| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9978586723768736, | |
| "eval_steps": 500, | |
| "global_step": 233, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 5.59375, | |
| "learning_rate": 4.977309719247571e-06, | |
| "log_odds_chosen": 0.06703004986047745, | |
| "log_odds_ratio": -0.7092211842536926, | |
| "logits/chosen": -3.2759811878204346, | |
| "logits/rejected": -3.2908482551574707, | |
| "logps/chosen": -0.7329679727554321, | |
| "logps/rejected": -0.7760596871376038, | |
| "loss": 0.6188, | |
| "nll_loss": 0.47388267517089844, | |
| "rewards/accuracies": 0.4937500059604645, | |
| "rewards/chosen": -0.1465936005115509, | |
| "rewards/margins": 0.008618330582976341, | |
| "rewards/rejected": -0.15521195530891418, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 5.21875, | |
| "learning_rate": 4.909650756062782e-06, | |
| "log_odds_chosen": 0.054611045867204666, | |
| "log_odds_ratio": -0.7257325053215027, | |
| "logits/chosen": -3.2224247455596924, | |
| "logits/rejected": -3.236621141433716, | |
| "logps/chosen": -0.7471343874931335, | |
| "logps/rejected": -0.7686465978622437, | |
| "loss": 0.5619, | |
| "nll_loss": 0.4292874336242676, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -0.1494268774986267, | |
| "rewards/margins": 0.004302443936467171, | |
| "rewards/rejected": -0.15372933447360992, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 4.7982512711416995e-06, | |
| "log_odds_chosen": 0.06513820588588715, | |
| "log_odds_ratio": -0.7152190208435059, | |
| "logits/chosen": -3.165383815765381, | |
| "logits/rejected": -3.1888585090637207, | |
| "logps/chosen": -0.7319885492324829, | |
| "logps/rejected": -0.7425517439842224, | |
| "loss": 0.5721, | |
| "nll_loss": 0.43641433119773865, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -0.14639770984649658, | |
| "rewards/margins": 0.0021126591600477695, | |
| "rewards/rejected": -0.14851033687591553, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 4.21875, | |
| "learning_rate": 4.64513341295515e-06, | |
| "log_odds_chosen": -0.017645269632339478, | |
| "log_odds_ratio": -0.7585808038711548, | |
| "logits/chosen": -3.1905665397644043, | |
| "logits/rejected": -3.186293125152588, | |
| "logps/chosen": -0.7048856019973755, | |
| "logps/rejected": -0.7017339468002319, | |
| "loss": 0.5925, | |
| "nll_loss": 0.4144836366176605, | |
| "rewards/accuracies": 0.4625000059604645, | |
| "rewards/chosen": -0.14097711443901062, | |
| "rewards/margins": -0.0006303312256932259, | |
| "rewards/rejected": -0.14034679532051086, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 4.453076611255507e-06, | |
| "log_odds_chosen": 0.14627456665039062, | |
| "log_odds_ratio": -0.6653262972831726, | |
| "logits/chosen": -3.1202988624572754, | |
| "logits/rejected": -3.1385226249694824, | |
| "logps/chosen": -0.6960284113883972, | |
| "logps/rejected": -0.7755590081214905, | |
| "loss": 0.5531, | |
| "nll_loss": 0.3655123710632324, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.139205664396286, | |
| "rewards/margins": 0.015906114131212234, | |
| "rewards/rejected": -0.15511178970336914, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 4.6875, | |
| "learning_rate": 4.22556712424355e-06, | |
| "log_odds_chosen": 0.12080030143260956, | |
| "log_odds_ratio": -0.6814600229263306, | |
| "logits/chosen": -3.0596089363098145, | |
| "logits/rejected": -3.056286573410034, | |
| "logps/chosen": -0.6838747262954712, | |
| "logps/rejected": -0.7580210566520691, | |
| "loss": 0.5745, | |
| "nll_loss": 0.4387393891811371, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.1367749273777008, | |
| "rewards/margins": 0.014829280786216259, | |
| "rewards/rejected": -0.15160420536994934, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 4.34375, | |
| "learning_rate": 3.9667347552265945e-06, | |
| "log_odds_chosen": 0.09098054468631744, | |
| "log_odds_ratio": -0.7007282376289368, | |
| "logits/chosen": -3.1298584938049316, | |
| "logits/rejected": -3.135943651199341, | |
| "logps/chosen": -0.7081547975540161, | |
| "logps/rejected": -0.7608820796012878, | |
| "loss": 0.6087, | |
| "nll_loss": 0.4499574303627014, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -0.14163094758987427, | |
| "rewards/margins": 0.0105454595759511, | |
| "rewards/rejected": -0.1521764099597931, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 4.40625, | |
| "learning_rate": 3.68127788750129e-06, | |
| "log_odds_chosen": 0.18564960360527039, | |
| "log_odds_ratio": -0.6482046842575073, | |
| "logits/chosen": -3.142078399658203, | |
| "logits/rejected": -3.145967960357666, | |
| "logps/chosen": -0.6542805433273315, | |
| "logps/rejected": -0.7562066912651062, | |
| "loss": 0.5855, | |
| "nll_loss": 0.40380844473838806, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -0.13085612654685974, | |
| "rewards/margins": 0.020385239273309708, | |
| "rewards/rejected": -0.15124136209487915, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 3.921875, | |
| "learning_rate": 3.3743781982447533e-06, | |
| "log_odds_chosen": 0.13925865292549133, | |
| "log_odds_ratio": -0.6794918775558472, | |
| "logits/chosen": -3.1532516479492188, | |
| "logits/rejected": -3.1458470821380615, | |
| "logps/chosen": -0.6936465501785278, | |
| "logps/rejected": -0.7801955938339233, | |
| "loss": 0.567, | |
| "nll_loss": 0.4436867833137512, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -0.1387293040752411, | |
| "rewards/margins": 0.01730981469154358, | |
| "rewards/rejected": -0.15603913366794586, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 3.75, | |
| "learning_rate": 3.0516065995466336e-06, | |
| "log_odds_chosen": 0.16556194424629211, | |
| "log_odds_ratio": -0.6641014218330383, | |
| "logits/chosen": -3.1563491821289062, | |
| "logits/rejected": -3.1614346504211426, | |
| "logps/chosen": -0.6500056982040405, | |
| "logps/rejected": -0.7318493127822876, | |
| "loss": 0.5861, | |
| "nll_loss": 0.45738714933395386, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.13000114262104034, | |
| "rewards/margins": 0.016368722543120384, | |
| "rewards/rejected": -0.14636987447738647, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 4.0625, | |
| "learning_rate": 2.7188221139616303e-06, | |
| "log_odds_chosen": 0.10007637739181519, | |
| "log_odds_ratio": -0.6944689750671387, | |
| "logits/chosen": -3.105386972427368, | |
| "logits/rejected": -3.1417195796966553, | |
| "logps/chosen": -0.7295275926589966, | |
| "logps/rejected": -0.7832143902778625, | |
| "loss": 0.5604, | |
| "nll_loss": 0.4221896231174469, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.1459055244922638, | |
| "rewards/margins": 0.010737359523773193, | |
| "rewards/rejected": -0.156642884016037, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 3.609375, | |
| "learning_rate": 2.3820655202161237e-06, | |
| "log_odds_chosen": 0.18686607480049133, | |
| "log_odds_ratio": -0.6535667777061462, | |
| "logits/chosen": -3.1217126846313477, | |
| "logits/rejected": -3.1293649673461914, | |
| "logps/chosen": -0.632846474647522, | |
| "logps/rejected": -0.726894736289978, | |
| "loss": 0.5714, | |
| "nll_loss": 0.4484385848045349, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -0.12656930088996887, | |
| "rewards/margins": 0.018809664994478226, | |
| "rewards/rejected": -0.1453789472579956, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 4.0, | |
| "learning_rate": 2.0474496996359676e-06, | |
| "log_odds_chosen": 0.12438281625509262, | |
| "log_odds_ratio": -0.6807030439376831, | |
| "logits/chosen": -3.114492177963257, | |
| "logits/rejected": -3.1283891201019287, | |
| "logps/chosen": -0.6877269148826599, | |
| "logps/rejected": -0.7388188242912292, | |
| "loss": 0.5653, | |
| "nll_loss": 0.407818466424942, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -0.1375453919172287, | |
| "rewards/margins": 0.010218393988907337, | |
| "rewards/rejected": -0.14776377379894257, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 4.375, | |
| "learning_rate": 1.7210486737516947e-06, | |
| "log_odds_chosen": 0.13585877418518066, | |
| "log_odds_ratio": -0.6894018650054932, | |
| "logits/chosen": -3.094355583190918, | |
| "logits/rejected": -3.100367784500122, | |
| "logps/chosen": -0.6505359411239624, | |
| "logps/rejected": -0.7270746231079102, | |
| "loss": 0.5667, | |
| "nll_loss": 0.38747844099998474, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -0.13010717928409576, | |
| "rewards/margins": 0.015307746827602386, | |
| "rewards/rejected": -0.14541491866111755, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 1.4087873472954638e-06, | |
| "log_odds_chosen": 0.14618463814258575, | |
| "log_odds_ratio": -0.6783791780471802, | |
| "logits/chosen": -3.124422788619995, | |
| "logits/rejected": -3.1180996894836426, | |
| "logps/chosen": -0.7014532089233398, | |
| "logps/rejected": -0.7804998159408569, | |
| "loss": 0.5371, | |
| "nll_loss": 0.3992606997489929, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.14029064774513245, | |
| "rewards/margins": 0.015809325501322746, | |
| "rewards/rejected": -0.15609999001026154, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 3.796875, | |
| "learning_rate": 1.116333957999608e-06, | |
| "log_odds_chosen": 0.06774308532476425, | |
| "log_odds_ratio": -0.7074956893920898, | |
| "logits/chosen": -3.1124911308288574, | |
| "logits/rejected": -3.1118929386138916, | |
| "logps/chosen": -0.7164414525032043, | |
| "logps/rejected": -0.7556790113449097, | |
| "loss": 0.5714, | |
| "nll_loss": 0.40609636902809143, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.14328829944133759, | |
| "rewards/margins": 0.007847519591450691, | |
| "rewards/rejected": -0.15113583207130432, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 8.48997185472226e-07, | |
| "log_odds_chosen": 0.14430537819862366, | |
| "log_odds_ratio": -0.6831859946250916, | |
| "logits/chosen": -3.1008052825927734, | |
| "logits/rejected": -3.118614912033081, | |
| "logps/chosen": -0.6754971742630005, | |
| "logps/rejected": -0.7525571584701538, | |
| "loss": 0.5597, | |
| "nll_loss": 0.41810712218284607, | |
| "rewards/accuracies": 0.550000011920929, | |
| "rewards/chosen": -0.13509942591190338, | |
| "rewards/margins": 0.015411997213959694, | |
| "rewards/rejected": -0.15051142871379852, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 4.65625, | |
| "learning_rate": 6.11629786852592e-07, | |
| "log_odds_chosen": 0.20049241185188293, | |
| "log_odds_ratio": -0.6564125418663025, | |
| "logits/chosen": -3.0964746475219727, | |
| "logits/rejected": -3.097423553466797, | |
| "logps/chosen": -0.6751464009284973, | |
| "logps/rejected": -0.7748547196388245, | |
| "loss": 0.5689, | |
| "nll_loss": 0.4084566533565521, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -0.13502928614616394, | |
| "rewards/margins": 0.019941668957471848, | |
| "rewards/rejected": -0.1549709588289261, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 3.8125, | |
| "learning_rate": 4.0854050847362966e-07, | |
| "log_odds_chosen": 0.1618964970111847, | |
| "log_odds_ratio": -0.6639573574066162, | |
| "logits/chosen": -3.1198277473449707, | |
| "logits/rejected": -3.1235454082489014, | |
| "logps/chosen": -0.6751562356948853, | |
| "logps/rejected": -0.7577996253967285, | |
| "loss": 0.5453, | |
| "nll_loss": 0.4503118097782135, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -0.13503125309944153, | |
| "rewards/margins": 0.016528675332665443, | |
| "rewards/rejected": -0.15155991911888123, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 3.78125, | |
| "learning_rate": 2.4341587253072035e-07, | |
| "log_odds_chosen": 0.21062397956848145, | |
| "log_odds_ratio": -0.6452255249023438, | |
| "logits/chosen": -3.0891025066375732, | |
| "logits/rejected": -3.077317714691162, | |
| "logps/chosen": -0.6431701183319092, | |
| "logps/rejected": -0.7490943670272827, | |
| "loss": 0.5505, | |
| "nll_loss": 0.3902386724948883, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -0.1286340355873108, | |
| "rewards/margins": 0.021184856072068214, | |
| "rewards/rejected": -0.14981889724731445, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 4.28125, | |
| "learning_rate": 1.1925325850281416e-07, | |
| "log_odds_chosen": 0.08836773782968521, | |
| "log_odds_ratio": -0.7011784315109253, | |
| "logits/chosen": -3.1118626594543457, | |
| "logits/rejected": -3.1287245750427246, | |
| "logps/chosen": -0.6649689674377441, | |
| "logps/rejected": -0.6961485743522644, | |
| "loss": 0.534, | |
| "nll_loss": 0.34850066900253296, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -0.13299378752708435, | |
| "rewards/margins": 0.00623591523617506, | |
| "rewards/rejected": -0.13922971487045288, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 3.984375, | |
| "learning_rate": 3.830649404690939e-08, | |
| "log_odds_chosen": 0.10791333764791489, | |
| "log_odds_ratio": -0.691511332988739, | |
| "logits/chosen": -3.0939254760742188, | |
| "logits/rejected": -3.1030945777893066, | |
| "logps/chosen": -0.6941600441932678, | |
| "logps/rejected": -0.7733976244926453, | |
| "loss": 0.5775, | |
| "nll_loss": 0.40806055068969727, | |
| "rewards/accuracies": 0.5249999761581421, | |
| "rewards/chosen": -0.1388320028781891, | |
| "rewards/margins": 0.015847522765398026, | |
| "rewards/rejected": -0.15467952191829681, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 4.15625, | |
| "learning_rate": 2.044943012210754e-09, | |
| "log_odds_chosen": 0.10530634969472885, | |
| "log_odds_ratio": -0.69704669713974, | |
| "logits/chosen": -3.1206929683685303, | |
| "logits/rejected": -3.117326259613037, | |
| "logps/chosen": -0.6853540539741516, | |
| "logps/rejected": -0.7511805295944214, | |
| "loss": 0.567, | |
| "nll_loss": 0.40571802854537964, | |
| "rewards/accuracies": 0.581250011920929, | |
| "rewards/chosen": -0.13707080483436584, | |
| "rewards/margins": 0.013165315613150597, | |
| "rewards/rejected": -0.150236114859581, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 233, | |
| "total_flos": 0.0, | |
| "train_loss": 0.5704265398017326, | |
| "train_runtime": 3228.2239, | |
| "train_samples_per_second": 4.624, | |
| "train_steps_per_second": 0.072 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 233, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "total_flos": 0.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |