| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9876543209876543, | |
| "eval_steps": 100, | |
| "global_step": 363, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0411522633744856, | |
| "grad_norm": 71.36946521074697, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": -2.7249937057495117, | |
| "logits/rejected": -2.7219715118408203, | |
| "logps/chosen": -289.096435546875, | |
| "logps/rejected": -212.59097290039062, | |
| "loss": 0.6888, | |
| "rewards/accuracies": 0.35624998807907104, | |
| "rewards/chosen": 0.027115171775221825, | |
| "rewards/margins": 0.011037254706025124, | |
| "rewards/rejected": 0.01607791893184185, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0823045267489712, | |
| "grad_norm": 68.75739482144014, | |
| "learning_rate": 1e-06, | |
| "logits/chosen": -2.673173666000366, | |
| "logits/rejected": -2.6852009296417236, | |
| "logps/chosen": -258.5091857910156, | |
| "logps/rejected": -228.7921905517578, | |
| "loss": 0.6519, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": 0.6803622841835022, | |
| "rewards/margins": 0.2561650276184082, | |
| "rewards/rejected": 0.424197256565094, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 43.9449007096878, | |
| "learning_rate": 9.995050530093366e-07, | |
| "logits/chosen": -2.5606446266174316, | |
| "logits/rejected": -2.555354595184326, | |
| "logps/chosen": -258.5283508300781, | |
| "logps/rejected": -217.637939453125, | |
| "loss": 0.5873, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": 1.7179749011993408, | |
| "rewards/margins": 0.7173956036567688, | |
| "rewards/rejected": 1.0005793571472168, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.1646090534979424, | |
| "grad_norm": 49.90030149803026, | |
| "learning_rate": 9.980211919274406e-07, | |
| "logits/chosen": -2.334833860397339, | |
| "logits/rejected": -2.3182854652404785, | |
| "logps/chosen": -234.5125732421875, | |
| "logps/rejected": -194.8851318359375, | |
| "loss": 0.6125, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": 1.7243343591690063, | |
| "rewards/margins": 0.8316472172737122, | |
| "rewards/rejected": 0.8926870226860046, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.205761316872428, | |
| "grad_norm": 43.554349506398026, | |
| "learning_rate": 9.955513544846204e-07, | |
| "logits/chosen": -2.12056303024292, | |
| "logits/rejected": -2.095937728881836, | |
| "logps/chosen": -284.00323486328125, | |
| "logps/rejected": -210.3358154296875, | |
| "loss": 0.5749, | |
| "rewards/accuracies": 0.8125, | |
| "rewards/chosen": 2.1191883087158203, | |
| "rewards/margins": 1.4548943042755127, | |
| "rewards/rejected": 0.6642940044403076, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.24691358024691357, | |
| "grad_norm": 45.05915140113881, | |
| "learning_rate": 9.921004304353147e-07, | |
| "logits/chosen": -2.04213547706604, | |
| "logits/rejected": -2.0172839164733887, | |
| "logps/chosen": -232.2016143798828, | |
| "logps/rejected": -217.5736846923828, | |
| "loss": 0.5989, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": 1.9440408945083618, | |
| "rewards/margins": 1.5185799598693848, | |
| "rewards/rejected": 0.4254608750343323, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.2880658436213992, | |
| "grad_norm": 40.00728614202134, | |
| "learning_rate": 9.876752518774164e-07, | |
| "logits/chosen": -2.0041847229003906, | |
| "logits/rejected": -1.9888496398925781, | |
| "logps/chosen": -255.5012969970703, | |
| "logps/rejected": -238.2528839111328, | |
| "loss": 0.6076, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": 1.3912312984466553, | |
| "rewards/margins": 1.0289623737335205, | |
| "rewards/rejected": 0.36226886510849, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.3292181069958848, | |
| "grad_norm": 42.181862044805364, | |
| "learning_rate": 9.822845797261675e-07, | |
| "logits/chosen": -2.024127244949341, | |
| "logits/rejected": -2.020592451095581, | |
| "logps/chosen": -249.13394165039062, | |
| "logps/rejected": -199.90975952148438, | |
| "loss": 0.5846, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": 1.3734517097473145, | |
| "rewards/margins": 0.8223851919174194, | |
| "rewards/rejected": 0.5510665774345398, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 41.15847921708812, | |
| "learning_rate": 9.759390863694029e-07, | |
| "logits/chosen": -2.0532474517822266, | |
| "logits/rejected": -1.9978084564208984, | |
| "logps/chosen": -256.01446533203125, | |
| "logps/rejected": -206.8267059326172, | |
| "loss": 0.5481, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 1.589166283607483, | |
| "rewards/margins": 1.2907274961471558, | |
| "rewards/rejected": 0.2984387278556824, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.411522633744856, | |
| "grad_norm": 42.63680924826028, | |
| "learning_rate": 9.68651334538488e-07, | |
| "logits/chosen": -2.034133195877075, | |
| "logits/rejected": -2.0025076866149902, | |
| "logps/chosen": -259.46942138671875, | |
| "logps/rejected": -229.2208251953125, | |
| "loss": 0.5652, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": 1.4298592805862427, | |
| "rewards/margins": 1.076907992362976, | |
| "rewards/rejected": 0.35295119881629944, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.45267489711934156, | |
| "grad_norm": 36.48668334468458, | |
| "learning_rate": 9.604357524367722e-07, | |
| "logits/chosen": -2.0932247638702393, | |
| "logits/rejected": -2.0437166690826416, | |
| "logps/chosen": -281.03289794921875, | |
| "logps/rejected": -227.46109008789062, | |
| "loss": 0.5437, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 1.40286123752594, | |
| "rewards/margins": 1.028884768486023, | |
| "rewards/rejected": 0.3739764094352722, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 35.36330599361053, | |
| "learning_rate": 9.513086051748067e-07, | |
| "logits/chosen": -2.1159732341766357, | |
| "logits/rejected": -2.078249931335449, | |
| "logps/chosen": -265.8070373535156, | |
| "logps/rejected": -214.79428100585938, | |
| "loss": 0.5166, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": 1.7003364562988281, | |
| "rewards/margins": 1.7438255548477173, | |
| "rewards/rejected": -0.0434890016913414, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.5349794238683128, | |
| "grad_norm": 31.36432376366485, | |
| "learning_rate": 9.412879625688742e-07, | |
| "logits/chosen": -2.183833599090576, | |
| "logits/rejected": -2.1385440826416016, | |
| "logps/chosen": -270.89263916015625, | |
| "logps/rejected": -203.67922973632812, | |
| "loss": 0.4848, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": 1.6083428859710693, | |
| "rewards/margins": 1.5229980945587158, | |
| "rewards/rejected": 0.08534489572048187, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.5761316872427984, | |
| "grad_norm": 35.68216693219843, | |
| "learning_rate": 9.303936633665839e-07, | |
| "logits/chosen": -2.3082363605499268, | |
| "logits/rejected": -2.2824604511260986, | |
| "logps/chosen": -255.9834747314453, | |
| "logps/rejected": -194.7764892578125, | |
| "loss": 0.5289, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": 1.1171067953109741, | |
| "rewards/margins": 1.4306641817092896, | |
| "rewards/rejected": -0.3135572373867035, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.6172839506172839, | |
| "grad_norm": 38.51565362073314, | |
| "learning_rate": 9.186472759703578e-07, | |
| "logits/chosen": -2.3410449028015137, | |
| "logits/rejected": -2.3213045597076416, | |
| "logps/chosen": -275.8757019042969, | |
| "logps/rejected": -213.70693969726562, | |
| "loss": 0.5387, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": 0.7660292387008667, | |
| "rewards/margins": 1.4290556907653809, | |
| "rewards/rejected": -0.6630264520645142, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.6584362139917695, | |
| "grad_norm": 30.908945588893605, | |
| "learning_rate": 9.060720557365682e-07, | |
| "logits/chosen": -2.3798623085021973, | |
| "logits/rejected": -2.378147602081299, | |
| "logps/chosen": -277.94622802734375, | |
| "logps/rejected": -228.6498565673828, | |
| "loss": 0.524, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": 0.9844567179679871, | |
| "rewards/margins": 1.5679962635040283, | |
| "rewards/rejected": -0.5835394859313965, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.6995884773662552, | |
| "grad_norm": 36.12667290276971, | |
| "learning_rate": 8.926928989348611e-07, | |
| "logits/chosen": -2.438974618911743, | |
| "logits/rejected": -2.4293782711029053, | |
| "logps/chosen": -264.4499816894531, | |
| "logps/rejected": -233.60958862304688, | |
| "loss": 0.5124, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": 0.7334972023963928, | |
| "rewards/margins": 1.825126051902771, | |
| "rewards/rejected": -1.091629147529602, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 31.43710612772888, | |
| "learning_rate": 8.785362934588233e-07, | |
| "logits/chosen": -2.4581363201141357, | |
| "logits/rejected": -2.4250473976135254, | |
| "logps/chosen": -272.42498779296875, | |
| "logps/rejected": -206.20614624023438, | |
| "loss": 0.5073, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": 1.6099742650985718, | |
| "rewards/margins": 1.7208999395370483, | |
| "rewards/rejected": -0.11092579364776611, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.7818930041152263, | |
| "grad_norm": 35.922757319188804, | |
| "learning_rate": 8.636302663855681e-07, | |
| "logits/chosen": -2.368760585784912, | |
| "logits/rejected": -2.3825132846832275, | |
| "logps/chosen": -247.90396118164062, | |
| "logps/rejected": -212.88232421875, | |
| "loss": 0.4971, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": 1.2823846340179443, | |
| "rewards/margins": 1.8866965770721436, | |
| "rewards/rejected": -0.6043121814727783, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.823045267489712, | |
| "grad_norm": 27.664598721354345, | |
| "learning_rate": 8.480043284880664e-07, | |
| "logits/chosen": -2.346686601638794, | |
| "logits/rejected": -2.317147970199585, | |
| "logps/chosen": -269.21417236328125, | |
| "logps/rejected": -233.6097412109375, | |
| "loss": 0.4674, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": 0.5551630258560181, | |
| "rewards/margins": 2.1603965759277344, | |
| "rewards/rejected": -1.6052335500717163, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.823045267489712, | |
| "eval_logits/chosen": -2.2803401947021484, | |
| "eval_logits/rejected": -2.256579875946045, | |
| "eval_logps/chosen": -257.0998229980469, | |
| "eval_logps/rejected": -231.74539184570312, | |
| "eval_loss": 0.4985389709472656, | |
| "eval_rewards/accuracies": 0.7939814925193787, | |
| "eval_rewards/chosen": 0.17793893814086914, | |
| "eval_rewards/margins": 1.7997103929519653, | |
| "eval_rewards/rejected": -1.6217713356018066, | |
| "eval_runtime": 230.2785, | |
| "eval_samples_per_second": 15.008, | |
| "eval_steps_per_second": 0.234, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.8641975308641975, | |
| "grad_norm": 33.43388986335041, | |
| "learning_rate": 8.316894158100727e-07, | |
| "logits/chosen": -2.238370895385742, | |
| "logits/rejected": -2.205950975418091, | |
| "logps/chosen": -270.1739807128906, | |
| "logps/rejected": -237.7426300048828, | |
| "loss": 0.5036, | |
| "rewards/accuracies": 0.824999988079071, | |
| "rewards/chosen": 0.2168927639722824, | |
| "rewards/margins": 2.0453083515167236, | |
| "rewards/rejected": -1.8284155130386353, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.9053497942386831, | |
| "grad_norm": 41.06626958250484, | |
| "learning_rate": 8.147178284193184e-07, | |
| "logits/chosen": -1.9968522787094116, | |
| "logits/rejected": -1.9477859735488892, | |
| "logps/chosen": -271.5672912597656, | |
| "logps/rejected": -244.5254364013672, | |
| "loss": 0.523, | |
| "rewards/accuracies": 0.8374999761581421, | |
| "rewards/chosen": -0.18483969569206238, | |
| "rewards/margins": 2.090688467025757, | |
| "rewards/rejected": -2.2755284309387207, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.9465020576131687, | |
| "grad_norm": 33.45568853055463, | |
| "learning_rate": 7.971231664602271e-07, | |
| "logits/chosen": -1.8657859563827515, | |
| "logits/rejected": -1.7577025890350342, | |
| "logps/chosen": -255.1681365966797, | |
| "logps/rejected": -235.93856811523438, | |
| "loss": 0.4781, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": 0.08642071485519409, | |
| "rewards/margins": 2.032249689102173, | |
| "rewards/rejected": -1.9458287954330444, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 31.32834367464404, | |
| "learning_rate": 7.789402636327525e-07, | |
| "logits/chosen": -1.7241904735565186, | |
| "logits/rejected": -1.6637340784072876, | |
| "logps/chosen": -269.67364501953125, | |
| "logps/rejected": -239.79965209960938, | |
| "loss": 0.4614, | |
| "rewards/accuracies": 0.8187500238418579, | |
| "rewards/chosen": 0.0743473693728447, | |
| "rewards/margins": 2.101712942123413, | |
| "rewards/rejected": -2.0273656845092773, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.02880658436214, | |
| "grad_norm": 18.313357022047114, | |
| "learning_rate": 7.602051182290381e-07, | |
| "logits/chosen": -1.5669622421264648, | |
| "logits/rejected": -1.4961906671524048, | |
| "logps/chosen": -270.39056396484375, | |
| "logps/rejected": -223.95706176757812, | |
| "loss": 0.321, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": 0.7599193453788757, | |
| "rewards/margins": 2.714322805404663, | |
| "rewards/rejected": -1.9544035196304321, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.0699588477366255, | |
| "grad_norm": 18.77066721006591, | |
| "learning_rate": 7.409548218644331e-07, | |
| "logits/chosen": -1.4371721744537354, | |
| "logits/rejected": -1.3102617263793945, | |
| "logps/chosen": -257.923095703125, | |
| "logps/rejected": -222.04959106445312, | |
| "loss": 0.1777, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.8628284931182861, | |
| "rewards/margins": 3.623333692550659, | |
| "rewards/rejected": -1.7605053186416626, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1111111111111112, | |
| "grad_norm": 14.482571733068447, | |
| "learning_rate": 7.212274860439576e-07, | |
| "logits/chosen": -1.4088728427886963, | |
| "logits/rejected": -1.3359241485595703, | |
| "logps/chosen": -252.8369140625, | |
| "logps/rejected": -247.0041046142578, | |
| "loss": 0.2246, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.8369052410125732, | |
| "rewards/margins": 3.677825450897217, | |
| "rewards/rejected": -1.840920090675354, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.1522633744855968, | |
| "grad_norm": 19.475514209975124, | |
| "learning_rate": 7.010621667096041e-07, | |
| "logits/chosen": -1.5916813611984253, | |
| "logits/rejected": -1.479448676109314, | |
| "logps/chosen": -254.99136352539062, | |
| "logps/rejected": -218.8384246826172, | |
| "loss": 0.2218, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 1.6720364093780518, | |
| "rewards/margins": 3.2526676654815674, | |
| "rewards/rejected": -1.5806310176849365, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.1934156378600824, | |
| "grad_norm": 24.87312122824749, | |
| "learning_rate": 6.804987869178539e-07, | |
| "logits/chosen": -1.7563555240631104, | |
| "logits/rejected": -1.6887686252593994, | |
| "logps/chosen": -241.65676879882812, | |
| "logps/rejected": -225.277099609375, | |
| "loss": 0.2373, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": 1.7571462392807007, | |
| "rewards/margins": 3.4047298431396484, | |
| "rewards/rejected": -1.6475833654403687, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.2345679012345678, | |
| "grad_norm": 22.485449779074028, | |
| "learning_rate": 6.5957805780049e-07, | |
| "logits/chosen": -1.889991044998169, | |
| "logits/rejected": -1.8203752040863037, | |
| "logps/chosen": -250.935302734375, | |
| "logps/rejected": -223.52401733398438, | |
| "loss": 0.2196, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 2.23984432220459, | |
| "rewards/margins": 3.5551295280456543, | |
| "rewards/rejected": -1.3152849674224854, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.2757201646090535, | |
| "grad_norm": 20.0119744226792, | |
| "learning_rate": 6.383413979651893e-07, | |
| "logits/chosen": -1.9477765560150146, | |
| "logits/rejected": -1.8932664394378662, | |
| "logps/chosen": -242.27685546875, | |
| "logps/rejected": -231.18991088867188, | |
| "loss": 0.2229, | |
| "rewards/accuracies": 0.918749988079071, | |
| "rewards/chosen": 1.800172209739685, | |
| "rewards/margins": 3.45011568069458, | |
| "rewards/rejected": -1.6499433517456055, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.316872427983539, | |
| "grad_norm": 20.63931604768156, | |
| "learning_rate": 6.168308514954602e-07, | |
| "logits/chosen": -1.973009705543518, | |
| "logits/rejected": -1.8899316787719727, | |
| "logps/chosen": -261.8257141113281, | |
| "logps/rejected": -258.97515869140625, | |
| "loss": 0.2121, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": 1.9451110363006592, | |
| "rewards/margins": 4.489598274230957, | |
| "rewards/rejected": -2.544487237930298, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.3580246913580247, | |
| "grad_norm": 23.779662167366467, | |
| "learning_rate": 5.950890047122741e-07, | |
| "logits/chosen": -1.9724878072738647, | |
| "logits/rejected": -1.9425151348114014, | |
| "logps/chosen": -260.43084716796875, | |
| "logps/rejected": -236.8948211669922, | |
| "loss": 0.2464, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 1.4944422245025635, | |
| "rewards/margins": 3.678725481033325, | |
| "rewards/rejected": -2.18428373336792, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 1.3991769547325104, | |
| "grad_norm": 15.224094688709425, | |
| "learning_rate": 5.731589018621776e-07, | |
| "logits/chosen": -1.9535115957260132, | |
| "logits/rejected": -1.8948615789413452, | |
| "logps/chosen": -252.6552276611328, | |
| "logps/rejected": -226.4263916015625, | |
| "loss": 0.2351, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": 1.834676742553711, | |
| "rewards/margins": 4.032426357269287, | |
| "rewards/rejected": -2.1977500915527344, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.4403292181069958, | |
| "grad_norm": 20.636053561561848, | |
| "learning_rate": 5.510839598988136e-07, | |
| "logits/chosen": -1.8348503112792969, | |
| "logits/rejected": -1.7934105396270752, | |
| "logps/chosen": -255.14895629882812, | |
| "logps/rejected": -232.3575897216797, | |
| "loss": 0.2069, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 2.0120339393615723, | |
| "rewards/margins": 3.713160276412964, | |
| "rewards/rejected": -1.7011263370513916, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 1.4814814814814814, | |
| "grad_norm": 21.755357371160876, | |
| "learning_rate": 5.289078825265572e-07, | |
| "logits/chosen": -1.7341606616973877, | |
| "logits/rejected": -1.6741468906402588, | |
| "logps/chosen": -237.35433959960938, | |
| "logps/rejected": -228.7030487060547, | |
| "loss": 0.234, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": 1.63128662109375, | |
| "rewards/margins": 3.659700393676758, | |
| "rewards/rejected": -2.028413772583008, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.522633744855967, | |
| "grad_norm": 23.810123453795516, | |
| "learning_rate": 5.066745736764489e-07, | |
| "logits/chosen": -1.635679841041565, | |
| "logits/rejected": -1.5873550176620483, | |
| "logps/chosen": -248.98135375976562, | |
| "logps/rejected": -240.08987426757812, | |
| "loss": 0.2576, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 1.4315288066864014, | |
| "rewards/margins": 3.4555141925811768, | |
| "rewards/rejected": -2.0239853858947754, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 1.5637860082304527, | |
| "grad_norm": 22.3759752093868, | |
| "learning_rate": 4.844280505857202e-07, | |
| "logits/chosen": -1.5894463062286377, | |
| "logits/rejected": -1.5013604164123535, | |
| "logps/chosen": -239.4411163330078, | |
| "logps/rejected": -219.7681121826172, | |
| "loss": 0.2732, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 1.9483649730682373, | |
| "rewards/margins": 3.714170455932617, | |
| "rewards/rejected": -1.7658058404922485, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.6049382716049383, | |
| "grad_norm": 21.666055935350588, | |
| "learning_rate": 4.6221235665299684e-07, | |
| "logits/chosen": -1.6968196630477905, | |
| "logits/rejected": -1.6124290227890015, | |
| "logps/chosen": -246.6077117919922, | |
| "logps/rejected": -233.7628631591797, | |
| "loss": 0.2689, | |
| "rewards/accuracies": 0.90625, | |
| "rewards/chosen": 2.1885757446289062, | |
| "rewards/margins": 3.544438600540161, | |
| "rewards/rejected": -1.3558627367019653, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.646090534979424, | |
| "grad_norm": 23.567423393969673, | |
| "learning_rate": 4.400714742417091e-07, | |
| "logits/chosen": -1.7539150714874268, | |
| "logits/rejected": -1.6715869903564453, | |
| "logps/chosen": -289.3243713378906, | |
| "logps/rejected": -238.78271484375, | |
| "loss": 0.2463, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 2.3782408237457275, | |
| "rewards/margins": 3.8166255950927734, | |
| "rewards/rejected": -1.438385248184204, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.646090534979424, | |
| "eval_logits/chosen": -1.7062827348709106, | |
| "eval_logits/rejected": -1.629170298576355, | |
| "eval_logps/chosen": -247.42041015625, | |
| "eval_logps/rejected": -227.5958709716797, | |
| "eval_loss": 0.5190241932868958, | |
| "eval_rewards/accuracies": 0.7962962985038757, | |
| "eval_rewards/chosen": 1.1458828449249268, | |
| "eval_rewards/margins": 2.3527021408081055, | |
| "eval_rewards/rejected": -1.2068192958831787, | |
| "eval_runtime": 228.0783, | |
| "eval_samples_per_second": 15.153, | |
| "eval_steps_per_second": 0.237, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.6872427983539096, | |
| "grad_norm": 20.05042831109418, | |
| "learning_rate": 4.180492376043371e-07, | |
| "logits/chosen": -1.7294807434082031, | |
| "logits/rejected": -1.6129295825958252, | |
| "logps/chosen": -239.91696166992188, | |
| "logps/rejected": -241.2155303955078, | |
| "loss": 0.2475, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.699279546737671, | |
| "rewards/margins": 3.706660747528076, | |
| "rewards/rejected": -2.007380962371826, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.7283950617283952, | |
| "grad_norm": 17.373566601078217, | |
| "learning_rate": 3.961892460998862e-07, | |
| "logits/chosen": -1.7376630306243896, | |
| "logits/rejected": -1.672767996788025, | |
| "logps/chosen": -259.5295104980469, | |
| "logps/rejected": -219.8362274169922, | |
| "loss": 0.2275, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 1.8958297967910767, | |
| "rewards/margins": 3.636307954788208, | |
| "rewards/rejected": -1.7404781579971313, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.7695473251028808, | |
| "grad_norm": 20.32259020536467, | |
| "learning_rate": 3.7453477787640077e-07, | |
| "logits/chosen": -1.6703641414642334, | |
| "logits/rejected": -1.6055065393447876, | |
| "logps/chosen": -259.04559326171875, | |
| "logps/rejected": -238.02713012695312, | |
| "loss": 0.2558, | |
| "rewards/accuracies": 0.893750011920929, | |
| "rewards/chosen": 1.8848392963409424, | |
| "rewards/margins": 3.780524492263794, | |
| "rewards/rejected": -1.8956845998764038, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.8106995884773662, | |
| "grad_norm": 20.716775450731596, | |
| "learning_rate": 3.531287041894075e-07, | |
| "logits/chosen": -1.636228322982788, | |
| "logits/rejected": -1.5927408933639526, | |
| "logps/chosen": -259.4163513183594, | |
| "logps/rejected": -262.77691650390625, | |
| "loss": 0.2641, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": 1.8492801189422607, | |
| "rewards/margins": 3.8553290367126465, | |
| "rewards/rejected": -2.0060486793518066, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.8518518518518519, | |
| "grad_norm": 20.26085395927115, | |
| "learning_rate": 3.320134045259192e-07, | |
| "logits/chosen": -1.6199842691421509, | |
| "logits/rejected": -1.5809019804000854, | |
| "logps/chosen": -261.5071716308594, | |
| "logps/rejected": -244.0452117919922, | |
| "loss": 0.2836, | |
| "rewards/accuracies": 0.887499988079071, | |
| "rewards/chosen": 1.7676365375518799, | |
| "rewards/margins": 3.8491673469543457, | |
| "rewards/rejected": -2.081530809402466, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.8930041152263375, | |
| "grad_norm": 19.9900109721012, | |
| "learning_rate": 3.112306827020377e-07, | |
| "logits/chosen": -1.6224733591079712, | |
| "logits/rejected": -1.5683706998825073, | |
| "logps/chosen": -246.66726684570312, | |
| "logps/rejected": -252.150634765625, | |
| "loss": 0.2967, | |
| "rewards/accuracies": 0.8999999761581421, | |
| "rewards/chosen": 1.2352790832519531, | |
| "rewards/margins": 3.3191657066345215, | |
| "rewards/rejected": -2.0838871002197266, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.934156378600823, | |
| "grad_norm": 20.679234729146177, | |
| "learning_rate": 2.90821684100261e-07, | |
| "logits/chosen": -1.665122628211975, | |
| "logits/rejected": -1.585533857345581, | |
| "logps/chosen": -258.1650390625, | |
| "logps/rejected": -238.0010223388672, | |
| "loss": 0.2521, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": 1.45806884765625, | |
| "rewards/margins": 3.821526288986206, | |
| "rewards/rejected": -2.363457202911377, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.9753086419753085, | |
| "grad_norm": 17.897922449348748, | |
| "learning_rate": 2.708268142103509e-07, | |
| "logits/chosen": -1.6568527221679688, | |
| "logits/rejected": -1.594029426574707, | |
| "logps/chosen": -249.9292449951172, | |
| "logps/rejected": -217.1236114501953, | |
| "loss": 0.2458, | |
| "rewards/accuracies": 0.9125000238418579, | |
| "rewards/chosen": 1.140490174293518, | |
| "rewards/margins": 3.4049384593963623, | |
| "rewards/rejected": -2.264448404312134, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.016460905349794, | |
| "grad_norm": 15.579483343495324, | |
| "learning_rate": 2.5128565863503e-07, | |
| "logits/chosen": -1.7464730739593506, | |
| "logits/rejected": -1.64523446559906, | |
| "logps/chosen": -269.5633544921875, | |
| "logps/rejected": -218.4349365234375, | |
| "loss": 0.1875, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 1.345157504081726, | |
| "rewards/margins": 3.8898367881774902, | |
| "rewards/rejected": -2.5446791648864746, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 2.05761316872428, | |
| "grad_norm": 15.642770624996952, | |
| "learning_rate": 2.3223690471888286e-07, | |
| "logits/chosen": -1.7972164154052734, | |
| "logits/rejected": -1.6923631429672241, | |
| "logps/chosen": -276.4811706542969, | |
| "logps/rejected": -239.2648468017578, | |
| "loss": 0.1218, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 1.7780349254608154, | |
| "rewards/margins": 4.118841171264648, | |
| "rewards/rejected": -2.3408069610595703, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0987654320987654, | |
| "grad_norm": 13.364305072324674, | |
| "learning_rate": 2.1371826495561613e-07, | |
| "logits/chosen": -1.8449236154556274, | |
| "logits/rejected": -1.7506535053253174, | |
| "logps/chosen": -255.83792114257812, | |
| "logps/rejected": -221.6796875, | |
| "loss": 0.146, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": 1.8460966348648071, | |
| "rewards/margins": 3.9246277809143066, | |
| "rewards/rejected": -2.078531265258789, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 2.139917695473251, | |
| "grad_norm": 13.06395689210594, | |
| "learning_rate": 1.9576640232531784e-07, | |
| "logits/chosen": -1.8692007064819336, | |
| "logits/rejected": -1.8045275211334229, | |
| "logps/chosen": -248.9095916748047, | |
| "logps/rejected": -250.84481811523438, | |
| "loss": 0.1171, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 1.9673175811767578, | |
| "rewards/margins": 4.323936462402344, | |
| "rewards/rejected": -2.356618642807007, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.1810699588477367, | |
| "grad_norm": 15.133332987736472, | |
| "learning_rate": 1.784168577095307e-07, | |
| "logits/chosen": -1.9296722412109375, | |
| "logits/rejected": -1.8828375339508057, | |
| "logps/chosen": -250.7962646484375, | |
| "logps/rejected": -228.93923950195312, | |
| "loss": 0.1322, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 2.0834712982177734, | |
| "rewards/margins": 3.928879499435425, | |
| "rewards/rejected": -1.8454080820083618, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 2.2222222222222223, | |
| "grad_norm": 13.466085492542144, | |
| "learning_rate": 1.6170397952784248e-07, | |
| "logits/chosen": -1.9489628076553345, | |
| "logits/rejected": -1.8797670602798462, | |
| "logps/chosen": -270.56427001953125, | |
| "logps/rejected": -242.9454803466797, | |
| "loss": 0.1229, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 2.427950143814087, | |
| "rewards/margins": 4.682461261749268, | |
| "rewards/rejected": -2.2545108795166016, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.263374485596708, | |
| "grad_norm": 14.794346267314218, | |
| "learning_rate": 1.4566085573529874e-07, | |
| "logits/chosen": -1.9156001806259155, | |
| "logits/rejected": -1.8757755756378174, | |
| "logps/chosen": -258.8504333496094, | |
| "logps/rejected": -229.5829315185547, | |
| "loss": 0.1305, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": 2.124898910522461, | |
| "rewards/margins": 4.520539283752441, | |
| "rewards/rejected": -2.3956406116485596, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 2.3045267489711936, | |
| "grad_norm": 14.6085524255932, | |
| "learning_rate": 1.3031924831526737e-07, | |
| "logits/chosen": -1.918760895729065, | |
| "logits/rejected": -1.8703607320785522, | |
| "logps/chosen": -261.5938415527344, | |
| "logps/rejected": -230.3494415283203, | |
| "loss": 0.1162, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": 1.9034366607666016, | |
| "rewards/margins": 4.663661003112793, | |
| "rewards/rejected": -2.7602241039276123, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.3456790123456788, | |
| "grad_norm": 16.362862237175147, | |
| "learning_rate": 1.1570953039744591e-07, | |
| "logits/chosen": -1.9305750131607056, | |
| "logits/rejected": -1.8696216344833374, | |
| "logps/chosen": -266.16680908203125, | |
| "logps/rejected": -258.2370910644531, | |
| "loss": 0.1186, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": 2.204184055328369, | |
| "rewards/margins": 4.997335433959961, | |
| "rewards/rejected": -2.7931509017944336, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 2.386831275720165, | |
| "grad_norm": 13.275572612341923, | |
| "learning_rate": 1.0186062612550616e-07, | |
| "logits/chosen": -1.9214690923690796, | |
| "logits/rejected": -1.8716766834259033, | |
| "logps/chosen": -252.57180786132812, | |
| "logps/rejected": -259.24224853515625, | |
| "loss": 0.12, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 1.912581205368042, | |
| "rewards/margins": 4.5273051261901855, | |
| "rewards/rejected": -2.6147236824035645, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.42798353909465, | |
| "grad_norm": 14.003480945619684, | |
| "learning_rate": 8.879995339342167e-08, | |
| "logits/chosen": -1.914181113243103, | |
| "logits/rejected": -1.8485758304595947, | |
| "logps/chosen": -248.25320434570312, | |
| "logps/rejected": -228.18118286132812, | |
| "loss": 0.1167, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.6209943294525146, | |
| "rewards/margins": 4.539933204650879, | |
| "rewards/rejected": -2.918938398361206, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 2.4691358024691357, | |
| "grad_norm": 13.39746651643324, | |
| "learning_rate": 7.655336956385155e-08, | |
| "logits/chosen": -1.936248540878296, | |
| "logits/rejected": -1.8758357763290405, | |
| "logps/chosen": -251.0574951171875, | |
| "logps/rejected": -252.95425415039062, | |
| "loss": 0.1311, | |
| "rewards/accuracies": 0.981249988079071, | |
| "rewards/chosen": 1.768341302871704, | |
| "rewards/margins": 4.591066360473633, | |
| "rewards/rejected": -2.8227250576019287, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.4691358024691357, | |
| "eval_logits/chosen": -1.9243203401565552, | |
| "eval_logits/rejected": -1.8631280660629272, | |
| "eval_logps/chosen": -251.93479919433594, | |
| "eval_logps/rejected": -234.54112243652344, | |
| "eval_loss": 0.5211819410324097, | |
| "eval_rewards/accuracies": 0.8194444179534912, | |
| "eval_rewards/chosen": 0.6944435238838196, | |
| "eval_rewards/margins": 2.59578800201416, | |
| "eval_rewards/rejected": -1.9013442993164062, | |
| "eval_runtime": 228.1654, | |
| "eval_samples_per_second": 15.147, | |
| "eval_steps_per_second": 0.237, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.5102880658436213, | |
| "grad_norm": 15.074747119995138, | |
| "learning_rate": 6.514512027604508e-08, | |
| "logits/chosen": -1.9279800653457642, | |
| "logits/rejected": -1.8792842626571655, | |
| "logps/chosen": -232.16232299804688, | |
| "logps/rejected": -224.8663330078125, | |
| "loss": 0.1173, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 1.4570283889770508, | |
| "rewards/margins": 4.1241984367370605, | |
| "rewards/rejected": -2.667170286178589, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 2.551440329218107, | |
| "grad_norm": 15.943407922179238, | |
| "learning_rate": 5.459779144461712e-08, | |
| "logits/chosen": -1.967230200767517, | |
| "logits/rejected": -1.8994722366333008, | |
| "logps/chosen": -251.5553436279297, | |
| "logps/rejected": -234.64218139648438, | |
| "loss": 0.132, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 1.8404948711395264, | |
| "rewards/margins": 4.55427885055542, | |
| "rewards/rejected": -2.7137837409973145, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.5925925925925926, | |
| "grad_norm": 15.791999145358414, | |
| "learning_rate": 4.49322645442266e-08, | |
| "logits/chosen": -1.9726388454437256, | |
| "logits/rejected": -1.9029220342636108, | |
| "logps/chosen": -226.0243377685547, | |
| "logps/rejected": -245.57943725585938, | |
| "loss": 0.1327, | |
| "rewards/accuracies": 0.9375, | |
| "rewards/chosen": 1.4504220485687256, | |
| "rewards/margins": 4.483643531799316, | |
| "rewards/rejected": -3.033221483230591, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 2.633744855967078, | |
| "grad_norm": 14.527344025713838, | |
| "learning_rate": 3.616767526868353e-08, | |
| "logits/chosen": -1.9656314849853516, | |
| "logits/rejected": -1.898186445236206, | |
| "logps/chosen": -268.8167419433594, | |
| "logps/rejected": -251.64340209960938, | |
| "loss": 0.1062, | |
| "rewards/accuracies": 0.9750000238418579, | |
| "rewards/chosen": 2.3940348625183105, | |
| "rewards/margins": 5.15994930267334, | |
| "rewards/rejected": -2.7659144401550293, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.674897119341564, | |
| "grad_norm": 14.467899944932638, | |
| "learning_rate": 2.8321375646333023e-08, | |
| "logits/chosen": -1.984684944152832, | |
| "logits/rejected": -1.905601143836975, | |
| "logps/chosen": -226.098876953125, | |
| "logps/rejected": -269.22723388671875, | |
| "loss": 0.1209, | |
| "rewards/accuracies": 0.9624999761581421, | |
| "rewards/chosen": 1.9057655334472656, | |
| "rewards/margins": 4.68411922454834, | |
| "rewards/rejected": -2.778353691101074, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 2.7160493827160495, | |
| "grad_norm": 14.107526535593529, | |
| "learning_rate": 2.1408899686718996e-08, | |
| "logits/chosen": -1.996860146522522, | |
| "logits/rejected": -1.8913567066192627, | |
| "logps/chosen": -248.2650909423828, | |
| "logps/rejected": -243.4825439453125, | |
| "loss": 0.1195, | |
| "rewards/accuracies": 0.956250011920929, | |
| "rewards/chosen": 1.7427318096160889, | |
| "rewards/margins": 4.723761558532715, | |
| "rewards/rejected": -2.981029987335205, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.757201646090535, | |
| "grad_norm": 16.668582895840217, | |
| "learning_rate": 1.5443932626538314e-08, | |
| "logits/chosen": -1.9676933288574219, | |
| "logits/rejected": -1.910146713256836, | |
| "logps/chosen": -238.7953338623047, | |
| "logps/rejected": -224.4933319091797, | |
| "loss": 0.15, | |
| "rewards/accuracies": 0.9312499761581421, | |
| "rewards/chosen": 1.859580636024475, | |
| "rewards/margins": 4.216904640197754, | |
| "rewards/rejected": -2.3573238849639893, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 2.7983539094650207, | |
| "grad_norm": 13.029805689567587, | |
| "learning_rate": 1.0438283835774387e-08, | |
| "logits/chosen": -1.9859317541122437, | |
| "logits/rejected": -1.8881919384002686, | |
| "logps/chosen": -242.4602508544922, | |
| "logps/rejected": -228.0737762451172, | |
| "loss": 0.1257, | |
| "rewards/accuracies": 0.925000011920929, | |
| "rewards/chosen": 1.7638639211654663, | |
| "rewards/margins": 4.470877170562744, | |
| "rewards/rejected": -2.7070131301879883, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.8395061728395063, | |
| "grad_norm": 13.50245071791209, | |
| "learning_rate": 6.401863437648481e-09, | |
| "logits/chosen": -1.9783008098602295, | |
| "logits/rejected": -1.8936630487442017, | |
| "logps/chosen": -262.051025390625, | |
| "logps/rejected": -244.21853637695312, | |
| "loss": 0.1265, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": 1.8696222305297852, | |
| "rewards/margins": 4.679049491882324, | |
| "rewards/rejected": -2.809427261352539, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 2.8806584362139915, | |
| "grad_norm": 19.77488068943749, | |
| "learning_rate": 3.3426626886769448e-09, | |
| "logits/chosen": -1.9724162817001343, | |
| "logits/rejected": -1.9013561010360718, | |
| "logps/chosen": -265.6155700683594, | |
| "logps/rejected": -258.1453552246094, | |
| "loss": 0.1582, | |
| "rewards/accuracies": 0.949999988079071, | |
| "rewards/chosen": 2.233098030090332, | |
| "rewards/margins": 4.871306419372559, | |
| "rewards/rejected": -2.6382088661193848, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.9218106995884776, | |
| "grad_norm": 16.657133866108477, | |
| "learning_rate": 1.2667381576779712e-09, | |
| "logits/chosen": -1.9556434154510498, | |
| "logits/rejected": -1.890546202659607, | |
| "logps/chosen": -237.84500122070312, | |
| "logps/rejected": -261.2818298339844, | |
| "loss": 0.1363, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": 1.9853665828704834, | |
| "rewards/margins": 5.180100440979004, | |
| "rewards/rejected": -3.194733142852783, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 2.962962962962963, | |
| "grad_norm": 12.623822906293494, | |
| "learning_rate": 1.7819973504940023e-10, | |
| "logits/chosen": -1.9709722995758057, | |
| "logits/rejected": -1.8710010051727295, | |
| "logps/chosen": -241.50997924804688, | |
| "logps/rejected": -266.9458923339844, | |
| "loss": 0.1258, | |
| "rewards/accuracies": 0.9437500238418579, | |
| "rewards/chosen": 1.947824478149414, | |
| "rewards/margins": 4.476650238037109, | |
| "rewards/rejected": -2.5288257598876953, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.9876543209876543, | |
| "step": 363, | |
| "total_flos": 4280357159436288.0, | |
| "train_loss": 0.30565077164941584, | |
| "train_runtime": 13036.7565, | |
| "train_samples_per_second": 7.158, | |
| "train_steps_per_second": 0.028 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 363, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 4280357159436288.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |