| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9984301412872841, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0020931449502878076, | |
| "grad_norm": 3.2714715375601737, | |
| "learning_rate": 1.0416666666666666e-08, | |
| "logits/chosen": -1.3119012117385864, | |
| "logits/rejected": -1.3116823434829712, | |
| "logps/chosen": -358.818603515625, | |
| "logps/rejected": -361.4830322265625, | |
| "loss": 0.6931, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.020931449502878074, | |
| "grad_norm": 3.1981499549192196, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -1.2483385801315308, | |
| "logits/rejected": -1.2390522956848145, | |
| "logps/chosen": -300.1659851074219, | |
| "logps/rejected": -278.4343566894531, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": 0.0003463309840299189, | |
| "rewards/margins": 0.00022707899915985763, | |
| "rewards/rejected": 0.0001192519994219765, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04186289900575615, | |
| "grad_norm": 2.9666481993391387, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -1.2118756771087646, | |
| "logits/rejected": -1.2122180461883545, | |
| "logps/chosen": -312.6712951660156, | |
| "logps/rejected": -281.015869140625, | |
| "loss": 0.6932, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": 0.0008083735592663288, | |
| "rewards/margins": -0.0004832709673792124, | |
| "rewards/rejected": 0.0012916445266455412, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06279434850863422, | |
| "grad_norm": 3.032286414272413, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -1.22763991355896, | |
| "logits/rejected": -1.2292324304580688, | |
| "logps/chosen": -289.8343811035156, | |
| "logps/rejected": -250.0952911376953, | |
| "loss": 0.6921, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": 0.007961934432387352, | |
| "rewards/margins": 0.0013618591474369168, | |
| "rewards/rejected": 0.006600075867027044, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0837257980115123, | |
| "grad_norm": 3.0530878838982813, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -1.2131128311157227, | |
| "logits/rejected": -1.2069206237792969, | |
| "logps/chosen": -258.03167724609375, | |
| "logps/rejected": -260.1754150390625, | |
| "loss": 0.6903, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": 0.022502344101667404, | |
| "rewards/margins": 0.004317447543144226, | |
| "rewards/rejected": 0.018184896558523178, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10465724751439037, | |
| "grad_norm": 3.05177048406134, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -1.200819730758667, | |
| "logits/rejected": -1.1967626810073853, | |
| "logps/chosen": -285.07757568359375, | |
| "logps/rejected": -270.8570556640625, | |
| "loss": 0.6863, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": 0.04324551671743393, | |
| "rewards/margins": 0.016647540032863617, | |
| "rewards/rejected": 0.026597972959280014, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12558869701726844, | |
| "grad_norm": 3.2417202808010526, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -1.2190834283828735, | |
| "logits/rejected": -1.21430242061615, | |
| "logps/chosen": -250.82080078125, | |
| "logps/rejected": -243.12191772460938, | |
| "loss": 0.6803, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": 0.06585284322500229, | |
| "rewards/margins": 0.030171776190400124, | |
| "rewards/rejected": 0.035681065171957016, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.14652014652014653, | |
| "grad_norm": 2.9783130818034245, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -1.2192996740341187, | |
| "logits/rejected": -1.2149441242218018, | |
| "logps/chosen": -292.11279296875, | |
| "logps/rejected": -281.4046325683594, | |
| "loss": 0.6726, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": 0.06224682927131653, | |
| "rewards/margins": 0.04184813052415848, | |
| "rewards/rejected": 0.020398706197738647, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1674515960230246, | |
| "grad_norm": 3.7566626119690905, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -1.2239179611206055, | |
| "logits/rejected": -1.2163951396942139, | |
| "logps/chosen": -324.00897216796875, | |
| "logps/rejected": -258.05609130859375, | |
| "loss": 0.6636, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": 0.050625842064619064, | |
| "rewards/margins": 0.07805721461772919, | |
| "rewards/rejected": -0.027431374415755272, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.18838304552590268, | |
| "grad_norm": 3.65927557077053, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -1.2626798152923584, | |
| "logits/rejected": -1.2355681657791138, | |
| "logps/chosen": -258.53778076171875, | |
| "logps/rejected": -272.76470947265625, | |
| "loss": 0.6454, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.028159942477941513, | |
| "rewards/margins": 0.10718291997909546, | |
| "rewards/rejected": -0.13534286618232727, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.20931449502878074, | |
| "grad_norm": 3.9637480940035683, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -1.2693599462509155, | |
| "logits/rejected": -1.2746694087982178, | |
| "logps/chosen": -307.8084411621094, | |
| "logps/rejected": -307.66632080078125, | |
| "loss": 0.6333, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.07238290458917618, | |
| "rewards/margins": 0.1892419159412384, | |
| "rewards/rejected": -0.26162484288215637, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.2302459445316588, | |
| "grad_norm": 5.324044543906808, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -1.2339061498641968, | |
| "logits/rejected": -1.2281516790390015, | |
| "logps/chosen": -300.7259216308594, | |
| "logps/rejected": -303.48419189453125, | |
| "loss": 0.6256, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.11372830718755722, | |
| "rewards/margins": 0.19473466277122498, | |
| "rewards/rejected": -0.3084629774093628, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25117739403453687, | |
| "grad_norm": 5.868949121433513, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -1.172503113746643, | |
| "logits/rejected": -1.194064974784851, | |
| "logps/chosen": -299.2550048828125, | |
| "logps/rejected": -319.26580810546875, | |
| "loss": 0.6076, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.3037038743495941, | |
| "rewards/margins": 0.24693575501441956, | |
| "rewards/rejected": -0.5506396293640137, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.272108843537415, | |
| "grad_norm": 5.532495930770894, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -1.2171634435653687, | |
| "logits/rejected": -1.2139819860458374, | |
| "logps/chosen": -306.67327880859375, | |
| "logps/rejected": -327.5093078613281, | |
| "loss": 0.5838, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.25290900468826294, | |
| "rewards/margins": 0.29776304960250854, | |
| "rewards/rejected": -0.550672173500061, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29304029304029305, | |
| "grad_norm": 9.379379641862023, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -1.2662270069122314, | |
| "logits/rejected": -1.2556852102279663, | |
| "logps/chosen": -338.9728088378906, | |
| "logps/rejected": -327.0832824707031, | |
| "loss": 0.5827, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.18600435554981232, | |
| "rewards/margins": 0.3997423052787781, | |
| "rewards/rejected": -0.585746705532074, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3139717425431711, | |
| "grad_norm": 7.900278401061318, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -1.2308932542800903, | |
| "logits/rejected": -1.2276499271392822, | |
| "logps/chosen": -320.3226318359375, | |
| "logps/rejected": -326.42791748046875, | |
| "loss": 0.582, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.39666932821273804, | |
| "rewards/margins": 0.3853836953639984, | |
| "rewards/rejected": -0.7820531129837036, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3349031920460492, | |
| "grad_norm": 7.301380914088517, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -1.2494581937789917, | |
| "logits/rejected": -1.2624256610870361, | |
| "logps/chosen": -312.92584228515625, | |
| "logps/rejected": -334.1939697265625, | |
| "loss": 0.5702, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.33961960673332214, | |
| "rewards/margins": 0.45542359352111816, | |
| "rewards/rejected": -0.7950432300567627, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.35583464154892724, | |
| "grad_norm": 6.951907491069847, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -1.1752383708953857, | |
| "logits/rejected": -1.176161766052246, | |
| "logps/chosen": -305.71905517578125, | |
| "logps/rejected": -318.2366638183594, | |
| "loss": 0.5717, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.44680055975914, | |
| "rewards/margins": 0.37692078948020935, | |
| "rewards/rejected": -0.8237212896347046, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.37676609105180536, | |
| "grad_norm": 7.66796629301651, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -1.2374298572540283, | |
| "logits/rejected": -1.2232410907745361, | |
| "logps/chosen": -314.8545227050781, | |
| "logps/rejected": -330.78082275390625, | |
| "loss": 0.5482, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.42410340905189514, | |
| "rewards/margins": 0.39998775720596313, | |
| "rewards/rejected": -0.8240911364555359, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.3976975405546834, | |
| "grad_norm": 9.347951343579599, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": -1.2337291240692139, | |
| "logits/rejected": -1.2184447050094604, | |
| "logps/chosen": -306.25042724609375, | |
| "logps/rejected": -344.9998474121094, | |
| "loss": 0.5552, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.4029362201690674, | |
| "rewards/margins": 0.4247189164161682, | |
| "rewards/rejected": -0.8276551365852356, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4186289900575615, | |
| "grad_norm": 20.34015507630146, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": -1.1851434707641602, | |
| "logits/rejected": -1.2023327350616455, | |
| "logps/chosen": -317.30950927734375, | |
| "logps/rejected": -374.68707275390625, | |
| "loss": 0.565, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.6525951027870178, | |
| "rewards/margins": 0.36915016174316406, | |
| "rewards/rejected": -1.021745204925537, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.43956043956043955, | |
| "grad_norm": 8.66046401969586, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -1.2199736833572388, | |
| "logits/rejected": -1.2007920742034912, | |
| "logps/chosen": -373.19403076171875, | |
| "logps/rejected": -374.0740051269531, | |
| "loss": 0.5549, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.6114795804023743, | |
| "rewards/margins": 0.5353660583496094, | |
| "rewards/rejected": -1.1468455791473389, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.4604918890633176, | |
| "grad_norm": 10.308255360757213, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -1.18425714969635, | |
| "logits/rejected": -1.211212396621704, | |
| "logps/chosen": -276.52880859375, | |
| "logps/rejected": -339.9971008300781, | |
| "loss": 0.5522, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.42873048782348633, | |
| "rewards/margins": 0.5462521910667419, | |
| "rewards/rejected": -0.9749826192855835, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48142333856619574, | |
| "grad_norm": 12.741083745076377, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -1.2262427806854248, | |
| "logits/rejected": -1.2177931070327759, | |
| "logps/chosen": -330.7462158203125, | |
| "logps/rejected": -376.4158630371094, | |
| "loss": 0.555, | |
| "rewards/accuracies": 0.793749988079071, | |
| "rewards/chosen": -0.7175010442733765, | |
| "rewards/margins": 0.6692829132080078, | |
| "rewards/rejected": -1.3867839574813843, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5023547880690737, | |
| "grad_norm": 9.964168424713547, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -1.229190468788147, | |
| "logits/rejected": -1.1953736543655396, | |
| "logps/chosen": -340.6610412597656, | |
| "logps/rejected": -378.64459228515625, | |
| "loss": 0.5203, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.5589435696601868, | |
| "rewards/margins": 0.5544083714485168, | |
| "rewards/rejected": -1.1133520603179932, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5232862375719518, | |
| "grad_norm": 9.304678892746303, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -1.237926959991455, | |
| "logits/rejected": -1.202215313911438, | |
| "logps/chosen": -335.1034240722656, | |
| "logps/rejected": -362.30694580078125, | |
| "loss": 0.5384, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.5443297624588013, | |
| "rewards/margins": 0.6573392748832703, | |
| "rewards/rejected": -1.2016689777374268, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.54421768707483, | |
| "grad_norm": 11.08571008923917, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -1.2713990211486816, | |
| "logits/rejected": -1.28865385055542, | |
| "logps/chosen": -339.6799011230469, | |
| "logps/rejected": -381.6545104980469, | |
| "loss": 0.5364, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.793778121471405, | |
| "rewards/margins": 0.5738049745559692, | |
| "rewards/rejected": -1.367583155632019, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.565149136577708, | |
| "grad_norm": 10.522304727677207, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": -1.226070761680603, | |
| "logits/rejected": -1.224416732788086, | |
| "logps/chosen": -342.8006591796875, | |
| "logps/rejected": -384.75091552734375, | |
| "loss": 0.5349, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.4287544786930084, | |
| "rewards/margins": 0.7136284112930298, | |
| "rewards/rejected": -1.1423828601837158, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.5860805860805861, | |
| "grad_norm": 10.244786469121673, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": -1.1765474081039429, | |
| "logits/rejected": -1.1918376684188843, | |
| "logps/chosen": -294.80682373046875, | |
| "logps/rejected": -329.14312744140625, | |
| "loss": 0.5451, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.3686367869377136, | |
| "rewards/margins": 0.5905336737632751, | |
| "rewards/rejected": -0.9591705203056335, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.6070120355834642, | |
| "grad_norm": 9.226838895321402, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": -1.2361958026885986, | |
| "logits/rejected": -1.2049939632415771, | |
| "logps/chosen": -316.0486145019531, | |
| "logps/rejected": -381.14862060546875, | |
| "loss": 0.5398, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -0.6723560690879822, | |
| "rewards/margins": 0.6183323860168457, | |
| "rewards/rejected": -1.2906882762908936, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6279434850863422, | |
| "grad_norm": 8.816959226535186, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": -1.209092140197754, | |
| "logits/rejected": -1.1889410018920898, | |
| "logps/chosen": -401.56768798828125, | |
| "logps/rejected": -416.4493713378906, | |
| "loss": 0.5146, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.7396571040153503, | |
| "rewards/margins": 0.6182384490966797, | |
| "rewards/rejected": -1.3578956127166748, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6488749345892203, | |
| "grad_norm": 8.036248628388176, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": -1.1924922466278076, | |
| "logits/rejected": -1.1694473028182983, | |
| "logps/chosen": -372.14892578125, | |
| "logps/rejected": -391.8318786621094, | |
| "loss": 0.5116, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.6560718417167664, | |
| "rewards/margins": 0.7323848009109497, | |
| "rewards/rejected": -1.3884565830230713, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6698063840920984, | |
| "grad_norm": 11.007790934706696, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": -1.2555673122406006, | |
| "logits/rejected": -1.2262015342712402, | |
| "logps/chosen": -311.16534423828125, | |
| "logps/rejected": -354.5485534667969, | |
| "loss": 0.5076, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -0.5307902097702026, | |
| "rewards/margins": 0.7380278706550598, | |
| "rewards/rejected": -1.2688181400299072, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6907378335949764, | |
| "grad_norm": 10.593080489103407, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": -1.2314331531524658, | |
| "logits/rejected": -1.2257896661758423, | |
| "logps/chosen": -330.34393310546875, | |
| "logps/rejected": -386.2178039550781, | |
| "loss": 0.5312, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.5327134728431702, | |
| "rewards/margins": 0.6293205618858337, | |
| "rewards/rejected": -1.162034034729004, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7116692830978545, | |
| "grad_norm": 9.439948657937165, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": -1.15424644947052, | |
| "logits/rejected": -1.16172456741333, | |
| "logps/chosen": -363.83575439453125, | |
| "logps/rejected": -402.63494873046875, | |
| "loss": 0.5245, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.6764324307441711, | |
| "rewards/margins": 0.6598538756370544, | |
| "rewards/rejected": -1.3362863063812256, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7326007326007326, | |
| "grad_norm": 9.012621184513685, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": -1.241715908050537, | |
| "logits/rejected": -1.2342867851257324, | |
| "logps/chosen": -327.93231201171875, | |
| "logps/rejected": -370.9751892089844, | |
| "loss": 0.5044, | |
| "rewards/accuracies": 0.7875000238418579, | |
| "rewards/chosen": -0.6473630666732788, | |
| "rewards/margins": 0.722842812538147, | |
| "rewards/rejected": -1.3702059984207153, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7535321821036107, | |
| "grad_norm": 9.66579260075727, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -1.2246743440628052, | |
| "logits/rejected": -1.2141767740249634, | |
| "logps/chosen": -381.3811950683594, | |
| "logps/rejected": -412.919921875, | |
| "loss": 0.5388, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -0.6807909607887268, | |
| "rewards/margins": 0.7211862802505493, | |
| "rewards/rejected": -1.4019771814346313, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7744636316064888, | |
| "grad_norm": 8.633284456509573, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": -1.2712879180908203, | |
| "logits/rejected": -1.254762053489685, | |
| "logps/chosen": -322.0973205566406, | |
| "logps/rejected": -358.21942138671875, | |
| "loss": 0.5202, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.5800286531448364, | |
| "rewards/margins": 0.6660745739936829, | |
| "rewards/rejected": -1.246103286743164, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7953950811093669, | |
| "grad_norm": 9.854809153275779, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": -1.1864761114120483, | |
| "logits/rejected": -1.1942225694656372, | |
| "logps/chosen": -318.14453125, | |
| "logps/rejected": -416.2574768066406, | |
| "loss": 0.5224, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -0.7182336449623108, | |
| "rewards/margins": 0.8087763786315918, | |
| "rewards/rejected": -1.5270097255706787, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8163265306122449, | |
| "grad_norm": 9.683226914416851, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": -1.2469178438186646, | |
| "logits/rejected": -1.2045327425003052, | |
| "logps/chosen": -398.12994384765625, | |
| "logps/rejected": -434.2908630371094, | |
| "loss": 0.5144, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.6139215230941772, | |
| "rewards/margins": 0.6766796112060547, | |
| "rewards/rejected": -1.290601134300232, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.837257980115123, | |
| "grad_norm": 10.323841015921616, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": -1.2089258432388306, | |
| "logits/rejected": -1.2096412181854248, | |
| "logps/chosen": -322.91912841796875, | |
| "logps/rejected": -366.2901611328125, | |
| "loss": 0.5247, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.593218982219696, | |
| "rewards/margins": 0.6565582752227783, | |
| "rewards/rejected": -1.2497771978378296, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.858189429618001, | |
| "grad_norm": 8.925114203702107, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": -1.2606794834136963, | |
| "logits/rejected": -1.2504216432571411, | |
| "logps/chosen": -372.19635009765625, | |
| "logps/rejected": -403.69183349609375, | |
| "loss": 0.5278, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.6300618052482605, | |
| "rewards/margins": 0.7462152242660522, | |
| "rewards/rejected": -1.3762768507003784, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8791208791208791, | |
| "grad_norm": 7.917460029763406, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": -1.1264762878417969, | |
| "logits/rejected": -1.1589303016662598, | |
| "logps/chosen": -339.3796691894531, | |
| "logps/rejected": -417.29217529296875, | |
| "loss": 0.5154, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -0.7496095895767212, | |
| "rewards/margins": 0.7165287137031555, | |
| "rewards/rejected": -1.466138243675232, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9000523286237572, | |
| "grad_norm": 8.226733820840508, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": -1.1936299800872803, | |
| "logits/rejected": -1.1583229303359985, | |
| "logps/chosen": -357.48248291015625, | |
| "logps/rejected": -410.64349365234375, | |
| "loss": 0.4945, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -0.6429892778396606, | |
| "rewards/margins": 0.7604572772979736, | |
| "rewards/rejected": -1.4034465551376343, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9209837781266352, | |
| "grad_norm": 9.626312612509446, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": -1.2455865144729614, | |
| "logits/rejected": -1.222106695175171, | |
| "logps/chosen": -331.60198974609375, | |
| "logps/rejected": -350.9930419921875, | |
| "loss": 0.5179, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.6523444056510925, | |
| "rewards/margins": 0.5979769825935364, | |
| "rewards/rejected": -1.250321388244629, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9419152276295133, | |
| "grad_norm": 9.381121387990033, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": -1.1223390102386475, | |
| "logits/rejected": -1.1296846866607666, | |
| "logps/chosen": -343.0498962402344, | |
| "logps/rejected": -405.4967041015625, | |
| "loss": 0.5146, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.70673668384552, | |
| "rewards/margins": 0.7429010272026062, | |
| "rewards/rejected": -1.449637770652771, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9628466771323915, | |
| "grad_norm": 10.172974951790811, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": -1.1808750629425049, | |
| "logits/rejected": -1.1635867357254028, | |
| "logps/chosen": -334.4978942871094, | |
| "logps/rejected": -361.5756530761719, | |
| "loss": 0.5223, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -0.6034250259399414, | |
| "rewards/margins": 0.697030782699585, | |
| "rewards/rejected": -1.3004556894302368, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9837781266352695, | |
| "grad_norm": 9.398707959258918, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": -1.2121164798736572, | |
| "logits/rejected": -1.1964446306228638, | |
| "logps/chosen": -357.72991943359375, | |
| "logps/rejected": -408.731689453125, | |
| "loss": 0.5015, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.605668842792511, | |
| "rewards/margins": 0.6676869988441467, | |
| "rewards/rejected": -1.2733557224273682, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9984301412872841, | |
| "eval_logits/chosen": -1.218731164932251, | |
| "eval_logits/rejected": -1.20514976978302, | |
| "eval_logps/chosen": -331.9158630371094, | |
| "eval_logps/rejected": -413.128173828125, | |
| "eval_loss": 0.524560809135437, | |
| "eval_rewards/accuracies": 0.7890625, | |
| "eval_rewards/chosen": -0.5787502527236938, | |
| "eval_rewards/margins": 0.7770635485649109, | |
| "eval_rewards/rejected": -1.35581374168396, | |
| "eval_runtime": 228.1084, | |
| "eval_samples_per_second": 8.768, | |
| "eval_steps_per_second": 0.14, | |
| "step": 477 | |
| }, | |
| { | |
| "epoch": 0.9984301412872841, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 0.5676461645642167, | |
| "train_runtime": 13701.4312, | |
| "train_samples_per_second": 4.462, | |
| "train_steps_per_second": 0.035 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |