen": 0.15109729766845703, | |
| "logits/rejected": 0.17160046100616455, | |
| "logps/chosen": -411.62109375, | |
| "logps/rejected": -422.8251953125, | |
| "loss": 0.1035, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.08082294464111328, | |
| "rewards/margins": 0.03992414474487305, | |
| "rewards/rejected": -0.12074708938598633, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.78221901815689e-06, | |
| "logits/chosen": 0.11030396074056625, | |
| "logits/rejected": 0.10951008647680283, | |
| "logps/chosen": -354.53839111328125, | |
| "logps/rejected": -379.4768371582031, | |
| "loss": 0.0944, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.07093557715415955, | |
| "rewards/margins": 0.02872433327138424, | |
| "rewards/rejected": -0.09965990483760834, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.7592005009687836e-06, | |
| "logits/chosen": 0.12368909269571304, | |
| "logits/rejected": 0.1551579087972641, | |
| "logps/chosen": -422.1844787597656, | |
| "logps/rejected": -442.5235290527344, | |
| "loss": 0.0942, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.09620015323162079, | |
| "rewards/margins": 0.03423624113202095, | |
| "rewards/rejected": -0.13043639063835144, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.735087047027525e-06, | |
| "logits/chosen": 0.09877191483974457, | |
| "logits/rejected": 0.1075328141450882, | |
| "logps/chosen": -345.7369689941406, | |
| "logps/rejected": -428.72320556640625, | |
| "loss": 0.1058, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.08192293345928192, | |
| "rewards/margins": 0.041438352316617966, | |
| "rewards/rejected": -0.12336127460002899, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.709890343081127e-06, | |
| "logits/chosen": 0.11233127117156982, | |
| "logits/rejected": 0.09415660798549652, | |
| "logps/chosen": -375.85552978515625, | |
| "logps/rejected": -400.69561767578125, | |
| "loss": 0.0912, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.07030915468931198, | |
| "rewards/margins": 0.028951376676559448, | |
| "rewards/rejected": -0.09926052391529083, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.6836226008820215e-06, | |
| "logits/chosen": 0.16574355959892273, | |
| "logits/rejected": 0.1816074252128601, | |
| "logps/chosen": -416.6439514160156, | |
| "logps/rejected": -431.9356994628906, | |
| "loss": 0.0968, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.06910283863544464, | |
| "rewards/margins": 0.021879449486732483, | |
| "rewards/rejected": -0.09098228812217712, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.656296551268547e-06, | |
| "logits/chosen": 0.15201489627361298, | |
| "logits/rejected": 0.18494170904159546, | |
| "logps/chosen": -382.65350341796875, | |
| "logps/rejected": -439.33502197265625, | |
| "loss": 0.0938, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.0796704962849617, | |
| "rewards/margins": 0.04547972232103348, | |
| "rewards/rejected": -0.12515023350715637, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.627925437994868e-06, | |
| "logits/chosen": 0.13324889540672302, | |
| "logits/rejected": 0.15600983798503876, | |
| "logps/chosen": -342.90924072265625, | |
| "logps/rejected": -418.2884216308594, | |
| "loss": 0.0905, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.08169208467006683, | |
| "rewards/margins": 0.04994327947497368, | |
| "rewards/rejected": -0.1316353678703308, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.598523011312296e-06, | |
| "logits/chosen": 0.15302890539169312, | |
| "logits/rejected": 0.15263435244560242, | |
| "logps/chosen": -399.0650939941406, | |
| "logps/rejected": -442.39080810546875, | |
| "loss": 0.0898, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.08143939077854156, | |
| "rewards/margins": 0.03956509754061699, | |
| "rewards/rejected": -0.12100447714328766, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.568103521305137e-06, | |
| "logits/chosen": 0.18960033357143402, | |
| "logits/rejected": 0.15051600337028503, | |
| "logps/chosen": -382.1358337402344, | |
| "logps/rejected": -444.8484802246094, | |
| "loss": 0.0959, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.07425881922245026, | |
| "rewards/margins": 0.04583179950714111, | |
| "rewards/rejected": -0.12009060382843018, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.536681710984296e-06, | |
| "logits/chosen": 0.14668092131614685, | |
| "logits/rejected": 0.20314042270183563, | |
| "logps/chosen": -405.0791015625, | |
| "logps/rejected": -454.7623596191406, | |
| "loss": 0.0927, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.07047592103481293, | |
| "rewards/margins": 0.05656527355313301, | |
| "rewards/rejected": -0.12704120576381683, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.5042728091419764e-06, | |
| "logits/chosen": 0.14518819749355316, | |
| "logits/rejected": 0.15771886706352234, | |
| "logps/chosen": -351.4585266113281, | |
| "logps/rejected": -418.5862731933594, | |
| "loss": 0.0903, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.06795553117990494, | |
| "rewards/margins": 0.03801413252949715, | |
| "rewards/rejected": -0.10596966743469238, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.470892522970937e-06, | |
| "logits/chosen": 0.12991555035114288, | |
| "logits/rejected": 0.13929405808448792, | |
| "logps/chosen": -369.97943115234375, | |
| "logps/rejected": -418.1014099121094, | |
| "loss": 0.0993, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.06941680610179901, | |
| "rewards/margins": 0.03608622029423714, | |
| "rewards/rejected": -0.10550303757190704, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.436557030451907e-06, | |
| "logits/chosen": 0.13183408975601196, | |
| "logits/rejected": 0.14138123393058777, | |
| "logps/chosen": -386.870849609375, | |
| "logps/rejected": -483.94854736328125, | |
| "loss": 0.1037, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.08128295838832855, | |
| "rewards/margins": 0.05732026696205139, | |
| "rewards/rejected": -0.13860321044921875, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.401282972512805e-06, | |
| "logits/chosen": 0.09972018748521805, | |
| "logits/rejected": 0.10249191522598267, | |
| "logps/chosen": -423.345703125, | |
| "logps/rejected": -447.601318359375, | |
| "loss": 0.1098, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.09919498860836029, | |
| "rewards/margins": 0.02907257154583931, | |
| "rewards/rejected": -0.1282675564289093, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.365087444963611e-06, | |
| "logits/chosen": 0.08806703239679337, | |
| "logits/rejected": 0.1713884323835373, | |
| "logps/chosen": -467.56182861328125, | |
| "logps/rejected": -477.89581298828125, | |
| "loss": 0.0904, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.09871919453144073, | |
| "rewards/margins": 0.04472241550683975, | |
| "rewards/rejected": -0.14344163239002228, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.327987990210763e-06, | |
| "logits/chosen": 0.126537024974823, | |
| "logits/rejected": 0.15067018568515778, | |
| "logps/chosen": -361.6537170410156, | |
| "logps/rejected": -433.5013122558594, | |
| "loss": 0.0984, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.08296778053045273, | |
| "rewards/margins": 0.053381673991680145, | |
| "rewards/rejected": -0.13634945452213287, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.290002588755112e-06, | |
| "logits/chosen": 0.14435867965221405, | |
| "logits/rejected": 0.18025091290473938, | |
| "logps/chosen": -383.94110107421875, | |
| "logps/rejected": -440.88446044921875, | |
| "loss": 0.0917, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.07438959181308746, | |
| "rewards/margins": 0.035205405205488205, | |
| "rewards/rejected": -0.10959500074386597, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.251149650477553e-06, | |
| "logits/chosen": 0.11688809096813202, | |
| "logits/rejected": 0.1297851949930191, | |
| "logps/chosen": -420.3448181152344, | |
| "logps/rejected": -450.2854919433594, | |
| "loss": 0.1182, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.10054901987314224, | |
| "rewards/margins": 0.025883013382554054, | |
| "rewards/rejected": -0.12643203139305115, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.211448005716547e-06, | |
| "logits/chosen": 0.140035018324852, | |
| "logits/rejected": 0.1768953651189804, | |
| "logps/chosen": -340.6114196777344, | |
| "logps/rejected": -433.100341796875, | |
| "loss": 0.1061, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.06748305261135101, | |
| "rewards/margins": 0.05944541096687317, | |
| "rewards/rejected": -0.12692847847938538, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.170916896141877e-06, | |
| "logits/chosen": 0.11086130142211914, | |
| "logits/rejected": 0.11766747385263443, | |
| "logps/chosen": -409.5714111328125, | |
| "logps/rejected": -471.7931213378906, | |
| "loss": 0.0909, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.08654967695474625, | |
| "rewards/margins": 0.052199166268110275, | |
| "rewards/rejected": -0.13874883949756622, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.129575965429036e-06, | |
| "logits/chosen": 0.1578402817249298, | |
| "logits/rejected": 0.12391265481710434, | |
| "logps/chosen": -406.35784912109375, | |
| "logps/rejected": -423.5259704589844, | |
| "loss": 0.0988, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.07460321485996246, | |
| "rewards/margins": 0.04027136042714119, | |
| "rewards/rejected": -0.11487456411123276, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.087445249738795e-06, | |
| "logits/chosen": 0.13769268989562988, | |
| "logits/rejected": 0.14535634219646454, | |
| "logps/chosen": -402.810791015625, | |
| "logps/rejected": -468.11578369140625, | |
| "loss": 0.0885, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.07266028970479965, | |
| "rewards/margins": 0.04542839899659157, | |
| "rewards/rejected": -0.11808867752552032, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.044545168006535e-06, | |
| "logits/chosen": 0.135701984167099, | |
| "logits/rejected": 0.15048106014728546, | |
| "logps/chosen": -367.40716552734375, | |
| "logps/rejected": -425.750732421875, | |
| "loss": 0.0933, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.06163717433810234, | |
| "rewards/margins": 0.0523550882935524, | |
| "rewards/rejected": -0.11399225890636444, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 4.000896512046079e-06, | |
| "logits/chosen": 0.06643296778202057, | |
| "logits/rejected": 0.11304454505443573, | |
| "logps/chosen": -382.0234069824219, | |
| "logps/rejected": -408.7083435058594, | |
| "loss": 0.0964, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.08725063502788544, | |
| "rewards/margins": 0.030442586168646812, | |
| "rewards/rejected": -0.117693230509758, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.956520436472794e-06, | |
| "logits/chosen": 0.09026265144348145, | |
| "logits/rejected": 0.14468029141426086, | |
| "logps/chosen": -425.26214599609375, | |
| "logps/rejected": -457.64990234375, | |
| "loss": 0.0923, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.09045656025409698, | |
| "rewards/margins": 0.050621867179870605, | |
| "rewards/rejected": -0.1410784274339676, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.911438448450867e-06, | |
| "logits/chosen": 0.09659644216299057, | |
| "logits/rejected": 0.10519146919250488, | |
| "logps/chosen": -409.36737060546875, | |
| "logps/rejected": -426.3362731933594, | |
| "loss": 0.0965, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.09249825030565262, | |
| "rewards/margins": 0.03923904523253441, | |
| "rewards/rejected": -0.13173729181289673, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.865672397269712e-06, | |
| "logits/chosen": 0.16088059544563293, | |
| "logits/rejected": 0.15578779578208923, | |
| "logps/chosen": -421.38134765625, | |
| "logps/rejected": -467.1226501464844, | |
| "loss": 0.103, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.09401367604732513, | |
| "rewards/margins": 0.037443626672029495, | |
| "rewards/rejected": -0.13145729899406433, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.819244463754566e-06, | |
| "logits/chosen": 0.1309213638305664, | |
| "logits/rejected": 0.183641254901886, | |
| "logps/chosen": -372.6539611816406, | |
| "logps/rejected": -422.17333984375, | |
| "loss": 0.1079, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.07998184859752655, | |
| "rewards/margins": 0.03546987846493721, | |
| "rewards/rejected": -0.11545171588659286, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7721771495164e-06, | |
| "logits/chosen": 0.18562278151512146, | |
| "logits/rejected": 0.17855581641197205, | |
| "logps/chosen": -368.06634521484375, | |
| "logps/rejected": -410.43157958984375, | |
| "loss": 0.0984, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.06547744572162628, | |
| "rewards/margins": 0.037668634206056595, | |
| "rewards/rejected": -0.10314607620239258, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7244932660463628e-06, | |
| "logits/chosen": 0.10567553341388702, | |
| "logits/rejected": 0.09214256703853607, | |
| "logps/chosen": -387.8129577636719, | |
| "logps/rejected": -428.2767028808594, | |
| "loss": 0.1038, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.08541779965162277, | |
| "rewards/margins": 0.04381578415632248, | |
| "rewards/rejected": -0.12923358380794525, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.676215923660037e-06, | |
| "logits/chosen": 0.1147516518831253, | |
| "logits/rejected": 0.13854558765888214, | |
| "logps/chosen": -388.88397216796875, | |
| "logps/rejected": -402.66162109375, | |
| "loss": 0.0979, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.08885400742292404, | |
| "rewards/margins": 0.03468873351812363, | |
| "rewards/rejected": -0.12354274094104767, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.6273685202968727e-06, | |
| "logits/chosen": 0.14652499556541443, | |
| "logits/rejected": 0.23127985000610352, | |
| "logps/chosen": -427.195556640625, | |
| "logps/rejected": -474.3085021972656, | |
| "loss": 0.1012, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.10611389577388763, | |
| "rewards/margins": 0.047405023127794266, | |
| "rewards/rejected": -0.1535189151763916, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.577974730180212e-06, | |
| "logits/chosen": 0.1295718103647232, | |
| "logits/rejected": 0.18317390978336334, | |
| "logps/chosen": -454.80938720703125, | |
| "logps/rejected": -480.46490478515625, | |
| "loss": 0.1008, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.09846655279397964, | |
| "rewards/margins": 0.038999587297439575, | |
| "rewards/rejected": -0.13746614754199982, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.528058492343423e-06, | |
| "logits/chosen": 0.11106806993484497, | |
| "logits/rejected": 0.1717728078365326, | |
| "logps/chosen": -415.18414306640625, | |
| "logps/rejected": -438.2942810058594, | |
| "loss": 0.1083, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.1040610671043396, | |
| "rewards/margins": 0.030150944367051125, | |
| "rewards/rejected": -0.13421201705932617, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.477643999027683e-06, | |
| "logits/chosen": 0.16643357276916504, | |
| "logits/rejected": 0.18508140742778778, | |
| "logps/chosen": -382.0606384277344, | |
| "logps/rejected": -410.14691162109375, | |
| "loss": 0.0969, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.08891114592552185, | |
| "rewards/margins": 0.03431970998644829, | |
| "rewards/rejected": -0.12323085963726044, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4267556839570467e-06, | |
| "logits/chosen": 0.11085295677185059, | |
| "logits/rejected": 0.11592169851064682, | |
| "logps/chosen": -380.14892578125, | |
| "logps/rejected": -436.01422119140625, | |
| "loss": 0.0928, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.08309037983417511, | |
| "rewards/margins": 0.050611067563295364, | |
| "rewards/rejected": -0.13370142877101898, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.375418210496473e-06, | |
| "logits/chosen": 0.14460065960884094, | |
| "logits/rejected": 0.17223772406578064, | |
| "logps/chosen": -410.51422119140625, | |
| "logps/rejected": -431.4485778808594, | |
| "loss": 0.1039, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.08633999526500702, | |
| "rewards/margins": 0.03471770137548447, | |
| "rewards/rejected": -0.12105770409107208, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3236564596985677e-06, | |
| "logits/chosen": 0.11466125398874283, | |
| "logits/rejected": 0.14213159680366516, | |
| "logps/chosen": -429.1424865722656, | |
| "logps/rejected": -489.776123046875, | |
| "loss": 0.083, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.08991511166095734, | |
| "rewards/margins": 0.04996410012245178, | |
| "rewards/rejected": -0.13987921178340912, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2714955182448044e-06, | |
| "logits/chosen": 0.06632956117391586, | |
| "logits/rejected": 0.10460522025823593, | |
| "logps/chosen": -355.56988525390625, | |
| "logps/rejected": -453.8470764160156, | |
| "loss": 0.1004, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.07422816753387451, | |
| "rewards/margins": 0.059829723089933395, | |
| "rewards/rejected": -0.1340578943490982, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.2189606662871017e-06, | |
| "logits/chosen": 0.1409912258386612, | |
| "logits/rejected": 0.14539703726768494, | |
| "logps/chosen": -390.64251708984375, | |
| "logps/rejected": -435.5365295410156, | |
| "loss": 0.0906, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.07141707837581635, | |
| "rewards/margins": 0.03945548087358475, | |
| "rewards/rejected": -0.1108725517988205, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1660773651956223e-06, | |
| "logits/chosen": 0.12456997483968735, | |
| "logits/rejected": 0.1372366100549698, | |
| "logps/chosen": -355.9635314941406, | |
| "logps/rejected": -383.43853759765625, | |
| "loss": 0.1004, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.07638130336999893, | |
| "rewards/margins": 0.037354812026023865, | |
| "rewards/rejected": -0.1137361153960228, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.1128712452187535e-06, | |
| "logits/chosen": 0.09267710149288177, | |
| "logits/rejected": 0.07987795025110245, | |
| "logps/chosen": -402.697509765625, | |
| "logps/rejected": -450.74786376953125, | |
| "loss": 0.0967, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.08393266797065735, | |
| "rewards/margins": 0.04311661794781685, | |
| "rewards/rejected": -0.1270492970943451, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0593680930612295e-06, | |
| "logits/chosen": 0.0781303271651268, | |
| "logits/rejected": 0.16469840705394745, | |
| "logps/chosen": -437.64013671875, | |
| "logps/rejected": -474.09661865234375, | |
| "loss": 0.1021, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.10180965811014175, | |
| "rewards/margins": 0.03583382070064545, | |
| "rewards/rejected": -0.1376434713602066, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.005593839386434e-06, | |
| "logits/chosen": 0.10138174146413803, | |
| "logits/rejected": 0.1595858335494995, | |
| "logps/chosen": -466.75811767578125, | |
| "logps/rejected": -494.52435302734375, | |
| "loss": 0.0913, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.09192174673080444, | |
| "rewards/margins": 0.048992156982421875, | |
| "rewards/rejected": -0.14091390371322632, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.951574546248928e-06, | |
| "logits/chosen": 0.09503944218158722, | |
| "logits/rejected": 0.1364067792892456, | |
| "logps/chosen": -458.38775634765625, | |
| "logps/rejected": -459.333251953125, | |
| "loss": 0.0985, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.09930691868066788, | |
| "rewards/margins": 0.029138978570699692, | |
| "rewards/rejected": -0.12844590842723846, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.897336394463301e-06, | |
| "logits/chosen": 0.10846900939941406, | |
| "logits/rejected": 0.09937886893749237, | |
| "logps/chosen": -424.815673828125, | |
| "logps/rejected": -459.7933044433594, | |
| "loss": 0.0941, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.08887843042612076, | |
| "rewards/margins": 0.05371922254562378, | |
| "rewards/rejected": -0.14259764552116394, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.842905670915462e-06, | |
| "logits/chosen": 0.1058458462357521, | |
| "logits/rejected": 0.0782022550702095, | |
| "logps/chosen": -434.9790954589844, | |
| "logps/rejected": -469.34454345703125, | |
| "loss": 0.0967, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.09677759557962418, | |
| "rewards/margins": 0.04304363578557968, | |
| "rewards/rejected": -0.13982121646404266, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.788308755822522e-06, | |
| "logits/chosen": 0.0967346802353859, | |
| "logits/rejected": 0.10881376266479492, | |
| "logps/chosen": -369.01995849609375, | |
| "logps/rejected": -449.07623291015625, | |
| "loss": 0.0926, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.09960204362869263, | |
| "rewards/margins": 0.04991160333156586, | |
| "rewards/rejected": -0.1495136320590973, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.733572109947444e-06, | |
| "logits/chosen": 0.10651399195194244, | |
| "logits/rejected": 0.128664031624794, | |
| "logps/chosen": -466.0299377441406, | |
| "logps/rejected": -499.25762939453125, | |
| "loss": 0.0972, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.1191006675362587, | |
| "rewards/margins": 0.044090814888477325, | |
| "rewards/rejected": -0.16319148242473602, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.6787222617746583e-06, | |
| "logits/chosen": 0.12026312202215195, | |
| "logits/rejected": 0.1642097681760788, | |
| "logps/chosen": -380.1086120605469, | |
| "logps/rejected": -439.4449157714844, | |
| "loss": 0.0884, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.09910561889410019, | |
| "rewards/margins": 0.04509486258029938, | |
| "rewards/rejected": -0.14420048892498016, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.62378579465285e-06, | |
| "logits/chosen": 0.12080010026693344, | |
| "logits/rejected": 0.15950220823287964, | |
| "logps/chosen": -441.6851501464844, | |
| "logps/rejected": -498.75152587890625, | |
| "loss": 0.0919, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.09932798892259598, | |
| "rewards/margins": 0.05052763968706131, | |
| "rewards/rejected": -0.14985564351081848, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.568789333911165e-06, | |
| "logits/chosen": 0.1327330321073532, | |
| "logits/rejected": 0.14310786128044128, | |
| "logps/chosen": -375.2086486816406, | |
| "logps/rejected": -488.7005310058594, | |
| "loss": 0.0957, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.08790336549282074, | |
| "rewards/margins": 0.0683864951133728, | |
| "rewards/rejected": -0.15628986060619354, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.5137595339550564e-06, | |
| "logits/chosen": 0.10466363281011581, | |
| "logits/rejected": 0.12011264264583588, | |
| "logps/chosen": -457.970458984375, | |
| "logps/rejected": -444.8792419433594, | |
| "loss": 0.0943, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.10931634902954102, | |
| "rewards/margins": 0.027462299913167953, | |
| "rewards/rejected": -0.13677865266799927, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.4587230653480577e-06, | |
| "logits/chosen": 0.07321979105472565, | |
| "logits/rejected": 0.09171778708696365, | |
| "logps/chosen": -435.8045349121094, | |
| "logps/rejected": -464.2123107910156, | |
| "loss": 0.0896, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.11067759990692139, | |
| "rewards/margins": 0.045581571757793427, | |
| "rewards/rejected": -0.15625916421413422, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.403706601885705e-06, | |
| "logits/chosen": 0.12501958012580872, | |
| "logits/rejected": 0.10027003288269043, | |
| "logps/chosen": -439.4681701660156, | |
| "logps/rejected": -450.8436584472656, | |
| "loss": 0.0988, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.09043081104755402, | |
| "rewards/margins": 0.03802182897925377, | |
| "rewards/rejected": -0.12845264375209808, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3487368076679095e-06, | |
| "logits/chosen": 0.03233950585126877, | |
| "logits/rejected": 0.10005395114421844, | |
| "logps/chosen": -423.36163330078125, | |
| "logps/rejected": -499.0625, | |
| "loss": 0.0823, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.10427310317754745, | |
| "rewards/margins": 0.07294373214244843, | |
| "rewards/rejected": -0.17721685767173767, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.2938403241760153e-06, | |
| "logits/chosen": 0.046649765223264694, | |
| "logits/rejected": 0.06434310227632523, | |
| "logps/chosen": -385.28753662109375, | |
| "logps/rejected": -428.0790100097656, | |
| "loss": 0.0921, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.097981296479702, | |
| "rewards/margins": 0.046177938580513, | |
| "rewards/rejected": -0.1441592425107956, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2390437573608313e-06, | |
| "logits/chosen": 0.04348721355199814, | |
| "logits/rejected": 0.09973670542240143, | |
| "logps/chosen": -419.55841064453125, | |
| "logps/rejected": -455.224853515625, | |
| "loss": 0.0924, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.11863915622234344, | |
| "rewards/margins": 0.03215927258133888, | |
| "rewards/rejected": -0.15079843997955322, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1843736647478733e-06, | |
| "logits/chosen": 0.12401044368743896, | |
| "logits/rejected": 0.11555816233158112, | |
| "logps/chosen": -461.777099609375, | |
| "logps/rejected": -446.89154052734375, | |
| "loss": 0.0958, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.10517005622386932, | |
| "rewards/margins": 0.03241740167140961, | |
| "rewards/rejected": -0.13758745789527893, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1298565425660842e-06, | |
| "logits/chosen": 0.12125201523303986, | |
| "logits/rejected": 0.13117629289627075, | |
| "logps/chosen": -443.516845703125, | |
| "logps/rejected": -483.62030029296875, | |
| "loss": 0.086, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.09896551072597504, | |
| "rewards/margins": 0.045016057789325714, | |
| "rewards/rejected": -0.14398157596588135, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0755188129062566e-06, | |
| "logits/chosen": 0.1026390790939331, | |
| "logits/rejected": 0.09158008545637131, | |
| "logps/chosen": -444.39862060546875, | |
| "logps/rejected": -479.50390625, | |
| "loss": 0.0983, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.10992302745580673, | |
| "rewards/margins": 0.03600381314754486, | |
| "rewards/rejected": -0.145926833152771, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.0213868109153843e-06, | |
| "logits/chosen": 0.0887732207775116, | |
| "logits/rejected": 0.13315047323703766, | |
| "logps/chosen": -451.1080627441406, | |
| "logps/rejected": -477.95477294921875, | |
| "loss": 0.0883, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.10277727991342545, | |
| "rewards/margins": 0.045780349522829056, | |
| "rewards/rejected": -0.1485576331615448, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.967486772033164e-06, | |
| "logits/chosen": 0.1042795404791832, | |
| "logits/rejected": 0.1479608118534088, | |
| "logps/chosen": -433.328125, | |
| "logps/rejected": -472.20904541015625, | |
| "loss": 0.0802, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.09643669426441193, | |
| "rewards/margins": 0.04762507602572441, | |
| "rewards/rejected": -0.14406177401542664, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.9138448192768072e-06, | |
| "logits/chosen": 0.08118772506713867, | |
| "logits/rejected": 0.09627708047628403, | |
| "logps/chosen": -424.64898681640625, | |
| "logps/rejected": -493.44476318359375, | |
| "loss": 0.0919, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.08744023740291595, | |
| "rewards/margins": 0.0591917559504509, | |
| "rewards/rejected": -0.14663198590278625, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8604869505803407e-06, | |
| "logits/chosen": 0.1322592794895172, | |
| "logits/rejected": 0.13863101601600647, | |
| "logps/chosen": -423.8280334472656, | |
| "logps/rejected": -469.0419921875, | |
| "loss": 0.0978, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.09548251330852509, | |
| "rewards/margins": 0.05783646181225777, | |
| "rewards/rejected": -0.15331897139549255, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.8074390261945418e-06, | |
| "logits/chosen": 0.10184427350759506, | |
| "logits/rejected": 0.12423571199178696, | |
| "logps/chosen": -377.3212585449219, | |
| "logps/rejected": -427.81317138671875, | |
| "loss": 0.0902, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.0946311205625534, | |
| "rewards/margins": 0.046779900789260864, | |
| "rewards/rejected": -0.14141102135181427, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.7547267561535836e-06, | |
| "logits/chosen": 0.06255786120891571, | |
| "logits/rejected": 0.12503619492053986, | |
| "logps/chosen": -418.79144287109375, | |
| "logps/rejected": -441.3435974121094, | |
| "loss": 0.0864, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.10577164590358734, | |
| "rewards/margins": 0.05561785027384758, | |
| "rewards/rejected": -0.16138949990272522, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.702375687814501e-06, | |
| "logits/chosen": 0.10921992361545563, | |
| "logits/rejected": 0.0866740271449089, | |
| "logps/chosen": -404.1749267578125, | |
| "logps/rejected": -476.277099609375, | |
| "loss": 0.0791, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.10695375502109528, | |
| "rewards/margins": 0.05870165675878525, | |
| "rewards/rejected": -0.16565541923046112, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.650411193475489e-06, | |
| "logits/chosen": 0.074369415640831, | |
| "logits/rejected": 0.12412852048873901, | |
| "logps/chosen": -420.69793701171875, | |
| "logps/rejected": -484.34161376953125, | |
| "loss": 0.0868, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.11432509124279022, | |
| "rewards/margins": 0.051989488303661346, | |
| "rewards/rejected": -0.16631458699703217, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5988584580790452e-06, | |
| "logits/chosen": 0.09442147612571716, | |
| "logits/rejected": 0.10659205913543701, | |
| "logps/chosen": -458.6270446777344, | |
| "logps/rejected": -523.6939697265625, | |
| "loss": 0.0705, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.10721229016780853, | |
| "rewards/margins": 0.05447040870785713, | |
| "rewards/rejected": -0.16168268024921417, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5477424670059227e-06, | |
| "logits/chosen": 0.08504728972911835, | |
| "logits/rejected": 0.10464467108249664, | |
| "logps/chosen": -423.983154296875, | |
| "logps/rejected": -449.3802185058594, | |
| "loss": 0.0851, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.11461503803730011, | |
| "rewards/margins": 0.03833803907036781, | |
| "rewards/rejected": -0.15295308828353882, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.4970879939657929e-06, | |
| "logits/chosen": 0.08304135501384735, | |
| "logits/rejected": 0.11797134578227997, | |
| "logps/chosen": -407.0821838378906, | |
| "logps/rejected": -456.46533203125, | |
| "loss": 0.0796, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.08808410167694092, | |
| "rewards/margins": 0.05510552600026131, | |
| "rewards/rejected": -0.14318962395191193, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.4469195889904983e-06, | |
| "logits/chosen": 0.044432397931814194, | |
| "logits/rejected": 0.10810061544179916, | |
| "logps/chosen": -429.70794677734375, | |
| "logps/rejected": -452.116943359375, | |
| "loss": 0.0926, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.10013709962368011, | |
| "rewards/margins": 0.03565113991498947, | |
| "rewards/rejected": -0.13578826189041138, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.3972615665357228e-06, | |
| "logits/chosen": 0.08212409913539886, | |
| "logits/rejected": 0.08114130049943924, | |
| "logps/chosen": -425.5025939941406, | |
| "logps/rejected": -518.18115234375, | |
| "loss": 0.0884, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.09838120639324188, | |
| "rewards/margins": 0.061208415776491165, | |
| "rewards/rejected": -0.15958961844444275, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3481379936968154e-06, | |
| "logits/chosen": 0.06021580845117569, | |
| "logits/rejected": 0.1053982749581337, | |
| "logps/chosen": -471.2576599121094, | |
| "logps/rejected": -499.8750915527344, | |
| "loss": 0.1021, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.10752852261066437, | |
| "rewards/margins": 0.04095206782221794, | |
| "rewards/rejected": -0.1484805792570114, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.2995726785445187e-06, | |
| "logits/chosen": 0.08110491931438446, | |
| "logits/rejected": 0.13657256960868835, | |
| "logps/chosen": -457.90234375, | |
| "logps/rejected": -445.5087890625, | |
| "loss": 0.0918, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.11181896924972534, | |
| "rewards/margins": 0.03636185824871063, | |
| "rewards/rejected": -0.14818084239959717, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2515891585862248e-06, | |
| "logits/chosen": 0.06536366045475006, | |
| "logits/rejected": 0.10223551094532013, | |
| "logps/chosen": -424.5037536621094, | |
| "logps/rejected": -465.29425048828125, | |
| "loss": 0.112, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.10022538900375366, | |
| "rewards/margins": 0.03513466566801071, | |
| "rewards/rejected": -0.13536006212234497, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.204210689358359e-06, | |
| "logits/chosen": 0.15191073715686798, | |
| "logits/rejected": 0.12366287410259247, | |
| "logps/chosen": -436.87823486328125, | |
| "logps/rejected": -426.62762451171875, | |
| "loss": 0.0811, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.09861338883638382, | |
| "rewards/margins": 0.033628519624471664, | |
| "rewards/rejected": -0.13224190473556519, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1574602331554352e-06, | |
| "logits/chosen": 0.05633612722158432, | |
| "logits/rejected": 0.04986396059393883, | |
| "logps/chosen": -424.77984619140625, | |
| "logps/rejected": -440.7940979003906, | |
| "loss": 0.1042, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.1184263676404953, | |
| "rewards/margins": 0.029354307800531387, | |
| "rewards/rejected": -0.1477806717157364, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.1113604479012202e-06, | |
| "logits/chosen": 0.04834729805588722, | |
| "logits/rejected": 0.079189732670784, | |
| "logps/chosen": -402.3204040527344, | |
| "logps/rejected": -482.1253967285156, | |
| "loss": 0.1046, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.10266365110874176, | |
| "rewards/margins": 0.04556725174188614, | |
| "rewards/rejected": -0.1482308954000473, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.0659336761674163e-06, | |
| "logits/chosen": 0.07221159338951111, | |
| "logits/rejected": 0.06988708674907684, | |
| "logps/chosen": -378.05255126953125, | |
| "logps/rejected": -401.8747863769531, | |
| "loss": 0.0872, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.09310795366764069, | |
| "rewards/margins": 0.04454265162348747, | |
| "rewards/rejected": -0.13765059411525726, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0212019343451889e-06, | |
| "logits/chosen": 0.051744282245635986, | |
| "logits/rejected": 0.05535260960459709, | |
| "logps/chosen": -367.1768798828125, | |
| "logps/rejected": -443.02960205078125, | |
| "loss": 0.0823, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.08505652099847794, | |
| "rewards/margins": 0.06588119268417358, | |
| "rewards/rejected": -0.15093770623207092, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.771869019747708e-07, | |
| "logits/chosen": 0.0807739868760109, | |
| "logits/rejected": 0.09035824239253998, | |
| "logps/chosen": -437.5768127441406, | |
| "logps/rejected": -521.4659423828125, | |
| "loss": 0.0918, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.11788342148065567, | |
| "rewards/margins": 0.07230274379253387, | |
| "rewards/rejected": -0.19018617272377014, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.339099112383263e-07, | |
| "logits/chosen": 0.08397279679775238, | |
| "logits/rejected": 0.12051109224557877, | |
| "logps/chosen": -429.1914978027344, | |
| "logps/rejected": -477.98968505859375, | |
| "loss": 0.0898, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.1165337786078453, | |
| "rewards/margins": 0.0657678097486496, | |
| "rewards/rejected": -0.1823015958070755, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.913919366211679e-07, | |
| "logits/chosen": 0.01225881278514862, | |
| "logits/rejected": 0.08070291578769684, | |
| "logps/chosen": -453.7749938964844, | |
| "logps/rejected": -491.718017578125, | |
| "loss": 0.0784, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.12794212996959686, | |
| "rewards/margins": 0.046219903975725174, | |
| "rewards/rejected": -0.17416204512119293, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.496535847463264e-07, | |
| "logits/chosen": 0.05466824769973755, | |
| "logits/rejected": 0.06951786577701569, | |
| "logps/chosen": -409.43731689453125, | |
| "logps/rejected": -478.67730712890625, | |
| "loss": 0.0878, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.0958373174071312, | |
| "rewards/margins": 0.06815023720264435, | |
| "rewards/rejected": -0.16398754715919495, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.08715084387417e-07, | |
| "logits/chosen": 0.0821441262960434, | |
| "logits/rejected": 0.09297174215316772, | |
| "logps/chosen": -443.9449768066406, | |
| "logps/rejected": -495.43304443359375, | |
| "loss": 0.0746, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.09165932238101959, | |
| "rewards/margins": 0.06980036199092865, | |
| "rewards/rejected": -0.16145966947078705, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.685962766646251e-07, | |
| "logits/chosen": 0.0996304377913475, | |
| "logits/rejected": 0.1416376680135727, | |
| "logps/chosen": -421.4239807128906, | |
| "logps/rejected": -448.3440856933594, | |
| "loss": 0.0854, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.1046370267868042, | |
| "rewards/margins": 0.04357679933309555, | |
| "rewards/rejected": -0.14821383357048035, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.29316605428573e-07, | |
| "logits/chosen": 0.050212837755680084, | |
| "logits/rejected": 0.08997879177331924, | |
| "logps/chosen": -405.666748046875, | |
| "logps/rejected": -481.518798828125, | |
| "loss": 0.0874, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.10379258543252945, | |
| "rewards/margins": 0.06835626810789108, | |
| "rewards/rejected": -0.17214885354042053, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 6.908951078367284e-07, | |
| "logits/chosen": 0.06391054391860962, | |
| "logits/rejected": 0.11078351736068726, | |
| "logps/chosen": -472.7290954589844, | |
| "logps/rejected": -525.4988403320312, | |
| "loss": 0.0804, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.11707256734371185, | |
| "rewards/margins": 0.06153646111488342, | |
| "rewards/rejected": -0.17860902845859528, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.533504051269185e-07, | |
| "logits/chosen": 0.09103193879127502, | |
| "logits/rejected": 0.1026821881532669, | |
| "logps/chosen": -356.58807373046875, | |
| "logps/rejected": -444.40869140625, | |
| "loss": 0.0813, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.08095113933086395, | |
| "rewards/margins": 0.059402190148830414, | |
| "rewards/rejected": -0.14035332202911377, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.167006935924164e-07, | |
| "logits/chosen": 0.07466550171375275, | |
| "logits/rejected": 0.08906082808971405, | |
| "logps/chosen": -419.82733154296875, | |
| "logps/rejected": -447.23419189453125, | |
| "loss": 0.0858, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.09368356317281723, | |
| "rewards/margins": 0.049560584127902985, | |
| "rewards/rejected": -0.14324414730072021, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.809637357629916e-07, | |
| "logits/chosen": 0.09017828106880188, | |
| "logits/rejected": 0.07519755512475967, | |
| "logps/chosen": -386.6206970214844, | |
| "logps/rejected": -444.78900146484375, | |
| "loss": 0.0884, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.09151438623666763, | |
| "rewards/margins": 0.0505540557205677, | |
| "rewards/rejected": -0.14206844568252563, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.461568517961763e-07, | |
| "logits/chosen": 0.022731300443410873, | |
| "logits/rejected": 0.03403092175722122, | |
| "logps/chosen": -406.1392822265625, | |
| "logps/rejected": -505.94366455078125, | |
| "loss": 0.0879, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.11663617938756943, | |
| "rewards/margins": 0.07096937298774719, | |
| "rewards/rejected": -0.18760555982589722, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.122969110829346e-07, | |
| "logits/chosen": 0.03181394562125206, | |
| "logits/rejected": 0.045416783541440964, | |
| "logps/chosen": -403.97100830078125, | |
| "logps/rejected": -524.0074462890625, | |
| "loss": 0.0895, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.11092919111251831, | |
| "rewards/margins": 0.07563652843236923, | |
| "rewards/rejected": -0.18656572699546814, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.794003240718004e-07, | |
| "logits/chosen": 0.09289757162332535, | |
| "logits/rejected": 0.09983552247285843, | |
| "logps/chosen": -467.2046813964844, | |
| "logps/rejected": -487.23760986328125, | |
| "loss": 0.095, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.10705260187387466, | |
| "rewards/margins": 0.050529349595308304, | |
| "rewards/rejected": -0.15758195519447327, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.4748303431543907e-07, | |
| "logits/chosen": 0.07837380468845367, | |
| "logits/rejected": 0.10405907779932022, | |
| "logps/chosen": -479.838134765625, | |
| "logps/rejected": -476.2897033691406, | |
| "loss": 0.0852, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.10937539488077164, | |
| "rewards/margins": 0.04146861657500267, | |
| "rewards/rejected": -0.150844007730484, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.1656051074350113e-07, | |
| "logits/chosen": 0.075148805975914, | |
| "logits/rejected": 0.08904428780078888, | |
| "logps/chosen": -405.589599609375, | |
| "logps/rejected": -448.927734375, | |
| "loss": 0.0901, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.09542171657085419, | |
| "rewards/margins": 0.03852013498544693, | |
| "rewards/rejected": -0.1339418590068817, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.866477401655e-07, | |
| "logits/chosen": 0.07279117405414581, | |
| "logits/rejected": 0.08455020189285278, | |
| "logps/chosen": -385.3067626953125, | |
| "logps/rejected": -440.7181701660156, | |
| "loss": 0.0822, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.09135736525058746, | |
| "rewards/margins": 0.04145896062254906, | |
| "rewards/rejected": -0.13281632959842682, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.577592200073518e-07, | |
| "logits/chosen": 0.05959012359380722, | |
| "logits/rejected": 0.04779018834233284, | |
| "logps/chosen": -411.390625, | |
| "logps/rejected": -462.2435607910156, | |
| "loss": 0.0844, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.09003390371799469, | |
| "rewards/margins": 0.03744681924581528, | |
| "rewards/rejected": -0.12748071551322937, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.2990895128510364e-07, | |
| "logits/chosen": 0.05087254196405411, | |
| "logits/rejected": 0.07081006467342377, | |
| "logps/chosen": -413.0889587402344, | |
| "logps/rejected": -446.05242919921875, | |
| "loss": 0.0881, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.09529262781143188, | |
| "rewards/margins": 0.045550812035799026, | |
| "rewards/rejected": -0.1408434361219406, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.0311043181924237e-07, | |
| "logits/chosen": 0.028466597199440002, | |
| "logits/rejected": 0.06887885928153992, | |
| "logps/chosen": -412.6106872558594, | |
| "logps/rejected": -454.3301696777344, | |
| "loss": 0.0969, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.10608142614364624, | |
| "rewards/margins": 0.02706090547144413, | |
| "rewards/rejected": -0.13314232230186462, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.773766496928898e-07, | |
| "logits/chosen": 0.06292058527469635, | |
| "logits/rejected": 0.08972954005002975, | |
| "logps/chosen": -411.333251953125, | |
| "logps/rejected": -474.076416015625, | |
| "loss": 0.0967, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.0901075154542923, | |
| "rewards/margins": 0.06678055226802826, | |
| "rewards/rejected": -0.15688805282115936, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.527200769570376e-07, | |
| "logits/chosen": 0.05736211687326431, | |
| "logits/rejected": 0.047862760722637177, | |
| "logps/chosen": -428.587890625, | |
| "logps/rejected": -487.86029052734375, | |
| "loss": 0.091, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.11271530389785767, | |
| "rewards/margins": 0.05474567413330078, | |
| "rewards/rejected": -0.16746096312999725, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.291526635858829e-07, | |
| "logits/chosen": 0.033168576657772064, | |
| "logits/rejected": 0.07321670651435852, | |
| "logps/chosen": -406.710205078125, | |
| "logps/rejected": -515.9356079101562, | |
| "loss": 0.0829, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.09568953514099121, | |
| "rewards/margins": 0.058294862508773804, | |
| "rewards/rejected": -0.1539844125509262, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.0668583168519608e-07, | |
| "logits/chosen": 0.047509074211120605, | |
| "logits/rejected": 0.06223832815885544, | |
| "logps/chosen": -388.98193359375, | |
| "logps/rejected": -471.21136474609375, | |
| "loss": 0.0839, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.09332209825515747, | |
| "rewards/margins": 0.04638640210032463, | |
| "rewards/rejected": -0.1397085040807724, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.8533046995651755e-07, | |
| "logits/chosen": 0.06117665767669678, | |
| "logits/rejected": 0.05806068703532219, | |
| "logps/chosen": -398.0728454589844, | |
| "logps/rejected": -476.2325744628906, | |
| "loss": 0.0928, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.10759732872247696, | |
| "rewards/margins": 0.06671231985092163, | |
| "rewards/rejected": -0.1743096262216568, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.6509692841987557e-07, | |
| "logits/chosen": 0.09554895013570786, | |
| "logits/rejected": 0.11183027923107147, | |
| "logps/chosen": -373.590576171875, | |
| "logps/rejected": -459.16046142578125, | |
| "loss": 0.0958, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.09168444573879242, | |
| "rewards/margins": 0.05873497575521469, | |
| "rewards/rejected": -0.1504194289445877, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.4599501339758043e-07, | |
| "logits/chosen": 0.14753518998622894, | |
| "logits/rejected": 0.15233811736106873, | |
| "logps/chosen": -391.5714416503906, | |
| "logps/rejected": -454.29901123046875, | |
| "loss": 0.0838, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.07346367090940475, | |
| "rewards/margins": 0.0692339688539505, | |
| "rewards/rejected": -0.14269763231277466, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.2803398276152236e-07, | |
| "logits/chosen": 0.0984094962477684, | |
| "logits/rejected": 0.158823162317276, | |
| "logps/chosen": -421.304931640625, | |
| "logps/rejected": -486.080810546875, | |
| "loss": 0.0911, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.10145977884531021, | |
| "rewards/margins": 0.05414135009050369, | |
| "rewards/rejected": -0.1556011140346527, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1122254144628308e-07, | |
| "logits/chosen": 0.03182172030210495, | |
| "logits/rejected": 0.06180509179830551, | |
| "logps/chosen": -370.4424743652344, | |
| "logps/rejected": -450.4090881347656, | |
| "loss": 0.0875, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.09365768730640411, | |
| "rewards/margins": 0.041229672729969025, | |
| "rewards/rejected": -0.13488735258579254, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.556883723023047e-08, | |
| "logits/chosen": 0.03198453038930893, | |
| "logits/rejected": 0.09605802595615387, | |
| "logps/chosen": -458.57366943359375, | |
| "logps/rejected": -517.7696533203125, | |
| "loss": 0.0928, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.10384533554315567, | |
| "rewards/margins": 0.06892184168100357, | |
| "rewards/rejected": -0.17276716232299805, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 8.108045678664133e-08, | |
| "logits/chosen": 0.046630989760160446, | |
| "logits/rejected": 0.11219167709350586, | |
| "logps/chosen": -429.394287109375, | |
| "logps/rejected": -471.2891540527344, | |
| "loss": 0.0851, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.10102847963571548, | |
| "rewards/margins": 0.04838350787758827, | |
| "rewards/rejected": -0.14941199123859406, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 6.776442200677136e-08, | |
| "logits/chosen": 0.07663191854953766, | |
| "logits/rejected": 0.08462060987949371, | |
| "logps/chosen": -418.231201171875, | |
| "logps/rejected": -460.2142639160156, | |
| "loss": 0.0978, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.08978335559368134, | |
| "rewards/margins": 0.05147411674261093, | |
| "rewards/rejected": -0.14125746488571167, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.562718659664651e-08, | |
| "logits/chosen": 0.05606203153729439, | |
| "logits/rejected": 0.03971995785832405, | |
| "logps/chosen": -446.9505920410156, | |
| "logps/rejected": -492.7408752441406, | |
| "loss": 0.1003, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.11127449572086334, | |
| "rewards/margins": 0.049640461802482605, | |
| "rewards/rejected": -0.16091494262218475, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 4.467463294923091e-08, | |
| "logits/chosen": 0.07213045656681061, | |
| "logits/rejected": 0.11403784900903702, | |
| "logps/chosen": -408.98272705078125, | |
| "logps/rejected": -505.5140686035156, | |
| "loss": 0.0818, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.10484303534030914, | |
| "rewards/margins": 0.07243579626083374, | |
| "rewards/rejected": -0.17727883160114288, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.491206929348595e-08, | |
| "logits/chosen": 0.09976326674222946, | |
| "logits/rejected": 0.11376956850290298, | |
| "logps/chosen": -462.38946533203125, | |
| "logps/rejected": -540.771484375, | |
| "loss": 0.083, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.11446479707956314, | |
| "rewards/margins": 0.06572078168392181, | |
| "rewards/rejected": -0.18018554151058197, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.634422712170209e-08, | |
| "logits/chosen": 0.06767505407333374, | |
| "logits/rejected": 0.08407372236251831, | |
| "logps/chosen": -445.62646484375, | |
| "logps/rejected": -456.46685791015625, | |
| "loss": 0.0938, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.11092151701450348, | |
| "rewards/margins": 0.028619635850191116, | |
| "rewards/rejected": -0.1395411491394043, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 1.8975258896347383e-08, | |
| "logits/chosen": 0.03723181039094925, | |
| "logits/rejected": 0.07889989018440247, | |
| "logps/chosen": -400.19281005859375, | |
| "logps/rejected": -445.9078063964844, | |
| "loss": 0.1039, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.09558527171611786, | |
| "rewards/margins": 0.049675554037094116, | |
| "rewards/rejected": -0.14526084065437317, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.2808736037550618e-08, | |
| "logits/chosen": 0.022983688861131668, | |
| "logits/rejected": 0.07783511281013489, | |
| "logps/chosen": -435.6316833496094, | |
| "logps/rejected": -445.15118408203125, | |
| "loss": 0.0877, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.104395292699337, | |
| "rewards/margins": 0.03368845209479332, | |
| "rewards/rejected": -0.13808374106884003, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 7.847647192186226e-09, | |
| "logits/chosen": 0.08826269954442978, | |
| "logits/rejected": 0.08017078787088394, | |
| "logps/chosen": -440.181884765625, | |
| "logps/rejected": -492.92510986328125, | |
| "loss": 0.0749, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.09225556254386902, | |
| "rewards/margins": 0.06564488261938095, | |
| "rewards/rejected": -0.15790043771266937, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 4.094396785408461e-09, | |
| "logits/chosen": 0.050535012036561966, | |
| "logits/rejected": 0.09702993929386139, | |
| "logps/chosen": -396.10906982421875, | |
| "logps/rejected": -434.914794921875, | |
| "loss": 0.0964, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.09921000152826309, | |
| "rewards/margins": 0.04018466919660568, | |
| "rewards/rejected": -0.13939465582370758, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 1.5508038553291304e-09, | |
| "logits/chosen": 0.06177855283021927, | |
| "logits/rejected": 0.08348623663187027, | |
| "logps/chosen": -436.14166259765625, | |
| "logps/rejected": -509.82373046875, | |
| "loss": 0.0862, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.09413688629865646, | |
| "rewards/margins": 0.05878259614109993, | |
| "rewards/rejected": -0.15291950106620789, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 2.1810117140974584e-10, | |
| "logits/chosen": 0.024281805381178856, | |
| "logits/rejected": 0.09278794378042221, | |
| "logps/chosen": -491.00299072265625, | |
| "logps/rejected": -472.14349365234375, | |
| "loss": 0.1018, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.11561523377895355, | |
| "rewards/margins": 0.04595177248120308, | |
| "rewards/rejected": -0.16156700253486633, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1586, | |
| "total_flos": 0.0, | |
| "train_loss": 0.09621614837856943, | |
| "train_runtime": 21999.9134, | |
| "train_samples_per_second": 3.461, | |
| "train_steps_per_second": 0.072 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1586, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |