| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9994111874386653, | |
| "eval_steps": 500, | |
| "global_step": 1273, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0, | |
| "learning_rate": 3.90625e-08, | |
| "logits/chosen": 0.571427047252655, | |
| "logits/rejected": 0.7314685583114624, | |
| "logps/chosen": -269.3661193847656, | |
| "logps/rejected": -211.47052001953125, | |
| "loss": 0.0442, | |
| "rewards/accuracies": 0.0, | |
| "rewards/chosen": 0.0, | |
| "rewards/margins": 0.0, | |
| "rewards/rejected": 0.0, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.01, | |
| "learning_rate": 3.90625e-07, | |
| "logits/chosen": 0.693711519241333, | |
| "logits/rejected": 0.7037237882614136, | |
| "logps/chosen": -279.3418884277344, | |
| "logps/rejected": -237.24127197265625, | |
| "loss": 0.0615, | |
| "rewards/accuracies": 0.3888888955116272, | |
| "rewards/chosen": 0.00013600506645161659, | |
| "rewards/margins": 9.131695696851239e-05, | |
| "rewards/rejected": 4.4688102207146585e-05, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 7.8125e-07, | |
| "logits/chosen": 0.6187659502029419, | |
| "logits/rejected": 0.6341584324836731, | |
| "logps/chosen": -269.4906921386719, | |
| "logps/rejected": -235.9794464111328, | |
| "loss": 0.0505, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": 0.00019002554472535849, | |
| "rewards/margins": 0.0005181421292945743, | |
| "rewards/rejected": -0.0003281166427768767, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "learning_rate": 1.1718750000000001e-06, | |
| "logits/chosen": 0.7198699712753296, | |
| "logits/rejected": 0.7689396739006042, | |
| "logps/chosen": -262.8460998535156, | |
| "logps/rejected": -238.15896606445312, | |
| "loss": 0.0612, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": 0.0001914335007313639, | |
| "rewards/margins": -9.696001507109031e-05, | |
| "rewards/rejected": 0.0002883935230784118, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "learning_rate": 1.5625e-06, | |
| "logits/chosen": 0.6236015558242798, | |
| "logits/rejected": 0.663158655166626, | |
| "logps/chosen": -271.2866516113281, | |
| "logps/rejected": -256.9141540527344, | |
| "loss": 0.0451, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -0.0013797181891277432, | |
| "rewards/margins": 0.0008465956198051572, | |
| "rewards/rejected": -0.0022263135761022568, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "learning_rate": 1.953125e-06, | |
| "logits/chosen": 0.6586641073226929, | |
| "logits/rejected": 0.7124195098876953, | |
| "logps/chosen": -277.61431884765625, | |
| "logps/rejected": -245.189697265625, | |
| "loss": 0.0533, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.0009271189337596297, | |
| "rewards/margins": 0.0012959478190168738, | |
| "rewards/rejected": -0.0022230667527765036, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.3437500000000002e-06, | |
| "logits/chosen": 0.6583735346794128, | |
| "logits/rejected": 0.7763499021530151, | |
| "logps/chosen": -295.41766357421875, | |
| "logps/rejected": -274.49810791015625, | |
| "loss": 0.0564, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.003529118839651346, | |
| "rewards/margins": 0.0007951221195980906, | |
| "rewards/rejected": -0.004324241541326046, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "learning_rate": 2.7343750000000004e-06, | |
| "logits/chosen": 0.7303035855293274, | |
| "logits/rejected": 0.7642933130264282, | |
| "logps/chosen": -274.9570617675781, | |
| "logps/rejected": -242.8776092529297, | |
| "loss": 0.0458, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -0.005638590082526207, | |
| "rewards/margins": 0.0016975027974694967, | |
| "rewards/rejected": -0.00733609264716506, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "learning_rate": 3.125e-06, | |
| "logits/chosen": 0.6080405712127686, | |
| "logits/rejected": 0.7384923696517944, | |
| "logps/chosen": -301.12823486328125, | |
| "logps/rejected": -243.5036163330078, | |
| "loss": 0.0526, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.009938476607203484, | |
| "rewards/margins": 0.004849700257182121, | |
| "rewards/rejected": -0.01478817779570818, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "learning_rate": 3.5156250000000003e-06, | |
| "logits/chosen": 0.733268141746521, | |
| "logits/rejected": 0.7006837129592896, | |
| "logps/chosen": -306.02166748046875, | |
| "logps/rejected": -309.63238525390625, | |
| "loss": 0.046, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.01198151707649231, | |
| "rewards/margins": 0.005548244807869196, | |
| "rewards/rejected": -0.017529763281345367, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "learning_rate": 3.90625e-06, | |
| "logits/chosen": 0.6403722763061523, | |
| "logits/rejected": 0.6439489126205444, | |
| "logps/chosen": -265.2840881347656, | |
| "logps/rejected": -230.4324493408203, | |
| "loss": 0.0536, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.018176386132836342, | |
| "rewards/margins": 0.005613826680928469, | |
| "rewards/rejected": -0.023790214210748672, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.296875e-06, | |
| "logits/chosen": 0.7128833532333374, | |
| "logits/rejected": 0.5932031869888306, | |
| "logps/chosen": -304.94171142578125, | |
| "logps/rejected": -258.391845703125, | |
| "loss": 0.0585, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.028187787160277367, | |
| "rewards/margins": 0.008357586339116096, | |
| "rewards/rejected": -0.03654537722468376, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.6875000000000004e-06, | |
| "logits/chosen": 0.6879287958145142, | |
| "logits/rejected": 0.7084887027740479, | |
| "logps/chosen": -298.180419921875, | |
| "logps/rejected": -226.73483276367188, | |
| "loss": 0.0596, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.04302723705768585, | |
| "rewards/margins": 0.012504307553172112, | |
| "rewards/rejected": -0.05553155019879341, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "learning_rate": 4.999962359300416e-06, | |
| "logits/chosen": 0.7091008424758911, | |
| "logits/rejected": 0.7590427994728088, | |
| "logps/chosen": -296.51300048828125, | |
| "logps/rejected": -253.19873046875, | |
| "loss": 0.0479, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.0437840037047863, | |
| "rewards/margins": 0.016129568219184875, | |
| "rewards/rejected": -0.059913575649261475, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "learning_rate": 4.998645053824218e-06, | |
| "logits/chosen": 0.6947115063667297, | |
| "logits/rejected": 0.6665563583374023, | |
| "logps/chosen": -335.786865234375, | |
| "logps/rejected": -286.8999938964844, | |
| "loss": 0.0472, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.048582445830106735, | |
| "rewards/margins": 0.025500774383544922, | |
| "rewards/rejected": -0.07408322393894196, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "learning_rate": 4.9954468466732145e-06, | |
| "logits/chosen": 0.6275844573974609, | |
| "logits/rejected": 0.6720636487007141, | |
| "logps/chosen": -296.78680419921875, | |
| "logps/rejected": -272.51873779296875, | |
| "loss": 0.0443, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.07314912974834442, | |
| "rewards/margins": 0.01452522724866867, | |
| "rewards/rejected": -0.08767435699701309, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.990370145357496e-06, | |
| "logits/chosen": 0.5851293802261353, | |
| "logits/rejected": 0.637703537940979, | |
| "logps/chosen": -273.9035339355469, | |
| "logps/rejected": -253.0749969482422, | |
| "loss": 0.0501, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.08011965453624725, | |
| "rewards/margins": 0.028026139363646507, | |
| "rewards/rejected": -0.10814578831195831, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "learning_rate": 4.983418771458684e-06, | |
| "logits/chosen": 0.6668591499328613, | |
| "logits/rejected": 0.6740854978561401, | |
| "logps/chosen": -326.5847473144531, | |
| "logps/rejected": -284.8072814941406, | |
| "loss": 0.0466, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.08650139719247818, | |
| "rewards/margins": 0.02333686873316765, | |
| "rewards/rejected": -0.10983824729919434, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "learning_rate": 4.97459795775315e-06, | |
| "logits/chosen": 0.6102133393287659, | |
| "logits/rejected": 0.6243614554405212, | |
| "logps/chosen": -325.43829345703125, | |
| "logps/rejected": -297.8895263671875, | |
| "loss": 0.0469, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.10013189166784286, | |
| "rewards/margins": 0.029614349827170372, | |
| "rewards/rejected": -0.12974625825881958, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "learning_rate": 4.963914344272961e-06, | |
| "logits/chosen": 0.5422903895378113, | |
| "logits/rejected": 0.6029689908027649, | |
| "logps/chosen": -304.6243591308594, | |
| "logps/rejected": -265.31231689453125, | |
| "loss": 0.0432, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.10887263715267181, | |
| "rewards/margins": 0.029181605204939842, | |
| "rewards/rejected": -0.1380542516708374, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.951375973307458e-06, | |
| "logits/chosen": 0.5463661551475525, | |
| "logits/rejected": 0.5838527679443359, | |
| "logps/chosen": -301.7105712890625, | |
| "logps/rejected": -276.5257568359375, | |
| "loss": 0.0474, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.11956067383289337, | |
| "rewards/margins": 0.026623660698533058, | |
| "rewards/rejected": -0.14618434011936188, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "learning_rate": 4.93699228334928e-06, | |
| "logits/chosen": 0.6322023868560791, | |
| "logits/rejected": 0.5507072806358337, | |
| "logps/chosen": -316.19110107421875, | |
| "logps/rejected": -265.67608642578125, | |
| "loss": 0.0423, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.11251308768987656, | |
| "rewards/margins": 0.04010463505983353, | |
| "rewards/rejected": -0.15261772274971008, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "learning_rate": 4.920774101989362e-06, | |
| "logits/chosen": 0.5577625036239624, | |
| "logits/rejected": 0.5779579877853394, | |
| "logps/chosen": -311.96624755859375, | |
| "logps/rejected": -279.3911437988281, | |
| "loss": 0.046, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.11701588332653046, | |
| "rewards/margins": 0.032017000019550323, | |
| "rewards/rejected": -0.14903289079666138, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.902733637766261e-06, | |
| "logits/chosen": 0.54422527551651, | |
| "logits/rejected": 0.5549877285957336, | |
| "logps/chosen": -326.54290771484375, | |
| "logps/rejected": -289.0083923339844, | |
| "loss": 0.0472, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.12084343284368515, | |
| "rewards/margins": 0.04505180940032005, | |
| "rewards/rejected": -0.1658952534198761, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "learning_rate": 4.882884470975954e-06, | |
| "logits/chosen": 0.4986969828605652, | |
| "logits/rejected": 0.5934550166130066, | |
| "logps/chosen": -280.1143798828125, | |
| "logps/rejected": -251.39358520507812, | |
| "loss": 0.0437, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.135459303855896, | |
| "rewards/margins": 0.034046441316604614, | |
| "rewards/rejected": -0.16950573027133942, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.861241543449015e-06, | |
| "logits/chosen": 0.5392612218856812, | |
| "logits/rejected": 0.577301025390625, | |
| "logps/chosen": -316.1842346191406, | |
| "logps/rejected": -293.55126953125, | |
| "loss": 0.0368, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.14570352435112, | |
| "rewards/margins": 0.02999885380268097, | |
| "rewards/rejected": -0.17570236325263977, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "learning_rate": 4.8378211473028755e-06, | |
| "logits/chosen": 0.5802425146102905, | |
| "logits/rejected": 0.5668612122535706, | |
| "logps/chosen": -284.0962829589844, | |
| "logps/rejected": -251.66818237304688, | |
| "loss": 0.0404, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.12709495425224304, | |
| "rewards/margins": 0.05544697493314743, | |
| "rewards/rejected": -0.18254193663597107, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 4.812640912677624e-06, | |
| "logits/chosen": 0.485908180475235, | |
| "logits/rejected": 0.5380377769470215, | |
| "logps/chosen": -273.9252624511719, | |
| "logps/rejected": -245.73794555664062, | |
| "loss": 0.0488, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.12756113708019257, | |
| "rewards/margins": 0.020655876025557518, | |
| "rewards/rejected": -0.14821702241897583, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "learning_rate": 4.785719794464596e-06, | |
| "logits/chosen": 0.43520134687423706, | |
| "logits/rejected": 0.5617900490760803, | |
| "logps/chosen": -343.7205505371094, | |
| "logps/rejected": -276.3643798828125, | |
| "loss": 0.0404, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.13573798537254333, | |
| "rewards/margins": 0.043094925582408905, | |
| "rewards/rejected": -0.17883291840553284, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "learning_rate": 4.757078058037722e-06, | |
| "logits/chosen": 0.4410250186920166, | |
| "logits/rejected": 0.5906217694282532, | |
| "logps/chosen": -331.30657958984375, | |
| "logps/rejected": -277.7027282714844, | |
| "loss": 0.0408, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.11059264838695526, | |
| "rewards/margins": 0.04809447377920151, | |
| "rewards/rejected": -0.15868714451789856, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.72673726399839e-06, | |
| "logits/chosen": 0.5095167756080627, | |
| "logits/rejected": 0.6414391398429871, | |
| "logps/chosen": -262.4388122558594, | |
| "logps/rejected": -231.455322265625, | |
| "loss": 0.0392, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.12092028558254242, | |
| "rewards/margins": 0.02815370261669159, | |
| "rewards/rejected": -0.1490740031003952, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "learning_rate": 4.694720251945298e-06, | |
| "logits/chosen": 0.5478841066360474, | |
| "logits/rejected": 0.4881111979484558, | |
| "logps/chosen": -321.4375305175781, | |
| "logps/rejected": -280.4942932128906, | |
| "loss": 0.0413, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.12603510916233063, | |
| "rewards/margins": 0.04496049880981445, | |
| "rewards/rejected": -0.17099562287330627, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "learning_rate": 4.661051123281528e-06, | |
| "logits/chosen": 0.5421676635742188, | |
| "logits/rejected": 0.5195788145065308, | |
| "logps/chosen": -291.16778564453125, | |
| "logps/rejected": -270.92779541015625, | |
| "loss": 0.0416, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.1264342963695526, | |
| "rewards/margins": 0.042602840811014175, | |
| "rewards/rejected": -0.1690371334552765, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "learning_rate": 4.6257552230717536e-06, | |
| "logits/chosen": 0.567541778087616, | |
| "logits/rejected": 0.5551782250404358, | |
| "logps/chosen": -292.7509765625, | |
| "logps/rejected": -310.8721618652344, | |
| "loss": 0.0372, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.15939810872077942, | |
| "rewards/margins": 0.0593896210193634, | |
| "rewards/rejected": -0.21878774464130402, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.588859120963282e-06, | |
| "logits/chosen": 0.44213318824768066, | |
| "logits/rejected": 0.4428648352622986, | |
| "logps/chosen": -260.6324768066406, | |
| "logps/rejected": -255.5272674560547, | |
| "loss": 0.0452, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.15107281506061554, | |
| "rewards/margins": 0.03870910778641701, | |
| "rewards/rejected": -0.18978191912174225, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.5503905911852435e-06, | |
| "logits/chosen": 0.5433288216590881, | |
| "logits/rejected": 0.5616317987442017, | |
| "logps/chosen": -283.3091735839844, | |
| "logps/rejected": -249.4799041748047, | |
| "loss": 0.0441, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.163166344165802, | |
| "rewards/margins": 0.03194192424416542, | |
| "rewards/rejected": -0.19510827958583832, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "learning_rate": 4.510378591641036e-06, | |
| "logits/chosen": 0.5152170658111572, | |
| "logits/rejected": 0.6420885324478149, | |
| "logps/chosen": -339.47576904296875, | |
| "logps/rejected": -327.2655334472656, | |
| "loss": 0.0351, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.16583441197872162, | |
| "rewards/margins": 0.04150914400815964, | |
| "rewards/rejected": -0.20734354853630066, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "learning_rate": 4.468853242109712e-06, | |
| "logits/chosen": 0.5886040925979614, | |
| "logits/rejected": 0.5353233814239502, | |
| "logps/chosen": -294.714599609375, | |
| "logps/rejected": -272.34649658203125, | |
| "loss": 0.0404, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.16863426566123962, | |
| "rewards/margins": 0.04221067577600479, | |
| "rewards/rejected": -0.2108449637889862, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "learning_rate": 4.42584580157276e-06, | |
| "logits/chosen": 0.4872845709323883, | |
| "logits/rejected": 0.6496473550796509, | |
| "logps/chosen": -322.4450988769531, | |
| "logps/rejected": -284.5151062011719, | |
| "loss": 0.0396, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.1785622537136078, | |
| "rewards/margins": 0.04358912259340286, | |
| "rewards/rejected": -0.22215136885643005, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.381388644683317e-06, | |
| "logits/chosen": 0.5238572359085083, | |
| "logits/rejected": 0.5852701663970947, | |
| "logps/chosen": -330.46746826171875, | |
| "logps/rejected": -284.7983093261719, | |
| "loss": 0.0413, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.1601935178041458, | |
| "rewards/margins": 0.04235311970114708, | |
| "rewards/rejected": -0.202546626329422, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "learning_rate": 4.33551523739555e-06, | |
| "logits/chosen": 0.5760276913642883, | |
| "logits/rejected": 0.6830281615257263, | |
| "logps/chosen": -317.65252685546875, | |
| "logps/rejected": -259.23590087890625, | |
| "loss": 0.0411, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.1589946150779724, | |
| "rewards/margins": 0.04796143248677254, | |
| "rewards/rejected": -0.20695605874061584, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "learning_rate": 4.288260111772535e-06, | |
| "logits/chosen": 0.6023763418197632, | |
| "logits/rejected": 0.6028311252593994, | |
| "logps/chosen": -301.0093994140625, | |
| "logps/rejected": -288.390625, | |
| "loss": 0.0399, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.15324203670024872, | |
| "rewards/margins": 0.06706853210926056, | |
| "rewards/rejected": -0.22031056880950928, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "learning_rate": 4.239658839991594e-06, | |
| "logits/chosen": 0.4882682263851166, | |
| "logits/rejected": 0.6300086975097656, | |
| "logps/chosen": -326.09796142578125, | |
| "logps/rejected": -292.12744140625, | |
| "loss": 0.0367, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.15980076789855957, | |
| "rewards/margins": 0.05597474053502083, | |
| "rewards/rejected": -0.2157754898071289, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "learning_rate": 4.189748007566686e-06, | |
| "logits/chosen": 0.6084526181221008, | |
| "logits/rejected": 0.6407297849655151, | |
| "logps/chosen": -295.1040954589844, | |
| "logps/rejected": -283.36199951171875, | |
| "loss": 0.0415, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.14586099982261658, | |
| "rewards/margins": 0.044940222054719925, | |
| "rewards/rejected": -0.1908012181520462, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.138565185807972e-06, | |
| "logits/chosen": 0.6028437614440918, | |
| "logits/rejected": 0.5474241971969604, | |
| "logps/chosen": -259.4440002441406, | |
| "logps/rejected": -262.1929931640625, | |
| "loss": 0.0457, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.13820604979991913, | |
| "rewards/margins": 0.04853081330657005, | |
| "rewards/rejected": -0.18673686683177948, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "learning_rate": 4.086148903539311e-06, | |
| "logits/chosen": 0.5923335552215576, | |
| "logits/rejected": 0.6619054079055786, | |
| "logps/chosen": -295.82440185546875, | |
| "logps/rejected": -260.25653076171875, | |
| "loss": 0.0387, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.14740963280200958, | |
| "rewards/margins": 0.04628303274512291, | |
| "rewards/rejected": -0.19369268417358398, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "learning_rate": 4.032538618094972e-06, | |
| "logits/chosen": 0.5214491486549377, | |
| "logits/rejected": 0.6175715923309326, | |
| "logps/chosen": -275.34234619140625, | |
| "logps/rejected": -263.62628173828125, | |
| "loss": 0.0375, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.1500253975391388, | |
| "rewards/margins": 0.058636367321014404, | |
| "rewards/rejected": -0.2086617648601532, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 3.977774685617386e-06, | |
| "logits/chosen": 0.5247635841369629, | |
| "logits/rejected": 0.5686976909637451, | |
| "logps/chosen": -331.62969970703125, | |
| "logps/rejected": -301.4044494628906, | |
| "loss": 0.0342, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.15346726775169373, | |
| "rewards/margins": 0.03650849312543869, | |
| "rewards/rejected": -0.18997573852539062, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.92189833067831e-06, | |
| "logits/chosen": 0.48843416571617126, | |
| "logits/rejected": 0.5862494707107544, | |
| "logps/chosen": -315.828857421875, | |
| "logps/rejected": -266.48931884765625, | |
| "loss": 0.0396, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.13729068636894226, | |
| "rewards/margins": 0.05156175419688225, | |
| "rewards/rejected": -0.18885241448879242, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "learning_rate": 3.864951615246261e-06, | |
| "logits/chosen": 0.5482813119888306, | |
| "logits/rejected": 0.5736523866653442, | |
| "logps/chosen": -312.53289794921875, | |
| "logps/rejected": -297.5661315917969, | |
| "loss": 0.0393, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.15364210307598114, | |
| "rewards/margins": 0.06609909236431122, | |
| "rewards/rejected": -0.21974118053913116, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "learning_rate": 3.806977407023581e-06, | |
| "logits/chosen": 0.6043223142623901, | |
| "logits/rejected": 0.5951470136642456, | |
| "logps/chosen": -287.01416015625, | |
| "logps/rejected": -266.211669921875, | |
| "loss": 0.0317, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.15719470381736755, | |
| "rewards/margins": 0.05330941826105118, | |
| "rewards/rejected": -0.21050409972667694, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "learning_rate": 3.7480193471769815e-06, | |
| "logits/chosen": 0.508133053779602, | |
| "logits/rejected": 0.5117858648300171, | |
| "logps/chosen": -303.5716857910156, | |
| "logps/rejected": -281.1479797363281, | |
| "loss": 0.0433, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.1687772274017334, | |
| "rewards/margins": 0.04205339774489403, | |
| "rewards/rejected": -0.21083059906959534, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "learning_rate": 3.6881218174858354e-06, | |
| "logits/chosen": 0.5807247161865234, | |
| "logits/rejected": 0.5294843912124634, | |
| "logps/chosen": -325.0351867675781, | |
| "logps/rejected": -303.7823791503906, | |
| "loss": 0.0406, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.17659473419189453, | |
| "rewards/margins": 0.052628111094236374, | |
| "rewards/rejected": -0.2292228490114212, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.627329906932964e-06, | |
| "logits/chosen": 0.5673174262046814, | |
| "logits/rejected": 0.5497801303863525, | |
| "logps/chosen": -339.43023681640625, | |
| "logps/rejected": -292.1569519042969, | |
| "loss": 0.039, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.1955457627773285, | |
| "rewards/margins": 0.05005335807800293, | |
| "rewards/rejected": -0.24559912085533142, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "learning_rate": 3.5656893777630686e-06, | |
| "logits/chosen": 0.5412625670433044, | |
| "logits/rejected": 0.6069876551628113, | |
| "logps/chosen": -285.06072998046875, | |
| "logps/rejected": -270.3983459472656, | |
| "loss": 0.038, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -0.1897256076335907, | |
| "rewards/margins": 0.0756111592054367, | |
| "rewards/rejected": -0.2653367817401886, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.503246631034345e-06, | |
| "logits/chosen": 0.5480440855026245, | |
| "logits/rejected": 0.5790668725967407, | |
| "logps/chosen": -278.4317932128906, | |
| "logps/rejected": -268.5929260253906, | |
| "loss": 0.0412, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -0.16996586322784424, | |
| "rewards/margins": 0.05341961234807968, | |
| "rewards/rejected": -0.22338548302650452, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "learning_rate": 3.440048671689219e-06, | |
| "logits/chosen": 0.5717700123786926, | |
| "logits/rejected": 0.6006878614425659, | |
| "logps/chosen": -311.20782470703125, | |
| "logps/rejected": -283.98321533203125, | |
| "loss": 0.0373, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.1866205632686615, | |
| "rewards/margins": 0.05927601456642151, | |
| "rewards/rejected": -0.24589654803276062, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "learning_rate": 3.3761430731705056e-06, | |
| "logits/chosen": 0.5225729942321777, | |
| "logits/rejected": 0.6431065797805786, | |
| "logps/chosen": -302.95758056640625, | |
| "logps/rejected": -277.3609313964844, | |
| "loss": 0.0322, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.18959183990955353, | |
| "rewards/margins": 0.055399276316165924, | |
| "rewards/rejected": -0.24499110877513885, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.311577941609604e-06, | |
| "logits/chosen": 0.6244104504585266, | |
| "logits/rejected": 0.5891777873039246, | |
| "logps/chosen": -266.7265319824219, | |
| "logps/rejected": -242.4050750732422, | |
| "loss": 0.043, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.18597814440727234, | |
| "rewards/margins": 0.04474898427724838, | |
| "rewards/rejected": -0.2307271510362625, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 3.2464018796137157e-06, | |
| "logits/chosen": 0.5826702117919922, | |
| "logits/rejected": 0.5641965270042419, | |
| "logps/chosen": -303.88970947265625, | |
| "logps/rejected": -268.25640869140625, | |
| "loss": 0.0352, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.17130930721759796, | |
| "rewards/margins": 0.037886954843997955, | |
| "rewards/rejected": -0.20919625461101532, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "learning_rate": 3.1806639496793245e-06, | |
| "logits/chosen": 0.6168242692947388, | |
| "logits/rejected": 0.6149759888648987, | |
| "logps/chosen": -285.6898193359375, | |
| "logps/rejected": -286.07379150390625, | |
| "loss": 0.0402, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.14284178614616394, | |
| "rewards/margins": 0.050834327936172485, | |
| "rewards/rejected": -0.19367611408233643, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "learning_rate": 3.114413637259484e-06, | |
| "logits/chosen": 0.5781451463699341, | |
| "logits/rejected": 0.6401593089103699, | |
| "logps/chosen": -303.3341064453125, | |
| "logps/rejected": -291.5891418457031, | |
| "loss": 0.0394, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.14258936047554016, | |
| "rewards/margins": 0.04603290185332298, | |
| "rewards/rejected": -0.18862226605415344, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 3.0477008135127247e-06, | |
| "logits/chosen": 0.5642613768577576, | |
| "logits/rejected": 0.4775732159614563, | |
| "logps/chosen": -308.5189208984375, | |
| "logps/rejected": -277.7254333496094, | |
| "loss": 0.0406, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.15040606260299683, | |
| "rewards/margins": 0.047120168805122375, | |
| "rewards/rejected": -0.1975262314081192, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "learning_rate": 2.980575697761603e-06, | |
| "logits/chosen": 0.5864249467849731, | |
| "logits/rejected": 0.585323691368103, | |
| "logps/chosen": -279.18060302734375, | |
| "logps/rejected": -274.70880126953125, | |
| "loss": 0.0431, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.1575286090373993, | |
| "rewards/margins": 0.05304408073425293, | |
| "rewards/rejected": -0.21057268977165222, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "learning_rate": 2.9130888196891755e-06, | |
| "logits/chosen": 0.4903879165649414, | |
| "logits/rejected": 0.5989735126495361, | |
| "logps/chosen": -272.58428955078125, | |
| "logps/rejected": -251.16531372070312, | |
| "loss": 0.0431, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.1445484459400177, | |
| "rewards/margins": 0.022020401433110237, | |
| "rewards/rejected": -0.1665688455104828, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "learning_rate": 2.845290981301834e-06, | |
| "logits/chosen": 0.5647401809692383, | |
| "logits/rejected": 0.5633600950241089, | |
| "logps/chosen": -294.28594970703125, | |
| "logps/rejected": -258.2326965332031, | |
| "loss": 0.0347, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.15951719880104065, | |
| "rewards/margins": 0.04437902569770813, | |
| "rewards/rejected": -0.20389623939990997, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "learning_rate": 2.7772332186871464e-06, | |
| "logits/chosen": 0.5554920434951782, | |
| "logits/rejected": 0.6650098562240601, | |
| "logps/chosen": -307.80804443359375, | |
| "logps/rejected": -271.55963134765625, | |
| "loss": 0.0397, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.1580692082643509, | |
| "rewards/margins": 0.052956365048885345, | |
| "rewards/rejected": -0.21102559566497803, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.708966763595493e-06, | |
| "logits/chosen": 0.5610272884368896, | |
| "logits/rejected": 0.6106783151626587, | |
| "logps/chosen": -309.99945068359375, | |
| "logps/rejected": -258.3206481933594, | |
| "loss": 0.0395, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.16326628625392914, | |
| "rewards/margins": 0.06232718378305435, | |
| "rewards/rejected": -0.22559349238872528, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 2.640543004874409e-06, | |
| "logits/chosen": 0.4919437766075134, | |
| "logits/rejected": 0.6173873543739319, | |
| "logps/chosen": -316.8961486816406, | |
| "logps/rejected": -278.5799255371094, | |
| "loss": 0.0408, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.15790703892707825, | |
| "rewards/margins": 0.05107961967587471, | |
| "rewards/rejected": -0.20898666977882385, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "learning_rate": 2.572013449784671e-06, | |
| "logits/chosen": 0.5977932810783386, | |
| "logits/rejected": 0.5793666839599609, | |
| "logps/chosen": -274.3351745605469, | |
| "logps/rejected": -260.75408935546875, | |
| "loss": 0.0364, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.16240599751472473, | |
| "rewards/margins": 0.04127690568566322, | |
| "rewards/rejected": -0.20368289947509766, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 2.503429685227245e-06, | |
| "logits/chosen": 0.5404385924339294, | |
| "logits/rejected": 0.5224953293800354, | |
| "logps/chosen": -333.06292724609375, | |
| "logps/rejected": -277.642333984375, | |
| "loss": 0.039, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.1504869908094406, | |
| "rewards/margins": 0.05367060750722885, | |
| "rewards/rejected": -0.20415759086608887, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "learning_rate": 2.434843338910286e-06, | |
| "logits/chosen": 0.5657359957695007, | |
| "logits/rejected": 0.6558262705802917, | |
| "logps/chosen": -338.18048095703125, | |
| "logps/rejected": -290.9669189453125, | |
| "loss": 0.0376, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.15950533747673035, | |
| "rewards/margins": 0.04541635885834694, | |
| "rewards/rejected": -0.2049216777086258, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.3663060404854155e-06, | |
| "logits/chosen": 0.5390611886978149, | |
| "logits/rejected": 0.5992618799209595, | |
| "logps/chosen": -287.187744140625, | |
| "logps/rejected": -266.4593505859375, | |
| "loss": 0.0403, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.16842734813690186, | |
| "rewards/margins": 0.05528571456670761, | |
| "rewards/rejected": -0.22371308505535126, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "learning_rate": 2.2978693826825406e-06, | |
| "logits/chosen": 0.5962172746658325, | |
| "logits/rejected": 0.618035614490509, | |
| "logps/chosen": -325.6087951660156, | |
| "logps/rejected": -282.9885559082031, | |
| "loss": 0.0349, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.1736389696598053, | |
| "rewards/margins": 0.05205898359417915, | |
| "rewards/rejected": -0.22569794952869415, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "learning_rate": 2.2295848824724612e-06, | |
| "logits/chosen": 0.5612713098526001, | |
| "logits/rejected": 0.5915429592132568, | |
| "logps/chosen": -273.39300537109375, | |
| "logps/rejected": -248.25247192382812, | |
| "loss": 0.0474, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.1872544288635254, | |
| "rewards/margins": 0.06460748612880707, | |
| "rewards/rejected": -0.25186192989349365, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "learning_rate": 2.1615039422865136e-06, | |
| "logits/chosen": 0.5421571731567383, | |
| "logits/rejected": 0.613379955291748, | |
| "logps/chosen": -293.1100769042969, | |
| "logps/rejected": -261.9356994628906, | |
| "loss": 0.0435, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.18293720483779907, | |
| "rewards/margins": 0.05960521101951599, | |
| "rewards/rejected": -0.24254241585731506, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0936778113224253e-06, | |
| "logits/chosen": 0.6936706900596619, | |
| "logits/rejected": 0.6555561423301697, | |
| "logps/chosen": -301.51629638671875, | |
| "logps/rejected": -261.8467102050781, | |
| "loss": 0.0408, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.18969768285751343, | |
| "rewards/margins": 0.055843252688646317, | |
| "rewards/rejected": -0.24554093182086945, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "learning_rate": 2.0261575469655304e-06, | |
| "logits/chosen": 0.5248943567276001, | |
| "logits/rejected": 0.548130214214325, | |
| "logps/chosen": -300.0675048828125, | |
| "logps/rejected": -273.765869140625, | |
| "loss": 0.0363, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.19320233166217804, | |
| "rewards/margins": 0.05187714099884033, | |
| "rewards/rejected": -0.24507947266101837, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "learning_rate": 1.9589939763543693e-06, | |
| "logits/chosen": 0.6453245282173157, | |
| "logits/rejected": 0.48234623670578003, | |
| "logps/chosen": -298.9971008300781, | |
| "logps/rejected": -260.1150207519531, | |
| "loss": 0.0408, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.19145645201206207, | |
| "rewards/margins": 0.05295104905962944, | |
| "rewards/rejected": -0.24440748989582062, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "learning_rate": 1.8922376581196107e-06, | |
| "logits/chosen": 0.6130558252334595, | |
| "logits/rejected": 0.5399459004402161, | |
| "logps/chosen": -336.5523376464844, | |
| "logps/rejected": -287.4131774902344, | |
| "loss": 0.0434, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.19996806979179382, | |
| "rewards/margins": 0.041921354830265045, | |
| "rewards/rejected": -0.24188943207263947, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "learning_rate": 1.8259388443250993e-06, | |
| "logits/chosen": 0.48818960785865784, | |
| "logits/rejected": 0.5838159918785095, | |
| "logps/chosen": -295.0948791503906, | |
| "logps/rejected": -259.95452880859375, | |
| "loss": 0.0422, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.16656538844108582, | |
| "rewards/margins": 0.04429946839809418, | |
| "rewards/rejected": -0.21086485683918, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.760147442639679e-06, | |
| "logits/chosen": 0.4529576301574707, | |
| "logits/rejected": 0.6206762194633484, | |
| "logps/chosen": -310.306396484375, | |
| "logps/rejected": -250.24331665039062, | |
| "loss": 0.0406, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.18583405017852783, | |
| "rewards/margins": 0.061879683285951614, | |
| "rewards/rejected": -0.24771372973918915, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 1.6949129787682628e-06, | |
| "logits/chosen": 0.4996159076690674, | |
| "logits/rejected": 0.5564543008804321, | |
| "logps/chosen": -348.6169738769531, | |
| "logps/rejected": -285.4058532714844, | |
| "loss": 0.0494, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.18937376141548157, | |
| "rewards/margins": 0.05144171044230461, | |
| "rewards/rejected": -0.24081549048423767, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "learning_rate": 1.6302845591704348e-06, | |
| "logits/chosen": 0.5382525324821472, | |
| "logits/rejected": 0.5930293202400208, | |
| "logps/chosen": -290.4473571777344, | |
| "logps/rejected": -286.49615478515625, | |
| "loss": 0.0448, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.17880043387413025, | |
| "rewards/margins": 0.05653693526983261, | |
| "rewards/rejected": -0.23533737659454346, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "learning_rate": 1.5663108340946465e-06, | |
| "logits/chosen": 0.5701301097869873, | |
| "logits/rejected": 0.6285163760185242, | |
| "logps/chosen": -288.7821960449219, | |
| "logps/rejected": -275.9776611328125, | |
| "loss": 0.0354, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.18739423155784607, | |
| "rewards/margins": 0.05587046220898628, | |
| "rewards/rejected": -0.24326470494270325, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "learning_rate": 1.5030399609558364e-06, | |
| "logits/chosen": 0.4821529984474182, | |
| "logits/rejected": 0.545883059501648, | |
| "logps/chosen": -295.86456298828125, | |
| "logps/rejected": -284.3241271972656, | |
| "loss": 0.0417, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.20055215060710907, | |
| "rewards/margins": 0.049835216253995895, | |
| "rewards/rejected": -0.25038737058639526, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.4405195680840357e-06, | |
| "logits/chosen": 0.5690888166427612, | |
| "logits/rejected": 0.6206023097038269, | |
| "logps/chosen": -282.1586608886719, | |
| "logps/rejected": -248.159423828125, | |
| "loss": 0.0371, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.17809875309467316, | |
| "rewards/margins": 0.05816931277513504, | |
| "rewards/rejected": -0.2362680435180664, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "learning_rate": 1.378796718871252e-06, | |
| "logits/chosen": 0.5121451616287231, | |
| "logits/rejected": 0.6095593571662903, | |
| "logps/chosen": -339.46673583984375, | |
| "logps/rejected": -275.1758728027344, | |
| "loss": 0.0394, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.18696454167366028, | |
| "rewards/margins": 0.05620839074254036, | |
| "rewards/rejected": -0.24317291378974915, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "learning_rate": 1.3179178763436302e-06, | |
| "logits/chosen": 0.48259830474853516, | |
| "logits/rejected": 0.6230058073997498, | |
| "logps/chosen": -290.6861267089844, | |
| "logps/rejected": -267.58428955078125, | |
| "loss": 0.0312, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.1768282651901245, | |
| "rewards/margins": 0.061559922993183136, | |
| "rewards/rejected": -0.23838821053504944, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "learning_rate": 1.2579288681855364e-06, | |
| "logits/chosen": 0.6349326968193054, | |
| "logits/rejected": 0.5690282583236694, | |
| "logps/chosen": -283.8800964355469, | |
| "logps/rejected": -267.09417724609375, | |
| "loss": 0.0435, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.19100067019462585, | |
| "rewards/margins": 0.05224493891000748, | |
| "rewards/rejected": -0.24324557185173035, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1988748522419163e-06, | |
| "logits/chosen": 0.5791734457015991, | |
| "logits/rejected": 0.6113063097000122, | |
| "logps/chosen": -307.1993408203125, | |
| "logps/rejected": -278.4551696777344, | |
| "loss": 0.0402, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.18477237224578857, | |
| "rewards/margins": 0.057056255638599396, | |
| "rewards/rejected": -0.24182860553264618, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "learning_rate": 1.1408002825248842e-06, | |
| "logits/chosen": 0.5439066886901855, | |
| "logits/rejected": 0.5941084027290344, | |
| "logps/chosen": -329.396240234375, | |
| "logps/rejected": -295.2839660644531, | |
| "loss": 0.0331, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.1919744461774826, | |
| "rewards/margins": 0.052353404462337494, | |
| "rewards/rejected": -0.2443278580904007, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "learning_rate": 1.0837488757501369e-06, | |
| "logits/chosen": 0.5491483211517334, | |
| "logits/rejected": 0.6100614666938782, | |
| "logps/chosen": -292.9952392578125, | |
| "logps/rejected": -277.0441589355469, | |
| "loss": 0.0356, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.17984095215797424, | |
| "rewards/margins": 0.042489904910326004, | |
| "rewards/rejected": -0.22233083844184875, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 1.027763578428379e-06, | |
| "logits/chosen": 0.6163350939750671, | |
| "logits/rejected": 0.6365456581115723, | |
| "logps/chosen": -264.69342041015625, | |
| "logps/rejected": -243.10806274414062, | |
| "loss": 0.0417, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.19993123412132263, | |
| "rewards/margins": 0.046492286026477814, | |
| "rewards/rejected": -0.24642352759838104, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "learning_rate": 9.728865345365379e-07, | |
| "logits/chosen": 0.5641825199127197, | |
| "logits/rejected": 0.606396496295929, | |
| "logps/chosen": -331.7998962402344, | |
| "logps/rejected": -272.21807861328125, | |
| "loss": 0.0449, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.19182895123958588, | |
| "rewards/margins": 0.047703493386507034, | |
| "rewards/rejected": -0.239532470703125, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 9.191590537930975e-07, | |
| "logits/chosen": 0.5647382736206055, | |
| "logits/rejected": 0.5931411981582642, | |
| "logps/chosen": -301.1795959472656, | |
| "logps/rejected": -249.8610076904297, | |
| "loss": 0.0382, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.1782718151807785, | |
| "rewards/margins": 0.065487340092659, | |
| "rewards/rejected": -0.2437591254711151, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "learning_rate": 8.666215805614373e-07, | |
| "logits/chosen": 0.5855926871299744, | |
| "logits/rejected": 0.6292095184326172, | |
| "logps/chosen": -314.7355041503906, | |
| "logps/rejected": -275.4202575683594, | |
| "loss": 0.0394, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.20517146587371826, | |
| "rewards/margins": 0.041300006210803986, | |
| "rewards/rejected": -0.24647144973278046, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "learning_rate": 8.153136634045844e-07, | |
| "logits/chosen": 0.5548075437545776, | |
| "logits/rejected": 0.6082533001899719, | |
| "logps/chosen": -298.9161071777344, | |
| "logps/rejected": -263.01263427734375, | |
| "loss": 0.0347, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.20182470977306366, | |
| "rewards/margins": 0.05161082744598389, | |
| "rewards/rejected": -0.25343555212020874, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "learning_rate": 7.652739253142915e-07, | |
| "logits/chosen": 0.6577678322792053, | |
| "logits/rejected": 0.7113202810287476, | |
| "logps/chosen": -305.44976806640625, | |
| "logps/rejected": -270.2418212890625, | |
| "loss": 0.0368, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.18101605772972107, | |
| "rewards/margins": 0.06050257757306099, | |
| "rewards/rejected": -0.24151864647865295, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "learning_rate": 7.165400346368648e-07, | |
| "logits/chosen": 0.6242294311523438, | |
| "logits/rejected": 0.6780456304550171, | |
| "logps/chosen": -332.4653625488281, | |
| "logps/rejected": -270.494384765625, | |
| "loss": 0.0428, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.18956415355205536, | |
| "rewards/margins": 0.056781280785799026, | |
| "rewards/rejected": -0.24634544551372528, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.691486767176092e-07, | |
| "logits/chosen": 0.5437681078910828, | |
| "logits/rejected": 0.5767791867256165, | |
| "logps/chosen": -271.98846435546875, | |
| "logps/rejected": -275.7591247558594, | |
| "loss": 0.042, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.15761049091815948, | |
| "rewards/margins": 0.060720719397068024, | |
| "rewards/rejected": -0.2183312177658081, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "learning_rate": 6.231355262852529e-07, | |
| "logits/chosen": 0.5681263208389282, | |
| "logits/rejected": 0.5934298038482666, | |
| "logps/chosen": -362.6823425292969, | |
| "logps/rejected": -315.0286560058594, | |
| "loss": 0.0427, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.19303514063358307, | |
| "rewards/margins": 0.03555145114660263, | |
| "rewards/rejected": -0.2285865992307663, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 5.785352205971275e-07, | |
| "logits/chosen": 0.5157848596572876, | |
| "logits/rejected": 0.6102535724639893, | |
| "logps/chosen": -334.3918151855469, | |
| "logps/rejected": -284.1390075683594, | |
| "loss": 0.0442, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.1846066266298294, | |
| "rewards/margins": 0.06148224323987961, | |
| "rewards/rejected": -0.24608883261680603, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "learning_rate": 5.353813333653287e-07, | |
| "logits/chosen": 0.539936363697052, | |
| "logits/rejected": 0.6743307709693909, | |
| "logps/chosen": -305.96270751953125, | |
| "logps/rejected": -280.2798767089844, | |
| "loss": 0.0339, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.1911364048719406, | |
| "rewards/margins": 0.05630878359079361, | |
| "rewards/rejected": -0.24744519591331482, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.937063494834774e-07, | |
| "logits/chosen": 0.5440152883529663, | |
| "logits/rejected": 0.5379213094711304, | |
| "logps/chosen": -341.1717224121094, | |
| "logps/rejected": -311.44549560546875, | |
| "loss": 0.0393, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -0.18748323619365692, | |
| "rewards/margins": 0.0409691259264946, | |
| "rewards/rejected": -0.22845235466957092, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 4.5354164057310857e-07, | |
| "logits/chosen": 0.5548704862594604, | |
| "logits/rejected": 0.689899742603302, | |
| "logps/chosen": -312.20428466796875, | |
| "logps/rejected": -294.8197021484375, | |
| "loss": 0.0314, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -0.18130271136760712, | |
| "rewards/margins": 0.04764419421553612, | |
| "rewards/rejected": -0.22894692420959473, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "learning_rate": 4.1491744136810066e-07, | |
| "logits/chosen": 0.4567073881626129, | |
| "logits/rejected": 0.574688732624054, | |
| "logps/chosen": -294.9923400878906, | |
| "logps/rejected": -275.3056945800781, | |
| "loss": 0.0326, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.1727767288684845, | |
| "rewards/margins": 0.04382780194282532, | |
| "rewards/rejected": -0.216604545712471, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "learning_rate": 3.7786282695491313e-07, | |
| "logits/chosen": 0.6358944773674011, | |
| "logits/rejected": 0.6744691133499146, | |
| "logps/chosen": -316.3371276855469, | |
| "logps/rejected": -270.2391662597656, | |
| "loss": 0.037, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.1606210619211197, | |
| "rewards/margins": 0.05917732045054436, | |
| "rewards/rejected": -0.21979837119579315, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "learning_rate": 3.4240569088577564e-07, | |
| "logits/chosen": 0.5884599089622498, | |
| "logits/rejected": 0.6169819831848145, | |
| "logps/chosen": -307.93701171875, | |
| "logps/rejected": -279.9734191894531, | |
| "loss": 0.0364, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.17860962450504303, | |
| "rewards/margins": 0.05303224176168442, | |
| "rewards/rejected": -0.23164185881614685, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 3.0857272418129136e-07, | |
| "logits/chosen": 0.47011032700538635, | |
| "logits/rejected": 0.595490574836731, | |
| "logps/chosen": -283.87786865234375, | |
| "logps/rejected": -263.67724609375, | |
| "loss": 0.0334, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.17268426716327667, | |
| "rewards/margins": 0.049144428223371506, | |
| "rewards/rejected": -0.22182869911193848, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 2.7638939523827956e-07, | |
| "logits/chosen": 0.6203007698059082, | |
| "logits/rejected": 0.6173182129859924, | |
| "logps/chosen": -283.423828125, | |
| "logps/rejected": -276.0241394042969, | |
| "loss": 0.0371, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.17366403341293335, | |
| "rewards/margins": 0.06443198025226593, | |
| "rewards/rejected": -0.23809599876403809, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "learning_rate": 2.4587993065795983e-07, | |
| "logits/chosen": 0.6205240488052368, | |
| "logits/rejected": 0.6329373121261597, | |
| "logps/chosen": -326.6821594238281, | |
| "logps/rejected": -297.9998474121094, | |
| "loss": 0.0429, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.18457256257534027, | |
| "rewards/margins": 0.05959055945277214, | |
| "rewards/rejected": -0.2441631555557251, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "learning_rate": 2.170672970089291e-07, | |
| "logits/chosen": 0.5973344445228577, | |
| "logits/rejected": 0.6125074028968811, | |
| "logps/chosen": -318.9927978515625, | |
| "logps/rejected": -277.43475341796875, | |
| "loss": 0.0384, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.17308750748634338, | |
| "rewards/margins": 0.03716738894581795, | |
| "rewards/rejected": -0.21025490760803223, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.8997318353864673e-07, | |
| "logits/chosen": 0.5647025108337402, | |
| "logits/rejected": 0.6136811971664429, | |
| "logps/chosen": -306.97698974609375, | |
| "logps/rejected": -276.1070251464844, | |
| "loss": 0.0394, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.1707993894815445, | |
| "rewards/margins": 0.05791207030415535, | |
| "rewards/rejected": -0.22871145606040955, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "learning_rate": 1.6461798584644944e-07, | |
| "logits/chosen": 0.6306375861167908, | |
| "logits/rejected": 0.677390456199646, | |
| "logps/chosen": -310.4776306152344, | |
| "logps/rejected": -283.60137939453125, | |
| "loss": 0.0358, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.17375603318214417, | |
| "rewards/margins": 0.06438947468996048, | |
| "rewards/rejected": -0.23814550042152405, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "learning_rate": 1.4102079053038454e-07, | |
| "logits/chosen": 0.5880599021911621, | |
| "logits/rejected": 0.5578813552856445, | |
| "logps/chosen": -300.5694580078125, | |
| "logps/rejected": -258.61004638671875, | |
| "loss": 0.0377, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.1807607263326645, | |
| "rewards/margins": 0.037232767790555954, | |
| "rewards/rejected": -0.21799346804618835, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 1.1919936081941585e-07, | |
| "logits/chosen": 0.5770727396011353, | |
| "logits/rejected": 0.6810525059700012, | |
| "logps/chosen": -277.33453369140625, | |
| "logps/rejected": -288.65362548828125, | |
| "loss": 0.036, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.1911756694316864, | |
| "rewards/margins": 0.05452916771173477, | |
| "rewards/rejected": -0.24570484459400177, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "learning_rate": 9.917012320182245e-08, | |
| "logits/chosen": 0.5721954703330994, | |
| "logits/rejected": 0.6260513067245483, | |
| "logps/chosen": -284.9855651855469, | |
| "logps/rejected": -293.6813659667969, | |
| "loss": 0.0456, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.1743406057357788, | |
| "rewards/margins": 0.05407222360372543, | |
| "rewards/rejected": -0.22841282188892365, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 8.094815505985315e-08, | |
| "logits/chosen": 0.6424626111984253, | |
| "logits/rejected": 0.5862604975700378, | |
| "logps/chosen": -305.09735107421875, | |
| "logps/rejected": -268.40179443359375, | |
| "loss": 0.0364, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.17467457056045532, | |
| "rewards/margins": 0.03977496176958084, | |
| "rewards/rejected": -0.21444952487945557, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "learning_rate": 6.454717331994542e-08, | |
| "logits/chosen": 0.6076738238334656, | |
| "logits/rejected": 0.5825602412223816, | |
| "logps/chosen": -316.4405212402344, | |
| "logps/rejected": -298.938232421875, | |
| "loss": 0.0406, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.18210789561271667, | |
| "rewards/margins": 0.03575897961854935, | |
| "rewards/rejected": -0.2178668975830078, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "learning_rate": 4.9979524127052595e-08, | |
| "logits/chosen": 0.573727011680603, | |
| "logits/rejected": 0.6119558811187744, | |
| "logps/chosen": -308.22174072265625, | |
| "logps/rejected": -276.9308776855469, | |
| "loss": 0.0392, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.16624239087104797, | |
| "rewards/margins": 0.043992962688207626, | |
| "rewards/rejected": -0.2102353572845459, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "learning_rate": 3.725617355085476e-08, | |
| "logits/chosen": 0.5791876316070557, | |
| "logits/rejected": 0.6267887353897095, | |
| "logps/chosen": -314.5113830566406, | |
| "logps/rejected": -280.70892333984375, | |
| "loss": 0.039, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.20197847485542297, | |
| "rewards/margins": 0.04940859228372574, | |
| "rewards/rejected": -0.2513870596885681, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "learning_rate": 2.63866993308437e-08, | |
| "logits/chosen": 0.5729714632034302, | |
| "logits/rejected": 0.6838079690933228, | |
| "logps/chosen": -325.7432861328125, | |
| "logps/rejected": -262.1849060058594, | |
| "loss": 0.0383, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.16999685764312744, | |
| "rewards/margins": 0.05414390563964844, | |
| "rewards/rejected": -0.22414080798625946, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.737928366650099e-08, | |
| "logits/chosen": 0.537979006767273, | |
| "logits/rejected": 0.6233702898025513, | |
| "logps/chosen": -347.04754638671875, | |
| "logps/rejected": -299.84564208984375, | |
| "loss": 0.0378, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -0.1912025660276413, | |
| "rewards/margins": 0.057091712951660156, | |
| "rewards/rejected": -0.24829427897930145, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "learning_rate": 1.0240707057995735e-08, | |
| "logits/chosen": 0.5721354484558105, | |
| "logits/rejected": 0.6452512741088867, | |
| "logps/chosen": -291.0641784667969, | |
| "logps/rejected": -280.39947509765625, | |
| "loss": 0.0384, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.18037080764770508, | |
| "rewards/margins": 0.05573723465204239, | |
| "rewards/rejected": -0.23610803484916687, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "learning_rate": 4.976343202034717e-09, | |
| "logits/chosen": 0.5639461278915405, | |
| "logits/rejected": 0.6287204027175903, | |
| "logps/chosen": -303.62811279296875, | |
| "logps/rejected": -280.3854675292969, | |
| "loss": 0.0316, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.18784946203231812, | |
| "rewards/margins": 0.043084483593702316, | |
| "rewards/rejected": -0.23093394935131073, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "learning_rate": 1.5901549467139953e-09, | |
| "logits/chosen": 0.4898532032966614, | |
| "logits/rejected": 0.5764004588127136, | |
| "logps/chosen": -281.58013916015625, | |
| "logps/rejected": -252.1945343017578, | |
| "loss": 0.0345, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.1634509116411209, | |
| "rewards/margins": 0.06314502656459808, | |
| "rewards/rejected": -0.2265959233045578, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "learning_rate": 8.469130840960127e-11, | |
| "logits/chosen": 0.6247270107269287, | |
| "logits/rejected": 0.6422919034957886, | |
| "logps/chosen": -290.09881591796875, | |
| "logps/rejected": -274.88592529296875, | |
| "loss": 0.0406, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.16841378808021545, | |
| "rewards/margins": 0.0537291057407856, | |
| "rewards/rejected": -0.22214286029338837, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 1273, | |
| "total_flos": 0.0, | |
| "train_loss": 0.04122506890033327, | |
| "train_runtime": 12850.1249, | |
| "train_samples_per_second": 4.758, | |
| "train_steps_per_second": 0.099 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1273, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 100, | |
| "total_flos": 0.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |