| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.998691442030882, | |
| "eval_steps": 500, | |
| "global_step": 477, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010468463752944255, | |
| "grad_norm": 15.176345540381961, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": -2.7708451747894287, | |
| "logits/rejected": -2.7284560203552246, | |
| "logps/chosen": -1.0280715227127075, | |
| "logps/rejected": -1.1731975078582764, | |
| "loss": 2.0284, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.0280715227127075, | |
| "rewards/margins": 0.14512601494789124, | |
| "rewards/rejected": -1.1731975078582764, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 13.877817908572124, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.7508482933044434, | |
| "logits/rejected": -2.725215196609497, | |
| "logps/chosen": -0.9954730868339539, | |
| "logps/rejected": -1.071991205215454, | |
| "loss": 2.002, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -0.9954730868339539, | |
| "rewards/margins": 0.0765179991722107, | |
| "rewards/rejected": -1.071991205215454, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 13.915470437591527, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": -2.730905294418335, | |
| "logits/rejected": -2.697235584259033, | |
| "logps/chosen": -0.9896604418754578, | |
| "logps/rejected": -1.1480659246444702, | |
| "loss": 1.9874, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.9896604418754578, | |
| "rewards/margins": 0.15840543806552887, | |
| "rewards/rejected": -1.1480659246444702, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 13.632313250241245, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.8031985759735107, | |
| "logits/rejected": -2.7098159790039062, | |
| "logps/chosen": -1.022659420967102, | |
| "logps/rejected": -1.1234548091888428, | |
| "loss": 2.0265, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -1.022659420967102, | |
| "rewards/margins": 0.10079552233219147, | |
| "rewards/rejected": -1.1234548091888428, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 14.107067164134795, | |
| "learning_rate": 2.604166666666667e-07, | |
| "logits/chosen": -2.799668073654175, | |
| "logits/rejected": -2.7763009071350098, | |
| "logps/chosen": -1.0291378498077393, | |
| "logps/rejected": -1.086358904838562, | |
| "loss": 2.0092, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -1.0291378498077393, | |
| "rewards/margins": 0.0572209358215332, | |
| "rewards/rejected": -1.086358904838562, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 15.048800842708138, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.7735042572021484, | |
| "logits/rejected": -2.731670379638672, | |
| "logps/chosen": -0.9846929311752319, | |
| "logps/rejected": -1.0904910564422607, | |
| "loss": 2.0118, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.9846929311752319, | |
| "rewards/margins": 0.10579802095890045, | |
| "rewards/rejected": -1.0904910564422607, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 14.073259274006203, | |
| "learning_rate": 3.645833333333333e-07, | |
| "logits/chosen": -2.777062177658081, | |
| "logits/rejected": -2.690617084503174, | |
| "logps/chosen": -1.0319979190826416, | |
| "logps/rejected": -1.1718026399612427, | |
| "loss": 1.9903, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.0319979190826416, | |
| "rewards/margins": 0.13980472087860107, | |
| "rewards/rejected": -1.1718026399612427, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 19.704149826171133, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.7942392826080322, | |
| "logits/rejected": -2.697408437728882, | |
| "logps/chosen": -1.011974573135376, | |
| "logps/rejected": -1.2997562885284424, | |
| "loss": 1.9462, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.011974573135376, | |
| "rewards/margins": 0.28778162598609924, | |
| "rewards/rejected": -1.2997562885284424, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 15.255628868525834, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": -2.805002450942993, | |
| "logits/rejected": -2.7587850093841553, | |
| "logps/chosen": -0.963559627532959, | |
| "logps/rejected": -1.1423972845077515, | |
| "loss": 1.9511, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.963559627532959, | |
| "rewards/margins": 0.17883776128292084, | |
| "rewards/rejected": -1.1423972845077515, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 20.885465849075064, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -2.761730194091797, | |
| "logits/rejected": -2.693643093109131, | |
| "logps/chosen": -1.0126588344573975, | |
| "logps/rejected": -1.3037517070770264, | |
| "loss": 1.9653, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0126588344573975, | |
| "rewards/margins": 0.2910929024219513, | |
| "rewards/rejected": -1.3037517070770264, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 16.254626281156803, | |
| "learning_rate": 4.996716052911017e-07, | |
| "logits/chosen": -2.7590675354003906, | |
| "logits/rejected": -2.723374843597412, | |
| "logps/chosen": -1.0168770551681519, | |
| "logps/rejected": -1.268347978591919, | |
| "loss": 1.9612, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.0168770551681519, | |
| "rewards/margins": 0.25147074460983276, | |
| "rewards/rejected": -1.268347978591919, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 16.222896740457696, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -2.809765338897705, | |
| "logits/rejected": -2.7633471488952637, | |
| "logps/chosen": -1.0493232011795044, | |
| "logps/rejected": -1.2437658309936523, | |
| "loss": 1.9157, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0493232011795044, | |
| "rewards/margins": 0.1944427341222763, | |
| "rewards/rejected": -1.2437658309936523, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 15.145405551888192, | |
| "learning_rate": 4.980652179769217e-07, | |
| "logits/chosen": -2.8015551567077637, | |
| "logits/rejected": -2.6827645301818848, | |
| "logps/chosen": -1.0214020013809204, | |
| "logps/rejected": -1.5461061000823975, | |
| "loss": 1.9028, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.0214020013809204, | |
| "rewards/margins": 0.524704098701477, | |
| "rewards/rejected": -1.5461061000823975, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 20.7647978251983, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -2.6961591243743896, | |
| "logits/rejected": -2.6422698497772217, | |
| "logps/chosen": -1.0481222867965698, | |
| "logps/rejected": -1.3449742794036865, | |
| "loss": 1.9095, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.0481222867965698, | |
| "rewards/margins": 0.29685190320014954, | |
| "rewards/rejected": -1.3449742794036865, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 29.67977958863066, | |
| "learning_rate": 4.951291206355559e-07, | |
| "logits/chosen": -2.7507481575012207, | |
| "logits/rejected": -2.6747162342071533, | |
| "logps/chosen": -0.9930871725082397, | |
| "logps/rejected": -1.354509711265564, | |
| "loss": 1.8891, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.9930871725082397, | |
| "rewards/margins": 0.36142244935035706, | |
| "rewards/rejected": -1.354509711265564, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 22.015004964913427, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -2.709141254425049, | |
| "logits/rejected": -2.6435017585754395, | |
| "logps/chosen": -0.9975327253341675, | |
| "logps/rejected": -1.4075746536254883, | |
| "loss": 1.9339, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.9975327253341675, | |
| "rewards/margins": 0.4100419878959656, | |
| "rewards/rejected": -1.4075746536254883, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 22.42041956383337, | |
| "learning_rate": 4.908790517010636e-07, | |
| "logits/chosen": -2.7113754749298096, | |
| "logits/rejected": -2.6602649688720703, | |
| "logps/chosen": -0.9959508776664734, | |
| "logps/rejected": -1.3768701553344727, | |
| "loss": 1.9555, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.9959508776664734, | |
| "rewards/margins": 0.3809191882610321, | |
| "rewards/rejected": -1.3768701553344727, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 21.065770503793715, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -2.689040184020996, | |
| "logits/rejected": -2.6611132621765137, | |
| "logps/chosen": -1.0262080430984497, | |
| "logps/rejected": -1.4091228246688843, | |
| "loss": 1.8793, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0262080430984497, | |
| "rewards/margins": 0.38291481137275696, | |
| "rewards/rejected": -1.4091228246688843, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 31.8448042024589, | |
| "learning_rate": 4.853377929214243e-07, | |
| "logits/chosen": -2.7642765045166016, | |
| "logits/rejected": -2.6598925590515137, | |
| "logps/chosen": -1.0126625299453735, | |
| "logps/rejected": -1.3561102151870728, | |
| "loss": 1.9092, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.0126625299453735, | |
| "rewards/margins": 0.34344789385795593, | |
| "rewards/rejected": -1.3561102151870728, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 25.05097837239479, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -2.6988279819488525, | |
| "logits/rejected": -2.597813844680786, | |
| "logps/chosen": -1.0824997425079346, | |
| "logps/rejected": -1.5376174449920654, | |
| "loss": 1.9333, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.0824997425079346, | |
| "rewards/margins": 0.45511776208877563, | |
| "rewards/rejected": -1.5376174449920654, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 26.762526766885564, | |
| "learning_rate": 4.785350472409791e-07, | |
| "logits/chosen": -2.709327220916748, | |
| "logits/rejected": -2.6909279823303223, | |
| "logps/chosen": -0.9367496371269226, | |
| "logps/rejected": -1.3381364345550537, | |
| "loss": 1.8988, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.9367496371269226, | |
| "rewards/margins": 0.40138688683509827, | |
| "rewards/rejected": -1.3381364345550537, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 24.759397812808643, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -2.785827398300171, | |
| "logits/rejected": -2.7306325435638428, | |
| "logps/chosen": -1.0359995365142822, | |
| "logps/rejected": -1.4377586841583252, | |
| "loss": 1.8787, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.0359995365142822, | |
| "rewards/margins": 0.40175899863243103, | |
| "rewards/rejected": -1.4377586841583252, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 22.275954156633127, | |
| "learning_rate": 4.70507279583015e-07, | |
| "logits/chosen": -2.734710693359375, | |
| "logits/rejected": -2.67667818069458, | |
| "logps/chosen": -1.0577763319015503, | |
| "logps/rejected": -1.6186084747314453, | |
| "loss": 1.875, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.0577763319015503, | |
| "rewards/margins": 0.560832142829895, | |
| "rewards/rejected": -1.6186084747314453, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 31.253706152387053, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -2.747194766998291, | |
| "logits/rejected": -2.7159297466278076, | |
| "logps/chosen": -1.061531662940979, | |
| "logps/rejected": -1.4542839527130127, | |
| "loss": 1.8991, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.061531662940979, | |
| "rewards/margins": 0.39275234937667847, | |
| "rewards/rejected": -1.4542839527130127, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 24.012313672135768, | |
| "learning_rate": 4.612975213859487e-07, | |
| "logits/chosen": -2.809326648712158, | |
| "logits/rejected": -2.786160707473755, | |
| "logps/chosen": -1.0037329196929932, | |
| "logps/rejected": -1.2904164791107178, | |
| "loss": 1.9171, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.0037329196929932, | |
| "rewards/margins": 0.28668367862701416, | |
| "rewards/rejected": -1.2904164791107178, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 27.105267597546742, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -2.7714219093322754, | |
| "logits/rejected": -2.7451789379119873, | |
| "logps/chosen": -1.0449589490890503, | |
| "logps/rejected": -1.2745070457458496, | |
| "loss": 1.9054, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0449589490890503, | |
| "rewards/margins": 0.22954817116260529, | |
| "rewards/rejected": -1.2745070457458496, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 34.82290559167317, | |
| "learning_rate": 4.5095513994085974e-07, | |
| "logits/chosen": -2.783876419067383, | |
| "logits/rejected": -2.67991304397583, | |
| "logps/chosen": -1.1126677989959717, | |
| "logps/rejected": -1.8345043659210205, | |
| "loss": 1.8811, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -1.1126677989959717, | |
| "rewards/margins": 0.7218365669250488, | |
| "rewards/rejected": -1.8345043659210205, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 25.33210409531245, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -2.8026270866394043, | |
| "logits/rejected": -2.7463619709014893, | |
| "logps/chosen": -1.0125950574874878, | |
| "logps/rejected": -1.5290297269821167, | |
| "loss": 1.815, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.0125950574874878, | |
| "rewards/margins": 0.5164345502853394, | |
| "rewards/rejected": -1.5290297269821167, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 25.346225453192677, | |
| "learning_rate": 4.395355737667985e-07, | |
| "logits/chosen": -2.775475025177002, | |
| "logits/rejected": -2.719919204711914, | |
| "logps/chosen": -0.9933168292045593, | |
| "logps/rejected": -1.3926680088043213, | |
| "loss": 1.8662, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.9933168292045593, | |
| "rewards/margins": 0.3993511199951172, | |
| "rewards/rejected": -1.3926680088043213, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 25.05252990164211, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -2.8103270530700684, | |
| "logits/rejected": -2.7926547527313232, | |
| "logps/chosen": -0.9610552787780762, | |
| "logps/rejected": -1.3762198686599731, | |
| "loss": 1.8355, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -0.9610552787780762, | |
| "rewards/margins": 0.4151645302772522, | |
| "rewards/rejected": -1.3762198686599731, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 27.37807184294238, | |
| "learning_rate": 4.271000354423425e-07, | |
| "logits/chosen": -2.7627320289611816, | |
| "logits/rejected": -2.735795497894287, | |
| "logps/chosen": -1.078357458114624, | |
| "logps/rejected": -1.435006856918335, | |
| "loss": 1.8878, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.078357458114624, | |
| "rewards/margins": 0.3566494882106781, | |
| "rewards/rejected": -1.435006856918335, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 31.453776336135004, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -2.744175434112549, | |
| "logits/rejected": -2.718656063079834, | |
| "logps/chosen": -1.1089026927947998, | |
| "logps/rejected": -1.5105451345443726, | |
| "loss": 1.893, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.1089026927947998, | |
| "rewards/margins": 0.4016422629356384, | |
| "rewards/rejected": -1.5105451345443726, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 25.427022864790626, | |
| "learning_rate": 4.137151834863213e-07, | |
| "logits/chosen": -2.7565693855285645, | |
| "logits/rejected": -2.6773271560668945, | |
| "logps/chosen": -0.9760361909866333, | |
| "logps/rejected": -1.4186967611312866, | |
| "loss": 1.8368, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.9760361909866333, | |
| "rewards/margins": 0.4426605701446533, | |
| "rewards/rejected": -1.4186967611312866, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 25.0313872723157, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -2.7782142162323, | |
| "logits/rejected": -2.706545829772949, | |
| "logps/chosen": -1.0534151792526245, | |
| "logps/rejected": -1.450411081314087, | |
| "loss": 1.8695, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0534151792526245, | |
| "rewards/margins": 0.39699599146842957, | |
| "rewards/rejected": -1.450411081314087, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 22.99470630834903, | |
| "learning_rate": 3.994527650465352e-07, | |
| "logits/chosen": -2.7585906982421875, | |
| "logits/rejected": -2.6691181659698486, | |
| "logps/chosen": -1.0702944993972778, | |
| "logps/rejected": -1.4621968269348145, | |
| "loss": 1.8958, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0702944993972778, | |
| "rewards/margins": 0.39190229773521423, | |
| "rewards/rejected": -1.4621968269348145, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 25.69330171727391, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -2.740964889526367, | |
| "logits/rejected": -2.7247314453125, | |
| "logps/chosen": -1.0360349416732788, | |
| "logps/rejected": -1.5588700771331787, | |
| "loss": 1.8488, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0360349416732788, | |
| "rewards/margins": 0.5228349566459656, | |
| "rewards/rejected": -1.5588700771331787, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 26.206907546340837, | |
| "learning_rate": 3.8438923131177237e-07, | |
| "logits/chosen": -2.7401633262634277, | |
| "logits/rejected": -2.637202501296997, | |
| "logps/chosen": -1.0996856689453125, | |
| "logps/rejected": -1.7466999292373657, | |
| "loss": 1.8192, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.0996856689453125, | |
| "rewards/margins": 0.6470142602920532, | |
| "rewards/rejected": -1.7466999292373657, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 27.521588311939052, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": -2.743375301361084, | |
| "logits/rejected": -2.69993257522583, | |
| "logps/chosen": -1.086106300354004, | |
| "logps/rejected": -1.5462232828140259, | |
| "loss": 1.9254, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.086106300354004, | |
| "rewards/margins": 0.4601168632507324, | |
| "rewards/rejected": -1.5462232828140259, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 27.446157471379898, | |
| "learning_rate": 3.6860532770864005e-07, | |
| "logits/chosen": -2.684860944747925, | |
| "logits/rejected": -2.650524616241455, | |
| "logps/chosen": -1.0963340997695923, | |
| "logps/rejected": -1.4702621698379517, | |
| "loss": 1.8975, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0963340997695923, | |
| "rewards/margins": 0.3739281892776489, | |
| "rewards/rejected": -1.4702621698379517, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 24.602486055800032, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": -2.7188162803649902, | |
| "logits/rejected": -2.712054967880249, | |
| "logps/chosen": -1.0557594299316406, | |
| "logps/rejected": -1.4215106964111328, | |
| "loss": 1.8755, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.0557594299316406, | |
| "rewards/margins": 0.36575111746788025, | |
| "rewards/rejected": -1.4215106964111328, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 27.200176778512795, | |
| "learning_rate": 3.5218566107988867e-07, | |
| "logits/chosen": -2.6477105617523193, | |
| "logits/rejected": -2.5724520683288574, | |
| "logps/chosen": -0.9946632385253906, | |
| "logps/rejected": -1.4712902307510376, | |
| "loss": 1.8403, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.9946632385253906, | |
| "rewards/margins": 0.4766268730163574, | |
| "rewards/rejected": -1.4712902307510376, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 36.0559472403249, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -2.657755136489868, | |
| "logits/rejected": -2.61205792427063, | |
| "logps/chosen": -1.0976860523223877, | |
| "logps/rejected": -1.4535160064697266, | |
| "loss": 1.8803, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.0976860523223877, | |
| "rewards/margins": 0.35582998394966125, | |
| "rewards/rejected": -1.4535160064697266, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 22.419086184833272, | |
| "learning_rate": 3.3521824616429284e-07, | |
| "logits/chosen": -2.690389633178711, | |
| "logits/rejected": -2.669494152069092, | |
| "logps/chosen": -0.9983667135238647, | |
| "logps/rejected": -1.5212194919586182, | |
| "loss": 1.8208, | |
| "rewards/accuracies": 0.800000011920929, | |
| "rewards/chosen": -0.9983667135238647, | |
| "rewards/margins": 0.5228530168533325, | |
| "rewards/rejected": -1.5212194919586182, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 30.606607589341355, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -2.7263293266296387, | |
| "logits/rejected": -2.649702787399292, | |
| "logps/chosen": -1.0452382564544678, | |
| "logps/rejected": -1.5576632022857666, | |
| "loss": 1.8431, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0452382564544678, | |
| "rewards/margins": 0.5124249458312988, | |
| "rewards/rejected": -1.5576632022857666, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 27.672871069441197, | |
| "learning_rate": 3.1779403380910425e-07, | |
| "logits/chosen": -2.7160162925720215, | |
| "logits/rejected": -2.6121773719787598, | |
| "logps/chosen": -1.0075920820236206, | |
| "logps/rejected": -1.576601505279541, | |
| "loss": 1.8271, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.0075920820236206, | |
| "rewards/margins": 0.5690094232559204, | |
| "rewards/rejected": -1.576601505279541, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 25.08698759897845, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -2.7268662452697754, | |
| "logits/rejected": -2.6340160369873047, | |
| "logps/chosen": -1.0295268297195435, | |
| "logps/rejected": -1.7138006687164307, | |
| "loss": 1.8333, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.0295268297195435, | |
| "rewards/margins": 0.6842738389968872, | |
| "rewards/rejected": -1.7138006687164307, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 24.9727125758065, | |
| "learning_rate": 3.000064234440111e-07, | |
| "logits/chosen": -2.7256393432617188, | |
| "logits/rejected": -2.67500638961792, | |
| "logps/chosen": -1.0034698247909546, | |
| "logps/rejected": -1.5250154733657837, | |
| "loss": 1.8603, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.0034698247909546, | |
| "rewards/margins": 0.5215455889701843, | |
| "rewards/rejected": -1.5250154733657837, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 23.709236995530965, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -2.667459011077881, | |
| "logits/rejected": -2.622892379760742, | |
| "logps/chosen": -0.998893141746521, | |
| "logps/rejected": -1.5178442001342773, | |
| "loss": 1.8707, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.998893141746521, | |
| "rewards/margins": 0.5189512372016907, | |
| "rewards/rejected": -1.5178442001342773, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 23.55356072022147, | |
| "learning_rate": 2.8195076242990116e-07, | |
| "logits/chosen": -2.6935830116271973, | |
| "logits/rejected": -2.5973896980285645, | |
| "logps/chosen": -1.1026113033294678, | |
| "logps/rejected": -1.5293716192245483, | |
| "loss": 1.8533, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.1026113033294678, | |
| "rewards/margins": 0.4267602562904358, | |
| "rewards/rejected": -1.5293716192245483, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 33.32048093164131, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -2.5986247062683105, | |
| "logits/rejected": -2.545224905014038, | |
| "logps/chosen": -1.019071340560913, | |
| "logps/rejected": -1.5590027570724487, | |
| "loss": 1.8581, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.019071340560913, | |
| "rewards/margins": 0.5399314761161804, | |
| "rewards/rejected": -1.5590027570724487, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 24.969288713955503, | |
| "learning_rate": 2.6372383496608186e-07, | |
| "logits/chosen": -2.647831678390503, | |
| "logits/rejected": -2.5525565147399902, | |
| "logps/chosen": -1.0960899591445923, | |
| "logps/rejected": -1.7084300518035889, | |
| "loss": 1.7984, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.0960899591445923, | |
| "rewards/margins": 0.6123402714729309, | |
| "rewards/rejected": -1.7084300518035889, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 24.978593158166646, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -2.6099724769592285, | |
| "logits/rejected": -2.503551959991455, | |
| "logps/chosen": -0.94499671459198, | |
| "logps/rejected": -1.6384252309799194, | |
| "loss": 1.8396, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.94499671459198, | |
| "rewards/margins": 0.6934284567832947, | |
| "rewards/rejected": -1.6384252309799194, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 27.898305291163492, | |
| "learning_rate": 2.454233432955807e-07, | |
| "logits/chosen": -2.661271095275879, | |
| "logits/rejected": -2.6026952266693115, | |
| "logps/chosen": -1.0296385288238525, | |
| "logps/rejected": -1.382299780845642, | |
| "loss": 1.8101, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.0296385288238525, | |
| "rewards/margins": 0.3526611924171448, | |
| "rewards/rejected": -1.382299780845642, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 35.74021390544709, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": -2.6500651836395264, | |
| "logits/rejected": -2.5953967571258545, | |
| "logps/chosen": -1.1358740329742432, | |
| "logps/rejected": -1.5125305652618408, | |
| "loss": 1.8849, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.1358740329742432, | |
| "rewards/margins": 0.376656711101532, | |
| "rewards/rejected": -1.5125305652618408, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 25.432184339952187, | |
| "learning_rate": 2.2714738398943308e-07, | |
| "logits/chosen": -2.7251439094543457, | |
| "logits/rejected": -2.6094861030578613, | |
| "logps/chosen": -1.018061637878418, | |
| "logps/rejected": -1.627947449684143, | |
| "loss": 1.8328, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.018061637878418, | |
| "rewards/margins": 0.6098858118057251, | |
| "rewards/rejected": -1.627947449684143, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 23.87099932259008, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": -2.746220827102661, | |
| "logits/rejected": -2.6774742603302, | |
| "logps/chosen": -0.9781770706176758, | |
| "logps/rejected": -1.3979579210281372, | |
| "loss": 1.8368, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.9781770706176758, | |
| "rewards/margins": 0.4197807312011719, | |
| "rewards/rejected": -1.3979579210281372, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 26.340595120664386, | |
| "learning_rate": 2.089939221172446e-07, | |
| "logits/chosen": -2.6835036277770996, | |
| "logits/rejected": -2.6886842250823975, | |
| "logps/chosen": -1.0875818729400635, | |
| "logps/rejected": -1.583901047706604, | |
| "loss": 1.8233, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.0875818729400635, | |
| "rewards/margins": 0.49631914496421814, | |
| "rewards/rejected": -1.583901047706604, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 29.695224998335657, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": -2.7280991077423096, | |
| "logits/rejected": -2.6861159801483154, | |
| "logps/chosen": -1.0748018026351929, | |
| "logps/rejected": -1.6494481563568115, | |
| "loss": 1.8022, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.0748018026351929, | |
| "rewards/margins": 0.5746463537216187, | |
| "rewards/rejected": -1.6494481563568115, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 29.67617284412597, | |
| "learning_rate": 1.9106026612264315e-07, | |
| "logits/chosen": -2.7601213455200195, | |
| "logits/rejected": -2.6855008602142334, | |
| "logps/chosen": -1.0902409553527832, | |
| "logps/rejected": -1.6281505823135376, | |
| "loss": 1.8578, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.0902409553527832, | |
| "rewards/margins": 0.5379096269607544, | |
| "rewards/rejected": -1.6281505823135376, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 27.162465615999373, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": -2.66314435005188, | |
| "logits/rejected": -2.6144497394561768, | |
| "logps/chosen": -1.0360552072525024, | |
| "logps/rejected": -1.5152963399887085, | |
| "loss": 1.8344, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.0360552072525024, | |
| "rewards/margins": 0.47924113273620605, | |
| "rewards/rejected": -1.5152963399887085, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 31.001026793963305, | |
| "learning_rate": 1.7344254621846017e-07, | |
| "logits/chosen": -2.6350555419921875, | |
| "logits/rejected": -2.5956974029541016, | |
| "logps/chosen": -1.0529329776763916, | |
| "logps/rejected": -1.6219732761383057, | |
| "loss": 1.8196, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0529329776763916, | |
| "rewards/margins": 0.5690402388572693, | |
| "rewards/rejected": -1.6219732761383057, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 27.186267806704134, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": -2.646355390548706, | |
| "logits/rejected": -2.649137258529663, | |
| "logps/chosen": -0.9727001190185547, | |
| "logps/rejected": -1.5282304286956787, | |
| "loss": 1.8154, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -0.9727001190185547, | |
| "rewards/margins": 0.5555303692817688, | |
| "rewards/rejected": -1.5282304286956787, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 33.037660975429695, | |
| "learning_rate": 1.562351990976095e-07, | |
| "logits/chosen": -2.666133403778076, | |
| "logits/rejected": -2.6141715049743652, | |
| "logps/chosen": -0.9768770933151245, | |
| "logps/rejected": -1.6563736200332642, | |
| "loss": 1.819, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -0.9768770933151245, | |
| "rewards/margins": 0.6794965863227844, | |
| "rewards/rejected": -1.6563736200332642, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 37.124618715177, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": -2.6524291038513184, | |
| "logits/rejected": -2.6353344917297363, | |
| "logps/chosen": -1.0379709005355835, | |
| "logps/rejected": -1.5560357570648193, | |
| "loss": 1.7814, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.0379709005355835, | |
| "rewards/margins": 0.5180650353431702, | |
| "rewards/rejected": -1.5560357570648193, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 29.546811615709533, | |
| "learning_rate": 1.3953046172178413e-07, | |
| "logits/chosen": -2.6703402996063232, | |
| "logits/rejected": -2.5877861976623535, | |
| "logps/chosen": -1.0295711755752563, | |
| "logps/rejected": -1.6383397579193115, | |
| "loss": 1.7933, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0295711755752563, | |
| "rewards/margins": 0.6087687611579895, | |
| "rewards/rejected": -1.6383397579193115, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 27.130577131118542, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": -2.6374616622924805, | |
| "logits/rejected": -2.5912885665893555, | |
| "logps/chosen": -1.0082510709762573, | |
| "logps/rejected": -1.3762824535369873, | |
| "loss": 1.8378, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0082510709762573, | |
| "rewards/margins": 0.36803150177001953, | |
| "rewards/rejected": -1.3762824535369873, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 26.495064774873583, | |
| "learning_rate": 1.2341787690142435e-07, | |
| "logits/chosen": -2.751105308532715, | |
| "logits/rejected": -2.673309564590454, | |
| "logps/chosen": -0.9556055068969727, | |
| "logps/rejected": -1.5157105922698975, | |
| "loss": 1.8293, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.9556055068969727, | |
| "rewards/margins": 0.56010502576828, | |
| "rewards/rejected": -1.5157105922698975, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 26.629021066262204, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": -2.6683545112609863, | |
| "logits/rejected": -2.600292921066284, | |
| "logps/chosen": -1.1952508687973022, | |
| "logps/rejected": -1.6229918003082275, | |
| "loss": 1.8878, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.1952508687973022, | |
| "rewards/margins": 0.4277407228946686, | |
| "rewards/rejected": -1.6229918003082275, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 34.902511984640945, | |
| "learning_rate": 1.0798381331721107e-07, | |
| "logits/chosen": -2.702152729034424, | |
| "logits/rejected": -2.656289577484131, | |
| "logps/chosen": -1.0419316291809082, | |
| "logps/rejected": -1.6374496221542358, | |
| "loss": 1.8455, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.0419316291809082, | |
| "rewards/margins": 0.5955179929733276, | |
| "rewards/rejected": -1.6374496221542358, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 30.383520283042905, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": -2.6637139320373535, | |
| "logits/rejected": -2.640423536300659, | |
| "logps/chosen": -1.0635532140731812, | |
| "logps/rejected": -1.55173659324646, | |
| "loss": 1.8762, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.0635532140731812, | |
| "rewards/margins": 0.4881834387779236, | |
| "rewards/rejected": -1.55173659324646, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 28.202614006747883, | |
| "learning_rate": 9.331100255592436e-08, | |
| "logits/chosen": -2.672135591506958, | |
| "logits/rejected": -2.614642858505249, | |
| "logps/chosen": -0.9285499453544617, | |
| "logps/rejected": -1.6109098196029663, | |
| "loss": 1.7749, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -0.9285499453544617, | |
| "rewards/margins": 0.6823598146438599, | |
| "rewards/rejected": -1.6109098196029663, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 23.48852239281184, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -2.701026439666748, | |
| "logits/rejected": -2.645697832107544, | |
| "logps/chosen": -0.9583543539047241, | |
| "logps/rejected": -1.4256336688995361, | |
| "loss": 1.7696, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.9583543539047241, | |
| "rewards/margins": 0.46727919578552246, | |
| "rewards/rejected": -1.4256336688995361, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 29.4764609142287, | |
| "learning_rate": 7.947809564230445e-08, | |
| "logits/chosen": -2.6285886764526367, | |
| "logits/rejected": -2.5465035438537598, | |
| "logps/chosen": -1.07431960105896, | |
| "logps/rejected": -1.6673939228057861, | |
| "loss": 1.854, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.07431960105896, | |
| "rewards/margins": 0.5930741429328918, | |
| "rewards/rejected": -1.6673939228057861, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 32.53243633682995, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": -2.672539234161377, | |
| "logits/rejected": -2.612757682800293, | |
| "logps/chosen": -1.0804438591003418, | |
| "logps/rejected": -1.7286722660064697, | |
| "loss": 1.8244, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.0804438591003418, | |
| "rewards/margins": 0.6482284069061279, | |
| "rewards/rejected": -1.7286722660064697, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 26.086052446415493, | |
| "learning_rate": 6.655924144404906e-08, | |
| "logits/chosen": -2.656698226928711, | |
| "logits/rejected": -2.681257724761963, | |
| "logps/chosen": -1.0645740032196045, | |
| "logps/rejected": -1.5776240825653076, | |
| "loss": 1.8108, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.0645740032196045, | |
| "rewards/margins": 0.5130501985549927, | |
| "rewards/rejected": -1.5776240825653076, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 29.301563375037563, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": -2.6668169498443604, | |
| "logits/rejected": -2.6132140159606934, | |
| "logps/chosen": -1.0790109634399414, | |
| "logps/rejected": -1.5554571151733398, | |
| "loss": 1.782, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0790109634399414, | |
| "rewards/margins": 0.47644609212875366, | |
| "rewards/rejected": -1.5554571151733398, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 24.668280208736856, | |
| "learning_rate": 5.4623689209832484e-08, | |
| "logits/chosen": -2.677908420562744, | |
| "logits/rejected": -2.607048511505127, | |
| "logps/chosen": -1.0273312330245972, | |
| "logps/rejected": -1.5443226099014282, | |
| "loss": 1.7947, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0273312330245972, | |
| "rewards/margins": 0.5169912576675415, | |
| "rewards/rejected": -1.5443226099014282, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 27.191462805016396, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": -2.7038891315460205, | |
| "logits/rejected": -2.6414623260498047, | |
| "logps/chosen": -1.0105482339859009, | |
| "logps/rejected": -1.7210209369659424, | |
| "loss": 1.7734, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.0105482339859009, | |
| "rewards/margins": 0.7104724645614624, | |
| "rewards/rejected": -1.7210209369659424, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 31.43323955162571, | |
| "learning_rate": 4.373541737087263e-08, | |
| "logits/chosen": -2.636841058731079, | |
| "logits/rejected": -2.556833267211914, | |
| "logps/chosen": -1.0063461065292358, | |
| "logps/rejected": -1.5682775974273682, | |
| "loss": 1.8075, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -1.0063461065292358, | |
| "rewards/margins": 0.5619313716888428, | |
| "rewards/rejected": -1.5682775974273682, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 39.225726689241675, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": -2.6019065380096436, | |
| "logits/rejected": -2.5374197959899902, | |
| "logps/chosen": -0.9893720746040344, | |
| "logps/rejected": -1.5351721048355103, | |
| "loss": 1.8109, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -0.9893720746040344, | |
| "rewards/margins": 0.5458000898361206, | |
| "rewards/rejected": -1.5351721048355103, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 42.21715206184894, | |
| "learning_rate": 3.3952790595787986e-08, | |
| "logits/chosen": -2.6662096977233887, | |
| "logits/rejected": -2.62931489944458, | |
| "logps/chosen": -1.0822862386703491, | |
| "logps/rejected": -1.6576595306396484, | |
| "loss": 1.8405, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0822862386703491, | |
| "rewards/margins": 0.5753732323646545, | |
| "rewards/rejected": -1.6576595306396484, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 35.10610264317854, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": -2.678342819213867, | |
| "logits/rejected": -2.634387493133545, | |
| "logps/chosen": -0.9690961837768555, | |
| "logps/rejected": -1.4596083164215088, | |
| "loss": 1.8145, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.9690961837768555, | |
| "rewards/margins": 0.4905119836330414, | |
| "rewards/rejected": -1.4596083164215088, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 27.87180494304035, | |
| "learning_rate": 2.5328246937043525e-08, | |
| "logits/chosen": -2.702768564224243, | |
| "logits/rejected": -2.636719226837158, | |
| "logps/chosen": -1.0192323923110962, | |
| "logps/rejected": -1.6251684427261353, | |
| "loss": 1.77, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0192323923110962, | |
| "rewards/margins": 0.6059359312057495, | |
| "rewards/rejected": -1.6251684427261353, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 25.60878775790235, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": -2.6637589931488037, | |
| "logits/rejected": -2.6058096885681152, | |
| "logps/chosen": -1.0046476125717163, | |
| "logps/rejected": -1.5859066247940063, | |
| "loss": 1.8408, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0046476125717163, | |
| "rewards/margins": 0.5812588930130005, | |
| "rewards/rejected": -1.5859066247940063, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 34.39189970937955, | |
| "learning_rate": 1.7908016745981856e-08, | |
| "logits/chosen": -2.6160888671875, | |
| "logits/rejected": -2.5980353355407715, | |
| "logps/chosen": -1.1391981840133667, | |
| "logps/rejected": -1.5944236516952515, | |
| "loss": 1.8251, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1391981840133667, | |
| "rewards/margins": 0.4552256464958191, | |
| "rewards/rejected": -1.5944236516952515, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 33.16075597765061, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": -2.6214401721954346, | |
| "logits/rejected": -2.570112705230713, | |
| "logps/chosen": -1.0948253870010376, | |
| "logps/rejected": -1.6313488483428955, | |
| "loss": 1.8008, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0948253870010376, | |
| "rewards/margins": 0.5365234613418579, | |
| "rewards/rejected": -1.6313488483428955, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 33.93315468108953, | |
| "learning_rate": 1.1731874863145142e-08, | |
| "logits/chosen": -2.6755282878875732, | |
| "logits/rejected": -2.637303590774536, | |
| "logps/chosen": -1.0409172773361206, | |
| "logps/rejected": -1.7293403148651123, | |
| "loss": 1.8172, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.0409172773361206, | |
| "rewards/margins": 0.6884229779243469, | |
| "rewards/rejected": -1.7293403148651123, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 32.567789743092646, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": -2.617215871810913, | |
| "logits/rejected": -2.5979318618774414, | |
| "logps/chosen": -1.0106241703033447, | |
| "logps/rejected": -1.5670998096466064, | |
| "loss": 1.8428, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.0106241703033447, | |
| "rewards/margins": 0.5564756989479065, | |
| "rewards/rejected": -1.5670998096466064, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 30.42876501735674, | |
| "learning_rate": 6.832927412229017e-09, | |
| "logits/chosen": -2.681229591369629, | |
| "logits/rejected": -2.612400531768799, | |
| "logps/chosen": -1.06952965259552, | |
| "logps/rejected": -1.5481148958206177, | |
| "loss": 1.8604, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.06952965259552, | |
| "rewards/margins": 0.4785851836204529, | |
| "rewards/rejected": -1.5481148958206177, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 27.381238378090455, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": -2.579571485519409, | |
| "logits/rejected": -2.544320821762085, | |
| "logps/chosen": -1.0337889194488525, | |
| "logps/rejected": -1.5412318706512451, | |
| "loss": 1.7854, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.0337889194488525, | |
| "rewards/margins": 0.5074428915977478, | |
| "rewards/rejected": -1.5412318706512451, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 31.78517016150814, | |
| "learning_rate": 3.2374343405217884e-09, | |
| "logits/chosen": -2.6410415172576904, | |
| "logits/rejected": -2.5741779804229736, | |
| "logps/chosen": -1.1349095106124878, | |
| "logps/rejected": -1.5367697477340698, | |
| "loss": 1.8237, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1349095106124878, | |
| "rewards/margins": 0.40186023712158203, | |
| "rewards/rejected": -1.5367697477340698, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 29.1380242330253, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": -2.6333887577056885, | |
| "logits/rejected": -2.594615936279297, | |
| "logps/chosen": -0.9948427081108093, | |
| "logps/rejected": -1.598851203918457, | |
| "loss": 1.7821, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.9948427081108093, | |
| "rewards/margins": 0.6040085554122925, | |
| "rewards/rejected": -1.598851203918457, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 34.467412911376215, | |
| "learning_rate": 9.64668657069706e-10, | |
| "logits/chosen": -2.69240140914917, | |
| "logits/rejected": -2.6058619022369385, | |
| "logps/chosen": -1.087109923362732, | |
| "logps/rejected": -1.6203782558441162, | |
| "loss": 1.8436, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.087109923362732, | |
| "rewards/margins": 0.5332685112953186, | |
| "rewards/rejected": -1.6203782558441162, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 27.439754854696236, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": -2.6679489612579346, | |
| "logits/rejected": -2.603280544281006, | |
| "logps/chosen": -0.9998431205749512, | |
| "logps/rejected": -1.458508849143982, | |
| "loss": 1.8712, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.9998431205749512, | |
| "rewards/margins": 0.45866554975509644, | |
| "rewards/rejected": -1.458508849143982, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 31.225346161878143, | |
| "learning_rate": 2.6813123097352287e-11, | |
| "logits/chosen": -2.714961290359497, | |
| "logits/rejected": -2.6193995475769043, | |
| "logps/chosen": -1.041459083557129, | |
| "logps/rejected": -1.5316205024719238, | |
| "loss": 1.8186, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.041459083557129, | |
| "rewards/margins": 0.4901614189147949, | |
| "rewards/rejected": -1.5316205024719238, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 1.8633759126723188, | |
| "train_runtime": 7578.2385, | |
| "train_samples_per_second": 8.067, | |
| "train_steps_per_second": 0.063 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000000, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |