| { | |
| "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": 11.782889401902718, | |
| "learning_rate": 6.25e-08, | |
| "logits/chosen": -1.444485068321228, | |
| "logits/rejected": -1.4456722736358643, | |
| "logps/chosen": -7.9825921058654785, | |
| "logps/rejected": -8.156225204467773, | |
| "loss": 8.9796, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.9825921058654785, | |
| "rewards/margins": 0.17363198101520538, | |
| "rewards/rejected": -8.156225204467773, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 9.749361718413306, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": -1.447454810142517, | |
| "logits/rejected": -1.4387584924697876, | |
| "logps/chosen": -8.047009468078613, | |
| "logps/rejected": -7.960066795349121, | |
| "loss": 8.9813, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -8.047009468078613, | |
| "rewards/margins": -0.08694207668304443, | |
| "rewards/rejected": -7.960066795349121, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 16.533988717004068, | |
| "learning_rate": 1.875e-07, | |
| "logits/chosen": -1.4474663734436035, | |
| "logits/rejected": -1.4442191123962402, | |
| "logps/chosen": -7.851595401763916, | |
| "logps/rejected": -7.866987705230713, | |
| "loss": 8.8899, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.851595401763916, | |
| "rewards/margins": 0.01539215724915266, | |
| "rewards/rejected": -7.866987705230713, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 13.917496227050558, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": -1.440216064453125, | |
| "logits/rejected": -1.4452197551727295, | |
| "logps/chosen": -8.178640365600586, | |
| "logps/rejected": -8.201952934265137, | |
| "loss": 9.0475, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -8.178640365600586, | |
| "rewards/margins": 0.023312047123908997, | |
| "rewards/rejected": -8.201952934265137, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 13.358220692601913, | |
| "learning_rate": 3.125e-07, | |
| "logits/chosen": -1.474110722541809, | |
| "logits/rejected": -1.463666558265686, | |
| "logps/chosen": -8.079231262207031, | |
| "logps/rejected": -7.98193883895874, | |
| "loss": 9.124, | |
| "rewards/accuracies": 0.4375, | |
| "rewards/chosen": -8.079231262207031, | |
| "rewards/margins": -0.09729210287332535, | |
| "rewards/rejected": -7.98193883895874, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 11.375823739582524, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": -1.4473092555999756, | |
| "logits/rejected": -1.4344959259033203, | |
| "logps/chosen": -7.780773162841797, | |
| "logps/rejected": -7.703455448150635, | |
| "loss": 9.0197, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -7.780773162841797, | |
| "rewards/margins": -0.07731723040342331, | |
| "rewards/rejected": -7.703455448150635, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 10.213017154182484, | |
| "learning_rate": 4.3749999999999994e-07, | |
| "logits/chosen": -1.4583995342254639, | |
| "logits/rejected": -1.431770920753479, | |
| "logps/chosen": -8.027624130249023, | |
| "logps/rejected": -7.8937225341796875, | |
| "loss": 8.9843, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -8.027624130249023, | |
| "rewards/margins": -0.13390299677848816, | |
| "rewards/rejected": -7.8937225341796875, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 10.12652288345569, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": -1.4447615146636963, | |
| "logits/rejected": -1.458698034286499, | |
| "logps/chosen": -7.983005523681641, | |
| "logps/rejected": -8.174285888671875, | |
| "loss": 9.0094, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -7.983005523681641, | |
| "rewards/margins": 0.19128072261810303, | |
| "rewards/rejected": -8.174285888671875, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 10.9885005835532, | |
| "learning_rate": 5.625e-07, | |
| "logits/chosen": -1.4630662202835083, | |
| "logits/rejected": -1.4628698825836182, | |
| "logps/chosen": -8.03730583190918, | |
| "logps/rejected": -7.831875801086426, | |
| "loss": 8.9878, | |
| "rewards/accuracies": 0.45625001192092896, | |
| "rewards/chosen": -8.03730583190918, | |
| "rewards/margins": -0.20542971789836884, | |
| "rewards/rejected": -7.831875801086426, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 13.872196323961617, | |
| "learning_rate": 5.999678242522831e-07, | |
| "logits/chosen": -1.4442825317382812, | |
| "logits/rejected": -1.4613512754440308, | |
| "logps/chosen": -8.217935562133789, | |
| "logps/rejected": -8.252190589904785, | |
| "loss": 9.0757, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -8.217935562133789, | |
| "rewards/margins": 0.03425510972738266, | |
| "rewards/rejected": -8.252190589904785, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 10.905494395813982, | |
| "learning_rate": 5.996059263493219e-07, | |
| "logits/chosen": -1.4492484331130981, | |
| "logits/rejected": -1.4467532634735107, | |
| "logps/chosen": -8.046092987060547, | |
| "logps/rejected": -8.062843322753906, | |
| "loss": 9.1036, | |
| "rewards/accuracies": 0.4312500059604645, | |
| "rewards/chosen": -8.046092987060547, | |
| "rewards/margins": 0.01675090566277504, | |
| "rewards/rejected": -8.062843322753906, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 15.995330684554988, | |
| "learning_rate": 5.988423976115163e-07, | |
| "logits/chosen": -1.443290114402771, | |
| "logits/rejected": -1.4562170505523682, | |
| "logps/chosen": -8.026491165161133, | |
| "logps/rejected": -8.317246437072754, | |
| "loss": 8.9008, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -8.026491165161133, | |
| "rewards/margins": 0.29075488448143005, | |
| "rewards/rejected": -8.317246437072754, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 24.861886587620123, | |
| "learning_rate": 5.976782615723061e-07, | |
| "logits/chosen": -1.392534613609314, | |
| "logits/rejected": -1.4108682870864868, | |
| "logps/chosen": -7.828791618347168, | |
| "logps/rejected": -8.337072372436523, | |
| "loss": 8.934, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -7.828791618347168, | |
| "rewards/margins": 0.5082817673683167, | |
| "rewards/rejected": -8.337072372436523, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 35.209412870115344, | |
| "learning_rate": 5.961150787913738e-07, | |
| "logits/chosen": -1.39071524143219, | |
| "logits/rejected": -1.3853540420532227, | |
| "logps/chosen": -7.945198059082031, | |
| "logps/rejected": -8.038311004638672, | |
| "loss": 8.9653, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -7.945198059082031, | |
| "rewards/margins": 0.0931134819984436, | |
| "rewards/rejected": -8.038311004638672, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 12.413941901156766, | |
| "learning_rate": 5.941549447626671e-07, | |
| "logits/chosen": -1.3913167715072632, | |
| "logits/rejected": -1.3984179496765137, | |
| "logps/chosen": -7.823273658752441, | |
| "logps/rejected": -7.864768981933594, | |
| "loss": 8.9142, | |
| "rewards/accuracies": 0.48124998807907104, | |
| "rewards/chosen": -7.823273658752441, | |
| "rewards/margins": 0.04149458184838295, | |
| "rewards/rejected": -7.864768981933594, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 21.221667512587725, | |
| "learning_rate": 5.918004871053251e-07, | |
| "logits/chosen": -1.3923091888427734, | |
| "logits/rejected": -1.4085341691970825, | |
| "logps/chosen": -7.852835178375244, | |
| "logps/rejected": -7.9230217933654785, | |
| "loss": 8.9088, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -7.852835178375244, | |
| "rewards/margins": 0.07018764317035675, | |
| "rewards/rejected": -7.9230217933654785, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 12.603711372215182, | |
| "learning_rate": 5.890548620412763e-07, | |
| "logits/chosen": -1.4011937379837036, | |
| "logits/rejected": -1.39864182472229, | |
| "logps/chosen": -7.970945835113525, | |
| "logps/rejected": -8.160429000854492, | |
| "loss": 9.0488, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -7.970945835113525, | |
| "rewards/margins": 0.18948234617710114, | |
| "rewards/rejected": -8.160429000854492, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 13.164098047063113, | |
| "learning_rate": 5.859217501642258e-07, | |
| "logits/chosen": -1.375800371170044, | |
| "logits/rejected": -1.389070749282837, | |
| "logps/chosen": -7.946028232574463, | |
| "logps/rejected": -8.130967140197754, | |
| "loss": 9.0141, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -7.946028232574463, | |
| "rewards/margins": 0.18493881821632385, | |
| "rewards/rejected": -8.130967140197754, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 11.129043830781203, | |
| "learning_rate": 5.824053515057091e-07, | |
| "logits/chosen": -1.384723424911499, | |
| "logits/rejected": -1.3767420053482056, | |
| "logps/chosen": -8.055198669433594, | |
| "logps/rejected": -7.921385288238525, | |
| "loss": 9.0835, | |
| "rewards/accuracies": 0.4124999940395355, | |
| "rewards/chosen": -8.055198669433594, | |
| "rewards/margins": -0.13381320238113403, | |
| "rewards/rejected": -7.921385288238525, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 19.959846628166616, | |
| "learning_rate": 5.785103799048218e-07, | |
| "logits/chosen": -1.4132357835769653, | |
| "logits/rejected": -1.418881893157959, | |
| "logps/chosen": -8.033044815063477, | |
| "logps/rejected": -8.07997989654541, | |
| "loss": 9.0153, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -8.033044815063477, | |
| "rewards/margins": 0.04693456366658211, | |
| "rewards/rejected": -8.07997989654541, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 12.843923135972, | |
| "learning_rate": 5.742420566891749e-07, | |
| "logits/chosen": -1.413010835647583, | |
| "logits/rejected": -1.4074172973632812, | |
| "logps/chosen": -7.718166351318359, | |
| "logps/rejected": -7.9243879318237305, | |
| "loss": 8.9445, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -7.718166351318359, | |
| "rewards/margins": 0.206221342086792, | |
| "rewards/rejected": -7.9243879318237305, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 13.825481188162163, | |
| "learning_rate": 5.696061036755478e-07, | |
| "logits/chosen": -1.4453760385513306, | |
| "logits/rejected": -1.4452683925628662, | |
| "logps/chosen": -7.982637882232666, | |
| "logps/rejected": -8.220747947692871, | |
| "loss": 9.0144, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -7.982637882232666, | |
| "rewards/margins": 0.23810970783233643, | |
| "rewards/rejected": -8.220747947692871, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 1525.1356967991103, | |
| "learning_rate": 5.64608735499618e-07, | |
| "logits/chosen": -1.3860673904418945, | |
| "logits/rejected": -1.3894257545471191, | |
| "logps/chosen": -7.8776044845581055, | |
| "logps/rejected": -8.189804077148438, | |
| "loss": 8.9598, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -7.8776044845581055, | |
| "rewards/margins": 0.31219929456710815, | |
| "rewards/rejected": -8.189804077148438, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 13.913132246096096, | |
| "learning_rate": 5.592566512850545e-07, | |
| "logits/chosen": -1.3590507507324219, | |
| "logits/rejected": -1.3622348308563232, | |
| "logps/chosen": -8.100934982299805, | |
| "logps/rejected": -8.155590057373047, | |
| "loss": 8.9501, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -8.100934982299805, | |
| "rewards/margins": 0.054654598236083984, | |
| "rewards/rejected": -8.155590057373047, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 14.57715484351377, | |
| "learning_rate": 5.535570256631384e-07, | |
| "logits/chosen": -1.4173157215118408, | |
| "logits/rejected": -1.411921739578247, | |
| "logps/chosen": -8.191034317016602, | |
| "logps/rejected": -8.077339172363281, | |
| "loss": 9.0651, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -8.191034317016602, | |
| "rewards/margins": -0.11369502544403076, | |
| "rewards/rejected": -8.077339172363281, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 14.54742842440625, | |
| "learning_rate": 5.475174991549528e-07, | |
| "logits/chosen": -1.37632417678833, | |
| "logits/rejected": -1.3858749866485596, | |
| "logps/chosen": -8.046875953674316, | |
| "logps/rejected": -8.172870635986328, | |
| "loss": 8.9777, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -8.046875953674316, | |
| "rewards/margins": 0.12599456310272217, | |
| "rewards/rejected": -8.172870635986328, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 18.256408848890032, | |
| "learning_rate": 5.411461679290317e-07, | |
| "logits/chosen": -1.3864247798919678, | |
| "logits/rejected": -1.4004995822906494, | |
| "logps/chosen": -7.979268550872803, | |
| "logps/rejected": -8.406595230102539, | |
| "loss": 8.9672, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -7.979268550872803, | |
| "rewards/margins": 0.4273262023925781, | |
| "rewards/rejected": -8.406595230102539, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 14.27521931097187, | |
| "learning_rate": 5.34451572948201e-07, | |
| "logits/chosen": -1.4093233346939087, | |
| "logits/rejected": -1.4172067642211914, | |
| "logps/chosen": -7.903810977935791, | |
| "logps/rejected": -7.975949287414551, | |
| "loss": 8.9533, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -7.903810977935791, | |
| "rewards/margins": 0.07213909924030304, | |
| "rewards/rejected": -7.975949287414551, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 10.523105376926537, | |
| "learning_rate": 5.274426885201582e-07, | |
| "logits/chosen": -1.4147297143936157, | |
| "logits/rejected": -1.4396823644638062, | |
| "logps/chosen": -7.8977460861206055, | |
| "logps/rejected": -8.05931568145752, | |
| "loss": 8.915, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -7.8977460861206055, | |
| "rewards/margins": 0.16156847774982452, | |
| "rewards/rejected": -8.05931568145752, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 14.122907500033074, | |
| "learning_rate": 5.201289102671411e-07, | |
| "logits/chosen": -1.4332246780395508, | |
| "logits/rejected": -1.436842679977417, | |
| "logps/chosen": -7.895875453948975, | |
| "logps/rejected": -8.0299072265625, | |
| "loss": 8.9785, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.895875453948975, | |
| "rewards/margins": 0.13403132557868958, | |
| "rewards/rejected": -8.0299072265625, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 12.92310774863363, | |
| "learning_rate": 5.12520042530811e-07, | |
| "logits/chosen": -1.402719259262085, | |
| "logits/rejected": -1.3787992000579834, | |
| "logps/chosen": -7.979246616363525, | |
| "logps/rejected": -7.966032981872559, | |
| "loss": 9.0256, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -7.979246616363525, | |
| "rewards/margins": -0.013212683610618114, | |
| "rewards/rejected": -7.966032981872559, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 15.237628673130487, | |
| "learning_rate": 5.046262852292346e-07, | |
| "logits/chosen": -1.3872135877609253, | |
| "logits/rejected": -1.395935297012329, | |
| "logps/chosen": -8.034635543823242, | |
| "logps/rejected": -8.069303512573242, | |
| "loss": 9.0268, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -8.034635543823242, | |
| "rewards/margins": 0.03466759994626045, | |
| "rewards/rejected": -8.069303512573242, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 11.298592435998462, | |
| "learning_rate": 4.964582201835856e-07, | |
| "logits/chosen": -1.396750569343567, | |
| "logits/rejected": -1.3891570568084717, | |
| "logps/chosen": -7.99398946762085, | |
| "logps/rejected": -8.040716171264648, | |
| "loss": 9.0073, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -7.99398946762085, | |
| "rewards/margins": 0.04672648385167122, | |
| "rewards/rejected": -8.040716171264648, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 12.492415372530475, | |
| "learning_rate": 4.880267969328908e-07, | |
| "logits/chosen": -1.3683674335479736, | |
| "logits/rejected": -1.3726252317428589, | |
| "logps/chosen": -8.114925384521484, | |
| "logps/rejected": -8.097586631774902, | |
| "loss": 9.0856, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -8.114925384521484, | |
| "rewards/margins": -0.01733933761715889, | |
| "rewards/rejected": -8.097586631774902, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 12.610496367889976, | |
| "learning_rate": 4.793433180558423e-07, | |
| "logits/chosen": -1.3843915462493896, | |
| "logits/rejected": -1.3853034973144531, | |
| "logps/chosen": -7.956766605377197, | |
| "logps/rejected": -7.944356441497803, | |
| "loss": 9.0054, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -7.956766605377197, | |
| "rewards/margins": -0.012410154566168785, | |
| "rewards/rejected": -7.944356441497803, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 15.598692092405715, | |
| "learning_rate": 4.704194240193467e-07, | |
| "logits/chosen": -1.3554438352584839, | |
| "logits/rejected": -1.372804880142212, | |
| "logps/chosen": -8.031749725341797, | |
| "logps/rejected": -8.155205726623535, | |
| "loss": 8.9878, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -8.031749725341797, | |
| "rewards/margins": 0.12345610558986664, | |
| "rewards/rejected": -8.155205726623535, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 13.474501957199323, | |
| "learning_rate": 4.6126707757412686e-07, | |
| "logits/chosen": -1.3345744609832764, | |
| "logits/rejected": -1.3397581577301025, | |
| "logps/chosen": -7.977494716644287, | |
| "logps/rejected": -8.02932357788086, | |
| "loss": 8.9482, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -7.977494716644287, | |
| "rewards/margins": 0.051828037947416306, | |
| "rewards/rejected": -8.02932357788086, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 11.929724403265839, | |
| "learning_rate": 4.5189854771829086e-07, | |
| "logits/chosen": -1.3528499603271484, | |
| "logits/rejected": -1.3492704629898071, | |
| "logps/chosen": -7.803788661956787, | |
| "logps/rejected": -7.93734073638916, | |
| "loss": 8.9516, | |
| "rewards/accuracies": 0.44999998807907104, | |
| "rewards/chosen": -7.803788661956787, | |
| "rewards/margins": 0.1335521936416626, | |
| "rewards/rejected": -7.93734073638916, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 14.327437395286285, | |
| "learning_rate": 4.4232639325036807e-07, | |
| "logits/chosen": -1.3263393640518188, | |
| "logits/rejected": -1.3331449031829834, | |
| "logps/chosen": -8.183530807495117, | |
| "logps/rejected": -8.074382781982422, | |
| "loss": 9.054, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -8.183530807495117, | |
| "rewards/margins": -0.10914800316095352, | |
| "rewards/rejected": -8.074382781982422, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 12.623357323327125, | |
| "learning_rate": 4.32563445933859e-07, | |
| "logits/chosen": -1.3866218328475952, | |
| "logits/rejected": -1.376103401184082, | |
| "logps/chosen": -7.869284152984619, | |
| "logps/rejected": -7.980343818664551, | |
| "loss": 9.0216, | |
| "rewards/accuracies": 0.4749999940395355, | |
| "rewards/chosen": -7.869284152984619, | |
| "rewards/margins": 0.11105932295322418, | |
| "rewards/rejected": -7.980343818664551, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 15.673764218634288, | |
| "learning_rate": 4.226227932958664e-07, | |
| "logits/chosen": -1.3467977046966553, | |
| "logits/rejected": -1.3465808629989624, | |
| "logps/chosen": -7.946604251861572, | |
| "logps/rejected": -8.12873363494873, | |
| "loss": 8.9418, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -7.946604251861572, | |
| "rewards/margins": 0.18213000893592834, | |
| "rewards/rejected": -8.12873363494873, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 20.82547017360473, | |
| "learning_rate": 4.1251776108286854e-07, | |
| "logits/chosen": -1.3276244401931763, | |
| "logits/rejected": -1.3366806507110596, | |
| "logps/chosen": -7.942746639251709, | |
| "logps/rejected": -8.075704574584961, | |
| "loss": 8.992, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -7.942746639251709, | |
| "rewards/margins": 0.13295890390872955, | |
| "rewards/rejected": -8.075704574584961, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 11.77567830972404, | |
| "learning_rate": 4.022618953971514e-07, | |
| "logits/chosen": -1.3542811870574951, | |
| "logits/rejected": -1.3621467351913452, | |
| "logps/chosen": -7.741019248962402, | |
| "logps/rejected": -8.169224739074707, | |
| "loss": 8.9028, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -7.741019248962402, | |
| "rewards/margins": 0.42820531129837036, | |
| "rewards/rejected": -8.169224739074707, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 13.792748846310712, | |
| "learning_rate": 3.918689445378477e-07, | |
| "logits/chosen": -1.3647044897079468, | |
| "logits/rejected": -1.3888493776321411, | |
| "logps/chosen": -7.679605960845947, | |
| "logps/rejected": -7.820864677429199, | |
| "loss": 9.0059, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -7.679605960845947, | |
| "rewards/margins": 0.14125962555408478, | |
| "rewards/rejected": -7.820864677429199, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 10.698537268464346, | |
| "learning_rate": 3.813528405709251e-07, | |
| "logits/chosen": -1.3668994903564453, | |
| "logits/rejected": -1.370476484298706, | |
| "logps/chosen": -7.723212242126465, | |
| "logps/rejected": -7.974145412445068, | |
| "loss": 8.9131, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -7.723212242126465, | |
| "rewards/margins": 0.25093379616737366, | |
| "rewards/rejected": -7.974145412445068, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 12.476277662413903, | |
| "learning_rate": 3.707276806528282e-07, | |
| "logits/chosen": -1.37067449092865, | |
| "logits/rejected": -1.3700437545776367, | |
| "logps/chosen": -8.093690872192383, | |
| "logps/rejected": -8.251599311828613, | |
| "loss": 9.068, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -8.093690872192383, | |
| "rewards/margins": 0.157908633351326, | |
| "rewards/rejected": -8.251599311828613, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 12.703214615987921, | |
| "learning_rate": 3.6000770813281334e-07, | |
| "logits/chosen": -1.3918092250823975, | |
| "logits/rejected": -1.3941457271575928, | |
| "logps/chosen": -7.891854286193848, | |
| "logps/rejected": -8.121790885925293, | |
| "loss": 8.9911, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -7.891854286193848, | |
| "rewards/margins": 0.22993668913841248, | |
| "rewards/rejected": -8.121790885925293, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 18.912862114031174, | |
| "learning_rate": 3.4920729345930654e-07, | |
| "logits/chosen": -1.3598334789276123, | |
| "logits/rejected": -1.3656227588653564, | |
| "logps/chosen": -7.972811698913574, | |
| "logps/rejected": -8.120051383972168, | |
| "loss": 9.0708, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -7.972811698913574, | |
| "rewards/margins": 0.14723989367485046, | |
| "rewards/rejected": -8.120051383972168, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 15.322600609417346, | |
| "learning_rate": 3.383409149158814e-07, | |
| "logits/chosen": -1.3441493511199951, | |
| "logits/rejected": -1.3492319583892822, | |
| "logps/chosen": -8.092975616455078, | |
| "logps/rejected": -8.160036087036133, | |
| "loss": 8.9194, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -8.092975616455078, | |
| "rewards/margins": 0.06706006824970245, | |
| "rewards/rejected": -8.160036087036133, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 16.724538535729355, | |
| "learning_rate": 3.2742313921268035e-07, | |
| "logits/chosen": -1.3152296543121338, | |
| "logits/rejected": -1.3239524364471436, | |
| "logps/chosen": -7.889418601989746, | |
| "logps/rejected": -8.20849323272705, | |
| "loss": 8.8184, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -7.889418601989746, | |
| "rewards/margins": 0.31907448172569275, | |
| "rewards/rejected": -8.20849323272705, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 12.327867536896116, | |
| "learning_rate": 3.1646860195929825e-07, | |
| "logits/chosen": -1.3065917491912842, | |
| "logits/rejected": -1.3107439279556274, | |
| "logps/chosen": -8.116486549377441, | |
| "logps/rejected": -8.308655738830566, | |
| "loss": 8.9949, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -8.116486549377441, | |
| "rewards/margins": 0.19216908514499664, | |
| "rewards/rejected": -8.308655738830566, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 14.17754725379555, | |
| "learning_rate": 3.054919880453032e-07, | |
| "logits/chosen": -1.246124029159546, | |
| "logits/rejected": -1.2508999109268188, | |
| "logps/chosen": -7.7648186683654785, | |
| "logps/rejected": -8.22431755065918, | |
| "loss": 8.941, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -7.7648186683654785, | |
| "rewards/margins": 0.4594977796077728, | |
| "rewards/rejected": -8.22431755065918, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 11.969966746660198, | |
| "learning_rate": 2.9450801195469686e-07, | |
| "logits/chosen": -1.3018732070922852, | |
| "logits/rejected": -1.3149497509002686, | |
| "logps/chosen": -7.904818058013916, | |
| "logps/rejected": -8.152360916137695, | |
| "loss": 8.9657, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -7.904818058013916, | |
| "rewards/margins": 0.24754443764686584, | |
| "rewards/rejected": -8.152360916137695, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 14.47186665684816, | |
| "learning_rate": 2.835313980407017e-07, | |
| "logits/chosen": -1.3108150959014893, | |
| "logits/rejected": -1.288703441619873, | |
| "logps/chosen": -8.249927520751953, | |
| "logps/rejected": -8.318041801452637, | |
| "loss": 9.0073, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -8.249927520751953, | |
| "rewards/margins": 0.06811434030532837, | |
| "rewards/rejected": -8.318041801452637, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 26.602745593974163, | |
| "learning_rate": 2.7257686078731973e-07, | |
| "logits/chosen": -1.337909460067749, | |
| "logits/rejected": -1.348547339439392, | |
| "logps/chosen": -7.881032466888428, | |
| "logps/rejected": -8.068848609924316, | |
| "loss": 8.8981, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -7.881032466888428, | |
| "rewards/margins": 0.18781575560569763, | |
| "rewards/rejected": -8.068848609924316, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 14.906273538361356, | |
| "learning_rate": 2.6165908508411857e-07, | |
| "logits/chosen": -1.3503994941711426, | |
| "logits/rejected": -1.3676143884658813, | |
| "logps/chosen": -7.861943244934082, | |
| "logps/rejected": -8.101309776306152, | |
| "loss": 8.9213, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -7.861943244934082, | |
| "rewards/margins": 0.23936741054058075, | |
| "rewards/rejected": -8.101309776306152, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 14.643252229490672, | |
| "learning_rate": 2.5079270654069354e-07, | |
| "logits/chosen": -1.3024542331695557, | |
| "logits/rejected": -1.3081843852996826, | |
| "logps/chosen": -7.836719512939453, | |
| "logps/rejected": -8.08849048614502, | |
| "loss": 8.8721, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -7.836719512939453, | |
| "rewards/margins": 0.251770943403244, | |
| "rewards/rejected": -8.08849048614502, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 12.350106404715637, | |
| "learning_rate": 2.399922918671867e-07, | |
| "logits/chosen": -1.337571620941162, | |
| "logits/rejected": -1.3552089929580688, | |
| "logps/chosen": -7.821458339691162, | |
| "logps/rejected": -8.146204948425293, | |
| "loss": 8.9032, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -7.821458339691162, | |
| "rewards/margins": 0.3247470557689667, | |
| "rewards/rejected": -8.146204948425293, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 13.544262102627407, | |
| "learning_rate": 2.2927231934717176e-07, | |
| "logits/chosen": -1.331067442893982, | |
| "logits/rejected": -1.3430246114730835, | |
| "logps/chosen": -7.9300737380981445, | |
| "logps/rejected": -8.060845375061035, | |
| "loss": 8.9735, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -7.9300737380981445, | |
| "rewards/margins": 0.1307719349861145, | |
| "rewards/rejected": -8.060845375061035, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 12.721635836612304, | |
| "learning_rate": 2.1864715942907487e-07, | |
| "logits/chosen": -1.299328088760376, | |
| "logits/rejected": -1.3065472841262817, | |
| "logps/chosen": -7.961094856262207, | |
| "logps/rejected": -8.206907272338867, | |
| "loss": 8.9027, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -7.961094856262207, | |
| "rewards/margins": 0.24581179022789001, | |
| "rewards/rejected": -8.206907272338867, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 12.545308362098348, | |
| "learning_rate": 2.081310554621522e-07, | |
| "logits/chosen": -1.3111393451690674, | |
| "logits/rejected": -1.335069179534912, | |
| "logps/chosen": -8.182366371154785, | |
| "logps/rejected": -8.390935897827148, | |
| "loss": 9.0314, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -8.182366371154785, | |
| "rewards/margins": 0.20856896042823792, | |
| "rewards/rejected": -8.390935897827148, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 15.575558913925711, | |
| "learning_rate": 1.9773810460284862e-07, | |
| "logits/chosen": -1.3477294445037842, | |
| "logits/rejected": -1.3550546169281006, | |
| "logps/chosen": -8.07054328918457, | |
| "logps/rejected": -8.061942100524902, | |
| "loss": 9.0612, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -8.07054328918457, | |
| "rewards/margins": -0.008599767461419106, | |
| "rewards/rejected": -8.061942100524902, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 20.765211421302535, | |
| "learning_rate": 1.874822389171314e-07, | |
| "logits/chosen": -1.3256926536560059, | |
| "logits/rejected": -1.3409112691879272, | |
| "logps/chosen": -7.885645389556885, | |
| "logps/rejected": -8.124526977539062, | |
| "loss": 8.8864, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -7.885645389556885, | |
| "rewards/margins": 0.2388812005519867, | |
| "rewards/rejected": -8.124526977539062, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 14.208218028523063, | |
| "learning_rate": 1.7737720670413356e-07, | |
| "logits/chosen": -1.344118595123291, | |
| "logits/rejected": -1.336096167564392, | |
| "logps/chosen": -8.159255981445312, | |
| "logps/rejected": -8.0567045211792, | |
| "loss": 8.9837, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -8.159255981445312, | |
| "rewards/margins": -0.10255154222249985, | |
| "rewards/rejected": -8.0567045211792, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 14.61227257116642, | |
| "learning_rate": 1.6743655406614095e-07, | |
| "logits/chosen": -1.340541958808899, | |
| "logits/rejected": -1.3474371433258057, | |
| "logps/chosen": -8.056330680847168, | |
| "logps/rejected": -8.348928451538086, | |
| "loss": 8.9222, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -8.056330680847168, | |
| "rewards/margins": 0.29259705543518066, | |
| "rewards/rejected": -8.348928451538086, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 13.778075151913542, | |
| "learning_rate": 1.5767360674963198e-07, | |
| "logits/chosen": -1.3218133449554443, | |
| "logits/rejected": -1.3337442874908447, | |
| "logps/chosen": -7.961134910583496, | |
| "logps/rejected": -7.996614933013916, | |
| "loss": 9.0247, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -7.961134910583496, | |
| "rewards/margins": 0.035479746758937836, | |
| "rewards/rejected": -7.996614933013916, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 13.684086792814428, | |
| "learning_rate": 1.4810145228170922e-07, | |
| "logits/chosen": -1.3398381471633911, | |
| "logits/rejected": -1.3437585830688477, | |
| "logps/chosen": -7.856637001037598, | |
| "logps/rejected": -8.111886978149414, | |
| "loss": 8.8913, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -7.856637001037598, | |
| "rewards/margins": 0.25525030493736267, | |
| "rewards/rejected": -8.111886978149414, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 15.39649445200101, | |
| "learning_rate": 1.3873292242587306e-07, | |
| "logits/chosen": -1.3376450538635254, | |
| "logits/rejected": -1.3476964235305786, | |
| "logps/chosen": -8.228338241577148, | |
| "logps/rejected": -8.340727806091309, | |
| "loss": 9.0269, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -8.228338241577148, | |
| "rewards/margins": 0.11239071190357208, | |
| "rewards/rejected": -8.340727806091309, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 15.302013253785537, | |
| "learning_rate": 1.295805759806533e-07, | |
| "logits/chosen": -1.3724461793899536, | |
| "logits/rejected": -1.3841075897216797, | |
| "logps/chosen": -8.054750442504883, | |
| "logps/rejected": -8.403682708740234, | |
| "loss": 9.0089, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -8.054750442504883, | |
| "rewards/margins": 0.3489326238632202, | |
| "rewards/rejected": -8.403682708740234, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 18.608453972243662, | |
| "learning_rate": 1.2065668194415777e-07, | |
| "logits/chosen": -1.3417284488677979, | |
| "logits/rejected": -1.3348530530929565, | |
| "logps/chosen": -7.915482997894287, | |
| "logps/rejected": -8.044729232788086, | |
| "loss": 8.9016, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -7.915482997894287, | |
| "rewards/margins": 0.12924641370773315, | |
| "rewards/rejected": -8.044729232788086, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 14.900748845819772, | |
| "learning_rate": 1.1197320306710923e-07, | |
| "logits/chosen": -1.3621351718902588, | |
| "logits/rejected": -1.3541442155838013, | |
| "logps/chosen": -8.007196426391602, | |
| "logps/rejected": -7.965734004974365, | |
| "loss": 8.9062, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -8.007196426391602, | |
| "rewards/margins": -0.04146287590265274, | |
| "rewards/rejected": -7.965734004974365, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 11.569520650790327, | |
| "learning_rate": 1.035417798164145e-07, | |
| "logits/chosen": -1.3260619640350342, | |
| "logits/rejected": -1.3356263637542725, | |
| "logps/chosen": -7.753990173339844, | |
| "logps/rejected": -8.039525985717773, | |
| "loss": 8.8536, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -7.753990173339844, | |
| "rewards/margins": 0.2855362296104431, | |
| "rewards/rejected": -8.039525985717773, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 13.480030507608214, | |
| "learning_rate": 9.537371477076535e-08, | |
| "logits/chosen": -1.2944828271865845, | |
| "logits/rejected": -1.2956254482269287, | |
| "logps/chosen": -7.926826477050781, | |
| "logps/rejected": -7.9895477294921875, | |
| "loss": 8.9487, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -7.926826477050781, | |
| "rewards/margins": 0.06272158026695251, | |
| "rewards/rejected": -7.9895477294921875, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 14.634365970472302, | |
| "learning_rate": 8.747995746918898e-08, | |
| "logits/chosen": -1.3467233180999756, | |
| "logits/rejected": -1.3351846933364868, | |
| "logps/chosen": -8.043527603149414, | |
| "logps/rejected": -8.186015129089355, | |
| "loss": 8.9627, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -8.043527603149414, | |
| "rewards/margins": 0.14248715341091156, | |
| "rewards/rejected": -8.186015129089355, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 28.38170473677795, | |
| "learning_rate": 7.987108973285888e-08, | |
| "logits/chosen": -1.3258306980133057, | |
| "logits/rejected": -1.3155487775802612, | |
| "logps/chosen": -8.005027770996094, | |
| "logps/rejected": -8.246636390686035, | |
| "loss": 8.9413, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -8.005027770996094, | |
| "rewards/margins": 0.24160809814929962, | |
| "rewards/rejected": -8.246636390686035, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 13.711915418794124, | |
| "learning_rate": 7.255731147984174e-08, | |
| "logits/chosen": -1.3438084125518799, | |
| "logits/rejected": -1.297163963317871, | |
| "logps/chosen": -8.208559036254883, | |
| "logps/rejected": -8.363499641418457, | |
| "loss": 8.942, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -8.208559036254883, | |
| "rewards/margins": 0.15493938326835632, | |
| "rewards/rejected": -8.363499641418457, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 13.642711731891415, | |
| "learning_rate": 6.554842705179898e-08, | |
| "logits/chosen": -1.3352845907211304, | |
| "logits/rejected": -1.3314430713653564, | |
| "logps/chosen": -8.112469673156738, | |
| "logps/rejected": -8.209820747375488, | |
| "loss": 8.9588, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -8.112469673156738, | |
| "rewards/margins": 0.09735036641359329, | |
| "rewards/rejected": -8.209820747375488, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 14.269345053816819, | |
| "learning_rate": 5.885383207096832e-08, | |
| "logits/chosen": -1.3467012643814087, | |
| "logits/rejected": -1.3490493297576904, | |
| "logps/chosen": -7.833376884460449, | |
| "logps/rejected": -8.030352592468262, | |
| "loss": 8.8689, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -7.833376884460449, | |
| "rewards/margins": 0.1969761848449707, | |
| "rewards/rejected": -8.030352592468262, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 16.14618551872646, | |
| "learning_rate": 5.2482500845047165e-08, | |
| "logits/chosen": -1.3177175521850586, | |
| "logits/rejected": -1.3296372890472412, | |
| "logps/chosen": -7.635066032409668, | |
| "logps/rejected": -7.791895866394043, | |
| "loss": 8.9076, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -7.635066032409668, | |
| "rewards/margins": 0.15682990849018097, | |
| "rewards/rejected": -7.791895866394043, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 12.278193076130206, | |
| "learning_rate": 4.644297433686162e-08, | |
| "logits/chosen": -1.3246910572052002, | |
| "logits/rejected": -1.315019965171814, | |
| "logps/chosen": -7.837827205657959, | |
| "logps/rejected": -7.908313751220703, | |
| "loss": 8.951, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -7.837827205657959, | |
| "rewards/margins": 0.0704866498708725, | |
| "rewards/rejected": -7.908313751220703, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 12.786235556241849, | |
| "learning_rate": 4.074334871494558e-08, | |
| "logits/chosen": -1.3545995950698853, | |
| "logits/rejected": -1.3624496459960938, | |
| "logps/chosen": -8.024687767028809, | |
| "logps/rejected": -8.172109603881836, | |
| "loss": 8.9198, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -8.024687767028809, | |
| "rewards/margins": 0.1474229097366333, | |
| "rewards/rejected": -8.172109603881836, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 30.758577870183032, | |
| "learning_rate": 3.5391264500382e-08, | |
| "logits/chosen": -1.360478401184082, | |
| "logits/rejected": -1.3552910089492798, | |
| "logps/chosen": -7.844922065734863, | |
| "logps/rejected": -7.852625846862793, | |
| "loss": 8.8997, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -7.844922065734863, | |
| "rewards/margins": 0.0077047706581652164, | |
| "rewards/rejected": -7.852625846862793, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 14.260374307768236, | |
| "learning_rate": 3.0393896324452226e-08, | |
| "logits/chosen": -1.372036337852478, | |
| "logits/rejected": -1.3762390613555908, | |
| "logps/chosen": -7.982748508453369, | |
| "logps/rejected": -8.225188255310059, | |
| "loss": 8.9748, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -7.982748508453369, | |
| "rewards/margins": 0.24244041740894318, | |
| "rewards/rejected": -8.225188255310059, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 14.09879602927017, | |
| "learning_rate": 2.5757943310825026e-08, | |
| "logits/chosen": -1.3225996494293213, | |
| "logits/rejected": -1.3161330223083496, | |
| "logps/chosen": -7.865872859954834, | |
| "logps/rejected": -7.931491851806641, | |
| "loss": 8.9372, | |
| "rewards/accuracies": 0.5, | |
| "rewards/chosen": -7.865872859954834, | |
| "rewards/margins": 0.06561894714832306, | |
| "rewards/rejected": -7.931491851806641, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 13.351401390808332, | |
| "learning_rate": 2.148962009517823e-08, | |
| "logits/chosen": -1.342071771621704, | |
| "logits/rejected": -1.337024450302124, | |
| "logps/chosen": -8.03447151184082, | |
| "logps/rejected": -8.085325241088867, | |
| "loss": 8.9767, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -8.03447151184082, | |
| "rewards/margins": 0.050852321088314056, | |
| "rewards/rejected": -8.085325241088867, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 16.02345328859732, | |
| "learning_rate": 1.759464849429082e-08, | |
| "logits/chosen": -1.3405394554138184, | |
| "logits/rejected": -1.3419816493988037, | |
| "logps/chosen": -7.878898620605469, | |
| "logps/rejected": -8.002215385437012, | |
| "loss": 8.9292, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -7.878898620605469, | |
| "rewards/margins": 0.123316690325737, | |
| "rewards/rejected": -8.002215385437012, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 13.971661978504134, | |
| "learning_rate": 1.4078249835774169e-08, | |
| "logits/chosen": -1.3646373748779297, | |
| "logits/rejected": -1.3699538707733154, | |
| "logps/chosen": -7.937603950500488, | |
| "logps/rejected": -8.069661140441895, | |
| "loss": 8.8372, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -7.937603950500488, | |
| "rewards/margins": 0.13205692172050476, | |
| "rewards/rejected": -8.069661140441895, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 13.636923891581842, | |
| "learning_rate": 1.0945137958723705e-08, | |
| "logits/chosen": -1.3303980827331543, | |
| "logits/rejected": -1.3274564743041992, | |
| "logps/chosen": -8.00455379486084, | |
| "logps/rejected": -8.096671104431152, | |
| "loss": 8.9997, | |
| "rewards/accuracies": 0.48750001192092896, | |
| "rewards/chosen": -8.00455379486084, | |
| "rewards/margins": 0.09211695194244385, | |
| "rewards/rejected": -8.096671104431152, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 14.66331138432002, | |
| "learning_rate": 8.19951289467482e-09, | |
| "logits/chosen": -1.3527616262435913, | |
| "logits/rejected": -1.352975606918335, | |
| "logps/chosen": -7.898123741149902, | |
| "logps/rejected": -8.020647048950195, | |
| "loss": 8.9114, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -7.898123741149902, | |
| "rewards/margins": 0.1225227564573288, | |
| "rewards/rejected": -8.020647048950195, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 34.64920022108061, | |
| "learning_rate": 5.84505523733293e-09, | |
| "logits/chosen": -1.3027703762054443, | |
| "logits/rejected": -1.2922091484069824, | |
| "logps/chosen": -8.017878532409668, | |
| "logps/rejected": -8.019991874694824, | |
| "loss": 9.0038, | |
| "rewards/accuracies": 0.46875, | |
| "rewards/chosen": -8.017878532409668, | |
| "rewards/margins": 0.0021121830213814974, | |
| "rewards/rejected": -8.019991874694824, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 14.096689301269398, | |
| "learning_rate": 3.8849212086261466e-09, | |
| "logits/chosen": -1.3568954467773438, | |
| "logits/rejected": -1.345536231994629, | |
| "logps/chosen": -7.817251682281494, | |
| "logps/rejected": -8.18480110168457, | |
| "loss": 8.9022, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -7.817251682281494, | |
| "rewards/margins": 0.3675496578216553, | |
| "rewards/rejected": -8.18480110168457, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 21.948748802651522, | |
| "learning_rate": 2.3217384276938756e-09, | |
| "logits/chosen": -1.3387937545776367, | |
| "logits/rejected": -1.349258542060852, | |
| "logps/chosen": -7.9868292808532715, | |
| "logps/rejected": -8.197335243225098, | |
| "loss": 8.8854, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -7.9868292808532715, | |
| "rewards/margins": 0.21050508320331573, | |
| "rewards/rejected": -8.197335243225098, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 12.715751305789052, | |
| "learning_rate": 1.1576023884836472e-09, | |
| "logits/chosen": -1.3674533367156982, | |
| "logits/rejected": -1.3665874004364014, | |
| "logps/chosen": -8.10934066772461, | |
| "logps/rejected": -8.27099323272705, | |
| "loss": 8.9853, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -8.10934066772461, | |
| "rewards/margins": 0.16165266931056976, | |
| "rewards/rejected": -8.27099323272705, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 13.20358280327505, | |
| "learning_rate": 3.940736506780395e-10, | |
| "logits/chosen": -1.348550796508789, | |
| "logits/rejected": -1.3657060861587524, | |
| "logps/chosen": -7.707891941070557, | |
| "logps/rejected": -7.990015983581543, | |
| "loss": 8.9804, | |
| "rewards/accuracies": 0.543749988079071, | |
| "rewards/chosen": -7.707891941070557, | |
| "rewards/margins": 0.2821243703365326, | |
| "rewards/rejected": -7.990015983581543, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 17.668181816444864, | |
| "learning_rate": 3.2175747716822744e-11, | |
| "logits/chosen": -1.3433798551559448, | |
| "logits/rejected": -1.3304665088653564, | |
| "logps/chosen": -8.101046562194824, | |
| "logps/rejected": -8.15410041809082, | |
| "loss": 8.9813, | |
| "rewards/accuracies": 0.5375000238418579, | |
| "rewards/chosen": -8.101046562194824, | |
| "rewards/margins": 0.05305204540491104, | |
| "rewards/rejected": -8.15410041809082, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 8.967987340451286, | |
| "train_runtime": 8184.2286, | |
| "train_samples_per_second": 7.47, | |
| "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 | |
| } | |