| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 100, | |
| "global_step": 478, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 1.0416666666666666e-08, | |
| "logits/chosen": -2.8386030197143555, | |
| "logits/rejected": -2.823939323425293, | |
| "logps/chosen": -324.3727722167969, | |
| "logps/rejected": -231.64634704589844, | |
| "loss": 0.1053, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.8247900009155273, | |
| "logits/rejected": -2.7507708072662354, | |
| "logps/chosen": -275.7712707519531, | |
| "logps/rejected": -253.4366912841797, | |
| "loss": 0.1049, | |
| "rewards/accuracies": 0.4305555522441864, | |
| "rewards/chosen": -0.00010182532423641533, | |
| "rewards/margins": -0.0002283406356582418, | |
| "rewards/rejected": 0.00012651534052565694, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.7973473072052, | |
| "logits/rejected": -2.7797961235046387, | |
| "logps/chosen": -261.93072509765625, | |
| "logps/rejected": -257.0435791015625, | |
| "loss": 0.1048, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": 0.00023481599055230618, | |
| "rewards/margins": 0.0013333772076293826, | |
| "rewards/rejected": -0.0010985612170770764, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.7847647666931152, | |
| "logits/rejected": -2.778301239013672, | |
| "logps/chosen": -295.0087890625, | |
| "logps/rejected": -259.0351257324219, | |
| "loss": 0.1026, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": 0.0006326682632789016, | |
| "rewards/margins": 0.008715003728866577, | |
| "rewards/rejected": -0.008082335814833641, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.8043212890625, | |
| "logits/rejected": -2.773319959640503, | |
| "logps/chosen": -284.64312744140625, | |
| "logps/rejected": -264.53228759765625, | |
| "loss": 0.1011, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.007321106735616922, | |
| "rewards/margins": 0.018408337607979774, | |
| "rewards/rejected": -0.02572944387793541, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999733114418725e-07, | |
| "logits/chosen": -2.779865264892578, | |
| "logits/rejected": -2.7175371646881104, | |
| "logps/chosen": -283.73773193359375, | |
| "logps/rejected": -253.7095184326172, | |
| "loss": 0.0944, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -0.013570243492722511, | |
| "rewards/margins": 0.056820280849933624, | |
| "rewards/rejected": -0.07039052993059158, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.990398100856366e-07, | |
| "logits/chosen": -2.805795192718506, | |
| "logits/rejected": -2.78430438041687, | |
| "logps/chosen": -293.5281677246094, | |
| "logps/rejected": -261.317626953125, | |
| "loss": 0.0893, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.06878744065761566, | |
| "rewards/margins": 0.07276646792888641, | |
| "rewards/rejected": -0.14155389368534088, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.967775735898179e-07, | |
| "logits/chosen": -2.7008605003356934, | |
| "logits/rejected": -2.6654582023620605, | |
| "logps/chosen": -272.4254455566406, | |
| "logps/rejected": -264.991455078125, | |
| "loss": 0.0818, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -0.12376339733600616, | |
| "rewards/margins": 0.13984574377536774, | |
| "rewards/rejected": -0.2636091411113739, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.931986719649298e-07, | |
| "logits/chosen": -2.7169277667999268, | |
| "logits/rejected": -2.700068950653076, | |
| "logps/chosen": -293.3257751464844, | |
| "logps/rejected": -285.71270751953125, | |
| "loss": 0.0738, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.2549767792224884, | |
| "rewards/margins": 0.13869795203208923, | |
| "rewards/rejected": -0.39367473125457764, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.883222001996351e-07, | |
| "logits/chosen": -2.7454025745391846, | |
| "logits/rejected": -2.721019983291626, | |
| "logps/chosen": -322.88800048828125, | |
| "logps/rejected": -335.7534484863281, | |
| "loss": 0.0572, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.3772469460964203, | |
| "rewards/margins": 0.30307695269584656, | |
| "rewards/rejected": -0.6803240180015564, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.821741763807186e-07, | |
| "logits/chosen": -2.6386475563049316, | |
| "logits/rejected": -2.6311068534851074, | |
| "logps/chosen": -348.7519226074219, | |
| "logps/rejected": -341.0511169433594, | |
| "loss": 0.0517, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.5442075729370117, | |
| "rewards/margins": 0.2245761901140213, | |
| "rewards/rejected": -0.7687837481498718, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "eval_logits/chosen": -2.752265691757202, | |
| "eval_logits/rejected": -2.7319908142089844, | |
| "eval_logps/chosen": -320.34552001953125, | |
| "eval_logps/rejected": -359.42059326171875, | |
| "eval_loss": 0.04265438765287399, | |
| "eval_rewards/accuracies": 0.703125, | |
| "eval_rewards/chosen": -0.6330589056015015, | |
| "eval_rewards/margins": 0.38761410117149353, | |
| "eval_rewards/rejected": -1.0206730365753174, | |
| "eval_runtime": 53.2642, | |
| "eval_samples_per_second": 37.549, | |
| "eval_steps_per_second": 0.601, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.747874028753375e-07, | |
| "logits/chosen": -2.745008945465088, | |
| "logits/rejected": -2.6902968883514404, | |
| "logps/chosen": -342.7307434082031, | |
| "logps/rejected": -338.11322021484375, | |
| "loss": 0.0397, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.6825590133666992, | |
| "rewards/margins": 0.41266337037086487, | |
| "rewards/rejected": -1.0952222347259521, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.662012913161997e-07, | |
| "logits/chosen": -2.6888184547424316, | |
| "logits/rejected": -2.6683411598205566, | |
| "logps/chosen": -343.5830078125, | |
| "logps/rejected": -382.2731628417969, | |
| "loss": 0.044, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.8013314008712769, | |
| "rewards/margins": 0.41239532828330994, | |
| "rewards/rejected": -1.2137267589569092, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5646165232345103e-07, | |
| "logits/chosen": -2.690122127532959, | |
| "logits/rejected": -2.673074245452881, | |
| "logps/chosen": -349.1435546875, | |
| "logps/rejected": -378.6903381347656, | |
| "loss": 0.0367, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.8628436326980591, | |
| "rewards/margins": 0.46390500664711, | |
| "rewards/rejected": -1.3267487287521362, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.456204510851956e-07, | |
| "logits/chosen": -2.6322216987609863, | |
| "logits/rejected": -2.615206480026245, | |
| "logps/chosen": -335.52008056640625, | |
| "logps/rejected": -349.85638427734375, | |
| "loss": 0.0411, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.8122199773788452, | |
| "rewards/margins": 0.3598429262638092, | |
| "rewards/rejected": -1.1720629930496216, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.337355301007335e-07, | |
| "logits/chosen": -2.546915292739868, | |
| "logits/rejected": -2.5243468284606934, | |
| "logps/chosen": -353.551025390625, | |
| "logps/rejected": -377.07989501953125, | |
| "loss": 0.0368, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.8309852480888367, | |
| "rewards/margins": 0.3872799277305603, | |
| "rewards/rejected": -1.2182650566101074, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.2087030056579986e-07, | |
| "logits/chosen": -2.516842842102051, | |
| "logits/rejected": -2.509202480316162, | |
| "logps/chosen": -396.78466796875, | |
| "logps/rejected": -430.03662109375, | |
| "loss": 0.0276, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.324944257736206, | |
| "rewards/margins": 0.2987218201160431, | |
| "rewards/rejected": -1.6236661672592163, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.070934040463998e-07, | |
| "logits/chosen": -2.486632823944092, | |
| "logits/rejected": -2.4413046836853027, | |
| "logps/chosen": -419.40582275390625, | |
| "logps/rejected": -413.1378479003906, | |
| "loss": 0.0252, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.4800517559051514, | |
| "rewards/margins": 0.3067554235458374, | |
| "rewards/rejected": -1.7868072986602783, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.9247834624635404e-07, | |
| "logits/chosen": -2.510317325592041, | |
| "logits/rejected": -2.5057549476623535, | |
| "logps/chosen": -372.1962585449219, | |
| "logps/rejected": -414.73858642578125, | |
| "loss": 0.0272, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.3809983730316162, | |
| "rewards/margins": 0.39020317792892456, | |
| "rewards/rejected": -1.771201491355896, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7710310482256523e-07, | |
| "logits/chosen": -2.5819058418273926, | |
| "logits/rejected": -2.580766439437866, | |
| "logps/chosen": -367.2718505859375, | |
| "logps/rejected": -395.99261474609375, | |
| "loss": 0.0276, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1196916103363037, | |
| "rewards/margins": 0.3849378228187561, | |
| "rewards/rejected": -1.504629373550415, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.610497133404795e-07, | |
| "logits/chosen": -2.5476508140563965, | |
| "logits/rejected": -2.5647332668304443, | |
| "logps/chosen": -385.6960754394531, | |
| "logps/rejected": -413.998291015625, | |
| "loss": 0.0305, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0930745601654053, | |
| "rewards/margins": 0.34844768047332764, | |
| "rewards/rejected": -1.4415223598480225, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "eval_logits/chosen": -2.579733371734619, | |
| "eval_logits/rejected": -2.5648751258850098, | |
| "eval_logps/chosen": -366.0565490722656, | |
| "eval_logps/rejected": -424.7881164550781, | |
| "eval_loss": 0.029730383306741714, | |
| "eval_rewards/accuracies": 0.72265625, | |
| "eval_rewards/chosen": -1.0901691913604736, | |
| "eval_rewards/margins": 0.5841794013977051, | |
| "eval_rewards/rejected": -1.6743484735488892, | |
| "eval_runtime": 53.2173, | |
| "eval_samples_per_second": 37.582, | |
| "eval_steps_per_second": 0.601, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.4440382358952115e-07, | |
| "logits/chosen": -2.49862003326416, | |
| "logits/rejected": -2.4916858673095703, | |
| "logps/chosen": -370.3592834472656, | |
| "logps/rejected": -407.2472229003906, | |
| "loss": 0.0299, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.1277533769607544, | |
| "rewards/margins": 0.5519038438796997, | |
| "rewards/rejected": -1.679657220840454, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.272542485937368e-07, | |
| "logits/chosen": -2.5778369903564453, | |
| "logits/rejected": -2.558171510696411, | |
| "logps/chosen": -380.861328125, | |
| "logps/rejected": -401.4359130859375, | |
| "loss": 0.0296, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1397533416748047, | |
| "rewards/margins": 0.4481385350227356, | |
| "rewards/rejected": -1.587891936302185, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.096924887558854e-07, | |
| "logits/chosen": -2.5747199058532715, | |
| "logits/rejected": -2.535149097442627, | |
| "logps/chosen": -425.2906188964844, | |
| "logps/rejected": -415.38677978515625, | |
| "loss": 0.0278, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.116990327835083, | |
| "rewards/margins": 0.44105783104896545, | |
| "rewards/rejected": -1.5580482482910156, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9181224366319943e-07, | |
| "logits/chosen": -2.4556021690368652, | |
| "logits/rejected": -2.422232151031494, | |
| "logps/chosen": -395.6624755859375, | |
| "logps/rejected": -402.35894775390625, | |
| "loss": 0.0272, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.2475910186767578, | |
| "rewards/margins": 0.4241558909416199, | |
| "rewards/rejected": -1.671746850013733, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7370891215954565e-07, | |
| "logits/chosen": -2.432041645050049, | |
| "logits/rejected": -2.4213993549346924, | |
| "logps/chosen": -398.3387451171875, | |
| "logps/rejected": -416.9798889160156, | |
| "loss": 0.0264, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.2123913764953613, | |
| "rewards/margins": 0.5389462113380432, | |
| "rewards/rejected": -1.7513376474380493, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.55479083351317e-07, | |
| "logits/chosen": -2.477323532104492, | |
| "logits/rejected": -2.4589943885803223, | |
| "logps/chosen": -404.10638427734375, | |
| "logps/rejected": -437.82330322265625, | |
| "loss": 0.0289, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -1.1455386877059937, | |
| "rewards/margins": 0.6531764268875122, | |
| "rewards/rejected": -1.7987149953842163, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.3722002126275822e-07, | |
| "logits/chosen": -2.392612934112549, | |
| "logits/rejected": -2.3851914405822754, | |
| "logps/chosen": -389.7043151855469, | |
| "logps/rejected": -410.2840881347656, | |
| "loss": 0.0276, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2070677280426025, | |
| "rewards/margins": 0.4757324755191803, | |
| "rewards/rejected": -1.68280029296875, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.19029145890313e-07, | |
| "logits/chosen": -2.4097766876220703, | |
| "logits/rejected": -2.387056350708008, | |
| "logps/chosen": -399.73663330078125, | |
| "logps/rejected": -442.6588439941406, | |
| "loss": 0.0236, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.4455411434173584, | |
| "rewards/margins": 0.6014262437820435, | |
| "rewards/rejected": -2.046967029571533, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 2.0100351342479216e-07, | |
| "logits/chosen": -2.4165215492248535, | |
| "logits/rejected": -2.3698172569274902, | |
| "logps/chosen": -431.76922607421875, | |
| "logps/rejected": -444.64080810546875, | |
| "loss": 0.025, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.4643802642822266, | |
| "rewards/margins": 0.5780037045478821, | |
| "rewards/rejected": -2.0423836708068848, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.8323929841460178e-07, | |
| "logits/chosen": -2.457998275756836, | |
| "logits/rejected": -2.436868190765381, | |
| "logps/chosen": -404.76409912109375, | |
| "logps/rejected": -460.781005859375, | |
| "loss": 0.0258, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1497269868850708, | |
| "rewards/margins": 0.6188378930091858, | |
| "rewards/rejected": -1.7685649394989014, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "eval_logits/chosen": -2.438358783721924, | |
| "eval_logits/rejected": -2.4246902465820312, | |
| "eval_logps/chosen": -377.3524475097656, | |
| "eval_logps/rejected": -444.54278564453125, | |
| "eval_loss": 0.02741866372525692, | |
| "eval_rewards/accuracies": 0.71875, | |
| "eval_rewards/chosen": -1.2031283378601074, | |
| "eval_rewards/margins": 0.668766438961029, | |
| "eval_rewards/rejected": -1.8718948364257812, | |
| "eval_runtime": 53.272, | |
| "eval_samples_per_second": 37.543, | |
| "eval_steps_per_second": 0.601, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6583128063291573e-07, | |
| "logits/chosen": -2.4320359230041504, | |
| "logits/rejected": -2.4065451622009277, | |
| "logps/chosen": -390.3728332519531, | |
| "logps/rejected": -426.6919860839844, | |
| "loss": 0.025, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.2841016054153442, | |
| "rewards/margins": 0.5093994736671448, | |
| "rewards/rejected": -1.7935012578964233, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.488723393865766e-07, | |
| "logits/chosen": -2.36087965965271, | |
| "logits/rejected": -2.3556385040283203, | |
| "logps/chosen": -418.49493408203125, | |
| "logps/rejected": -450.12139892578125, | |
| "loss": 0.0232, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.3573024272918701, | |
| "rewards/margins": 0.4884418547153473, | |
| "rewards/rejected": -1.8457443714141846, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3245295796480788e-07, | |
| "logits/chosen": -2.3685812950134277, | |
| "logits/rejected": -2.3668787479400635, | |
| "logps/chosen": -373.538330078125, | |
| "logps/rejected": -445.90625, | |
| "loss": 0.025, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.239685297012329, | |
| "rewards/margins": 0.6630219221115112, | |
| "rewards/rejected": -1.9027073383331299, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1666074087171627e-07, | |
| "logits/chosen": -2.38093900680542, | |
| "logits/rejected": -2.3856873512268066, | |
| "logps/chosen": -420.5098571777344, | |
| "logps/rejected": -464.68243408203125, | |
| "loss": 0.0268, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.2994890213012695, | |
| "rewards/margins": 0.645952045917511, | |
| "rewards/rejected": -1.9454412460327148, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.0157994641835734e-07, | |
| "logits/chosen": -2.4149842262268066, | |
| "logits/rejected": -2.3791468143463135, | |
| "logps/chosen": -378.4749450683594, | |
| "logps/rejected": -417.32305908203125, | |
| "loss": 0.0256, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.2759901285171509, | |
| "rewards/margins": 0.5479945540428162, | |
| "rewards/rejected": -1.8239845037460327, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.729103716819111e-08, | |
| "logits/chosen": -2.410066604614258, | |
| "logits/rejected": -2.374561309814453, | |
| "logps/chosen": -394.406494140625, | |
| "logps/rejected": -416.733642578125, | |
| "loss": 0.0275, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.3009366989135742, | |
| "rewards/margins": 0.5371383428573608, | |
| "rewards/rejected": -1.838075041770935, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.387025063449081e-08, | |
| "logits/chosen": -2.397771120071411, | |
| "logits/rejected": -2.362600326538086, | |
| "logps/chosen": -412.84332275390625, | |
| "logps/rejected": -450.31048583984375, | |
| "loss": 0.0228, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.2637946605682373, | |
| "rewards/margins": 0.5941613912582397, | |
| "rewards/rejected": -1.8579561710357666, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.138919252022435e-08, | |
| "logits/chosen": -2.4274802207946777, | |
| "logits/rejected": -2.420586347579956, | |
| "logps/chosen": -411.314208984375, | |
| "logps/rejected": -444.90869140625, | |
| "loss": 0.0251, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.256594181060791, | |
| "rewards/margins": 0.5403616428375244, | |
| "rewards/rejected": -1.7969558238983154, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.991445467064689e-08, | |
| "logits/chosen": -2.3289995193481445, | |
| "logits/rejected": -2.306408405303955, | |
| "logps/chosen": -396.445556640625, | |
| "logps/rejected": -452.8165588378906, | |
| "loss": 0.024, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.269308090209961, | |
| "rewards/margins": 0.6829138994216919, | |
| "rewards/rejected": -1.9522218704223633, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.9507259776993954e-08, | |
| "logits/chosen": -2.41839337348938, | |
| "logits/rejected": -2.396681547164917, | |
| "logps/chosen": -400.5920715332031, | |
| "logps/rejected": -426.2327575683594, | |
| "loss": 0.0234, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2493975162506104, | |
| "rewards/margins": 0.4738248288631439, | |
| "rewards/rejected": -1.7232223749160767, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "eval_logits/chosen": -2.4251461029052734, | |
| "eval_logits/rejected": -2.4111008644104004, | |
| "eval_logps/chosen": -376.6192321777344, | |
| "eval_logps/rejected": -444.92230224609375, | |
| "eval_loss": 0.027030915021896362, | |
| "eval_rewards/accuracies": 0.7265625, | |
| "eval_rewards/chosen": -1.1957957744598389, | |
| "eval_rewards/margins": 0.6798944473266602, | |
| "eval_rewards/rejected": -1.8756903409957886, | |
| "eval_runtime": 53.2021, | |
| "eval_samples_per_second": 37.593, | |
| "eval_steps_per_second": 0.601, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.022313472693447e-08, | |
| "logits/chosen": -2.412111759185791, | |
| "logits/rejected": -2.3663253784179688, | |
| "logps/chosen": -387.081787109375, | |
| "logps/rejected": -392.51336669921875, | |
| "loss": 0.0248, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.1241955757141113, | |
| "rewards/margins": 0.5859023332595825, | |
| "rewards/rejected": -1.7100979089736938, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.2111614344599684e-08, | |
| "logits/chosen": -2.420088291168213, | |
| "logits/rejected": -2.3837363719940186, | |
| "logps/chosen": -418.43719482421875, | |
| "logps/rejected": -451.14996337890625, | |
| "loss": 0.0256, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.2756112813949585, | |
| "rewards/margins": 0.7357032895088196, | |
| "rewards/rejected": -2.011314868927002, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.521597710086439e-08, | |
| "logits/chosen": -2.376132011413574, | |
| "logits/rejected": -2.380525827407837, | |
| "logps/chosen": -408.94781494140625, | |
| "logps/rejected": -458.7674865722656, | |
| "loss": 0.0244, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.34002685546875, | |
| "rewards/margins": 0.5449159741401672, | |
| "rewards/rejected": -1.8849427700042725, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.57301420397924e-09, | |
| "logits/chosen": -2.360464334487915, | |
| "logits/rejected": -2.3589634895324707, | |
| "logps/chosen": -394.01226806640625, | |
| "logps/rejected": -432.554443359375, | |
| "loss": 0.026, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.3313329219818115, | |
| "rewards/margins": 0.5157779455184937, | |
| "rewards/rejected": -1.8471107482910156, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 5.212833302556258e-09, | |
| "logits/chosen": -2.3685426712036133, | |
| "logits/rejected": -2.3834328651428223, | |
| "logps/chosen": -390.4405212402344, | |
| "logps/rejected": -432.5458068847656, | |
| "loss": 0.0248, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.3699880838394165, | |
| "rewards/margins": 0.4800630509853363, | |
| "rewards/rejected": -1.8500511646270752, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.158697848236607e-09, | |
| "logits/chosen": -2.3582141399383545, | |
| "logits/rejected": -2.3275790214538574, | |
| "logps/chosen": -391.3001403808594, | |
| "logps/rejected": -415.3433532714844, | |
| "loss": 0.0249, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.381425380706787, | |
| "rewards/margins": 0.4009224474430084, | |
| "rewards/rejected": -1.7823479175567627, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 4.269029751107489e-10, | |
| "logits/chosen": -2.3417115211486816, | |
| "logits/rejected": -2.3047232627868652, | |
| "logps/chosen": -387.1515197753906, | |
| "logps/rejected": -449.08837890625, | |
| "loss": 0.0252, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.2849957942962646, | |
| "rewards/margins": 0.6696574091911316, | |
| "rewards/rejected": -1.9546531438827515, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 478, | |
| "total_flos": 0.0, | |
| "train_loss": 0.04001489771085803, | |
| "train_runtime": 4360.6473, | |
| "train_samples_per_second": 14.019, | |
| "train_steps_per_second": 0.11 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 478, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |