| { | |
| "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": 12.869433218842474, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": -2.7303454875946045, | |
| "logits/rejected": -2.6748926639556885, | |
| "logps/chosen": -1.0383304357528687, | |
| "logps/rejected": -1.093923807144165, | |
| "loss": 2.0262, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -1.0383304357528687, | |
| "rewards/margins": 0.05559344217181206, | |
| "rewards/rejected": -1.093923807144165, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 12.963970831519292, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.8506417274475098, | |
| "logits/rejected": -2.7654807567596436, | |
| "logps/chosen": -1.025138020515442, | |
| "logps/rejected": -1.1677943468093872, | |
| "loss": 2.0292, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.025138020515442, | |
| "rewards/margins": 0.1426563411951065, | |
| "rewards/rejected": -1.1677943468093872, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 15.099715830730236, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": -2.77595853805542, | |
| "logits/rejected": -2.7310924530029297, | |
| "logps/chosen": -1.0620777606964111, | |
| "logps/rejected": -1.1130949258804321, | |
| "loss": 2.0243, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.0620777606964111, | |
| "rewards/margins": 0.051017045974731445, | |
| "rewards/rejected": -1.1130949258804321, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 14.125874639614404, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.798926830291748, | |
| "logits/rejected": -2.6963000297546387, | |
| "logps/chosen": -1.026979684829712, | |
| "logps/rejected": -1.1289321184158325, | |
| "loss": 2.0224, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -1.026979684829712, | |
| "rewards/margins": 0.1019524484872818, | |
| "rewards/rejected": -1.1289321184158325, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 12.739023934080043, | |
| "learning_rate": 2.604166666666667e-07, | |
| "logits/chosen": -2.797764301300049, | |
| "logits/rejected": -2.6895253658294678, | |
| "logps/chosen": -0.9821299314498901, | |
| "logps/rejected": -1.1730941534042358, | |
| "loss": 1.9952, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.9821299314498901, | |
| "rewards/margins": 0.19096432626247406, | |
| "rewards/rejected": -1.1730941534042358, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 18.287014857226495, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.7723119258880615, | |
| "logits/rejected": -2.7337851524353027, | |
| "logps/chosen": -1.0000275373458862, | |
| "logps/rejected": -1.0346484184265137, | |
| "loss": 1.9952, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.0000275373458862, | |
| "rewards/margins": 0.034620799124240875, | |
| "rewards/rejected": -1.0346484184265137, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 17.8360996219737, | |
| "learning_rate": 3.645833333333333e-07, | |
| "logits/chosen": -2.7883362770080566, | |
| "logits/rejected": -2.715696334838867, | |
| "logps/chosen": -1.0128945112228394, | |
| "logps/rejected": -1.225642442703247, | |
| "loss": 1.9748, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.0128945112228394, | |
| "rewards/margins": 0.2127479612827301, | |
| "rewards/rejected": -1.225642442703247, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 18.333373463486794, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.7703633308410645, | |
| "logits/rejected": -2.6918962001800537, | |
| "logps/chosen": -1.0058094263076782, | |
| "logps/rejected": -1.131239891052246, | |
| "loss": 1.9619, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.0058094263076782, | |
| "rewards/margins": 0.12543055415153503, | |
| "rewards/rejected": -1.131239891052246, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 15.971248778309374, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": -2.732913017272949, | |
| "logits/rejected": -2.7122368812561035, | |
| "logps/chosen": -0.9598785638809204, | |
| "logps/rejected": -1.1564022302627563, | |
| "loss": 1.9542, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -0.9598785638809204, | |
| "rewards/margins": 0.196523517370224, | |
| "rewards/rejected": -1.1564022302627563, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 23.366392169452176, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -2.672335624694824, | |
| "logits/rejected": -2.6771888732910156, | |
| "logps/chosen": -1.0208399295806885, | |
| "logps/rejected": -1.2945979833602905, | |
| "loss": 1.9581, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.0208399295806885, | |
| "rewards/margins": 0.2737579941749573, | |
| "rewards/rejected": -1.2945979833602905, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 19.929116348471684, | |
| "learning_rate": 4.996716052911017e-07, | |
| "logits/chosen": -2.7723307609558105, | |
| "logits/rejected": -2.747865676879883, | |
| "logps/chosen": -1.0856590270996094, | |
| "logps/rejected": -1.219327449798584, | |
| "loss": 1.9648, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.0856590270996094, | |
| "rewards/margins": 0.13366840779781342, | |
| "rewards/rejected": -1.219327449798584, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 23.73282923950602, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -2.7040231227874756, | |
| "logits/rejected": -2.6466102600097656, | |
| "logps/chosen": -1.0688912868499756, | |
| "logps/rejected": -1.260387659072876, | |
| "loss": 1.9501, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -1.0688912868499756, | |
| "rewards/margins": 0.19149653613567352, | |
| "rewards/rejected": -1.260387659072876, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 18.942833820850367, | |
| "learning_rate": 4.980652179769217e-07, | |
| "logits/chosen": -2.716078519821167, | |
| "logits/rejected": -2.7220449447631836, | |
| "logps/chosen": -1.027010440826416, | |
| "logps/rejected": -1.2260963916778564, | |
| "loss": 1.9536, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.027010440826416, | |
| "rewards/margins": 0.199085995554924, | |
| "rewards/rejected": -1.2260963916778564, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 22.28146945050057, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -2.7145752906799316, | |
| "logits/rejected": -2.6811351776123047, | |
| "logps/chosen": -1.074029564857483, | |
| "logps/rejected": -1.3044393062591553, | |
| "loss": 1.9325, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.074029564857483, | |
| "rewards/margins": 0.2304096519947052, | |
| "rewards/rejected": -1.3044393062591553, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 30.956539188115524, | |
| "learning_rate": 4.951291206355559e-07, | |
| "logits/chosen": -2.640056610107422, | |
| "logits/rejected": -2.5216145515441895, | |
| "logps/chosen": -1.0524249076843262, | |
| "logps/rejected": -1.358999490737915, | |
| "loss": 1.9439, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0524249076843262, | |
| "rewards/margins": 0.3065745234489441, | |
| "rewards/rejected": -1.358999490737915, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 19.173789722918567, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -2.5955734252929688, | |
| "logits/rejected": -2.5555953979492188, | |
| "logps/chosen": -1.0017749071121216, | |
| "logps/rejected": -1.4020613431930542, | |
| "loss": 1.9082, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0017749071121216, | |
| "rewards/margins": 0.4002866744995117, | |
| "rewards/rejected": -1.4020613431930542, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 26.77907962647508, | |
| "learning_rate": 4.908790517010636e-07, | |
| "logits/chosen": -2.730196475982666, | |
| "logits/rejected": -2.6316986083984375, | |
| "logps/chosen": -1.0233409404754639, | |
| "logps/rejected": -1.4025558233261108, | |
| "loss": 1.9293, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0233409404754639, | |
| "rewards/margins": 0.3792147934436798, | |
| "rewards/rejected": -1.4025558233261108, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 20.94451147733862, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -2.7333931922912598, | |
| "logits/rejected": -2.594924211502075, | |
| "logps/chosen": -1.0183618068695068, | |
| "logps/rejected": -1.4752748012542725, | |
| "loss": 1.903, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0183618068695068, | |
| "rewards/margins": 0.4569129943847656, | |
| "rewards/rejected": -1.4752748012542725, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 27.23027264774861, | |
| "learning_rate": 4.853377929214243e-07, | |
| "logits/chosen": -2.6479899883270264, | |
| "logits/rejected": -2.612231492996216, | |
| "logps/chosen": -0.9796420335769653, | |
| "logps/rejected": -1.2683651447296143, | |
| "loss": 1.9133, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.9796420335769653, | |
| "rewards/margins": 0.28872328996658325, | |
| "rewards/rejected": -1.2683651447296143, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 20.647765632695478, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -2.7929205894470215, | |
| "logits/rejected": -2.730536460876465, | |
| "logps/chosen": -1.0671679973602295, | |
| "logps/rejected": -1.3122754096984863, | |
| "loss": 1.9562, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -1.0671679973602295, | |
| "rewards/margins": 0.24510732293128967, | |
| "rewards/rejected": -1.3122754096984863, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 19.882099174165575, | |
| "learning_rate": 4.785350472409791e-07, | |
| "logits/chosen": -2.7290585041046143, | |
| "logits/rejected": -2.661957263946533, | |
| "logps/chosen": -1.1218279600143433, | |
| "logps/rejected": -1.4676848649978638, | |
| "loss": 1.8476, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.1218279600143433, | |
| "rewards/margins": 0.3458569347858429, | |
| "rewards/rejected": -1.4676848649978638, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 29.356563953547646, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -2.677428722381592, | |
| "logits/rejected": -2.587128162384033, | |
| "logps/chosen": -1.0471330881118774, | |
| "logps/rejected": -1.3834892511367798, | |
| "loss": 1.8785, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.0471330881118774, | |
| "rewards/margins": 0.33635613322257996, | |
| "rewards/rejected": -1.3834892511367798, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 22.70599557212694, | |
| "learning_rate": 4.70507279583015e-07, | |
| "logits/chosen": -2.699206590652466, | |
| "logits/rejected": -2.6752054691314697, | |
| "logps/chosen": -1.037027359008789, | |
| "logps/rejected": -1.2504864931106567, | |
| "loss": 1.8781, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.037027359008789, | |
| "rewards/margins": 0.21345916390419006, | |
| "rewards/rejected": -1.2504864931106567, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 22.718720973130083, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -2.7028520107269287, | |
| "logits/rejected": -2.581420660018921, | |
| "logps/chosen": -0.958320140838623, | |
| "logps/rejected": -1.4017829895019531, | |
| "loss": 1.8894, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.958320140838623, | |
| "rewards/margins": 0.44346290826797485, | |
| "rewards/rejected": -1.4017829895019531, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 27.839637111582167, | |
| "learning_rate": 4.612975213859487e-07, | |
| "logits/chosen": -2.5790719985961914, | |
| "logits/rejected": -2.534557342529297, | |
| "logps/chosen": -1.0435621738433838, | |
| "logps/rejected": -1.4774357080459595, | |
| "loss": 1.9045, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.0435621738433838, | |
| "rewards/margins": 0.43387335538864136, | |
| "rewards/rejected": -1.4774357080459595, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 24.64049193676725, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -2.6262755393981934, | |
| "logits/rejected": -2.5886640548706055, | |
| "logps/chosen": -1.1033378839492798, | |
| "logps/rejected": -1.5101052522659302, | |
| "loss": 1.8774, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1033378839492798, | |
| "rewards/margins": 0.40676751732826233, | |
| "rewards/rejected": -1.5101052522659302, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 24.054739646210816, | |
| "learning_rate": 4.5095513994085974e-07, | |
| "logits/chosen": -2.5852103233337402, | |
| "logits/rejected": -2.5114357471466064, | |
| "logps/chosen": -1.036170244216919, | |
| "logps/rejected": -1.5170302391052246, | |
| "loss": 1.8964, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.036170244216919, | |
| "rewards/margins": 0.4808600842952728, | |
| "rewards/rejected": -1.5170302391052246, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 32.17420726348548, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -2.5846521854400635, | |
| "logits/rejected": -2.540843963623047, | |
| "logps/chosen": -1.1467993259429932, | |
| "logps/rejected": -1.3851490020751953, | |
| "loss": 1.958, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -1.1467993259429932, | |
| "rewards/margins": 0.23834970593452454, | |
| "rewards/rejected": -1.3851490020751953, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 31.293592588837402, | |
| "learning_rate": 4.395355737667985e-07, | |
| "logits/chosen": -2.5842862129211426, | |
| "logits/rejected": -2.481231451034546, | |
| "logps/chosen": -1.019995927810669, | |
| "logps/rejected": -1.450305700302124, | |
| "loss": 1.8621, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.019995927810669, | |
| "rewards/margins": 0.43030983209609985, | |
| "rewards/rejected": -1.450305700302124, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 31.395757438350795, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -2.6442532539367676, | |
| "logits/rejected": -2.5464065074920654, | |
| "logps/chosen": -1.066561222076416, | |
| "logps/rejected": -1.4769625663757324, | |
| "loss": 1.8856, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.066561222076416, | |
| "rewards/margins": 0.4104014039039612, | |
| "rewards/rejected": -1.4769625663757324, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 54.99577539115471, | |
| "learning_rate": 4.271000354423425e-07, | |
| "logits/chosen": -2.584909200668335, | |
| "logits/rejected": -2.5346925258636475, | |
| "logps/chosen": -1.1342883110046387, | |
| "logps/rejected": -1.709241509437561, | |
| "loss": 1.8796, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.1342883110046387, | |
| "rewards/margins": 0.5749532580375671, | |
| "rewards/rejected": -1.709241509437561, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 62.43588869532461, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -2.5957446098327637, | |
| "logits/rejected": -2.523685932159424, | |
| "logps/chosen": -1.048396348953247, | |
| "logps/rejected": -1.6725311279296875, | |
| "loss": 1.8954, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.048396348953247, | |
| "rewards/margins": 0.6241349577903748, | |
| "rewards/rejected": -1.6725311279296875, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 33.35503026194792, | |
| "learning_rate": 4.137151834863213e-07, | |
| "logits/chosen": -2.730977773666382, | |
| "logits/rejected": -2.645996570587158, | |
| "logps/chosen": -1.0159308910369873, | |
| "logps/rejected": -1.359212040901184, | |
| "loss": 1.8352, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.0159308910369873, | |
| "rewards/margins": 0.343281090259552, | |
| "rewards/rejected": -1.359212040901184, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 32.04170211845891, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -2.7265422344207764, | |
| "logits/rejected": -2.597689151763916, | |
| "logps/chosen": -0.9698517918586731, | |
| "logps/rejected": -1.3197566270828247, | |
| "loss": 1.8688, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.9698517918586731, | |
| "rewards/margins": 0.349904865026474, | |
| "rewards/rejected": -1.3197566270828247, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 27.34421984947454, | |
| "learning_rate": 3.994527650465352e-07, | |
| "logits/chosen": -2.652785062789917, | |
| "logits/rejected": -2.5916075706481934, | |
| "logps/chosen": -0.9570480585098267, | |
| "logps/rejected": -1.3661720752716064, | |
| "loss": 1.8595, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.9570480585098267, | |
| "rewards/margins": 0.409123957157135, | |
| "rewards/rejected": -1.3661720752716064, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 35.429346147889426, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -2.6718568801879883, | |
| "logits/rejected": -2.613348960876465, | |
| "logps/chosen": -1.037642240524292, | |
| "logps/rejected": -1.553224802017212, | |
| "loss": 1.8512, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.037642240524292, | |
| "rewards/margins": 0.5155824422836304, | |
| "rewards/rejected": -1.553224802017212, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 21.700575234978658, | |
| "learning_rate": 3.8438923131177237e-07, | |
| "logits/chosen": -2.6914803981781006, | |
| "logits/rejected": -2.6440558433532715, | |
| "logps/chosen": -0.9652795791625977, | |
| "logps/rejected": -1.3787884712219238, | |
| "loss": 1.8051, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.9652795791625977, | |
| "rewards/margins": 0.41350871324539185, | |
| "rewards/rejected": -1.3787884712219238, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 32.670099112476585, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": -2.6735928058624268, | |
| "logits/rejected": -2.640225648880005, | |
| "logps/chosen": -1.0144473314285278, | |
| "logps/rejected": -1.4563645124435425, | |
| "loss": 1.8273, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0144473314285278, | |
| "rewards/margins": 0.4419172406196594, | |
| "rewards/rejected": -1.4563645124435425, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 25.238718639520766, | |
| "learning_rate": 3.6860532770864005e-07, | |
| "logits/chosen": -2.705747365951538, | |
| "logits/rejected": -2.6104602813720703, | |
| "logps/chosen": -1.0011558532714844, | |
| "logps/rejected": -1.5988446474075317, | |
| "loss": 1.8926, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.0011558532714844, | |
| "rewards/margins": 0.5976886749267578, | |
| "rewards/rejected": -1.5988446474075317, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 23.97929885445044, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": -2.644472122192383, | |
| "logits/rejected": -2.5559470653533936, | |
| "logps/chosen": -1.0497673749923706, | |
| "logps/rejected": -1.3750804662704468, | |
| "loss": 1.853, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0497673749923706, | |
| "rewards/margins": 0.32531291246414185, | |
| "rewards/rejected": -1.3750804662704468, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 24.184854862916072, | |
| "learning_rate": 3.5218566107988867e-07, | |
| "logits/chosen": -2.730355739593506, | |
| "logits/rejected": -2.683979034423828, | |
| "logps/chosen": -0.9922051429748535, | |
| "logps/rejected": -1.5398856401443481, | |
| "loss": 1.8724, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.9922051429748535, | |
| "rewards/margins": 0.5476802587509155, | |
| "rewards/rejected": -1.5398856401443481, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 26.69782735208681, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -2.5600266456604004, | |
| "logits/rejected": -2.5641884803771973, | |
| "logps/chosen": -1.0444707870483398, | |
| "logps/rejected": -1.4003455638885498, | |
| "loss": 1.8764, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0444707870483398, | |
| "rewards/margins": 0.35587459802627563, | |
| "rewards/rejected": -1.4003455638885498, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 39.9600540211152, | |
| "learning_rate": 3.3521824616429284e-07, | |
| "logits/chosen": -2.594709873199463, | |
| "logits/rejected": -2.5683107376098633, | |
| "logps/chosen": -1.126425862312317, | |
| "logps/rejected": -1.614018440246582, | |
| "loss": 1.8882, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.126425862312317, | |
| "rewards/margins": 0.4875926077365875, | |
| "rewards/rejected": -1.614018440246582, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 35.61682268118621, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -2.5961153507232666, | |
| "logits/rejected": -2.541783094406128, | |
| "logps/chosen": -1.102477788925171, | |
| "logps/rejected": -1.5613858699798584, | |
| "loss": 1.8517, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.102477788925171, | |
| "rewards/margins": 0.45890823006629944, | |
| "rewards/rejected": -1.5613858699798584, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 28.356766220400576, | |
| "learning_rate": 3.1779403380910425e-07, | |
| "logits/chosen": -2.6027655601501465, | |
| "logits/rejected": -2.4865376949310303, | |
| "logps/chosen": -1.0470962524414062, | |
| "logps/rejected": -1.5330448150634766, | |
| "loss": 1.8686, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.0470962524414062, | |
| "rewards/margins": 0.4859485626220703, | |
| "rewards/rejected": -1.5330448150634766, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 46.96042544932364, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -2.6151013374328613, | |
| "logits/rejected": -2.59317684173584, | |
| "logps/chosen": -1.0420851707458496, | |
| "logps/rejected": -1.637976050376892, | |
| "loss": 1.8322, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.0420851707458496, | |
| "rewards/margins": 0.5958911180496216, | |
| "rewards/rejected": -1.637976050376892, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 26.100391626514853, | |
| "learning_rate": 3.000064234440111e-07, | |
| "logits/chosen": -2.5428245067596436, | |
| "logits/rejected": -2.4851417541503906, | |
| "logps/chosen": -1.0481144189834595, | |
| "logps/rejected": -1.4148355722427368, | |
| "loss": 1.8385, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.0481144189834595, | |
| "rewards/margins": 0.3667212724685669, | |
| "rewards/rejected": -1.4148355722427368, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 24.14989026973158, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -2.537334680557251, | |
| "logits/rejected": -2.472512722015381, | |
| "logps/chosen": -1.0776469707489014, | |
| "logps/rejected": -1.6333630084991455, | |
| "loss": 1.8508, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.0776469707489014, | |
| "rewards/margins": 0.5557159185409546, | |
| "rewards/rejected": -1.6333630084991455, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 32.38845341862983, | |
| "learning_rate": 2.8195076242990116e-07, | |
| "logits/chosen": -2.6101131439208984, | |
| "logits/rejected": -2.5712387561798096, | |
| "logps/chosen": -0.9908210635185242, | |
| "logps/rejected": -1.6324228048324585, | |
| "loss": 1.8075, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -0.9908210635185242, | |
| "rewards/margins": 0.6416017413139343, | |
| "rewards/rejected": -1.6324228048324585, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 24.243213200791057, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -2.559380292892456, | |
| "logits/rejected": -2.4537248611450195, | |
| "logps/chosen": -1.0512077808380127, | |
| "logps/rejected": -1.6259485483169556, | |
| "loss": 1.8615, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.0512077808380127, | |
| "rewards/margins": 0.5747405290603638, | |
| "rewards/rejected": -1.6259485483169556, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 24.228357677990427, | |
| "learning_rate": 2.6372383496608186e-07, | |
| "logits/chosen": -2.654383420944214, | |
| "logits/rejected": -2.566788911819458, | |
| "logps/chosen": -1.0131515264511108, | |
| "logps/rejected": -1.4951848983764648, | |
| "loss": 1.8526, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.0131515264511108, | |
| "rewards/margins": 0.4820333421230316, | |
| "rewards/rejected": -1.4951848983764648, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 26.47263727780241, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -2.597144603729248, | |
| "logits/rejected": -2.572925090789795, | |
| "logps/chosen": -0.9740694761276245, | |
| "logps/rejected": -1.3979696035385132, | |
| "loss": 1.8422, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -0.9740694761276245, | |
| "rewards/margins": 0.4239000678062439, | |
| "rewards/rejected": -1.3979696035385132, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 22.019445712488015, | |
| "learning_rate": 2.454233432955807e-07, | |
| "logits/chosen": -2.612497329711914, | |
| "logits/rejected": -2.5739381313323975, | |
| "logps/chosen": -1.0217043161392212, | |
| "logps/rejected": -1.4334819316864014, | |
| "loss": 1.8467, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.0217043161392212, | |
| "rewards/margins": 0.4117775857448578, | |
| "rewards/rejected": -1.4334819316864014, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 31.0373082982515, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": -2.673020839691162, | |
| "logits/rejected": -2.6832852363586426, | |
| "logps/chosen": -1.1109861135482788, | |
| "logps/rejected": -1.4625297784805298, | |
| "loss": 1.8524, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.1109861135482788, | |
| "rewards/margins": 0.3515436351299286, | |
| "rewards/rejected": -1.4625297784805298, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 29.601053617457204, | |
| "learning_rate": 2.2714738398943308e-07, | |
| "logits/chosen": -2.6691536903381348, | |
| "logits/rejected": -2.587501049041748, | |
| "logps/chosen": -1.0023844242095947, | |
| "logps/rejected": -1.6990854740142822, | |
| "loss": 1.7818, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.0023844242095947, | |
| "rewards/margins": 0.6967008113861084, | |
| "rewards/rejected": -1.6990854740142822, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 23.304378897117243, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": -2.7478485107421875, | |
| "logits/rejected": -2.6739115715026855, | |
| "logps/chosen": -0.967931866645813, | |
| "logps/rejected": -1.6212472915649414, | |
| "loss": 1.8289, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.967931866645813, | |
| "rewards/margins": 0.6533153653144836, | |
| "rewards/rejected": -1.6212472915649414, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 30.765932632186015, | |
| "learning_rate": 2.089939221172446e-07, | |
| "logits/chosen": -2.683306932449341, | |
| "logits/rejected": -2.599849224090576, | |
| "logps/chosen": -1.079810380935669, | |
| "logps/rejected": -1.5674939155578613, | |
| "loss": 1.8595, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.079810380935669, | |
| "rewards/margins": 0.4876834750175476, | |
| "rewards/rejected": -1.5674939155578613, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 29.814903147144946, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": -2.6533594131469727, | |
| "logits/rejected": -2.5845093727111816, | |
| "logps/chosen": -1.0454188585281372, | |
| "logps/rejected": -1.3878004550933838, | |
| "loss": 1.824, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0454188585281372, | |
| "rewards/margins": 0.3423817753791809, | |
| "rewards/rejected": -1.3878004550933838, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 27.45666647767728, | |
| "learning_rate": 1.9106026612264315e-07, | |
| "logits/chosen": -2.6601853370666504, | |
| "logits/rejected": -2.6058852672576904, | |
| "logps/chosen": -1.038104772567749, | |
| "logps/rejected": -1.3897347450256348, | |
| "loss": 1.8319, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.038104772567749, | |
| "rewards/margins": 0.3516300320625305, | |
| "rewards/rejected": -1.3897347450256348, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 30.25718786399205, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": -2.5553667545318604, | |
| "logits/rejected": -2.451190233230591, | |
| "logps/chosen": -1.108663558959961, | |
| "logps/rejected": -1.5200892686843872, | |
| "loss": 1.8653, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.108663558959961, | |
| "rewards/margins": 0.4114256501197815, | |
| "rewards/rejected": -1.5200892686843872, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 30.020985917091327, | |
| "learning_rate": 1.7344254621846017e-07, | |
| "logits/chosen": -2.640211343765259, | |
| "logits/rejected": -2.6043474674224854, | |
| "logps/chosen": -1.0029397010803223, | |
| "logps/rejected": -1.265366792678833, | |
| "loss": 1.8019, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.0029397010803223, | |
| "rewards/margins": 0.2624271512031555, | |
| "rewards/rejected": -1.265366792678833, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 35.00385114807507, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": -2.629307270050049, | |
| "logits/rejected": -2.4974796772003174, | |
| "logps/chosen": -1.0108293294906616, | |
| "logps/rejected": -1.5987032651901245, | |
| "loss": 1.825, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0108293294906616, | |
| "rewards/margins": 0.5878738164901733, | |
| "rewards/rejected": -1.5987032651901245, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 33.95334568069759, | |
| "learning_rate": 1.562351990976095e-07, | |
| "logits/chosen": -2.6276791095733643, | |
| "logits/rejected": -2.503952980041504, | |
| "logps/chosen": -1.0764460563659668, | |
| "logps/rejected": -1.5906479358673096, | |
| "loss": 1.8359, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0764460563659668, | |
| "rewards/margins": 0.514201819896698, | |
| "rewards/rejected": -1.5906479358673096, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 29.685480077784828, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": -2.592771291732788, | |
| "logits/rejected": -2.5038256645202637, | |
| "logps/chosen": -1.034784197807312, | |
| "logps/rejected": -1.6445497274398804, | |
| "loss": 1.7974, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.034784197807312, | |
| "rewards/margins": 0.6097655892372131, | |
| "rewards/rejected": -1.6445497274398804, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 45.93159466769048, | |
| "learning_rate": 1.3953046172178413e-07, | |
| "logits/chosen": -2.5624806880950928, | |
| "logits/rejected": -2.4602789878845215, | |
| "logps/chosen": -1.1025031805038452, | |
| "logps/rejected": -1.7038898468017578, | |
| "loss": 1.8729, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.1025031805038452, | |
| "rewards/margins": 0.6013867855072021, | |
| "rewards/rejected": -1.7038898468017578, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 31.678205608236812, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": -2.599463939666748, | |
| "logits/rejected": -2.5258090496063232, | |
| "logps/chosen": -0.9680983424186707, | |
| "logps/rejected": -1.6296237707138062, | |
| "loss": 1.8012, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.9680983424186707, | |
| "rewards/margins": 0.6615251302719116, | |
| "rewards/rejected": -1.6296237707138062, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 37.038121368656675, | |
| "learning_rate": 1.2341787690142435e-07, | |
| "logits/chosen": -2.5534796714782715, | |
| "logits/rejected": -2.517007827758789, | |
| "logps/chosen": -1.1220486164093018, | |
| "logps/rejected": -1.6694402694702148, | |
| "loss": 1.8584, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -1.1220486164093018, | |
| "rewards/margins": 0.5473915338516235, | |
| "rewards/rejected": -1.6694402694702148, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 32.99732563960218, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": -2.58585524559021, | |
| "logits/rejected": -2.4757778644561768, | |
| "logps/chosen": -1.0188167095184326, | |
| "logps/rejected": -1.71088445186615, | |
| "loss": 1.8167, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0188167095184326, | |
| "rewards/margins": 0.6920677423477173, | |
| "rewards/rejected": -1.71088445186615, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 31.150959005223893, | |
| "learning_rate": 1.0798381331721107e-07, | |
| "logits/chosen": -2.595191478729248, | |
| "logits/rejected": -2.491788387298584, | |
| "logps/chosen": -1.0228018760681152, | |
| "logps/rejected": -1.5643987655639648, | |
| "loss": 1.8471, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.0228018760681152, | |
| "rewards/margins": 0.5415968298912048, | |
| "rewards/rejected": -1.5643987655639648, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 32.974367073048874, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": -2.4937729835510254, | |
| "logits/rejected": -2.3299617767333984, | |
| "logps/chosen": -1.100921630859375, | |
| "logps/rejected": -1.6467349529266357, | |
| "loss": 1.8361, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.100921630859375, | |
| "rewards/margins": 0.5458132028579712, | |
| "rewards/rejected": -1.6467349529266357, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 28.926067639567258, | |
| "learning_rate": 9.331100255592436e-08, | |
| "logits/chosen": -2.4389522075653076, | |
| "logits/rejected": -2.374527931213379, | |
| "logps/chosen": -1.0656179189682007, | |
| "logps/rejected": -1.67294442653656, | |
| "loss": 1.8606, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0656179189682007, | |
| "rewards/margins": 0.6073265075683594, | |
| "rewards/rejected": -1.67294442653656, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 41.902021950847704, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -2.497872829437256, | |
| "logits/rejected": -2.4256420135498047, | |
| "logps/chosen": -1.0714236497879028, | |
| "logps/rejected": -1.6986573934555054, | |
| "loss": 1.8331, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.0714236497879028, | |
| "rewards/margins": 0.6272337436676025, | |
| "rewards/rejected": -1.6986573934555054, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 31.435500480488756, | |
| "learning_rate": 7.947809564230445e-08, | |
| "logits/chosen": -2.532473087310791, | |
| "logits/rejected": -2.366628646850586, | |
| "logps/chosen": -1.0359852313995361, | |
| "logps/rejected": -1.7668260335922241, | |
| "loss": 1.7899, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.0359852313995361, | |
| "rewards/margins": 0.730840802192688, | |
| "rewards/rejected": -1.7668260335922241, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 39.54146493100524, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": -2.484086513519287, | |
| "logits/rejected": -2.349870204925537, | |
| "logps/chosen": -1.0333083868026733, | |
| "logps/rejected": -1.6121408939361572, | |
| "loss": 1.7669, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.0333083868026733, | |
| "rewards/margins": 0.5788324475288391, | |
| "rewards/rejected": -1.6121408939361572, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 44.26258075929706, | |
| "learning_rate": 6.655924144404906e-08, | |
| "logits/chosen": -2.413093328475952, | |
| "logits/rejected": -2.3522191047668457, | |
| "logps/chosen": -0.994648277759552, | |
| "logps/rejected": -1.8310960531234741, | |
| "loss": 1.8098, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -0.994648277759552, | |
| "rewards/margins": 0.8364478945732117, | |
| "rewards/rejected": -1.8310960531234741, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 42.81392142197046, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": -2.5220797061920166, | |
| "logits/rejected": -2.5181660652160645, | |
| "logps/chosen": -1.0257781744003296, | |
| "logps/rejected": -1.565146803855896, | |
| "loss": 1.848, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.0257781744003296, | |
| "rewards/margins": 0.5393685102462769, | |
| "rewards/rejected": -1.565146803855896, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 29.963279293888323, | |
| "learning_rate": 5.4623689209832484e-08, | |
| "logits/chosen": -2.4084603786468506, | |
| "logits/rejected": -2.3918216228485107, | |
| "logps/chosen": -1.0294010639190674, | |
| "logps/rejected": -1.5665457248687744, | |
| "loss": 1.8117, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0294010639190674, | |
| "rewards/margins": 0.537144660949707, | |
| "rewards/rejected": -1.5665457248687744, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 33.04796978561497, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": -2.4773898124694824, | |
| "logits/rejected": -2.334437608718872, | |
| "logps/chosen": -1.0476093292236328, | |
| "logps/rejected": -1.5581729412078857, | |
| "loss": 1.8541, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.0476093292236328, | |
| "rewards/margins": 0.5105637311935425, | |
| "rewards/rejected": -1.5581729412078857, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 36.23962004411558, | |
| "learning_rate": 4.373541737087263e-08, | |
| "logits/chosen": -2.4543843269348145, | |
| "logits/rejected": -2.3622539043426514, | |
| "logps/chosen": -1.0004545450210571, | |
| "logps/rejected": -1.7118451595306396, | |
| "loss": 1.822, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0004545450210571, | |
| "rewards/margins": 0.7113906145095825, | |
| "rewards/rejected": -1.7118451595306396, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 35.25243486313715, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": -2.4065768718719482, | |
| "logits/rejected": -2.2744433879852295, | |
| "logps/chosen": -1.054172396659851, | |
| "logps/rejected": -1.5296390056610107, | |
| "loss": 1.8213, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.054172396659851, | |
| "rewards/margins": 0.4754667282104492, | |
| "rewards/rejected": -1.5296390056610107, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 37.565064056547584, | |
| "learning_rate": 3.3952790595787986e-08, | |
| "logits/chosen": -2.449446439743042, | |
| "logits/rejected": -2.3016552925109863, | |
| "logps/chosen": -1.0504591464996338, | |
| "logps/rejected": -1.7570114135742188, | |
| "loss": 1.8237, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0504591464996338, | |
| "rewards/margins": 0.7065523266792297, | |
| "rewards/rejected": -1.7570114135742188, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 32.90354747950104, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": -2.3721275329589844, | |
| "logits/rejected": -2.268225908279419, | |
| "logps/chosen": -1.1106208562850952, | |
| "logps/rejected": -1.9150855541229248, | |
| "loss": 1.8043, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1106208562850952, | |
| "rewards/margins": 0.8044648170471191, | |
| "rewards/rejected": -1.9150855541229248, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 56.846827175134244, | |
| "learning_rate": 2.5328246937043525e-08, | |
| "logits/chosen": -2.4376819133758545, | |
| "logits/rejected": -2.3333914279937744, | |
| "logps/chosen": -1.1099225282669067, | |
| "logps/rejected": -1.6782783269882202, | |
| "loss": 1.8252, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.1099225282669067, | |
| "rewards/margins": 0.568355917930603, | |
| "rewards/rejected": -1.6782783269882202, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 43.77757640842041, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": -2.4453635215759277, | |
| "logits/rejected": -2.2761971950531006, | |
| "logps/chosen": -1.050205111503601, | |
| "logps/rejected": -1.8707281351089478, | |
| "loss": 1.7931, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.050205111503601, | |
| "rewards/margins": 0.8205229640007019, | |
| "rewards/rejected": -1.8707281351089478, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 37.41494353520166, | |
| "learning_rate": 1.7908016745981856e-08, | |
| "logits/chosen": -2.448782444000244, | |
| "logits/rejected": -2.2901499271392822, | |
| "logps/chosen": -1.043718695640564, | |
| "logps/rejected": -1.739383339881897, | |
| "loss": 1.8267, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.043718695640564, | |
| "rewards/margins": 0.695664644241333, | |
| "rewards/rejected": -1.739383339881897, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 30.693244043593708, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": -2.4449446201324463, | |
| "logits/rejected": -2.294093608856201, | |
| "logps/chosen": -1.0893933773040771, | |
| "logps/rejected": -1.642295241355896, | |
| "loss": 1.7605, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.0893933773040771, | |
| "rewards/margins": 0.5529020428657532, | |
| "rewards/rejected": -1.642295241355896, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 33.12531830225382, | |
| "learning_rate": 1.1731874863145142e-08, | |
| "logits/chosen": -2.3656153678894043, | |
| "logits/rejected": -2.352541208267212, | |
| "logps/chosen": -0.9648414850234985, | |
| "logps/rejected": -1.5097295045852661, | |
| "loss": 1.7984, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.9648414850234985, | |
| "rewards/margins": 0.5448881983757019, | |
| "rewards/rejected": -1.5097295045852661, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 40.86500194765816, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": -2.48162579536438, | |
| "logits/rejected": -2.3563125133514404, | |
| "logps/chosen": -1.0542184114456177, | |
| "logps/rejected": -1.5571963787078857, | |
| "loss": 1.8252, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0542184114456177, | |
| "rewards/margins": 0.5029779672622681, | |
| "rewards/rejected": -1.5571963787078857, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 30.792748106701044, | |
| "learning_rate": 6.832927412229017e-09, | |
| "logits/chosen": -2.399744987487793, | |
| "logits/rejected": -2.232243061065674, | |
| "logps/chosen": -1.0769870281219482, | |
| "logps/rejected": -1.728468656539917, | |
| "loss": 1.7915, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0769870281219482, | |
| "rewards/margins": 0.6514816284179688, | |
| "rewards/rejected": -1.728468656539917, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 29.999766652539517, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": -2.459473133087158, | |
| "logits/rejected": -2.2956230640411377, | |
| "logps/chosen": -1.0055310726165771, | |
| "logps/rejected": -1.5959385633468628, | |
| "loss": 1.8126, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.0055310726165771, | |
| "rewards/margins": 0.5904075503349304, | |
| "rewards/rejected": -1.5959385633468628, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 36.992525127734154, | |
| "learning_rate": 3.2374343405217884e-09, | |
| "logits/chosen": -2.42592191696167, | |
| "logits/rejected": -2.3259921073913574, | |
| "logps/chosen": -1.0747435092926025, | |
| "logps/rejected": -1.6361202001571655, | |
| "loss": 1.8922, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.0747435092926025, | |
| "rewards/margins": 0.5613764524459839, | |
| "rewards/rejected": -1.6361202001571655, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 43.04775872660601, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": -2.488020896911621, | |
| "logits/rejected": -2.239854097366333, | |
| "logps/chosen": -0.9606490135192871, | |
| "logps/rejected": -1.82891845703125, | |
| "loss": 1.8115, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -0.9606490135192871, | |
| "rewards/margins": 0.8682695627212524, | |
| "rewards/rejected": -1.82891845703125, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 25.076712480507634, | |
| "learning_rate": 9.64668657069706e-10, | |
| "logits/chosen": -2.4506993293762207, | |
| "logits/rejected": -2.252009868621826, | |
| "logps/chosen": -1.0200697183609009, | |
| "logps/rejected": -1.8124902248382568, | |
| "loss": 1.8012, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.0200697183609009, | |
| "rewards/margins": 0.7924206852912903, | |
| "rewards/rejected": -1.8124902248382568, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 26.74704043050654, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": -2.3772401809692383, | |
| "logits/rejected": -2.2120604515075684, | |
| "logps/chosen": -0.9293681979179382, | |
| "logps/rejected": -1.7934061288833618, | |
| "loss": 1.7671, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -0.9293681979179382, | |
| "rewards/margins": 0.8640381097793579, | |
| "rewards/rejected": -1.7934061288833618, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 50.411989213964986, | |
| "learning_rate": 2.6813123097352287e-11, | |
| "logits/chosen": -2.4256701469421387, | |
| "logits/rejected": -2.3270206451416016, | |
| "logps/chosen": -0.9691774249076843, | |
| "logps/rejected": -1.615407943725586, | |
| "loss": 1.797, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.9691774249076843, | |
| "rewards/margins": 0.6462305784225464, | |
| "rewards/rejected": -1.615407943725586, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 1.8663299063716545, | |
| "train_runtime": 8277.1533, | |
| "train_samples_per_second": 7.386, | |
| "train_steps_per_second": 0.058 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 477, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |