| { | |
| "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": 11.773019635010494, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": -1.4459120035171509, | |
| "logits/rejected": -1.4473698139190674, | |
| "logps/chosen": -7.983118534088135, | |
| "logps/rejected": -8.156193733215332, | |
| "loss": 8.9797, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.983118534088135, | |
| "rewards/margins": 0.17307573556900024, | |
| "rewards/rejected": -8.156193733215332, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 9.709040224393025, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -1.4474694728851318, | |
| "logits/rejected": -1.438397765159607, | |
| "logps/chosen": -8.046891212463379, | |
| "logps/rejected": -7.96016788482666, | |
| "loss": 8.9815, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -8.046891212463379, | |
| "rewards/margins": -0.08672460168600082, | |
| "rewards/rejected": -7.96016788482666, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 16.89107530887862, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": -1.4477674961090088, | |
| "logits/rejected": -1.4442229270935059, | |
| "logps/chosen": -7.852339267730713, | |
| "logps/rejected": -7.866924285888672, | |
| "loss": 8.8904, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.852339267730713, | |
| "rewards/margins": 0.014585095457732677, | |
| "rewards/rejected": -7.866924285888672, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 14.261360231352997, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -1.4316000938415527, | |
| "logits/rejected": -1.435153841972351, | |
| "logps/chosen": -8.180787086486816, | |
| "logps/rejected": -8.204398155212402, | |
| "loss": 9.0475, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -8.180787086486816, | |
| "rewards/margins": 0.023611638695001602, | |
| "rewards/rejected": -8.204398155212402, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 13.453269431524436, | |
| "learning_rate": 2.604166666666667e-07, | |
| "logits/chosen": -1.479134440422058, | |
| "logits/rejected": -1.4693964719772339, | |
| "logps/chosen": -8.08868408203125, | |
| "logps/rejected": -7.990322113037109, | |
| "loss": 9.1249, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -8.08868408203125, | |
| "rewards/margins": -0.09836126863956451, | |
| "rewards/rejected": -7.990322113037109, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 11.378313696016495, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -1.447307825088501, | |
| "logits/rejected": -1.4355913400650024, | |
| "logps/chosen": -7.789179801940918, | |
| "logps/rejected": -7.710352897644043, | |
| "loss": 9.0197, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -7.789179801940918, | |
| "rewards/margins": -0.0788281261920929, | |
| "rewards/rejected": -7.710352897644043, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 10.492170948688017, | |
| "learning_rate": 3.645833333333333e-07, | |
| "logits/chosen": -1.4480352401733398, | |
| "logits/rejected": -1.421820044517517, | |
| "logps/chosen": -8.031718254089355, | |
| "logps/rejected": -7.896989345550537, | |
| "loss": 8.9858, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -8.031718254089355, | |
| "rewards/margins": -0.13472941517829895, | |
| "rewards/rejected": -7.896989345550537, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 9.438874148034836, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -1.4374274015426636, | |
| "logits/rejected": -1.4498326778411865, | |
| "logps/chosen": -7.972153663635254, | |
| "logps/rejected": -8.157554626464844, | |
| "loss": 9.013, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -7.972153663635254, | |
| "rewards/margins": 0.185401052236557, | |
| "rewards/rejected": -8.157554626464844, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 10.665489074479519, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": -1.4577468633651733, | |
| "logits/rejected": -1.45815110206604, | |
| "logps/chosen": -8.014823913574219, | |
| "logps/rejected": -7.8058881759643555, | |
| "loss": 8.9894, | |
| "rewards/accuracies": 0.45625001192092896, | |
| "rewards/chosen": -8.014823913574219, | |
| "rewards/margins": -0.20893602073192596, | |
| "rewards/rejected": -7.8058881759643555, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 14.935150052502319, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -1.4442975521087646, | |
| "logits/rejected": -1.461060881614685, | |
| "logps/chosen": -8.208638191223145, | |
| "logps/rejected": -8.235433578491211, | |
| "loss": 9.0806, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -8.208638191223145, | |
| "rewards/margins": 0.0267957653850317, | |
| "rewards/rejected": -8.235433578491211, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 10.181057969467895, | |
| "learning_rate": 4.996716052911017e-07, | |
| "logits/chosen": -1.4492734670639038, | |
| "logits/rejected": -1.4457294940948486, | |
| "logps/chosen": -8.069734573364258, | |
| "logps/rejected": -8.081002235412598, | |
| "loss": 9.1066, | |
| "rewards/accuracies": 0.42500001192092896, | |
| "rewards/chosen": -8.069734573364258, | |
| "rewards/margins": 0.011268272995948792, | |
| "rewards/rejected": -8.081002235412598, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 17.86320335028244, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -1.4542796611785889, | |
| "logits/rejected": -1.4671558141708374, | |
| "logps/chosen": -8.027522087097168, | |
| "logps/rejected": -8.309490203857422, | |
| "loss": 8.9006, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -8.027522087097168, | |
| "rewards/margins": 0.28196847438812256, | |
| "rewards/rejected": -8.309490203857422, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 31.178844889844076, | |
| "learning_rate": 4.980652179769217e-07, | |
| "logits/chosen": -1.4044889211654663, | |
| "logits/rejected": -1.4228571653366089, | |
| "logps/chosen": -7.809567451477051, | |
| "logps/rejected": -8.305734634399414, | |
| "loss": 8.9379, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -7.809567451477051, | |
| "rewards/margins": 0.49616700410842896, | |
| "rewards/rejected": -8.305734634399414, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 35.6438890099215, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -1.4016475677490234, | |
| "logits/rejected": -1.3958280086517334, | |
| "logps/chosen": -7.9446282386779785, | |
| "logps/rejected": -8.026212692260742, | |
| "loss": 8.9738, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -7.9446282386779785, | |
| "rewards/margins": 0.08158435672521591, | |
| "rewards/rejected": -8.026212692260742, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 12.429786415191892, | |
| "learning_rate": 4.951291206355559e-07, | |
| "logits/chosen": -1.4062070846557617, | |
| "logits/rejected": -1.412986159324646, | |
| "logps/chosen": -7.849791049957275, | |
| "logps/rejected": -7.881753444671631, | |
| "loss": 8.9235, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -7.849791049957275, | |
| "rewards/margins": 0.031962424516677856, | |
| "rewards/rejected": -7.881753444671631, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 19.088641406767135, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -1.4117752313613892, | |
| "logits/rejected": -1.4267140626907349, | |
| "logps/chosen": -7.871654510498047, | |
| "logps/rejected": -7.936394691467285, | |
| "loss": 8.9114, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -7.871654510498047, | |
| "rewards/margins": 0.06473957002162933, | |
| "rewards/rejected": -7.936394691467285, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 12.122887672693283, | |
| "learning_rate": 4.908790517010636e-07, | |
| "logits/chosen": -1.4204871654510498, | |
| "logits/rejected": -1.4174373149871826, | |
| "logps/chosen": -7.965147495269775, | |
| "logps/rejected": -8.144342422485352, | |
| "loss": 9.0544, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -7.965147495269775, | |
| "rewards/margins": 0.17919424176216125, | |
| "rewards/rejected": -8.144342422485352, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 13.676489394050206, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -1.394677996635437, | |
| "logits/rejected": -1.4083014726638794, | |
| "logps/chosen": -7.955393314361572, | |
| "logps/rejected": -8.132440567016602, | |
| "loss": 9.0209, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -7.955393314361572, | |
| "rewards/margins": 0.1770474910736084, | |
| "rewards/rejected": -8.132440567016602, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 11.55232614681137, | |
| "learning_rate": 4.853377929214243e-07, | |
| "logits/chosen": -1.4072471857070923, | |
| "logits/rejected": -1.3971055746078491, | |
| "logps/chosen": -8.058225631713867, | |
| "logps/rejected": -7.9064154624938965, | |
| "loss": 9.0891, | |
| "rewards/accuracies": 0.40625, | |
| "rewards/chosen": -8.058225631713867, | |
| "rewards/margins": -0.15181036293506622, | |
| "rewards/rejected": -7.9064154624938965, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 21.381914106448818, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -1.4291296005249023, | |
| "logits/rejected": -1.4334924221038818, | |
| "logps/chosen": -8.016979217529297, | |
| "logps/rejected": -8.049524307250977, | |
| "loss": 9.0182, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -8.016979217529297, | |
| "rewards/margins": 0.032544612884521484, | |
| "rewards/rejected": -8.049524307250977, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 12.15933543492837, | |
| "learning_rate": 4.785350472409791e-07, | |
| "logits/chosen": -1.4222753047943115, | |
| "logits/rejected": -1.4167351722717285, | |
| "logps/chosen": -7.735973358154297, | |
| "logps/rejected": -7.934323310852051, | |
| "loss": 8.9474, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -7.735973358154297, | |
| "rewards/margins": 0.19834883511066437, | |
| "rewards/rejected": -7.934323310852051, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 14.275715375180301, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -1.4484798908233643, | |
| "logits/rejected": -1.4482091665267944, | |
| "logps/chosen": -7.996209144592285, | |
| "logps/rejected": -8.223912239074707, | |
| "loss": 9.021, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -7.996209144592285, | |
| "rewards/margins": 0.22770316898822784, | |
| "rewards/rejected": -8.223912239074707, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 264.49732707332146, | |
| "learning_rate": 4.70507279583015e-07, | |
| "logits/chosen": -1.3956806659698486, | |
| "logits/rejected": -1.399662971496582, | |
| "logps/chosen": -7.871851921081543, | |
| "logps/rejected": -8.169795989990234, | |
| "loss": 8.9644, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -7.871851921081543, | |
| "rewards/margins": 0.2979447841644287, | |
| "rewards/rejected": -8.169795989990234, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 13.052538835040968, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -1.3686974048614502, | |
| "logits/rejected": -1.3719284534454346, | |
| "logps/chosen": -8.102027893066406, | |
| "logps/rejected": -8.149332046508789, | |
| "loss": 8.9524, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -8.102027893066406, | |
| "rewards/margins": 0.04730508476495743, | |
| "rewards/rejected": -8.149332046508789, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 13.986017227821636, | |
| "learning_rate": 4.612975213859487e-07, | |
| "logits/chosen": -1.4253513813018799, | |
| "logits/rejected": -1.419106125831604, | |
| "logps/chosen": -8.198808670043945, | |
| "logps/rejected": -8.076632499694824, | |
| "loss": 9.068, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -8.198808670043945, | |
| "rewards/margins": -0.12217602878808975, | |
| "rewards/rejected": -8.076632499694824, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 16.859825952779723, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -1.393610954284668, | |
| "logits/rejected": -1.4041980504989624, | |
| "logps/chosen": -8.033707618713379, | |
| "logps/rejected": -8.149084091186523, | |
| "loss": 8.9812, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -8.033707618713379, | |
| "rewards/margins": 0.11537569761276245, | |
| "rewards/rejected": -8.149084091186523, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 17.35250325551846, | |
| "learning_rate": 4.5095513994085974e-07, | |
| "logits/chosen": -1.3907815217971802, | |
| "logits/rejected": -1.4047492742538452, | |
| "logps/chosen": -7.996786594390869, | |
| "logps/rejected": -8.4099760055542, | |
| "loss": 8.9754, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -7.996786594390869, | |
| "rewards/margins": 0.41318947076797485, | |
| "rewards/rejected": -8.4099760055542, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 13.984769587358256, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -1.415302038192749, | |
| "logits/rejected": -1.4232456684112549, | |
| "logps/chosen": -7.9026665687561035, | |
| "logps/rejected": -7.970452785491943, | |
| "loss": 8.9574, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -7.9026665687561035, | |
| "rewards/margins": 0.06778701394796371, | |
| "rewards/rejected": -7.970452785491943, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 10.033908137947623, | |
| "learning_rate": 4.395355737667985e-07, | |
| "logits/chosen": -1.4233510494232178, | |
| "logits/rejected": -1.4498116970062256, | |
| "logps/chosen": -7.896154880523682, | |
| "logps/rejected": -8.049067497253418, | |
| "loss": 8.9238, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -7.896154880523682, | |
| "rewards/margins": 0.15291360020637512, | |
| "rewards/rejected": -8.049067497253418, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 12.948300110975106, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -1.4413058757781982, | |
| "logits/rejected": -1.4447827339172363, | |
| "logps/chosen": -7.901156425476074, | |
| "logps/rejected": -8.029939651489258, | |
| "loss": 8.9805, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.901156425476074, | |
| "rewards/margins": 0.12878116965293884, | |
| "rewards/rejected": -8.029939651489258, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 13.439201394209311, | |
| "learning_rate": 4.271000354423425e-07, | |
| "logits/chosen": -1.412276029586792, | |
| "logits/rejected": -1.389304280281067, | |
| "logps/chosen": -7.978514194488525, | |
| "logps/rejected": -7.953900337219238, | |
| "loss": 9.0293, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -7.978514194488525, | |
| "rewards/margins": -0.024613410234451294, | |
| "rewards/rejected": -7.953900337219238, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 14.950263562725501, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -1.3908809423446655, | |
| "logits/rejected": -1.4011869430541992, | |
| "logps/chosen": -8.05128288269043, | |
| "logps/rejected": -8.073431015014648, | |
| "loss": 9.0342, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -8.05128288269043, | |
| "rewards/margins": 0.022147323936223984, | |
| "rewards/rejected": -8.073431015014648, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 11.065140622134825, | |
| "learning_rate": 4.137151834863213e-07, | |
| "logits/chosen": -1.4063256978988647, | |
| "logits/rejected": -1.4003641605377197, | |
| "logps/chosen": -7.999358177185059, | |
| "logps/rejected": -8.034425735473633, | |
| "loss": 9.0115, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -7.999358177185059, | |
| "rewards/margins": 0.03506825119256973, | |
| "rewards/rejected": -8.034425735473633, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 12.669233243410716, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -1.3777008056640625, | |
| "logits/rejected": -1.3843880891799927, | |
| "logps/chosen": -8.11161994934082, | |
| "logps/rejected": -8.079643249511719, | |
| "loss": 9.088, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -8.11161994934082, | |
| "rewards/margins": -0.03197728842496872, | |
| "rewards/rejected": -8.079643249511719, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 11.420602500449908, | |
| "learning_rate": 3.994527650465352e-07, | |
| "logits/chosen": -1.4002071619033813, | |
| "logits/rejected": -1.4017866849899292, | |
| "logps/chosen": -7.9545183181762695, | |
| "logps/rejected": -7.929112434387207, | |
| "loss": 9.01, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -7.9545183181762695, | |
| "rewards/margins": -0.025406653061509132, | |
| "rewards/rejected": -7.929112434387207, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 15.567371494550915, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -1.379305124282837, | |
| "logits/rejected": -1.4001214504241943, | |
| "logps/chosen": -8.02379322052002, | |
| "logps/rejected": -8.138644218444824, | |
| "loss": 8.993, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -8.02379322052002, | |
| "rewards/margins": 0.11485125869512558, | |
| "rewards/rejected": -8.138644218444824, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 13.26955840349748, | |
| "learning_rate": 3.8438923131177237e-07, | |
| "logits/chosen": -1.345184564590454, | |
| "logits/rejected": -1.3495187759399414, | |
| "logps/chosen": -7.9985671043396, | |
| "logps/rejected": -8.032160758972168, | |
| "loss": 8.9564, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -7.9985671043396, | |
| "rewards/margins": 0.03359389677643776, | |
| "rewards/rejected": -8.032160758972168, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 12.477711396229129, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": -1.3648011684417725, | |
| "logits/rejected": -1.3609410524368286, | |
| "logps/chosen": -7.813986301422119, | |
| "logps/rejected": -7.937358856201172, | |
| "loss": 8.9559, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -7.813986301422119, | |
| "rewards/margins": 0.12337237596511841, | |
| "rewards/rejected": -7.937358856201172, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 14.974853102646863, | |
| "learning_rate": 3.6860532770864005e-07, | |
| "logits/chosen": -1.340057611465454, | |
| "logits/rejected": -1.3464308977127075, | |
| "logps/chosen": -8.168105125427246, | |
| "logps/rejected": -8.047191619873047, | |
| "loss": 9.0593, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -8.168105125427246, | |
| "rewards/margins": -0.12091370671987534, | |
| "rewards/rejected": -8.047191619873047, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 12.019056641062388, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": -1.3949248790740967, | |
| "logits/rejected": -1.3864567279815674, | |
| "logps/chosen": -7.904551029205322, | |
| "logps/rejected": -8.010711669921875, | |
| "loss": 9.0251, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -7.904551029205322, | |
| "rewards/margins": 0.10616078227758408, | |
| "rewards/rejected": -8.010711669921875, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 12.159976530620005, | |
| "learning_rate": 3.5218566107988867e-07, | |
| "logits/chosen": -1.3519536256790161, | |
| "logits/rejected": -1.3532142639160156, | |
| "logps/chosen": -7.96322774887085, | |
| "logps/rejected": -8.124198913574219, | |
| "loss": 8.9502, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.96322774887085, | |
| "rewards/margins": 0.16097018122673035, | |
| "rewards/rejected": -8.124198913574219, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 27.53866432550454, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -1.3502663373947144, | |
| "logits/rejected": -1.3595716953277588, | |
| "logps/chosen": -7.90906286239624, | |
| "logps/rejected": -8.031661033630371, | |
| "loss": 8.997, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -7.90906286239624, | |
| "rewards/margins": 0.12259860336780548, | |
| "rewards/rejected": -8.031661033630371, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 10.57208181223796, | |
| "learning_rate": 3.3521824616429284e-07, | |
| "logits/chosen": -1.3545321226119995, | |
| "logits/rejected": -1.363968849182129, | |
| "logps/chosen": -7.766638278961182, | |
| "logps/rejected": -8.186101913452148, | |
| "loss": 8.9091, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -7.766638278961182, | |
| "rewards/margins": 0.4194633960723877, | |
| "rewards/rejected": -8.186101913452148, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 12.898063816305818, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -1.3691070079803467, | |
| "logits/rejected": -1.3948938846588135, | |
| "logps/chosen": -7.7049560546875, | |
| "logps/rejected": -7.837745666503906, | |
| "loss": 9.0096, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -7.7049560546875, | |
| "rewards/margins": 0.13278940320014954, | |
| "rewards/rejected": -7.837745666503906, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 11.02542077899864, | |
| "learning_rate": 3.1779403380910425e-07, | |
| "logits/chosen": -1.3727288246154785, | |
| "logits/rejected": -1.3770244121551514, | |
| "logps/chosen": -7.688443660736084, | |
| "logps/rejected": -7.930489540100098, | |
| "loss": 8.9147, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -7.688443660736084, | |
| "rewards/margins": 0.24204584956169128, | |
| "rewards/rejected": -7.930489540100098, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 12.642936387401933, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -1.3643910884857178, | |
| "logits/rejected": -1.3646974563598633, | |
| "logps/chosen": -8.087334632873535, | |
| "logps/rejected": -8.227707862854004, | |
| "loss": 9.0751, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -8.087334632873535, | |
| "rewards/margins": 0.1403738558292389, | |
| "rewards/rejected": -8.227707862854004, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 11.653678081935743, | |
| "learning_rate": 3.000064234440111e-07, | |
| "logits/chosen": -1.3848552703857422, | |
| "logits/rejected": -1.3937251567840576, | |
| "logps/chosen": -7.939251899719238, | |
| "logps/rejected": -8.165653228759766, | |
| "loss": 8.9934, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -7.939251899719238, | |
| "rewards/margins": 0.2264014482498169, | |
| "rewards/rejected": -8.165653228759766, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 20.78939179590344, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -1.3638317584991455, | |
| "logits/rejected": -1.372854471206665, | |
| "logps/chosen": -7.975274562835693, | |
| "logps/rejected": -8.114542007446289, | |
| "loss": 9.0759, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -7.975274562835693, | |
| "rewards/margins": 0.13926678895950317, | |
| "rewards/rejected": -8.114542007446289, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 15.247620089410335, | |
| "learning_rate": 2.8195076242990116e-07, | |
| "logits/chosen": -1.3632863759994507, | |
| "logits/rejected": -1.368691086769104, | |
| "logps/chosen": -8.058801651000977, | |
| "logps/rejected": -8.10949420928955, | |
| "loss": 8.9298, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -8.058801651000977, | |
| "rewards/margins": 0.05069229006767273, | |
| "rewards/rejected": -8.10949420928955, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 16.020124654557804, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -1.335317611694336, | |
| "logits/rejected": -1.3454726934432983, | |
| "logps/chosen": -7.886631011962891, | |
| "logps/rejected": -8.195113182067871, | |
| "loss": 8.8247, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -7.886631011962891, | |
| "rewards/margins": 0.30848345160484314, | |
| "rewards/rejected": -8.195113182067871, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 13.208964203269911, | |
| "learning_rate": 2.6372383496608186e-07, | |
| "logits/chosen": -1.3199383020401, | |
| "logits/rejected": -1.3232190608978271, | |
| "logps/chosen": -8.123445510864258, | |
| "logps/rejected": -8.300798416137695, | |
| "loss": 8.9974, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -8.123445510864258, | |
| "rewards/margins": 0.17735353112220764, | |
| "rewards/rejected": -8.300798416137695, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 12.80879122998489, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -1.2541121244430542, | |
| "logits/rejected": -1.2609522342681885, | |
| "logps/chosen": -7.773028373718262, | |
| "logps/rejected": -8.211790084838867, | |
| "loss": 8.9532, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -7.773028373718262, | |
| "rewards/margins": 0.43876129388809204, | |
| "rewards/rejected": -8.211790084838867, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 11.386114149183353, | |
| "learning_rate": 2.454233432955807e-07, | |
| "logits/chosen": -1.2998675107955933, | |
| "logits/rejected": -1.3152759075164795, | |
| "logps/chosen": -7.926003932952881, | |
| "logps/rejected": -8.16883373260498, | |
| "loss": 8.972, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -7.926003932952881, | |
| "rewards/margins": 0.242829367518425, | |
| "rewards/rejected": -8.16883373260498, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 12.189434902359803, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": -1.311598539352417, | |
| "logits/rejected": -1.2913634777069092, | |
| "logps/chosen": -8.263054847717285, | |
| "logps/rejected": -8.317811965942383, | |
| "loss": 9.0177, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -8.263054847717285, | |
| "rewards/margins": 0.05475734546780586, | |
| "rewards/rejected": -8.317811965942383, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 26.512104480113177, | |
| "learning_rate": 2.2714738398943308e-07, | |
| "logits/chosen": -1.3503350019454956, | |
| "logits/rejected": -1.3630883693695068, | |
| "logps/chosen": -7.8860344886779785, | |
| "logps/rejected": -8.053815841674805, | |
| "loss": 8.9101, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -7.8860344886779785, | |
| "rewards/margins": 0.1677820235490799, | |
| "rewards/rejected": -8.053815841674805, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 13.657892445365238, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": -1.3591408729553223, | |
| "logits/rejected": -1.3795652389526367, | |
| "logps/chosen": -7.8742218017578125, | |
| "logps/rejected": -8.096671104431152, | |
| "loss": 8.926, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -7.8742218017578125, | |
| "rewards/margins": 0.222448468208313, | |
| "rewards/rejected": -8.096671104431152, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 13.125973843054254, | |
| "learning_rate": 2.089939221172446e-07, | |
| "logits/chosen": -1.3040008544921875, | |
| "logits/rejected": -1.3099054098129272, | |
| "logps/chosen": -7.843650817871094, | |
| "logps/rejected": -8.080728530883789, | |
| "loss": 8.8759, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -7.843650817871094, | |
| "rewards/margins": 0.23707744479179382, | |
| "rewards/rejected": -8.080728530883789, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 12.761067879239661, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": -1.3441493511199951, | |
| "logits/rejected": -1.3643635511398315, | |
| "logps/chosen": -7.810141086578369, | |
| "logps/rejected": -8.124038696289062, | |
| "loss": 8.9111, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -7.810141086578369, | |
| "rewards/margins": 0.31389790773391724, | |
| "rewards/rejected": -8.124038696289062, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 12.064219944453757, | |
| "learning_rate": 1.9106026612264315e-07, | |
| "logits/chosen": -1.3431895971298218, | |
| "logits/rejected": -1.3602094650268555, | |
| "logps/chosen": -7.9335832595825195, | |
| "logps/rejected": -8.049263000488281, | |
| "loss": 8.9788, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -7.9335832595825195, | |
| "rewards/margins": 0.11567918956279755, | |
| "rewards/rejected": -8.049263000488281, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 12.249297326029936, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": -1.3094807863235474, | |
| "logits/rejected": -1.3196088075637817, | |
| "logps/chosen": -7.965109825134277, | |
| "logps/rejected": -8.200112342834473, | |
| "loss": 8.9053, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.965109825134277, | |
| "rewards/margins": 0.23500177264213562, | |
| "rewards/rejected": -8.200112342834473, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 11.932717244582381, | |
| "learning_rate": 1.7344254621846017e-07, | |
| "logits/chosen": -1.317882776260376, | |
| "logits/rejected": -1.3457939624786377, | |
| "logps/chosen": -8.174354553222656, | |
| "logps/rejected": -8.366888046264648, | |
| "loss": 9.0377, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -8.174354553222656, | |
| "rewards/margins": 0.19253170490264893, | |
| "rewards/rejected": -8.366888046264648, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 14.13608210525545, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": -1.3378267288208008, | |
| "logits/rejected": -1.3463207483291626, | |
| "logps/chosen": -8.089614868164062, | |
| "logps/rejected": -8.071249008178711, | |
| "loss": 9.0688, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -8.089614868164062, | |
| "rewards/margins": -0.018364891409873962, | |
| "rewards/rejected": -8.071249008178711, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 32.011426029911085, | |
| "learning_rate": 1.562351990976095e-07, | |
| "logits/chosen": -1.330702304840088, | |
| "logits/rejected": -1.3502042293548584, | |
| "logps/chosen": -7.904356479644775, | |
| "logps/rejected": -8.121664047241211, | |
| "loss": 8.8938, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -7.904356479644775, | |
| "rewards/margins": 0.2173081338405609, | |
| "rewards/rejected": -8.121664047241211, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 12.807799082013823, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": -1.346669316291809, | |
| "logits/rejected": -1.3398444652557373, | |
| "logps/chosen": -8.155746459960938, | |
| "logps/rejected": -8.042104721069336, | |
| "loss": 8.9844, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -8.155746459960938, | |
| "rewards/margins": -0.11364219337701797, | |
| "rewards/rejected": -8.042104721069336, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 14.721102641186317, | |
| "learning_rate": 1.3953046172178413e-07, | |
| "logits/chosen": -1.3383146524429321, | |
| "logits/rejected": -1.3453185558319092, | |
| "logps/chosen": -8.06794548034668, | |
| "logps/rejected": -8.342490196228027, | |
| "loss": 8.9313, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -8.06794548034668, | |
| "rewards/margins": 0.27454426884651184, | |
| "rewards/rejected": -8.342490196228027, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 12.639814201884843, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": -1.3189464807510376, | |
| "logits/rejected": -1.3342128992080688, | |
| "logps/chosen": -7.980110168457031, | |
| "logps/rejected": -8.00477123260498, | |
| "loss": 9.0333, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.980110168457031, | |
| "rewards/margins": 0.024661889299750328, | |
| "rewards/rejected": -8.00477123260498, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 13.542292770349258, | |
| "learning_rate": 1.2341787690142435e-07, | |
| "logits/chosen": -1.3359485864639282, | |
| "logits/rejected": -1.3414697647094727, | |
| "logps/chosen": -7.863481044769287, | |
| "logps/rejected": -8.104010581970215, | |
| "loss": 8.8907, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -7.863481044769287, | |
| "rewards/margins": 0.24052973091602325, | |
| "rewards/rejected": -8.104010581970215, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 13.235795007023567, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": -1.3393312692642212, | |
| "logits/rejected": -1.3506433963775635, | |
| "logps/chosen": -8.217998504638672, | |
| "logps/rejected": -8.318530082702637, | |
| "loss": 9.0332, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -8.217998504638672, | |
| "rewards/margins": 0.1005309671163559, | |
| "rewards/rejected": -8.318530082702637, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 14.926426867380192, | |
| "learning_rate": 1.0798381331721107e-07, | |
| "logits/chosen": -1.3743078708648682, | |
| "logits/rejected": -1.390223503112793, | |
| "logps/chosen": -8.069305419921875, | |
| "logps/rejected": -8.406010627746582, | |
| "loss": 9.0149, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -8.069305419921875, | |
| "rewards/margins": 0.3367052972316742, | |
| "rewards/rejected": -8.406010627746582, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 15.925504397841232, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": -1.344575047492981, | |
| "logits/rejected": -1.3378257751464844, | |
| "logps/chosen": -7.946380615234375, | |
| "logps/rejected": -8.057405471801758, | |
| "loss": 8.9081, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -7.946380615234375, | |
| "rewards/margins": 0.11102403700351715, | |
| "rewards/rejected": -8.057405471801758, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 14.002913308986452, | |
| "learning_rate": 9.331100255592436e-08, | |
| "logits/chosen": -1.3642247915267944, | |
| "logits/rejected": -1.356299638748169, | |
| "logps/chosen": -8.019064903259277, | |
| "logps/rejected": -7.9631476402282715, | |
| "loss": 8.911, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -8.019064903259277, | |
| "rewards/margins": -0.055917102843523026, | |
| "rewards/rejected": -7.9631476402282715, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 11.189613386087895, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": -1.323051929473877, | |
| "logits/rejected": -1.3336186408996582, | |
| "logps/chosen": -7.749169826507568, | |
| "logps/rejected": -8.018518447875977, | |
| "loss": 8.86, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -7.749169826507568, | |
| "rewards/margins": 0.26934924721717834, | |
| "rewards/rejected": -8.018518447875977, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 12.73495657042311, | |
| "learning_rate": 7.947809564230445e-08, | |
| "logits/chosen": -1.299702525138855, | |
| "logits/rejected": -1.300682783126831, | |
| "logps/chosen": -7.912903785705566, | |
| "logps/rejected": -7.9577813148498535, | |
| "loss": 8.9548, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -7.912903785705566, | |
| "rewards/margins": 0.044878166168928146, | |
| "rewards/rejected": -7.9577813148498535, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 12.244582800127791, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": -1.3569344282150269, | |
| "logits/rejected": -1.3466898202896118, | |
| "logps/chosen": -8.036967277526855, | |
| "logps/rejected": -8.161175727844238, | |
| "loss": 8.9693, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -8.036967277526855, | |
| "rewards/margins": 0.12420801818370819, | |
| "rewards/rejected": -8.161175727844238, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 13.857374473897332, | |
| "learning_rate": 6.655924144404906e-08, | |
| "logits/chosen": -1.3418371677398682, | |
| "logits/rejected": -1.3312771320343018, | |
| "logps/chosen": -8.015717506408691, | |
| "logps/rejected": -8.231267929077148, | |
| "loss": 8.9495, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -8.015717506408691, | |
| "rewards/margins": 0.2155502289533615, | |
| "rewards/rejected": -8.231267929077148, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 12.96483071279786, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": -1.3451616764068604, | |
| "logits/rejected": -1.2988903522491455, | |
| "logps/chosen": -8.223435401916504, | |
| "logps/rejected": -8.359044075012207, | |
| "loss": 8.9575, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -8.223435401916504, | |
| "rewards/margins": 0.13560748100280762, | |
| "rewards/rejected": -8.359044075012207, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 13.685503251086777, | |
| "learning_rate": 5.4623689209832484e-08, | |
| "logits/chosen": -1.3425824642181396, | |
| "logits/rejected": -1.3416216373443604, | |
| "logps/chosen": -8.126443862915039, | |
| "logps/rejected": -8.207201957702637, | |
| "loss": 8.9674, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -8.126443862915039, | |
| "rewards/margins": 0.08075866848230362, | |
| "rewards/rejected": -8.207201957702637, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 13.422439706317645, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": -1.3507561683654785, | |
| "logits/rejected": -1.3538233041763306, | |
| "logps/chosen": -7.850442409515381, | |
| "logps/rejected": -8.027676582336426, | |
| "loss": 8.8745, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -7.850442409515381, | |
| "rewards/margins": 0.17723475396633148, | |
| "rewards/rejected": -8.027676582336426, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 17.927875137352093, | |
| "learning_rate": 4.373541737087263e-08, | |
| "logits/chosen": -1.3154563903808594, | |
| "logits/rejected": -1.3293858766555786, | |
| "logps/chosen": -7.652901649475098, | |
| "logps/rejected": -7.787165641784668, | |
| "loss": 8.9212, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -7.652901649475098, | |
| "rewards/margins": 0.13426539301872253, | |
| "rewards/rejected": -7.787165641784668, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 15.347923517709377, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": -1.3166577816009521, | |
| "logits/rejected": -1.308420181274414, | |
| "logps/chosen": -7.855776786804199, | |
| "logps/rejected": -7.910757541656494, | |
| "loss": 8.958, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -7.855776786804199, | |
| "rewards/margins": 0.05497948080301285, | |
| "rewards/rejected": -7.910757541656494, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 12.360735869108282, | |
| "learning_rate": 3.3952790595787986e-08, | |
| "logits/chosen": -1.3554697036743164, | |
| "logits/rejected": -1.361022710800171, | |
| "logps/chosen": -8.025907516479492, | |
| "logps/rejected": -8.157610893249512, | |
| "loss": 8.9282, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -8.025907516479492, | |
| "rewards/margins": 0.13170403242111206, | |
| "rewards/rejected": -8.157610893249512, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 31.497593173498622, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": -1.3601171970367432, | |
| "logits/rejected": -1.3568463325500488, | |
| "logps/chosen": -7.857785701751709, | |
| "logps/rejected": -7.846571922302246, | |
| "loss": 8.9122, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -7.857785701751709, | |
| "rewards/margins": -0.01121367234736681, | |
| "rewards/rejected": -7.846571922302246, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 14.186881098656178, | |
| "learning_rate": 2.5328246937043525e-08, | |
| "logits/chosen": -1.3695319890975952, | |
| "logits/rejected": -1.3772709369659424, | |
| "logps/chosen": -7.996212959289551, | |
| "logps/rejected": -8.215394973754883, | |
| "loss": 8.984, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -7.996212959289551, | |
| "rewards/margins": 0.21918316185474396, | |
| "rewards/rejected": -8.215394973754883, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 13.07091749141306, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": -1.320655107498169, | |
| "logits/rejected": -1.315647006034851, | |
| "logps/chosen": -7.8787431716918945, | |
| "logps/rejected": -7.92669677734375, | |
| "loss": 8.9426, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -7.8787431716918945, | |
| "rewards/margins": 0.047954294830560684, | |
| "rewards/rejected": -7.92669677734375, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 12.735545462610197, | |
| "learning_rate": 1.7908016745981856e-08, | |
| "logits/chosen": -1.3389335870742798, | |
| "logits/rejected": -1.335252285003662, | |
| "logps/chosen": -8.04029655456543, | |
| "logps/rejected": -8.07039737701416, | |
| "loss": 8.9855, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -8.04029655456543, | |
| "rewards/margins": 0.030100971460342407, | |
| "rewards/rejected": -8.07039737701416, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 15.247676128020851, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": -1.3410335779190063, | |
| "logits/rejected": -1.3450171947479248, | |
| "logps/chosen": -7.889002323150635, | |
| "logps/rejected": -7.991044044494629, | |
| "loss": 8.9385, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -7.889002323150635, | |
| "rewards/margins": 0.10204191505908966, | |
| "rewards/rejected": -7.991044044494629, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 13.85977210657239, | |
| "learning_rate": 1.1731874863145142e-08, | |
| "logits/chosen": -1.3579423427581787, | |
| "logits/rejected": -1.3657190799713135, | |
| "logps/chosen": -7.948614597320557, | |
| "logps/rejected": -8.05911636352539, | |
| "loss": 8.8473, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.948614597320557, | |
| "rewards/margins": 0.11050298064947128, | |
| "rewards/rejected": -8.05911636352539, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 13.136814821473001, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": -1.327292799949646, | |
| "logits/rejected": -1.3260419368743896, | |
| "logps/chosen": -8.011943817138672, | |
| "logps/rejected": -8.084588050842285, | |
| "loss": 9.0087, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -8.011943817138672, | |
| "rewards/margins": 0.07264318317174911, | |
| "rewards/rejected": -8.084588050842285, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 13.772252146186085, | |
| "learning_rate": 6.832927412229017e-09, | |
| "logits/chosen": -1.3517519235610962, | |
| "logits/rejected": -1.353088140487671, | |
| "logps/chosen": -7.9066948890686035, | |
| "logps/rejected": -8.016542434692383, | |
| "loss": 8.9143, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -7.9066948890686035, | |
| "rewards/margins": 0.10984818637371063, | |
| "rewards/rejected": -8.016542434692383, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 62.09067243328178, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": -1.3028905391693115, | |
| "logits/rejected": -1.2930048704147339, | |
| "logps/chosen": -8.025864601135254, | |
| "logps/rejected": -8.012077331542969, | |
| "loss": 9.0079, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -8.025864601135254, | |
| "rewards/margins": -0.013786206021904945, | |
| "rewards/rejected": -8.012077331542969, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 13.419716763500347, | |
| "learning_rate": 3.2374343405217884e-09, | |
| "logits/chosen": -1.3575193881988525, | |
| "logits/rejected": -1.351488709449768, | |
| "logps/chosen": -7.829765319824219, | |
| "logps/rejected": -8.183721542358398, | |
| "loss": 8.9117, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -7.829765319824219, | |
| "rewards/margins": 0.3539566695690155, | |
| "rewards/rejected": -8.183721542358398, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 22.929649885843165, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": -1.3399038314819336, | |
| "logits/rejected": -1.352718710899353, | |
| "logps/chosen": -7.996813774108887, | |
| "logps/rejected": -8.18346881866455, | |
| "loss": 8.8927, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -7.996813774108887, | |
| "rewards/margins": 0.18665429949760437, | |
| "rewards/rejected": -8.18346881866455, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 12.997334149917643, | |
| "learning_rate": 9.64668657069706e-10, | |
| "logits/chosen": -1.363525390625, | |
| "logits/rejected": -1.3654100894927979, | |
| "logps/chosen": -8.123039245605469, | |
| "logps/rejected": -8.254613876342773, | |
| "loss": 8.999, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -8.123039245605469, | |
| "rewards/margins": 0.13157425820827484, | |
| "rewards/rejected": -8.254613876342773, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 14.345805639596328, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": -1.3454748392105103, | |
| "logits/rejected": -1.3637425899505615, | |
| "logps/chosen": -7.719171047210693, | |
| "logps/rejected": -7.984010219573975, | |
| "loss": 8.9911, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -7.719171047210693, | |
| "rewards/margins": 0.26483961939811707, | |
| "rewards/rejected": -7.984010219573975, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 15.884627478532495, | |
| "learning_rate": 2.6813123097352287e-11, | |
| "logits/chosen": -1.3404934406280518, | |
| "logits/rejected": -1.329254388809204, | |
| "logps/chosen": -8.111069679260254, | |
| "logps/rejected": -8.152128219604492, | |
| "loss": 8.9896, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -8.111069679260254, | |
| "rewards/margins": 0.041059140115976334, | |
| "rewards/rejected": -8.152128219604492, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 8.9737869758526, | |
| "train_runtime": 8198.1649, | |
| "train_samples_per_second": 7.457, | |
| "train_steps_per_second": 0.058 | |
| } | |
| ], | |
| "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 | |
| } | |