| { | |
| "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": 22.273845933282402, | |
| "learning_rate": 5.208333333333333e-08, | |
| "logits/chosen": -2.770733118057251, | |
| "logits/rejected": -2.7281923294067383, | |
| "logps/chosen": -1.0278276205062866, | |
| "logps/rejected": -1.1733704805374146, | |
| "loss": 1.7067, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -1.0278276205062866, | |
| "rewards/margins": 0.14554281532764435, | |
| "rewards/rejected": -1.1733704805374146, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 19.65262772274273, | |
| "learning_rate": 1.0416666666666667e-07, | |
| "logits/chosen": -2.7513110637664795, | |
| "logits/rejected": -2.725446939468384, | |
| "logps/chosen": -0.9947978258132935, | |
| "logps/rejected": -1.0709822177886963, | |
| "loss": 1.7052, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -0.9947978258132935, | |
| "rewards/margins": 0.07618431746959686, | |
| "rewards/rejected": -1.0709822177886963, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 20.86060525388248, | |
| "learning_rate": 1.5624999999999999e-07, | |
| "logits/chosen": -2.7313146591186523, | |
| "logits/rejected": -2.697767496109009, | |
| "logps/chosen": -0.9808584451675415, | |
| "logps/rejected": -1.1414306163787842, | |
| "loss": 1.6534, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -0.9808584451675415, | |
| "rewards/margins": 0.16057202219963074, | |
| "rewards/rejected": -1.1414306163787842, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 18.455889774407584, | |
| "learning_rate": 2.0833333333333333e-07, | |
| "logits/chosen": -2.8040003776550293, | |
| "logits/rejected": -2.7113003730773926, | |
| "logps/chosen": -0.997894287109375, | |
| "logps/rejected": -1.098174810409546, | |
| "loss": 1.6781, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.997894287109375, | |
| "rewards/margins": 0.1002805083990097, | |
| "rewards/rejected": -1.098174810409546, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 17.169851595349584, | |
| "learning_rate": 2.604166666666667e-07, | |
| "logits/chosen": -2.8004283905029297, | |
| "logits/rejected": -2.777601718902588, | |
| "logps/chosen": -0.9775938987731934, | |
| "logps/rejected": -1.0392240285873413, | |
| "loss": 1.6816, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -0.9775938987731934, | |
| "rewards/margins": 0.061630137264728546, | |
| "rewards/rejected": -1.0392240285873413, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 17.60907134514165, | |
| "learning_rate": 3.1249999999999997e-07, | |
| "logits/chosen": -2.7743477821350098, | |
| "logits/rejected": -2.734339714050293, | |
| "logps/chosen": -0.8839586973190308, | |
| "logps/rejected": -0.9886126518249512, | |
| "loss": 1.7146, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.8839586973190308, | |
| "rewards/margins": 0.1046539768576622, | |
| "rewards/rejected": -0.9886126518249512, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 19.779646214864712, | |
| "learning_rate": 3.645833333333333e-07, | |
| "logits/chosen": -2.7794618606567383, | |
| "logits/rejected": -2.6966075897216797, | |
| "logps/chosen": -0.9109536409378052, | |
| "logps/rejected": -1.0334917306900024, | |
| "loss": 1.6813, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.9109536409378052, | |
| "rewards/margins": 0.12253817170858383, | |
| "rewards/rejected": -1.0334917306900024, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 17.435155033915933, | |
| "learning_rate": 4.1666666666666667e-07, | |
| "logits/chosen": -2.796264171600342, | |
| "logits/rejected": -2.706550121307373, | |
| "logps/chosen": -0.8977090120315552, | |
| "logps/rejected": -1.1700526475906372, | |
| "loss": 1.58, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -0.8977090120315552, | |
| "rewards/margins": 0.2723436951637268, | |
| "rewards/rejected": -1.1700526475906372, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 20.376725274104686, | |
| "learning_rate": 4.6874999999999996e-07, | |
| "logits/chosen": -2.7916207313537598, | |
| "logits/rejected": -2.7479615211486816, | |
| "logps/chosen": -0.921708881855011, | |
| "logps/rejected": -1.0657963752746582, | |
| "loss": 1.5991, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -0.921708881855011, | |
| "rewards/margins": 0.14408749341964722, | |
| "rewards/rejected": -1.0657963752746582, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 24.688089874039832, | |
| "learning_rate": 4.999731868769026e-07, | |
| "logits/chosen": -2.7189624309539795, | |
| "logits/rejected": -2.657292604446411, | |
| "logps/chosen": -0.9823735952377319, | |
| "logps/rejected": -1.2429628372192383, | |
| "loss": 1.5995, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.9823735952377319, | |
| "rewards/margins": 0.2605893015861511, | |
| "rewards/rejected": -1.2429628372192383, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 20.10805225669153, | |
| "learning_rate": 4.996716052911017e-07, | |
| "logits/chosen": -2.685032367706299, | |
| "logits/rejected": -2.6509592533111572, | |
| "logps/chosen": -1.0144091844558716, | |
| "logps/rejected": -1.2263678312301636, | |
| "loss": 1.6066, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -1.0144091844558716, | |
| "rewards/margins": 0.211958646774292, | |
| "rewards/rejected": -1.2263678312301636, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 23.40387121073354, | |
| "learning_rate": 4.990353313429303e-07, | |
| "logits/chosen": -2.717456817626953, | |
| "logits/rejected": -2.669968843460083, | |
| "logps/chosen": -1.071919322013855, | |
| "logps/rejected": -1.261800765991211, | |
| "loss": 1.5192, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.071919322013855, | |
| "rewards/margins": 0.18988142907619476, | |
| "rewards/rejected": -1.261800765991211, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 22.52237455325428, | |
| "learning_rate": 4.980652179769217e-07, | |
| "logits/chosen": -2.681735038757324, | |
| "logits/rejected": -2.568328857421875, | |
| "logps/chosen": -1.0160045623779297, | |
| "logps/rejected": -1.500519037246704, | |
| "loss": 1.5548, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.0160045623779297, | |
| "rewards/margins": 0.48451438546180725, | |
| "rewards/rejected": -1.500519037246704, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 29.253552274403233, | |
| "learning_rate": 4.967625656594781e-07, | |
| "logits/chosen": -2.5312721729278564, | |
| "logits/rejected": -2.480713129043579, | |
| "logps/chosen": -1.1166826486587524, | |
| "logps/rejected": -1.405466914176941, | |
| "loss": 1.5402, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.1166826486587524, | |
| "rewards/margins": 0.2887844145298004, | |
| "rewards/rejected": -1.405466914176941, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 31.175214085255675, | |
| "learning_rate": 4.951291206355559e-07, | |
| "logits/chosen": -2.567188024520874, | |
| "logits/rejected": -2.501024007797241, | |
| "logps/chosen": -1.0800148248672485, | |
| "logps/rejected": -1.4306553602218628, | |
| "loss": 1.5108, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0800148248672485, | |
| "rewards/margins": 0.3506406545639038, | |
| "rewards/rejected": -1.4306553602218628, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 23.89446565720433, | |
| "learning_rate": 4.93167072587771e-07, | |
| "logits/chosen": -2.5460524559020996, | |
| "logits/rejected": -2.4932408332824707, | |
| "logps/chosen": -1.083786964416504, | |
| "logps/rejected": -1.4139994382858276, | |
| "loss": 1.557, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.083786964416504, | |
| "rewards/margins": 0.3302125036716461, | |
| "rewards/rejected": -1.4139994382858276, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 35.04760138274803, | |
| "learning_rate": 4.908790517010636e-07, | |
| "logits/chosen": -2.555619239807129, | |
| "logits/rejected": -2.49599289894104, | |
| "logps/chosen": -1.027801275253296, | |
| "logps/rejected": -1.3224855661392212, | |
| "loss": 1.599, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.027801275253296, | |
| "rewards/margins": 0.29468411207199097, | |
| "rewards/rejected": -1.3224855661392212, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 24.7603865878364, | |
| "learning_rate": 4.882681251368548e-07, | |
| "logits/chosen": -2.4858529567718506, | |
| "logits/rejected": -2.458529472351074, | |
| "logps/chosen": -1.0844331979751587, | |
| "logps/rejected": -1.3636631965637207, | |
| "loss": 1.5441, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.0844331979751587, | |
| "rewards/margins": 0.27922993898391724, | |
| "rewards/rejected": -1.3636631965637207, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 31.023544965925865, | |
| "learning_rate": 4.853377929214243e-07, | |
| "logits/chosen": -2.50390887260437, | |
| "logits/rejected": -2.40189790725708, | |
| "logps/chosen": -1.1321955919265747, | |
| "logps/rejected": -1.4152982234954834, | |
| "loss": 1.5605, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.1321955919265747, | |
| "rewards/margins": 0.28310269117355347, | |
| "rewards/rejected": -1.4152982234954834, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 29.51747892906268, | |
| "learning_rate": 4.820919832540181e-07, | |
| "logits/chosen": -2.4223530292510986, | |
| "logits/rejected": -2.3233771324157715, | |
| "logps/chosen": -1.153446912765503, | |
| "logps/rejected": -1.5384576320648193, | |
| "loss": 1.5601, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.153446912765503, | |
| "rewards/margins": 0.3850107789039612, | |
| "rewards/rejected": -1.5384576320648193, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 30.939558746298886, | |
| "learning_rate": 4.785350472409791e-07, | |
| "logits/chosen": -2.3625779151916504, | |
| "logits/rejected": -2.3349967002868652, | |
| "logps/chosen": -1.1278746128082275, | |
| "logps/rejected": -1.5877573490142822, | |
| "loss": 1.4734, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.1278746128082275, | |
| "rewards/margins": 0.4598826766014099, | |
| "rewards/rejected": -1.5877573490142822, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 45.105416576816324, | |
| "learning_rate": 4.7467175306295647e-07, | |
| "logits/chosen": -2.367917537689209, | |
| "logits/rejected": -2.291100025177002, | |
| "logps/chosen": -1.2823549509048462, | |
| "logps/rejected": -1.7042672634124756, | |
| "loss": 1.5012, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.2823549509048462, | |
| "rewards/margins": 0.42191243171691895, | |
| "rewards/rejected": -1.7042672634124756, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 26.508438238557048, | |
| "learning_rate": 4.70507279583015e-07, | |
| "logits/chosen": -2.2769744396209717, | |
| "logits/rejected": -2.1966700553894043, | |
| "logps/chosen": -1.2124507427215576, | |
| "logps/rejected": -1.7381147146224976, | |
| "loss": 1.4243, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.2124507427215576, | |
| "rewards/margins": 0.5256639719009399, | |
| "rewards/rejected": -1.7381147146224976, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 36.57968166514704, | |
| "learning_rate": 4.6604720940421207e-07, | |
| "logits/chosen": -2.2149672508239746, | |
| "logits/rejected": -2.1592419147491455, | |
| "logps/chosen": -1.2770636081695557, | |
| "logps/rejected": -1.6595499515533447, | |
| "loss": 1.5083, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.2770636081695557, | |
| "rewards/margins": 0.3824862837791443, | |
| "rewards/rejected": -1.6595499515533447, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 30.33542228792605, | |
| "learning_rate": 4.612975213859487e-07, | |
| "logits/chosen": -2.2143490314483643, | |
| "logits/rejected": -2.177022695541382, | |
| "logps/chosen": -1.2711435556411743, | |
| "logps/rejected": -1.5601117610931396, | |
| "loss": 1.5047, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.2711435556411743, | |
| "rewards/margins": 0.28896820545196533, | |
| "rewards/rejected": -1.5601117610931396, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 31.4132000967526, | |
| "learning_rate": 4.5626458262912735e-07, | |
| "logits/chosen": -2.0743236541748047, | |
| "logits/rejected": -2.0040838718414307, | |
| "logps/chosen": -1.3241310119628906, | |
| "logps/rejected": -1.5913293361663818, | |
| "loss": 1.47, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.3241310119628906, | |
| "rewards/margins": 0.2671981453895569, | |
| "rewards/rejected": -1.5913293361663818, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 54.25569743941121, | |
| "learning_rate": 4.5095513994085974e-07, | |
| "logits/chosen": -1.8961169719696045, | |
| "logits/rejected": -1.7147821187973022, | |
| "logps/chosen": -1.5041354894638062, | |
| "logps/rejected": -2.126206398010254, | |
| "loss": 1.4478, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.5041354894638062, | |
| "rewards/margins": 0.6220710277557373, | |
| "rewards/rejected": -2.126206398010254, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 43.983222727893036, | |
| "learning_rate": 4.453763107901675e-07, | |
| "logits/chosen": -1.5513172149658203, | |
| "logits/rejected": -1.3982717990875244, | |
| "logps/chosen": -1.5197192430496216, | |
| "logps/rejected": -2.097294330596924, | |
| "loss": 1.3929, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.5197192430496216, | |
| "rewards/margins": 0.5775750279426575, | |
| "rewards/rejected": -2.097294330596924, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 28.469754678991702, | |
| "learning_rate": 4.395355737667985e-07, | |
| "logits/chosen": -1.4272167682647705, | |
| "logits/rejected": -1.1418006420135498, | |
| "logps/chosen": -1.4365965127944946, | |
| "logps/rejected": -1.9005510807037354, | |
| "loss": 1.442, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.4365965127944946, | |
| "rewards/margins": 0.46395477652549744, | |
| "rewards/rejected": -1.9005510807037354, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 41.18537586594856, | |
| "learning_rate": 4.3344075855595097e-07, | |
| "logits/chosen": -1.3107974529266357, | |
| "logits/rejected": -1.1333634853363037, | |
| "logps/chosen": -1.333939552307129, | |
| "logps/rejected": -1.8191452026367188, | |
| "loss": 1.3703, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.333939552307129, | |
| "rewards/margins": 0.4852055609226227, | |
| "rewards/rejected": -1.8191452026367188, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 43.62414965682813, | |
| "learning_rate": 4.271000354423425e-07, | |
| "logits/chosen": -0.9065883755683899, | |
| "logits/rejected": -0.6428723931312561, | |
| "logps/chosen": -1.6046466827392578, | |
| "logps/rejected": -2.088477373123169, | |
| "loss": 1.393, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -1.6046466827392578, | |
| "rewards/margins": 0.48383063077926636, | |
| "rewards/rejected": -2.088477373123169, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 58.236255602056396, | |
| "learning_rate": 4.2052190435769554e-07, | |
| "logits/chosen": -0.5077437162399292, | |
| "logits/rejected": -0.2604050636291504, | |
| "logps/chosen": -1.8002322912216187, | |
| "logps/rejected": -2.378002882003784, | |
| "loss": 1.3526, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -1.8002322912216187, | |
| "rewards/margins": 0.5777708292007446, | |
| "rewards/rejected": -2.378002882003784, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 37.97133552354845, | |
| "learning_rate": 4.137151834863213e-07, | |
| "logits/chosen": -0.9249609112739563, | |
| "logits/rejected": -0.7287668585777283, | |
| "logps/chosen": -1.729589819908142, | |
| "logps/rejected": -2.2937121391296387, | |
| "loss": 1.3021, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.729589819908142, | |
| "rewards/margins": 0.5641222596168518, | |
| "rewards/rejected": -2.2937121391296387, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 58.781010614086675, | |
| "learning_rate": 4.0668899744407567e-07, | |
| "logits/chosen": -1.0361697673797607, | |
| "logits/rejected": -0.7538817524909973, | |
| "logps/chosen": -1.8726847171783447, | |
| "logps/rejected": -2.472113847732544, | |
| "loss": 1.4168, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.8726847171783447, | |
| "rewards/margins": 0.5994290709495544, | |
| "rewards/rejected": -2.472113847732544, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 43.77533333071548, | |
| "learning_rate": 3.994527650465352e-07, | |
| "logits/chosen": -1.0444451570510864, | |
| "logits/rejected": -0.6541527509689331, | |
| "logps/chosen": -1.9884567260742188, | |
| "logps/rejected": -2.6078011989593506, | |
| "loss": 1.4179, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.9884567260742188, | |
| "rewards/margins": 0.6193446516990662, | |
| "rewards/rejected": -2.6078011989593506, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 39.30408636565025, | |
| "learning_rate": 3.920161866827889e-07, | |
| "logits/chosen": -1.1477383375167847, | |
| "logits/rejected": -0.9928515553474426, | |
| "logps/chosen": -2.031057596206665, | |
| "logps/rejected": -2.6500778198242188, | |
| "loss": 1.3292, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -2.031057596206665, | |
| "rewards/margins": 0.6190204620361328, | |
| "rewards/rejected": -2.6500778198242188, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 49.18397826699053, | |
| "learning_rate": 3.8438923131177237e-07, | |
| "logits/chosen": -1.3022983074188232, | |
| "logits/rejected": -1.1097471714019775, | |
| "logps/chosen": -2.1825501918792725, | |
| "logps/rejected": -2.8575053215026855, | |
| "loss": 1.3409, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -2.1825501918792725, | |
| "rewards/margins": 0.6749551892280579, | |
| "rewards/rejected": -2.8575053215026855, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 59.443069428032004, | |
| "learning_rate": 3.765821230985757e-07, | |
| "logits/chosen": -1.309631586074829, | |
| "logits/rejected": -1.1841962337493896, | |
| "logps/chosen": -2.1491549015045166, | |
| "logps/rejected": -2.7087011337280273, | |
| "loss": 1.4201, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -2.1491549015045166, | |
| "rewards/margins": 0.5595459938049316, | |
| "rewards/rejected": -2.7087011337280273, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 58.81635631771188, | |
| "learning_rate": 3.6860532770864005e-07, | |
| "logits/chosen": -0.9547362327575684, | |
| "logits/rejected": -0.6742358803749084, | |
| "logps/chosen": -2.285407066345215, | |
| "logps/rejected": -2.962996006011963, | |
| "loss": 1.3681, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -2.285407066345215, | |
| "rewards/margins": 0.6775889992713928, | |
| "rewards/rejected": -2.962996006011963, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 61.19133999845696, | |
| "learning_rate": 3.604695382782159e-07, | |
| "logits/chosen": -0.5743089914321899, | |
| "logits/rejected": -0.5020885467529297, | |
| "logps/chosen": -2.337622880935669, | |
| "logps/rejected": -2.9621822834014893, | |
| "loss": 1.3002, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -2.337622880935669, | |
| "rewards/margins": 0.6245594620704651, | |
| "rewards/rejected": -2.9621822834014893, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 65.1623208567972, | |
| "learning_rate": 3.5218566107988867e-07, | |
| "logits/chosen": -0.27199602127075195, | |
| "logits/rejected": 0.03203292936086655, | |
| "logps/chosen": -2.4965696334838867, | |
| "logps/rejected": -3.200573682785034, | |
| "loss": 1.2788, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -2.4965696334838867, | |
| "rewards/margins": 0.7040041089057922, | |
| "rewards/rejected": -3.200573682785034, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 81.87887955731682, | |
| "learning_rate": 3.4376480090239047e-07, | |
| "logits/chosen": -0.3396000266075134, | |
| "logits/rejected": -0.0072940438985824585, | |
| "logps/chosen": -2.886606216430664, | |
| "logps/rejected": -3.522667407989502, | |
| "loss": 1.3707, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -2.886606216430664, | |
| "rewards/margins": 0.6360612511634827, | |
| "rewards/rejected": -3.522667407989502, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 54.53418228916395, | |
| "learning_rate": 3.3521824616429284e-07, | |
| "logits/chosen": -0.3693031072616577, | |
| "logits/rejected": -0.05546772480010986, | |
| "logps/chosen": -2.6358675956726074, | |
| "logps/rejected": -3.3469290733337402, | |
| "loss": 1.2589, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -2.6358675956726074, | |
| "rewards/margins": 0.7110615968704224, | |
| "rewards/rejected": -3.3469290733337402, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 61.5068583100674, | |
| "learning_rate": 3.265574537815398e-07, | |
| "logits/chosen": -0.5512332320213318, | |
| "logits/rejected": -0.04919959604740143, | |
| "logps/chosen": -2.361748218536377, | |
| "logps/rejected": -3.2458934783935547, | |
| "loss": 1.2447, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -2.361748218536377, | |
| "rewards/margins": 0.8841449022293091, | |
| "rewards/rejected": -3.2458934783935547, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 62.18444756933803, | |
| "learning_rate": 3.1779403380910425e-07, | |
| "logits/chosen": -0.4573966860771179, | |
| "logits/rejected": 0.10396180301904678, | |
| "logps/chosen": -2.5001068115234375, | |
| "logps/rejected": -3.3893027305603027, | |
| "loss": 1.2645, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -2.5001068115234375, | |
| "rewards/margins": 0.8891960978507996, | |
| "rewards/rejected": -3.3893027305603027, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 69.58783982000695, | |
| "learning_rate": 3.0893973387735683e-07, | |
| "logits/chosen": -0.23679371178150177, | |
| "logits/rejected": 0.2879559099674225, | |
| "logps/chosen": -3.0258288383483887, | |
| "logps/rejected": -4.09028959274292, | |
| "loss": 1.1493, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -3.0258288383483887, | |
| "rewards/margins": 1.0644607543945312, | |
| "rewards/rejected": -4.09028959274292, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 65.85335888107983, | |
| "learning_rate": 3.000064234440111e-07, | |
| "logits/chosen": -0.09591199457645416, | |
| "logits/rejected": 0.3684416711330414, | |
| "logps/chosen": -2.861621379852295, | |
| "logps/rejected": -3.635110855102539, | |
| "loss": 1.2158, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -2.861621379852295, | |
| "rewards/margins": 0.7734889984130859, | |
| "rewards/rejected": -3.635110855102539, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 51.47234347289653, | |
| "learning_rate": 2.910060778827554e-07, | |
| "logits/chosen": -0.214645653963089, | |
| "logits/rejected": 0.15288104116916656, | |
| "logps/chosen": -2.2756009101867676, | |
| "logps/rejected": -3.0374677181243896, | |
| "loss": 1.2342, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -2.2756009101867676, | |
| "rewards/margins": 0.7618668079376221, | |
| "rewards/rejected": -3.0374677181243896, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 67.41003053954276, | |
| "learning_rate": 2.8195076242990116e-07, | |
| "logits/chosen": -0.233026385307312, | |
| "logits/rejected": 0.1556404083967209, | |
| "logps/chosen": -2.4698057174682617, | |
| "logps/rejected": -3.129883050918579, | |
| "loss": 1.2922, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -2.4698057174682617, | |
| "rewards/margins": 0.6600776314735413, | |
| "rewards/rejected": -3.129883050918579, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 70.04740584020762, | |
| "learning_rate": 2.7285261601056697e-07, | |
| "logits/chosen": -0.32400113344192505, | |
| "logits/rejected": 0.12020448595285416, | |
| "logps/chosen": -2.376567840576172, | |
| "logps/rejected": -3.175973653793335, | |
| "loss": 1.3275, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -2.376567840576172, | |
| "rewards/margins": 0.799405574798584, | |
| "rewards/rejected": -3.175973653793335, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 49.05614284788625, | |
| "learning_rate": 2.6372383496608186e-07, | |
| "logits/chosen": -0.07061919569969177, | |
| "logits/rejected": 0.49763208627700806, | |
| "logps/chosen": -2.67657470703125, | |
| "logps/rejected": -3.5274453163146973, | |
| "loss": 1.1737, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -2.67657470703125, | |
| "rewards/margins": 0.8508706092834473, | |
| "rewards/rejected": -3.5274453163146973, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 73.89578979282193, | |
| "learning_rate": 2.5457665670441937e-07, | |
| "logits/chosen": -0.18951067328453064, | |
| "logits/rejected": 0.2982691824436188, | |
| "logps/chosen": -2.413672924041748, | |
| "logps/rejected": -3.579845905303955, | |
| "loss": 1.2239, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -2.413672924041748, | |
| "rewards/margins": 1.166172981262207, | |
| "rewards/rejected": -3.579845905303955, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 54.444851596761815, | |
| "learning_rate": 2.454233432955807e-07, | |
| "logits/chosen": -0.05852239206433296, | |
| "logits/rejected": 0.3520205318927765, | |
| "logps/chosen": -2.7941486835479736, | |
| "logps/rejected": -3.663743495941162, | |
| "loss": 1.1459, | |
| "rewards/accuracies": 0.762499988079071, | |
| "rewards/chosen": -2.7941486835479736, | |
| "rewards/margins": 0.8695951700210571, | |
| "rewards/rejected": -3.663743495941162, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 79.13341556564741, | |
| "learning_rate": 2.3627616503391812e-07, | |
| "logits/chosen": 0.12447645515203476, | |
| "logits/rejected": 0.48517122864723206, | |
| "logps/chosen": -3.074282169342041, | |
| "logps/rejected": -3.778022289276123, | |
| "loss": 1.2406, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -3.074282169342041, | |
| "rewards/margins": 0.7037407159805298, | |
| "rewards/rejected": -3.778022289276123, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 65.77375872071654, | |
| "learning_rate": 2.2714738398943308e-07, | |
| "logits/chosen": 0.10315855592489243, | |
| "logits/rejected": 0.6628378629684448, | |
| "logps/chosen": -2.9850118160247803, | |
| "logps/rejected": -3.925368070602417, | |
| "loss": 1.1927, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -2.9850118160247803, | |
| "rewards/margins": 0.9403561353683472, | |
| "rewards/rejected": -3.925368070602417, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 68.7072404828179, | |
| "learning_rate": 2.1804923757009882e-07, | |
| "logits/chosen": 0.14654329419136047, | |
| "logits/rejected": 0.7450932264328003, | |
| "logps/chosen": -2.682091474533081, | |
| "logps/rejected": -3.504145860671997, | |
| "loss": 1.224, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -2.682091474533081, | |
| "rewards/margins": 0.8220542669296265, | |
| "rewards/rejected": -3.504145860671997, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 61.507759714231206, | |
| "learning_rate": 2.089939221172446e-07, | |
| "logits/chosen": 0.048857107758522034, | |
| "logits/rejected": 0.40524905920028687, | |
| "logps/chosen": -2.660057306289673, | |
| "logps/rejected": -3.4508941173553467, | |
| "loss": 1.2054, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -2.660057306289673, | |
| "rewards/margins": 0.7908371090888977, | |
| "rewards/rejected": -3.4508941173553467, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 68.54045280221443, | |
| "learning_rate": 1.9999357655598891e-07, | |
| "logits/chosen": 0.09905894100666046, | |
| "logits/rejected": 0.6364008784294128, | |
| "logps/chosen": -2.7884061336517334, | |
| "logps/rejected": -3.744817018508911, | |
| "loss": 1.1811, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -2.7884061336517334, | |
| "rewards/margins": 0.9564107060432434, | |
| "rewards/rejected": -3.744817018508911, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 69.13964326129287, | |
| "learning_rate": 1.9106026612264315e-07, | |
| "logits/chosen": 0.03760681673884392, | |
| "logits/rejected": 0.6295174956321716, | |
| "logps/chosen": -2.8681061267852783, | |
| "logps/rejected": -3.6628737449645996, | |
| "loss": 1.2487, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -2.8681061267852783, | |
| "rewards/margins": 0.7947677373886108, | |
| "rewards/rejected": -3.6628737449645996, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 68.24880568918918, | |
| "learning_rate": 1.8220596619089573e-07, | |
| "logits/chosen": 0.05892186239361763, | |
| "logits/rejected": 0.4875253140926361, | |
| "logps/chosen": -2.8865182399749756, | |
| "logps/rejected": -3.7262775897979736, | |
| "loss": 1.2566, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -2.8865182399749756, | |
| "rewards/margins": 0.8397592306137085, | |
| "rewards/rejected": -3.7262775897979736, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 63.303760044846996, | |
| "learning_rate": 1.7344254621846017e-07, | |
| "logits/chosen": 0.2723705768585205, | |
| "logits/rejected": 0.7761790752410889, | |
| "logps/chosen": -2.930652141571045, | |
| "logps/rejected": -3.947206974029541, | |
| "loss": 1.1113, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -2.930652141571045, | |
| "rewards/margins": 1.0165544748306274, | |
| "rewards/rejected": -3.947206974029541, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 74.87714108878664, | |
| "learning_rate": 1.647817538357072e-07, | |
| "logits/chosen": 0.40727514028549194, | |
| "logits/rejected": 0.6944658160209656, | |
| "logps/chosen": -2.684722900390625, | |
| "logps/rejected": -3.821376085281372, | |
| "loss": 1.1187, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -2.684722900390625, | |
| "rewards/margins": 1.1366530656814575, | |
| "rewards/rejected": -3.821376085281372, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 78.50549857938152, | |
| "learning_rate": 1.562351990976095e-07, | |
| "logits/chosen": 0.38405269384384155, | |
| "logits/rejected": 0.9985488653182983, | |
| "logps/chosen": -2.8799889087677, | |
| "logps/rejected": -4.003491401672363, | |
| "loss": 1.1249, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -2.8799889087677, | |
| "rewards/margins": 1.1235021352767944, | |
| "rewards/rejected": -4.003491401672363, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 80.43258089949055, | |
| "learning_rate": 1.478143389201113e-07, | |
| "logits/chosen": 0.6803662776947021, | |
| "logits/rejected": 0.8349927067756653, | |
| "logps/chosen": -2.9795212745666504, | |
| "logps/rejected": -3.9721953868865967, | |
| "loss": 1.1932, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -2.9795212745666504, | |
| "rewards/margins": 0.9926745295524597, | |
| "rewards/rejected": -3.9721953868865967, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 73.34968580085386, | |
| "learning_rate": 1.3953046172178413e-07, | |
| "logits/chosen": 0.5375081300735474, | |
| "logits/rejected": 1.211456060409546, | |
| "logps/chosen": -2.997009754180908, | |
| "logps/rejected": -3.9822990894317627, | |
| "loss": 1.1884, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -2.997009754180908, | |
| "rewards/margins": 0.9852892160415649, | |
| "rewards/rejected": -3.9822990894317627, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 67.18394255987927, | |
| "learning_rate": 1.3139467229135998e-07, | |
| "logits/chosen": 0.46189865469932556, | |
| "logits/rejected": 1.051397681236267, | |
| "logps/chosen": -2.7415473461151123, | |
| "logps/rejected": -3.6691060066223145, | |
| "loss": 1.1945, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -2.7415473461151123, | |
| "rewards/margins": 0.9275587201118469, | |
| "rewards/rejected": -3.6691060066223145, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 64.70715932336967, | |
| "learning_rate": 1.2341787690142435e-07, | |
| "logits/chosen": 0.2790513038635254, | |
| "logits/rejected": 0.8592203259468079, | |
| "logps/chosen": -2.72259783744812, | |
| "logps/rejected": -3.6879966259002686, | |
| "loss": 1.2498, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -2.72259783744812, | |
| "rewards/margins": 0.965398907661438, | |
| "rewards/rejected": -3.6879966259002686, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 61.63125001377349, | |
| "learning_rate": 1.1561076868822755e-07, | |
| "logits/chosen": 0.34762030839920044, | |
| "logits/rejected": 0.8974772691726685, | |
| "logps/chosen": -2.9279873371124268, | |
| "logps/rejected": -3.637186050415039, | |
| "loss": 1.1831, | |
| "rewards/accuracies": 0.731249988079071, | |
| "rewards/chosen": -2.9279873371124268, | |
| "rewards/margins": 0.7091987133026123, | |
| "rewards/rejected": -3.637186050415039, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 69.9867150019491, | |
| "learning_rate": 1.0798381331721107e-07, | |
| "logits/chosen": 0.42011842131614685, | |
| "logits/rejected": 0.9746605753898621, | |
| "logps/chosen": -2.681163787841797, | |
| "logps/rejected": -3.532362699508667, | |
| "loss": 1.1815, | |
| "rewards/accuracies": 0.71875, | |
| "rewards/chosen": -2.681163787841797, | |
| "rewards/margins": 0.8511991500854492, | |
| "rewards/rejected": -3.532362699508667, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 75.66340151463851, | |
| "learning_rate": 1.0054723495346482e-07, | |
| "logits/chosen": 0.5741020441055298, | |
| "logits/rejected": 0.8616167902946472, | |
| "logps/chosen": -2.8164191246032715, | |
| "logps/rejected": -3.6091017723083496, | |
| "loss": 1.2396, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -2.8164191246032715, | |
| "rewards/margins": 0.7926827669143677, | |
| "rewards/rejected": -3.6091017723083496, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 74.17047524574353, | |
| "learning_rate": 9.331100255592436e-08, | |
| "logits/chosen": 0.5733584761619568, | |
| "logits/rejected": 1.1586692333221436, | |
| "logps/chosen": -2.8743791580200195, | |
| "logps/rejected": -4.042096138000488, | |
| "loss": 1.1432, | |
| "rewards/accuracies": 0.84375, | |
| "rewards/chosen": -2.8743791580200195, | |
| "rewards/margins": 1.1677173376083374, | |
| "rewards/rejected": -4.042096138000488, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 73.14407398619424, | |
| "learning_rate": 8.628481651367875e-08, | |
| "logits/chosen": 0.29708194732666016, | |
| "logits/rejected": 0.8764042854309082, | |
| "logps/chosen": -3.0110630989074707, | |
| "logps/rejected": -3.7496750354766846, | |
| "loss": 1.1891, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -3.0110630989074707, | |
| "rewards/margins": 0.7386118173599243, | |
| "rewards/rejected": -3.7496750354766846, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 84.79816410361819, | |
| "learning_rate": 7.947809564230445e-08, | |
| "logits/chosen": 0.5876302719116211, | |
| "logits/rejected": 1.1461758613586426, | |
| "logps/chosen": -3.0007166862487793, | |
| "logps/rejected": -4.093684673309326, | |
| "loss": 1.2364, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -3.0007166862487793, | |
| "rewards/margins": 1.0929677486419678, | |
| "rewards/rejected": -4.093684673309326, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 76.4181977449788, | |
| "learning_rate": 7.289996455765748e-08, | |
| "logits/chosen": 0.7314077615737915, | |
| "logits/rejected": 1.2388323545455933, | |
| "logps/chosen": -3.2148640155792236, | |
| "logps/rejected": -4.329017639160156, | |
| "loss": 1.1397, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -3.2148640155792236, | |
| "rewards/margins": 1.1141533851623535, | |
| "rewards/rejected": -4.329017639160156, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 79.59752890916117, | |
| "learning_rate": 6.655924144404906e-08, | |
| "logits/chosen": 0.7917232513427734, | |
| "logits/rejected": 0.8404657244682312, | |
| "logps/chosen": -3.1535632610321045, | |
| "logps/rejected": -4.065492153167725, | |
| "loss": 1.117, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -3.1535632610321045, | |
| "rewards/margins": 0.9119284749031067, | |
| "rewards/rejected": -4.065492153167725, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 83.98695473040011, | |
| "learning_rate": 6.046442623320145e-08, | |
| "logits/chosen": 0.9315813779830933, | |
| "logits/rejected": 1.060347557067871, | |
| "logps/chosen": -3.2030982971191406, | |
| "logps/rejected": -4.173519134521484, | |
| "loss": 1.2339, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -3.2030982971191406, | |
| "rewards/margins": 0.9704204797744751, | |
| "rewards/rejected": -4.173519134521484, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 73.80815123078858, | |
| "learning_rate": 5.4623689209832484e-08, | |
| "logits/chosen": 0.6560049653053284, | |
| "logits/rejected": 1.1292173862457275, | |
| "logps/chosen": -3.0976970195770264, | |
| "logps/rejected": -4.078602313995361, | |
| "loss": 1.1986, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -3.0976970195770264, | |
| "rewards/margins": 0.9809052348136902, | |
| "rewards/rejected": -4.078602313995361, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 69.90570951076663, | |
| "learning_rate": 4.904486005914027e-08, | |
| "logits/chosen": 0.3861357569694519, | |
| "logits/rejected": 1.0230625867843628, | |
| "logps/chosen": -3.0694098472595215, | |
| "logps/rejected": -4.231044769287109, | |
| "loss": 1.12, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -3.0694098472595215, | |
| "rewards/margins": 1.161635160446167, | |
| "rewards/rejected": -4.231044769287109, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 71.22815378604164, | |
| "learning_rate": 4.373541737087263e-08, | |
| "logits/chosen": 0.35766056180000305, | |
| "logits/rejected": 1.0771682262420654, | |
| "logps/chosen": -2.899632692337036, | |
| "logps/rejected": -3.877995252609253, | |
| "loss": 1.1684, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -2.899632692337036, | |
| "rewards/margins": 0.9783622622489929, | |
| "rewards/rejected": -3.877995252609253, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 97.45446724327614, | |
| "learning_rate": 3.8702478614051345e-08, | |
| "logits/chosen": 0.5966876745223999, | |
| "logits/rejected": 1.0948045253753662, | |
| "logps/chosen": -2.8743724822998047, | |
| "logps/rejected": -3.8696448802948, | |
| "loss": 1.2227, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -2.8743724822998047, | |
| "rewards/margins": 0.9952724575996399, | |
| "rewards/rejected": -3.8696448802948, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 77.52250073936997, | |
| "learning_rate": 3.3952790595787986e-08, | |
| "logits/chosen": 0.5455753207206726, | |
| "logits/rejected": 0.8719021081924438, | |
| "logps/chosen": -3.101245880126953, | |
| "logps/rejected": -3.933109998703003, | |
| "loss": 1.2031, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -3.101245880126953, | |
| "rewards/margins": 0.8318643569946289, | |
| "rewards/rejected": -3.933109998703003, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 72.54119105678649, | |
| "learning_rate": 2.9492720416985e-08, | |
| "logits/chosen": 0.6000219583511353, | |
| "logits/rejected": 0.9464153051376343, | |
| "logps/chosen": -2.9545369148254395, | |
| "logps/rejected": -3.8140132427215576, | |
| "loss": 1.2519, | |
| "rewards/accuracies": 0.7749999761581421, | |
| "rewards/chosen": -2.9545369148254395, | |
| "rewards/margins": 0.8594767451286316, | |
| "rewards/rejected": -3.8140132427215576, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 57.55112621238877, | |
| "learning_rate": 2.5328246937043525e-08, | |
| "logits/chosen": 0.657649576663971, | |
| "logits/rejected": 1.3006607294082642, | |
| "logps/chosen": -2.929166078567505, | |
| "logps/rejected": -3.863920211791992, | |
| "loss": 1.1329, | |
| "rewards/accuracies": 0.768750011920929, | |
| "rewards/chosen": -2.929166078567505, | |
| "rewards/margins": 0.9347543716430664, | |
| "rewards/rejected": -3.863920211791992, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 67.15006871791722, | |
| "learning_rate": 2.1464952759020856e-08, | |
| "logits/chosen": 0.557062029838562, | |
| "logits/rejected": 0.9161618947982788, | |
| "logps/chosen": -2.8673954010009766, | |
| "logps/rejected": -3.752131938934326, | |
| "loss": 1.2637, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -2.8673954010009766, | |
| "rewards/margins": 0.8847367167472839, | |
| "rewards/rejected": -3.752131938934326, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 62.826606471582075, | |
| "learning_rate": 1.7908016745981856e-08, | |
| "logits/chosen": 0.6115803122520447, | |
| "logits/rejected": 1.1207200288772583, | |
| "logps/chosen": -2.983302354812622, | |
| "logps/rejected": -3.9271881580352783, | |
| "loss": 1.1823, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -2.983302354812622, | |
| "rewards/margins": 0.9438859820365906, | |
| "rewards/rejected": -3.9271881580352783, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 59.68175243856826, | |
| "learning_rate": 1.4662207078575684e-08, | |
| "logits/chosen": 0.5250687599182129, | |
| "logits/rejected": 1.0637626647949219, | |
| "logps/chosen": -2.930884599685669, | |
| "logps/rejected": -4.001204490661621, | |
| "loss": 1.1121, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -2.930884599685669, | |
| "rewards/margins": 1.0703200101852417, | |
| "rewards/rejected": -4.001204490661621, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 68.60209990488998, | |
| "learning_rate": 1.1731874863145142e-08, | |
| "logits/chosen": 0.5153384208679199, | |
| "logits/rejected": 0.7662239670753479, | |
| "logps/chosen": -2.8936798572540283, | |
| "logps/rejected": -3.7913997173309326, | |
| "loss": 1.1818, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -2.8936798572540283, | |
| "rewards/margins": 0.8977203369140625, | |
| "rewards/rejected": -3.7913997173309326, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 61.87029001897422, | |
| "learning_rate": 9.12094829893642e-09, | |
| "logits/chosen": 0.7237823009490967, | |
| "logits/rejected": 1.0966639518737793, | |
| "logps/chosen": -2.946139097213745, | |
| "logps/rejected": -4.009475231170654, | |
| "loss": 1.1835, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -2.946139097213745, | |
| "rewards/margins": 1.0633361339569092, | |
| "rewards/rejected": -4.009475231170654, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 64.1896852190105, | |
| "learning_rate": 6.832927412229017e-09, | |
| "logits/chosen": 0.6214802861213684, | |
| "logits/rejected": 1.0805151462554932, | |
| "logps/chosen": -3.0665371417999268, | |
| "logps/rejected": -4.000259876251221, | |
| "loss": 1.1936, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -3.0665371417999268, | |
| "rewards/margins": 0.9337228536605835, | |
| "rewards/rejected": -4.000259876251221, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 67.09074927088828, | |
| "learning_rate": 4.8708793644441086e-09, | |
| "logits/chosen": 0.6587177515029907, | |
| "logits/rejected": 0.9696787595748901, | |
| "logps/chosen": -2.8653368949890137, | |
| "logps/rejected": -3.8535594940185547, | |
| "loss": 1.1292, | |
| "rewards/accuracies": 0.78125, | |
| "rewards/chosen": -2.8653368949890137, | |
| "rewards/margins": 0.9882224798202515, | |
| "rewards/rejected": -3.8535594940185547, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 74.52088387896268, | |
| "learning_rate": 3.2374343405217884e-09, | |
| "logits/chosen": 0.6812163591384888, | |
| "logits/rejected": 1.1406841278076172, | |
| "logps/chosen": -2.954162120819092, | |
| "logps/rejected": -3.8345863819122314, | |
| "loss": 1.2136, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -2.954162120819092, | |
| "rewards/margins": 0.880423903465271, | |
| "rewards/rejected": -3.8345863819122314, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 57.70785985857215, | |
| "learning_rate": 1.9347820230782295e-09, | |
| "logits/chosen": 0.8318542242050171, | |
| "logits/rejected": 1.231487512588501, | |
| "logps/chosen": -2.9270339012145996, | |
| "logps/rejected": -3.9501945972442627, | |
| "loss": 1.1763, | |
| "rewards/accuracies": 0.75, | |
| "rewards/chosen": -2.9270339012145996, | |
| "rewards/margins": 1.0231605768203735, | |
| "rewards/rejected": -3.9501945972442627, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 66.60199874410606, | |
| "learning_rate": 9.64668657069706e-10, | |
| "logits/chosen": 0.6092048287391663, | |
| "logits/rejected": 1.2739137411117554, | |
| "logps/chosen": -2.9308724403381348, | |
| "logps/rejected": -4.049551963806152, | |
| "loss": 1.12, | |
| "rewards/accuracies": 0.8062499761581421, | |
| "rewards/chosen": -2.9308724403381348, | |
| "rewards/margins": 1.1186802387237549, | |
| "rewards/rejected": -4.049551963806152, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 72.73204833476578, | |
| "learning_rate": 3.2839470889836627e-10, | |
| "logits/chosen": 0.5203122496604919, | |
| "logits/rejected": 0.9083112478256226, | |
| "logps/chosen": -3.0056755542755127, | |
| "logps/rejected": -3.833085536956787, | |
| "loss": 1.1925, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -3.0056755542755127, | |
| "rewards/margins": 0.8274103999137878, | |
| "rewards/rejected": -3.833085536956787, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 81.94973190630537, | |
| "learning_rate": 2.6813123097352287e-11, | |
| "logits/chosen": 0.4619743824005127, | |
| "logits/rejected": 1.1849265098571777, | |
| "logps/chosen": -2.943962574005127, | |
| "logps/rejected": -3.8440794944763184, | |
| "loss": 1.2154, | |
| "rewards/accuracies": 0.737500011920929, | |
| "rewards/chosen": -2.943962574005127, | |
| "rewards/margins": 0.9001175761222839, | |
| "rewards/rejected": -3.8440794944763184, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 1.3288766333142166, | |
| "train_runtime": 7370.4172, | |
| "train_samples_per_second": 8.295, | |
| "train_steps_per_second": 0.065 | |
| } | |
| ], | |
| "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 | |
| } | |