| { | |
| "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": 38.22182319115289, | |
| "learning_rate": 6.25e-08, | |
| "logits/chosen": -0.7837417721748352, | |
| "logits/rejected": -0.8419772386550903, | |
| "logps/chosen": -1.1746927499771118, | |
| "logps/rejected": -1.3594377040863037, | |
| "loss": 2.1738, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.1746927499771118, | |
| "rewards/margins": 0.1847449541091919, | |
| "rewards/rejected": -1.3594377040863037, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.02093692750588851, | |
| "grad_norm": 15.829940538804763, | |
| "learning_rate": 1.25e-07, | |
| "logits/chosen": -0.8227180242538452, | |
| "logits/rejected": -0.8315876722335815, | |
| "logps/chosen": -1.1591246128082275, | |
| "logps/rejected": -1.2620265483856201, | |
| "loss": 2.1411, | |
| "rewards/accuracies": 0.5062500238418579, | |
| "rewards/chosen": -1.1591246128082275, | |
| "rewards/margins": 0.10290191322565079, | |
| "rewards/rejected": -1.2620265483856201, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.031405391258832765, | |
| "grad_norm": 22.77524777376306, | |
| "learning_rate": 1.875e-07, | |
| "logits/chosen": -0.7929477691650391, | |
| "logits/rejected": -0.8053458333015442, | |
| "logps/chosen": -1.1068744659423828, | |
| "logps/rejected": -1.3614892959594727, | |
| "loss": 2.1083, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.1068744659423828, | |
| "rewards/margins": 0.2546148896217346, | |
| "rewards/rejected": -1.3614892959594727, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04187385501177702, | |
| "grad_norm": 37.8845459916446, | |
| "learning_rate": 2.5e-07, | |
| "logits/chosen": -0.717773973941803, | |
| "logits/rejected": -0.80096834897995, | |
| "logps/chosen": -1.1600226163864136, | |
| "logps/rejected": -1.2640360593795776, | |
| "loss": 2.167, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.1600226163864136, | |
| "rewards/margins": 0.10401340574026108, | |
| "rewards/rejected": -1.2640360593795776, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05234231876472128, | |
| "grad_norm": 13.533858086681068, | |
| "learning_rate": 3.125e-07, | |
| "logits/chosen": -0.8047081232070923, | |
| "logits/rejected": -0.8094943165779114, | |
| "logps/chosen": -1.1760809421539307, | |
| "logps/rejected": -1.2413945198059082, | |
| "loss": 2.142, | |
| "rewards/accuracies": 0.518750011920929, | |
| "rewards/chosen": -1.1760809421539307, | |
| "rewards/margins": 0.06531357765197754, | |
| "rewards/rejected": -1.2413945198059082, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06281078251766553, | |
| "grad_norm": 46.12526967375131, | |
| "learning_rate": 3.75e-07, | |
| "logits/chosen": -0.7824481725692749, | |
| "logits/rejected": -0.8084238767623901, | |
| "logps/chosen": -1.1583902835845947, | |
| "logps/rejected": -1.2807166576385498, | |
| "loss": 2.1687, | |
| "rewards/accuracies": 0.512499988079071, | |
| "rewards/chosen": -1.1583902835845947, | |
| "rewards/margins": 0.12232650816440582, | |
| "rewards/rejected": -1.2807166576385498, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07327924627060979, | |
| "grad_norm": 20.51948865888897, | |
| "learning_rate": 4.3749999999999994e-07, | |
| "logits/chosen": -0.8100900650024414, | |
| "logits/rejected": -0.789565920829773, | |
| "logps/chosen": -1.142910361289978, | |
| "logps/rejected": -1.279740571975708, | |
| "loss": 2.1333, | |
| "rewards/accuracies": 0.5625, | |
| "rewards/chosen": -1.142910361289978, | |
| "rewards/margins": 0.136830136179924, | |
| "rewards/rejected": -1.279740571975708, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.08374771002355404, | |
| "grad_norm": 26.130472787451367, | |
| "learning_rate": 5e-07, | |
| "logits/chosen": -0.7920509576797485, | |
| "logits/rejected": -0.9067994356155396, | |
| "logps/chosen": -1.0735507011413574, | |
| "logps/rejected": -1.3959448337554932, | |
| "loss": 2.089, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.0735507011413574, | |
| "rewards/margins": 0.3223941922187805, | |
| "rewards/rejected": -1.3959448337554932, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0942161737764983, | |
| "grad_norm": 33.25286433207774, | |
| "learning_rate": 5.625e-07, | |
| "logits/chosen": -0.7628027200698853, | |
| "logits/rejected": -0.8258091807365417, | |
| "logps/chosen": -1.0904920101165771, | |
| "logps/rejected": -1.3169877529144287, | |
| "loss": 2.1077, | |
| "rewards/accuracies": 0.59375, | |
| "rewards/chosen": -1.0904920101165771, | |
| "rewards/margins": 0.2264958918094635, | |
| "rewards/rejected": -1.3169877529144287, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.10468463752944256, | |
| "grad_norm": 26.427906121657426, | |
| "learning_rate": 5.999678242522831e-07, | |
| "logits/chosen": -0.8172017931938171, | |
| "logits/rejected": -0.8944879770278931, | |
| "logps/chosen": -1.1694625616073608, | |
| "logps/rejected": -1.4630098342895508, | |
| "loss": 2.1326, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1694625616073608, | |
| "rewards/margins": 0.29354724287986755, | |
| "rewards/rejected": -1.4630098342895508, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.11515310128238682, | |
| "grad_norm": 83.98012535654864, | |
| "learning_rate": 5.996059263493219e-07, | |
| "logits/chosen": -0.8303399085998535, | |
| "logits/rejected": -0.8691905736923218, | |
| "logps/chosen": -1.1166013479232788, | |
| "logps/rejected": -1.3432626724243164, | |
| "loss": 2.1064, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -1.1166013479232788, | |
| "rewards/margins": 0.22666127979755402, | |
| "rewards/rejected": -1.3432626724243164, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.12562156503533106, | |
| "grad_norm": 20.73349404149956, | |
| "learning_rate": 5.988423976115163e-07, | |
| "logits/chosen": -0.8669672012329102, | |
| "logits/rejected": -0.9171522855758667, | |
| "logps/chosen": -1.2086117267608643, | |
| "logps/rejected": -1.3324425220489502, | |
| "loss": 2.0869, | |
| "rewards/accuracies": 0.5562499761581421, | |
| "rewards/chosen": -1.2086117267608643, | |
| "rewards/margins": 0.1238306537270546, | |
| "rewards/rejected": -1.3324425220489502, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.1360900287882753, | |
| "grad_norm": 21.610299966098676, | |
| "learning_rate": 5.976782615723061e-07, | |
| "logits/chosen": -0.8252432942390442, | |
| "logits/rejected": -0.900849461555481, | |
| "logps/chosen": -1.1040380001068115, | |
| "logps/rejected": -1.6288048028945923, | |
| "loss": 2.0719, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.1040380001068115, | |
| "rewards/margins": 0.524766743183136, | |
| "rewards/rejected": -1.6288048028945923, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.14655849254121958, | |
| "grad_norm": 53.60916157723815, | |
| "learning_rate": 5.961150787913738e-07, | |
| "logits/chosen": -0.8189405202865601, | |
| "logits/rejected": -0.8585717082023621, | |
| "logps/chosen": -1.1598970890045166, | |
| "logps/rejected": -1.423801302909851, | |
| "loss": 2.051, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.1598970890045166, | |
| "rewards/margins": 0.26390427350997925, | |
| "rewards/rejected": -1.423801302909851, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.15702695629416383, | |
| "grad_norm": 55.548769154194964, | |
| "learning_rate": 5.941549447626671e-07, | |
| "logits/chosen": -0.8151289224624634, | |
| "logits/rejected": -0.8495903015136719, | |
| "logps/chosen": -1.1472132205963135, | |
| "logps/rejected": -1.4964182376861572, | |
| "loss": 2.0657, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1472132205963135, | |
| "rewards/margins": 0.349204957485199, | |
| "rewards/rejected": -1.4964182376861572, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.16749542004710807, | |
| "grad_norm": 24.74010593915972, | |
| "learning_rate": 5.918004871053251e-07, | |
| "logits/chosen": -0.8290842175483704, | |
| "logits/rejected": -0.9109519720077515, | |
| "logps/chosen": -1.1479018926620483, | |
| "logps/rejected": -1.5490809679031372, | |
| "loss": 2.0988, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.1479018926620483, | |
| "rewards/margins": 0.40117907524108887, | |
| "rewards/rejected": -1.5490809679031372, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.17796388380005235, | |
| "grad_norm": 19.216616658463852, | |
| "learning_rate": 5.890548620412763e-07, | |
| "logits/chosen": -0.9178426861763, | |
| "logits/rejected": -0.9389097094535828, | |
| "logps/chosen": -1.1228134632110596, | |
| "logps/rejected": -1.4461188316345215, | |
| "loss": 2.0983, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1228134632110596, | |
| "rewards/margins": 0.3233054280281067, | |
| "rewards/rejected": -1.4461188316345215, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1884323475529966, | |
| "grad_norm": 9.550523787824096, | |
| "learning_rate": 5.859217501642258e-07, | |
| "logits/chosen": -0.8629040718078613, | |
| "logits/rejected": -0.9038488268852234, | |
| "logps/chosen": -1.1353037357330322, | |
| "logps/rejected": -1.4276666641235352, | |
| "loss": 2.0462, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.1353037357330322, | |
| "rewards/margins": 0.29236286878585815, | |
| "rewards/rejected": -1.4276666641235352, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.19890081130594087, | |
| "grad_norm": 16.55214910624781, | |
| "learning_rate": 5.824053515057091e-07, | |
| "logits/chosen": -0.879852294921875, | |
| "logits/rejected": -0.8908422589302063, | |
| "logps/chosen": -1.157950520515442, | |
| "logps/rejected": -1.457081913948059, | |
| "loss": 2.0807, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.157950520515442, | |
| "rewards/margins": 0.2991313636302948, | |
| "rewards/rejected": -1.457081913948059, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.2093692750588851, | |
| "grad_norm": 32.38005097808463, | |
| "learning_rate": 5.785103799048218e-07, | |
| "logits/chosen": -0.8940750360488892, | |
| "logits/rejected": -0.9557636380195618, | |
| "logps/chosen": -1.1859995126724243, | |
| "logps/rejected": -1.7212893962860107, | |
| "loss": 2.0735, | |
| "rewards/accuracies": 0.53125, | |
| "rewards/chosen": -1.1859995126724243, | |
| "rewards/margins": 0.5352898240089417, | |
| "rewards/rejected": -1.7212893962860107, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.21983773881182936, | |
| "grad_norm": 13.467954278767516, | |
| "learning_rate": 5.742420566891749e-07, | |
| "logits/chosen": -0.8789809942245483, | |
| "logits/rejected": -0.9158897399902344, | |
| "logps/chosen": -1.0777571201324463, | |
| "logps/rejected": -1.4952013492584229, | |
| "loss": 2.0438, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.0777571201324463, | |
| "rewards/margins": 0.4174441397190094, | |
| "rewards/rejected": -1.4952013492584229, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.23030620256477363, | |
| "grad_norm": 48.371034794873005, | |
| "learning_rate": 5.696061036755478e-07, | |
| "logits/chosen": -0.915793240070343, | |
| "logits/rejected": -0.9631668925285339, | |
| "logps/chosen": -1.081206202507019, | |
| "logps/rejected": -1.546509861946106, | |
| "loss": 2.0298, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.081206202507019, | |
| "rewards/margins": 0.4653037190437317, | |
| "rewards/rejected": -1.546509861946106, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.24077466631771788, | |
| "grad_norm": 33.44917537983854, | |
| "learning_rate": 5.64608735499618e-07, | |
| "logits/chosen": -0.8344041109085083, | |
| "logits/rejected": -0.9158002138137817, | |
| "logps/chosen": -1.1163157224655151, | |
| "logps/rejected": -1.6953636407852173, | |
| "loss": 2.0198, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1163157224655151, | |
| "rewards/margins": 0.5790477991104126, | |
| "rewards/rejected": -1.6953636407852173, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.2512431300706621, | |
| "grad_norm": 41.36153494880983, | |
| "learning_rate": 5.592566512850545e-07, | |
| "logits/chosen": -0.8542771339416504, | |
| "logits/rejected": -0.8888319134712219, | |
| "logps/chosen": -1.2859517335891724, | |
| "logps/rejected": -1.5670356750488281, | |
| "loss": 2.1007, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.2859517335891724, | |
| "rewards/margins": 0.28108420968055725, | |
| "rewards/rejected": -1.5670356750488281, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.26171159382360637, | |
| "grad_norm": 29.2467879102086, | |
| "learning_rate": 5.535570256631384e-07, | |
| "logits/chosen": -0.9180340766906738, | |
| "logits/rejected": -0.8835474252700806, | |
| "logps/chosen": -1.1059439182281494, | |
| "logps/rejected": -1.4397169351577759, | |
| "loss": 2.0475, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1059439182281494, | |
| "rewards/margins": 0.33377301692962646, | |
| "rewards/rejected": -1.4397169351577759, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.2721800575765506, | |
| "grad_norm": 47.965843324490926, | |
| "learning_rate": 5.475174991549528e-07, | |
| "logits/chosen": -0.8605352640151978, | |
| "logits/rejected": -0.9204422235488892, | |
| "logps/chosen": -1.076321005821228, | |
| "logps/rejected": -1.291337251663208, | |
| "loss": 2.0689, | |
| "rewards/accuracies": 0.574999988079071, | |
| "rewards/chosen": -1.076321005821228, | |
| "rewards/margins": 0.21501627564430237, | |
| "rewards/rejected": -1.291337251663208, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.2826485213294949, | |
| "grad_norm": 25.643927984267705, | |
| "learning_rate": 5.411461679290317e-07, | |
| "logits/chosen": -0.8433247804641724, | |
| "logits/rejected": -0.8802833557128906, | |
| "logps/chosen": -1.1742990016937256, | |
| "logps/rejected": -1.8538631200790405, | |
| "loss": 2.0085, | |
| "rewards/accuracies": 0.7562500238418579, | |
| "rewards/chosen": -1.1742990016937256, | |
| "rewards/margins": 0.6795639395713806, | |
| "rewards/rejected": -1.8538631200790405, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.29311698508243916, | |
| "grad_norm": 24.64563141309484, | |
| "learning_rate": 5.34451572948201e-07, | |
| "logits/chosen": -0.8815560340881348, | |
| "logits/rejected": -0.9196213483810425, | |
| "logps/chosen": -1.1883355379104614, | |
| "logps/rejected": -1.751808524131775, | |
| "loss": 1.9914, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.1883355379104614, | |
| "rewards/margins": 0.5634733438491821, | |
| "rewards/rejected": -1.751808524131775, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3035854488353834, | |
| "grad_norm": 16.805987111465818, | |
| "learning_rate": 5.274426885201582e-07, | |
| "logits/chosen": -0.8858097195625305, | |
| "logits/rejected": -0.9691437482833862, | |
| "logps/chosen": -1.1200135946273804, | |
| "logps/rejected": -1.5147252082824707, | |
| "loss": 2.0408, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1200135946273804, | |
| "rewards/margins": 0.3947116434574127, | |
| "rewards/rejected": -1.5147252082824707, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.31405391258832765, | |
| "grad_norm": 18.868227485230445, | |
| "learning_rate": 5.201289102671411e-07, | |
| "logits/chosen": -0.8802644610404968, | |
| "logits/rejected": -0.8881145715713501, | |
| "logps/chosen": -1.0411317348480225, | |
| "logps/rejected": -1.462885856628418, | |
| "loss": 2.002, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.0411317348480225, | |
| "rewards/margins": 0.42175430059432983, | |
| "rewards/rejected": -1.462885856628418, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3245223763412719, | |
| "grad_norm": 23.821140021157394, | |
| "learning_rate": 5.12520042530811e-07, | |
| "logits/chosen": -0.8837698101997375, | |
| "logits/rejected": -0.903447151184082, | |
| "logps/chosen": -1.1391239166259766, | |
| "logps/rejected": -1.514580249786377, | |
| "loss": 2.0007, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.1391239166259766, | |
| "rewards/margins": 0.3754563331604004, | |
| "rewards/rejected": -1.514580249786377, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.33499084009421615, | |
| "grad_norm": 25.07335170173153, | |
| "learning_rate": 5.046262852292346e-07, | |
| "logits/chosen": -0.8691936731338501, | |
| "logits/rejected": -0.8966501951217651, | |
| "logps/chosen": -1.1949414014816284, | |
| "logps/rejected": -1.6663414239883423, | |
| "loss": 2.0357, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.1949414014816284, | |
| "rewards/margins": 0.4713999629020691, | |
| "rewards/rejected": -1.6663414239883423, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.34545930384716045, | |
| "grad_norm": 15.930498315332532, | |
| "learning_rate": 4.964582201835856e-07, | |
| "logits/chosen": -0.9137632250785828, | |
| "logits/rejected": -0.9423500895500183, | |
| "logps/chosen": -1.1093122959136963, | |
| "logps/rejected": -1.6846201419830322, | |
| "loss": 2.0194, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.1093122959136963, | |
| "rewards/margins": 0.5753079652786255, | |
| "rewards/rejected": -1.6846201419830322, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.3559277676001047, | |
| "grad_norm": 14.533074422786573, | |
| "learning_rate": 4.880267969328908e-07, | |
| "logits/chosen": -0.8995744585990906, | |
| "logits/rejected": -0.9443628191947937, | |
| "logps/chosen": -1.1981232166290283, | |
| "logps/rejected": -1.643414855003357, | |
| "loss": 2.0262, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1981232166290283, | |
| "rewards/margins": 0.4452916085720062, | |
| "rewards/rejected": -1.643414855003357, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.36639623135304894, | |
| "grad_norm": 21.913922705783488, | |
| "learning_rate": 4.793433180558423e-07, | |
| "logits/chosen": -0.8912171125411987, | |
| "logits/rejected": -0.9618469476699829, | |
| "logps/chosen": -1.182425856590271, | |
| "logps/rejected": -1.684942603111267, | |
| "loss": 2.0155, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.182425856590271, | |
| "rewards/margins": 0.5025165677070618, | |
| "rewards/rejected": -1.684942603111267, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.3768646951059932, | |
| "grad_norm": 27.73808093722701, | |
| "learning_rate": 4.704194240193467e-07, | |
| "logits/chosen": -0.8934853672981262, | |
| "logits/rejected": -0.9609943628311157, | |
| "logps/chosen": -1.1971694231033325, | |
| "logps/rejected": -1.6838546991348267, | |
| "loss": 2.0284, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.1971694231033325, | |
| "rewards/margins": 0.4866850972175598, | |
| "rewards/rejected": -1.6838546991348267, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.38733315885893743, | |
| "grad_norm": 30.41505287850658, | |
| "learning_rate": 4.6126707757412686e-07, | |
| "logits/chosen": -0.8480084538459778, | |
| "logits/rejected": -0.9141961336135864, | |
| "logps/chosen": -1.1976919174194336, | |
| "logps/rejected": -1.9417314529418945, | |
| "loss": 1.9467, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.1976919174194336, | |
| "rewards/margins": 0.7440396547317505, | |
| "rewards/rejected": -1.9417314529418945, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.39780162261188173, | |
| "grad_norm": 34.17710511802847, | |
| "learning_rate": 4.5189854771829086e-07, | |
| "logits/chosen": -0.963171124458313, | |
| "logits/rejected": -0.9760686755180359, | |
| "logps/chosen": -1.2360587120056152, | |
| "logps/rejected": -1.665792465209961, | |
| "loss": 2.0894, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.2360587120056152, | |
| "rewards/margins": 0.42973384261131287, | |
| "rewards/rejected": -1.665792465209961, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.408270086364826, | |
| "grad_norm": 39.854588709056486, | |
| "learning_rate": 4.4232639325036807e-07, | |
| "logits/chosen": -0.9235826730728149, | |
| "logits/rejected": -0.9407492876052856, | |
| "logps/chosen": -1.2383768558502197, | |
| "logps/rejected": -1.6111949682235718, | |
| "loss": 2.0333, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.2383768558502197, | |
| "rewards/margins": 0.3728182315826416, | |
| "rewards/rejected": -1.6111949682235718, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.4187385501177702, | |
| "grad_norm": 46.073511129529436, | |
| "learning_rate": 4.32563445933859e-07, | |
| "logits/chosen": -0.9877565503120422, | |
| "logits/rejected": -0.9876266717910767, | |
| "logps/chosen": -1.2531481981277466, | |
| "logps/rejected": -1.6756317615509033, | |
| "loss": 2.0456, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.2531481981277466, | |
| "rewards/margins": 0.4224833846092224, | |
| "rewards/rejected": -1.6756317615509033, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.42920701387071447, | |
| "grad_norm": 46.11695411093734, | |
| "learning_rate": 4.226227932958664e-07, | |
| "logits/chosen": -0.9007834196090698, | |
| "logits/rejected": -0.9248853921890259, | |
| "logps/chosen": -1.0488381385803223, | |
| "logps/rejected": -1.6470975875854492, | |
| "loss": 1.9781, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.0488381385803223, | |
| "rewards/margins": 0.5982593297958374, | |
| "rewards/rejected": -1.6470975875854492, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.4396754776236587, | |
| "grad_norm": 39.48440428645483, | |
| "learning_rate": 4.1251776108286854e-07, | |
| "logits/chosen": -0.9045936465263367, | |
| "logits/rejected": -0.9374397397041321, | |
| "logps/chosen": -1.2596890926361084, | |
| "logps/rejected": -1.6047645807266235, | |
| "loss": 2.0371, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.2596890926361084, | |
| "rewards/margins": 0.345075398683548, | |
| "rewards/rejected": -1.6047645807266235, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.45014394137660296, | |
| "grad_norm": 39.165174769151264, | |
| "learning_rate": 4.022618953971514e-07, | |
| "logits/chosen": -0.8771007657051086, | |
| "logits/rejected": -0.9414412379264832, | |
| "logps/chosen": -1.1611162424087524, | |
| "logps/rejected": -1.7325302362442017, | |
| "loss": 2.0219, | |
| "rewards/accuracies": 0.7437499761581421, | |
| "rewards/chosen": -1.1611162424087524, | |
| "rewards/margins": 0.5714138746261597, | |
| "rewards/rejected": -1.7325302362442017, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.46061240512954726, | |
| "grad_norm": 40.26532103065248, | |
| "learning_rate": 3.918689445378477e-07, | |
| "logits/chosen": -0.9068889617919922, | |
| "logits/rejected": -0.9496709108352661, | |
| "logps/chosen": -1.1995134353637695, | |
| "logps/rejected": -1.7156946659088135, | |
| "loss": 1.987, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.1995134353637695, | |
| "rewards/margins": 0.5161812901496887, | |
| "rewards/rejected": -1.7156946659088135, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.4710808688824915, | |
| "grad_norm": 17.4001686828919, | |
| "learning_rate": 3.813528405709251e-07, | |
| "logits/chosen": -0.9098321795463562, | |
| "logits/rejected": -0.9786120653152466, | |
| "logps/chosen": -1.088384985923767, | |
| "logps/rejected": -1.7232511043548584, | |
| "loss": 1.9565, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.088384985923767, | |
| "rewards/margins": 0.6348661780357361, | |
| "rewards/rejected": -1.7232511043548584, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.48154933263543576, | |
| "grad_norm": 32.68221233140228, | |
| "learning_rate": 3.707276806528282e-07, | |
| "logits/chosen": -0.9034180641174316, | |
| "logits/rejected": -0.963924765586853, | |
| "logps/chosen": -1.1403090953826904, | |
| "logps/rejected": -1.8087886571884155, | |
| "loss": 1.9732, | |
| "rewards/accuracies": 0.625, | |
| "rewards/chosen": -1.1403090953826904, | |
| "rewards/margins": 0.6684795618057251, | |
| "rewards/rejected": -1.8087886571884155, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.49201779638838, | |
| "grad_norm": 33.604375971044966, | |
| "learning_rate": 3.6000770813281334e-07, | |
| "logits/chosen": -0.9114705920219421, | |
| "logits/rejected": -0.9957945942878723, | |
| "logps/chosen": -1.124693512916565, | |
| "logps/rejected": -1.6200597286224365, | |
| "loss": 2.0076, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.124693512916565, | |
| "rewards/margins": 0.4953663945198059, | |
| "rewards/rejected": -1.6200597286224365, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.5024862601413242, | |
| "grad_norm": 33.58172238687118, | |
| "learning_rate": 3.4920729345930654e-07, | |
| "logits/chosen": -0.8984516263008118, | |
| "logits/rejected": -0.9524177312850952, | |
| "logps/chosen": -1.136643409729004, | |
| "logps/rejected": -1.6992219686508179, | |
| "loss": 2.0428, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.136643409729004, | |
| "rewards/margins": 0.562578558921814, | |
| "rewards/rejected": -1.6992219686508179, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.5129547238942685, | |
| "grad_norm": 36.50989911591013, | |
| "learning_rate": 3.383409149158814e-07, | |
| "logits/chosen": -0.9273117184638977, | |
| "logits/rejected": -0.9691912531852722, | |
| "logps/chosen": -1.255492091178894, | |
| "logps/rejected": -1.654199242591858, | |
| "loss": 2.0089, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.255492091178894, | |
| "rewards/margins": 0.39870715141296387, | |
| "rewards/rejected": -1.654199242591858, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.5234231876472127, | |
| "grad_norm": 56.18660552674704, | |
| "learning_rate": 3.2742313921268035e-07, | |
| "logits/chosen": -0.8531250953674316, | |
| "logits/rejected": -0.9333058595657349, | |
| "logps/chosen": -1.1693034172058105, | |
| "logps/rejected": -1.7800544500350952, | |
| "loss": 2.0302, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.1693034172058105, | |
| "rewards/margins": 0.6107511520385742, | |
| "rewards/rejected": -1.7800544500350952, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.533891651400157, | |
| "grad_norm": 32.73686669905504, | |
| "learning_rate": 3.1646860195929825e-07, | |
| "logits/chosen": -0.8804293870925903, | |
| "logits/rejected": -0.9355670809745789, | |
| "logps/chosen": -1.2591540813446045, | |
| "logps/rejected": -1.830005407333374, | |
| "loss": 1.9653, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.2591540813446045, | |
| "rewards/margins": 0.5708513855934143, | |
| "rewards/rejected": -1.830005407333374, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.5443601151531012, | |
| "grad_norm": 48.758378863327145, | |
| "learning_rate": 3.054919880453032e-07, | |
| "logits/chosen": -0.8192955255508423, | |
| "logits/rejected": -0.8804057240486145, | |
| "logps/chosen": -1.1422827243804932, | |
| "logps/rejected": -1.8388431072235107, | |
| "loss": 2.0149, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.1422827243804932, | |
| "rewards/margins": 0.6965604424476624, | |
| "rewards/rejected": -1.8388431072235107, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.5548285789060455, | |
| "grad_norm": 57.73699922463872, | |
| "learning_rate": 2.9450801195469686e-07, | |
| "logits/chosen": -0.8590672612190247, | |
| "logits/rejected": -0.929406464099884, | |
| "logps/chosen": -1.2068861722946167, | |
| "logps/rejected": -1.5913540124893188, | |
| "loss": 1.9765, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.2068861722946167, | |
| "rewards/margins": 0.3844676911830902, | |
| "rewards/rejected": -1.5913540124893188, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.5652970426589898, | |
| "grad_norm": 39.05494180222134, | |
| "learning_rate": 2.835313980407017e-07, | |
| "logits/chosen": -0.8784425854682922, | |
| "logits/rejected": -0.8780919313430786, | |
| "logps/chosen": -1.257610559463501, | |
| "logps/rejected": -1.6434276103973389, | |
| "loss": 2.012, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.257610559463501, | |
| "rewards/margins": 0.3858169615268707, | |
| "rewards/rejected": -1.6434276103973389, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.575765506411934, | |
| "grad_norm": 12.369775833139514, | |
| "learning_rate": 2.7257686078731973e-07, | |
| "logits/chosen": -0.8967874646186829, | |
| "logits/rejected": -0.9520025253295898, | |
| "logps/chosen": -1.1508904695510864, | |
| "logps/rejected": -1.829810380935669, | |
| "loss": 1.99, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.1508904695510864, | |
| "rewards/margins": 0.678919792175293, | |
| "rewards/rejected": -1.829810380935669, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.5862339701648783, | |
| "grad_norm": 28.730474706137716, | |
| "learning_rate": 2.6165908508411857e-07, | |
| "logits/chosen": -0.9400002360343933, | |
| "logits/rejected": -1.015257477760315, | |
| "logps/chosen": -1.0943539142608643, | |
| "logps/rejected": -1.5329124927520752, | |
| "loss": 1.9819, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.0943539142608643, | |
| "rewards/margins": 0.4385586678981781, | |
| "rewards/rejected": -1.5329124927520752, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.5967024339178225, | |
| "grad_norm": 45.97759565415205, | |
| "learning_rate": 2.5079270654069354e-07, | |
| "logits/chosen": -0.8827983736991882, | |
| "logits/rejected": -0.9212683439254761, | |
| "logps/chosen": -1.213191270828247, | |
| "logps/rejected": -1.7406049966812134, | |
| "loss": 1.9873, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.213191270828247, | |
| "rewards/margins": 0.5274137258529663, | |
| "rewards/rejected": -1.7406049966812134, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.6071708976707668, | |
| "grad_norm": 23.809737226091617, | |
| "learning_rate": 2.399922918671867e-07, | |
| "logits/chosen": -0.8965229988098145, | |
| "logits/rejected": -1.0004279613494873, | |
| "logps/chosen": -1.1851826906204224, | |
| "logps/rejected": -1.8194730281829834, | |
| "loss": 1.9564, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.1851826906204224, | |
| "rewards/margins": 0.6342904567718506, | |
| "rewards/rejected": -1.8194730281829834, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.6176393614237111, | |
| "grad_norm": 48.93055509881971, | |
| "learning_rate": 2.2927231934717176e-07, | |
| "logits/chosen": -0.8713434338569641, | |
| "logits/rejected": -0.9567992091178894, | |
| "logps/chosen": -1.2080659866333008, | |
| "logps/rejected": -1.8678998947143555, | |
| "loss": 2.0124, | |
| "rewards/accuracies": 0.6812499761581421, | |
| "rewards/chosen": -1.2080659866333008, | |
| "rewards/margins": 0.6598338484764099, | |
| "rewards/rejected": -1.8678998947143555, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.6281078251766553, | |
| "grad_norm": 16.728041846610992, | |
| "learning_rate": 2.1864715942907487e-07, | |
| "logits/chosen": -0.8935406804084778, | |
| "logits/rejected": -0.9455572366714478, | |
| "logps/chosen": -1.2053089141845703, | |
| "logps/rejected": -1.6934458017349243, | |
| "loss": 1.9953, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.2053089141845703, | |
| "rewards/margins": 0.4881366789340973, | |
| "rewards/rejected": -1.6934458017349243, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.6385762889295996, | |
| "grad_norm": 64.72385101970457, | |
| "learning_rate": 2.081310554621522e-07, | |
| "logits/chosen": -0.8762944936752319, | |
| "logits/rejected": -0.9710057973861694, | |
| "logps/chosen": -1.21980881690979, | |
| "logps/rejected": -1.7974154949188232, | |
| "loss": 1.9564, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.21980881690979, | |
| "rewards/margins": 0.5776065587997437, | |
| "rewards/rejected": -1.7974154949188232, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.6490447526825438, | |
| "grad_norm": 30.32440992803243, | |
| "learning_rate": 1.9773810460284862e-07, | |
| "logits/chosen": -0.896551251411438, | |
| "logits/rejected": -0.925498366355896, | |
| "logps/chosen": -1.1150288581848145, | |
| "logps/rejected": -1.6414196491241455, | |
| "loss": 1.9559, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.1150288581848145, | |
| "rewards/margins": 0.5263907313346863, | |
| "rewards/rejected": -1.6414196491241455, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.6595132164354881, | |
| "grad_norm": 43.26797640360581, | |
| "learning_rate": 1.874822389171314e-07, | |
| "logits/chosen": -0.8692461848258972, | |
| "logits/rejected": -0.9539564251899719, | |
| "logps/chosen": -1.1112568378448486, | |
| "logps/rejected": -1.8892109394073486, | |
| "loss": 1.9671, | |
| "rewards/accuracies": 0.7250000238418579, | |
| "rewards/chosen": -1.1112568378448486, | |
| "rewards/margins": 0.7779541015625, | |
| "rewards/rejected": -1.8892109394073486, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.6699816801884323, | |
| "grad_norm": 27.417692518273306, | |
| "learning_rate": 1.7737720670413356e-07, | |
| "logits/chosen": -0.9024019241333008, | |
| "logits/rejected": -0.8824315071105957, | |
| "logps/chosen": -1.209397315979004, | |
| "logps/rejected": -1.7458854913711548, | |
| "loss": 1.9292, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.209397315979004, | |
| "rewards/margins": 0.5364881753921509, | |
| "rewards/rejected": -1.7458854913711548, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.6804501439413766, | |
| "grad_norm": 29.061168159916782, | |
| "learning_rate": 1.6743655406614095e-07, | |
| "logits/chosen": -0.8877676725387573, | |
| "logits/rejected": -0.9611382484436035, | |
| "logps/chosen": -1.1630074977874756, | |
| "logps/rejected": -1.8193352222442627, | |
| "loss": 1.9418, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.1630074977874756, | |
| "rewards/margins": 0.6563276052474976, | |
| "rewards/rejected": -1.8193352222442627, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.6909186076943209, | |
| "grad_norm": 21.820711826590742, | |
| "learning_rate": 1.5767360674963198e-07, | |
| "logits/chosen": -0.8649999499320984, | |
| "logits/rejected": -0.9277146458625793, | |
| "logps/chosen": -1.1468570232391357, | |
| "logps/rejected": -1.590583086013794, | |
| "loss": 1.9873, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.1468570232391357, | |
| "rewards/margins": 0.4437260031700134, | |
| "rewards/rejected": -1.590583086013794, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.7013870714472651, | |
| "grad_norm": 26.357123343274285, | |
| "learning_rate": 1.4810145228170922e-07, | |
| "logits/chosen": -0.8826435804367065, | |
| "logits/rejected": -0.9004185795783997, | |
| "logps/chosen": -1.1120072603225708, | |
| "logps/rejected": -1.5835678577423096, | |
| "loss": 1.9829, | |
| "rewards/accuracies": 0.706250011920929, | |
| "rewards/chosen": -1.1120072603225708, | |
| "rewards/margins": 0.4715605676174164, | |
| "rewards/rejected": -1.5835678577423096, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.7118555352002094, | |
| "grad_norm": 16.25494913759618, | |
| "learning_rate": 1.3873292242587306e-07, | |
| "logits/chosen": -0.8764508962631226, | |
| "logits/rejected": -0.9635313153266907, | |
| "logps/chosen": -1.2850573062896729, | |
| "logps/rejected": -1.7332632541656494, | |
| "loss": 2.0161, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.2850573062896729, | |
| "rewards/margins": 0.44820576906204224, | |
| "rewards/rejected": -1.7332632541656494, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.7223239989531536, | |
| "grad_norm": 46.740346613290164, | |
| "learning_rate": 1.295805759806533e-07, | |
| "logits/chosen": -0.8967689275741577, | |
| "logits/rejected": -0.9520794153213501, | |
| "logps/chosen": -1.177972435951233, | |
| "logps/rejected": -1.7682632207870483, | |
| "loss": 1.9908, | |
| "rewards/accuracies": 0.643750011920929, | |
| "rewards/chosen": -1.177972435951233, | |
| "rewards/margins": 0.5902906656265259, | |
| "rewards/rejected": -1.7682632207870483, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.7327924627060979, | |
| "grad_norm": 24.33920868401554, | |
| "learning_rate": 1.2065668194415777e-07, | |
| "logits/chosen": -0.9185832738876343, | |
| "logits/rejected": -0.9162608981132507, | |
| "logps/chosen": -1.1997076272964478, | |
| "logps/rejected": -1.6665418148040771, | |
| "loss": 2.0251, | |
| "rewards/accuracies": 0.6000000238418579, | |
| "rewards/chosen": -1.1997076272964478, | |
| "rewards/margins": 0.4668343663215637, | |
| "rewards/rejected": -1.6665418148040771, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.7432609264590422, | |
| "grad_norm": 17.97599943277886, | |
| "learning_rate": 1.1197320306710923e-07, | |
| "logits/chosen": -0.9116529226303101, | |
| "logits/rejected": -0.951018214225769, | |
| "logps/chosen": -1.0667304992675781, | |
| "logps/rejected": -1.7582483291625977, | |
| "loss": 1.93, | |
| "rewards/accuracies": 0.7124999761581421, | |
| "rewards/chosen": -1.0667304992675781, | |
| "rewards/margins": 0.6915179491043091, | |
| "rewards/rejected": -1.7582483291625977, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.7537293902119864, | |
| "grad_norm": 24.4350634478838, | |
| "learning_rate": 1.035417798164145e-07, | |
| "logits/chosen": -0.8609287142753601, | |
| "logits/rejected": -0.9163888096809387, | |
| "logps/chosen": -1.0682001113891602, | |
| "logps/rejected": -1.5872647762298584, | |
| "loss": 1.9091, | |
| "rewards/accuracies": 0.6625000238418579, | |
| "rewards/chosen": -1.0682001113891602, | |
| "rewards/margins": 0.5190645456314087, | |
| "rewards/rejected": -1.5872647762298584, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.7641978539649307, | |
| "grad_norm": 20.39985667399638, | |
| "learning_rate": 9.537371477076535e-08, | |
| "logits/chosen": -0.8938928842544556, | |
| "logits/rejected": -0.9040834307670593, | |
| "logps/chosen": -1.2582809925079346, | |
| "logps/rejected": -1.840842843055725, | |
| "loss": 2.0073, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.2582809925079346, | |
| "rewards/margins": 0.5825616121292114, | |
| "rewards/rejected": -1.840842843055725, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.7746663177178749, | |
| "grad_norm": 34.173481784767304, | |
| "learning_rate": 8.747995746918898e-08, | |
| "logits/chosen": -0.8829482197761536, | |
| "logits/rejected": -0.907356858253479, | |
| "logps/chosen": -1.230694055557251, | |
| "logps/rejected": -1.8831084966659546, | |
| "loss": 1.9603, | |
| "rewards/accuracies": 0.6875, | |
| "rewards/chosen": -1.230694055557251, | |
| "rewards/margins": 0.6524146199226379, | |
| "rewards/rejected": -1.8831084966659546, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.7851347814708192, | |
| "grad_norm": 47.37722808156103, | |
| "learning_rate": 7.987108973285888e-08, | |
| "logits/chosen": -0.9488266706466675, | |
| "logits/rejected": -0.9228233098983765, | |
| "logps/chosen": -1.203776240348816, | |
| "logps/rejected": -1.7508113384246826, | |
| "loss": 1.9796, | |
| "rewards/accuracies": 0.6187499761581421, | |
| "rewards/chosen": -1.203776240348816, | |
| "rewards/margins": 0.5470350980758667, | |
| "rewards/rejected": -1.7508113384246826, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.7956032452237635, | |
| "grad_norm": 30.92892535857735, | |
| "learning_rate": 7.255731147984174e-08, | |
| "logits/chosen": -0.9628175497055054, | |
| "logits/rejected": -0.8987346887588501, | |
| "logps/chosen": -1.2509288787841797, | |
| "logps/rejected": -1.6936590671539307, | |
| "loss": 1.9474, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.2509288787841797, | |
| "rewards/margins": 0.4427300989627838, | |
| "rewards/rejected": -1.6936590671539307, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.8060717089767077, | |
| "grad_norm": 14.25191792086823, | |
| "learning_rate": 6.554842705179898e-08, | |
| "logits/chosen": -0.8933061361312866, | |
| "logits/rejected": -0.9177101254463196, | |
| "logps/chosen": -1.1689434051513672, | |
| "logps/rejected": -1.6953961849212646, | |
| "loss": 1.9767, | |
| "rewards/accuracies": 0.5687500238418579, | |
| "rewards/chosen": -1.1689434051513672, | |
| "rewards/margins": 0.5264527797698975, | |
| "rewards/rejected": -1.6953961849212646, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.816540172729652, | |
| "grad_norm": 24.50848840983097, | |
| "learning_rate": 5.885383207096832e-08, | |
| "logits/chosen": -0.8765498995780945, | |
| "logits/rejected": -0.9643741846084595, | |
| "logps/chosen": -1.1704505681991577, | |
| "logps/rejected": -1.8245474100112915, | |
| "loss": 1.9596, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.1704505681991577, | |
| "rewards/margins": 0.654096782207489, | |
| "rewards/rejected": -1.8245474100112915, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.8270086364825961, | |
| "grad_norm": 49.84454621889519, | |
| "learning_rate": 5.2482500845047165e-08, | |
| "logits/chosen": -0.876386821269989, | |
| "logits/rejected": -0.9504436254501343, | |
| "logps/chosen": -1.1685442924499512, | |
| "logps/rejected": -1.7621757984161377, | |
| "loss": 1.9645, | |
| "rewards/accuracies": 0.6937500238418579, | |
| "rewards/chosen": -1.1685442924499512, | |
| "rewards/margins": 0.5936316251754761, | |
| "rewards/rejected": -1.7621757984161377, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.8374771002355405, | |
| "grad_norm": 24.717108157273426, | |
| "learning_rate": 4.644297433686162e-08, | |
| "logits/chosen": -0.8654069900512695, | |
| "logits/rejected": -0.8900250196456909, | |
| "logps/chosen": -1.1046303510665894, | |
| "logps/rejected": -1.7381641864776611, | |
| "loss": 1.9371, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1046303510665894, | |
| "rewards/margins": 0.6335335969924927, | |
| "rewards/rejected": -1.7381641864776611, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.8479455639884846, | |
| "grad_norm": 129.2852712689259, | |
| "learning_rate": 4.074334871494558e-08, | |
| "logits/chosen": -0.9184934496879578, | |
| "logits/rejected": -0.9239559173583984, | |
| "logps/chosen": -1.269236445426941, | |
| "logps/rejected": -1.8909661769866943, | |
| "loss": 1.9919, | |
| "rewards/accuracies": 0.65625, | |
| "rewards/chosen": -1.269236445426941, | |
| "rewards/margins": 0.6217294931411743, | |
| "rewards/rejected": -1.8909661769866943, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.8584140277414289, | |
| "grad_norm": 35.460771935934545, | |
| "learning_rate": 3.5391264500382e-08, | |
| "logits/chosen": -0.8921190500259399, | |
| "logits/rejected": -0.9240585565567017, | |
| "logps/chosen": -1.1124510765075684, | |
| "logps/rejected": -1.6030313968658447, | |
| "loss": 1.9615, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.1124510765075684, | |
| "rewards/margins": 0.49058032035827637, | |
| "rewards/rejected": -1.6030313968658447, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.8688824914943732, | |
| "grad_norm": 30.93762572614218, | |
| "learning_rate": 3.0393896324452226e-08, | |
| "logits/chosen": -0.9354831576347351, | |
| "logits/rejected": -1.0134633779525757, | |
| "logps/chosen": -1.14380943775177, | |
| "logps/rejected": -1.7913806438446045, | |
| "loss": 1.9275, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.14380943775177, | |
| "rewards/margins": 0.6475714445114136, | |
| "rewards/rejected": -1.7913806438446045, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.8793509552473174, | |
| "grad_norm": 23.73127172846191, | |
| "learning_rate": 2.5757943310825026e-08, | |
| "logits/chosen": -0.8869434595108032, | |
| "logits/rejected": -0.8954530954360962, | |
| "logps/chosen": -1.1472705602645874, | |
| "logps/rejected": -1.7492564916610718, | |
| "loss": 1.9917, | |
| "rewards/accuracies": 0.606249988079071, | |
| "rewards/chosen": -1.1472705602645874, | |
| "rewards/margins": 0.6019860506057739, | |
| "rewards/rejected": -1.7492564916610718, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.8898194190002617, | |
| "grad_norm": 39.148819343565215, | |
| "learning_rate": 2.148962009517823e-08, | |
| "logits/chosen": -0.8858118057250977, | |
| "logits/rejected": -0.9047883152961731, | |
| "logps/chosen": -1.2842466831207275, | |
| "logps/rejected": -1.7744776010513306, | |
| "loss": 1.9499, | |
| "rewards/accuracies": 0.612500011920929, | |
| "rewards/chosen": -1.2842466831207275, | |
| "rewards/margins": 0.49023112654685974, | |
| "rewards/rejected": -1.7744776010513306, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.9002878827532059, | |
| "grad_norm": 26.336696705653907, | |
| "learning_rate": 1.759464849429082e-08, | |
| "logits/chosen": -0.8909379243850708, | |
| "logits/rejected": -0.9339486360549927, | |
| "logps/chosen": -1.2071397304534912, | |
| "logps/rejected": -1.798255205154419, | |
| "loss": 1.9358, | |
| "rewards/accuracies": 0.6312500238418579, | |
| "rewards/chosen": -1.2071397304534912, | |
| "rewards/margins": 0.5911155939102173, | |
| "rewards/rejected": -1.798255205154419, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.9107563465061502, | |
| "grad_norm": 25.46620061376669, | |
| "learning_rate": 1.4078249835774169e-08, | |
| "logits/chosen": -0.9604193568229675, | |
| "logits/rejected": -0.983184814453125, | |
| "logps/chosen": -1.148258924484253, | |
| "logps/rejected": -1.949751615524292, | |
| "loss": 1.9302, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.148258924484253, | |
| "rewards/margins": 0.8014928102493286, | |
| "rewards/rejected": -1.949751615524292, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.9212248102590945, | |
| "grad_norm": 21.043141257775048, | |
| "learning_rate": 1.0945137958723705e-08, | |
| "logits/chosen": -0.894646167755127, | |
| "logits/rejected": -0.9200958013534546, | |
| "logps/chosen": -1.208785891532898, | |
| "logps/rejected": -1.7293192148208618, | |
| "loss": 2.0013, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.208785891532898, | |
| "rewards/margins": 0.5205334424972534, | |
| "rewards/rejected": -1.7293192148208618, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.9316932740120387, | |
| "grad_norm": 80.48128617194537, | |
| "learning_rate": 8.19951289467482e-09, | |
| "logits/chosen": -0.8895615339279175, | |
| "logits/rejected": -0.8950116038322449, | |
| "logps/chosen": -1.185006856918335, | |
| "logps/rejected": -1.7131178379058838, | |
| "loss": 2.011, | |
| "rewards/accuracies": 0.6499999761581421, | |
| "rewards/chosen": -1.185006856918335, | |
| "rewards/margins": 0.5281107425689697, | |
| "rewards/rejected": -1.7131178379058838, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.942161737764983, | |
| "grad_norm": 35.00393348206555, | |
| "learning_rate": 5.84505523733293e-09, | |
| "logits/chosen": -0.8778541684150696, | |
| "logits/rejected": -0.8582246899604797, | |
| "logps/chosen": -1.2027300596237183, | |
| "logps/rejected": -1.719333291053772, | |
| "loss": 1.9411, | |
| "rewards/accuracies": 0.699999988079071, | |
| "rewards/chosen": -1.2027300596237183, | |
| "rewards/margins": 0.5166033506393433, | |
| "rewards/rejected": -1.719333291053772, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.9526302015179272, | |
| "grad_norm": 45.63097391162078, | |
| "learning_rate": 3.8849212086261466e-09, | |
| "logits/chosen": -0.8674596548080444, | |
| "logits/rejected": -0.8951441645622253, | |
| "logps/chosen": -1.3365813493728638, | |
| "logps/rejected": -1.6624677181243896, | |
| "loss": 1.9961, | |
| "rewards/accuracies": 0.637499988079071, | |
| "rewards/chosen": -1.3365813493728638, | |
| "rewards/margins": 0.325886607170105, | |
| "rewards/rejected": -1.6624677181243896, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.9630986652708715, | |
| "grad_norm": 25.00020706198223, | |
| "learning_rate": 2.3217384276938756e-09, | |
| "logits/chosen": -0.9242098927497864, | |
| "logits/rejected": -0.9427459836006165, | |
| "logps/chosen": -1.098924160003662, | |
| "logps/rejected": -1.7310899496078491, | |
| "loss": 1.9237, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.098924160003662, | |
| "rewards/margins": 0.632165789604187, | |
| "rewards/rejected": -1.7310899496078491, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.9735671290238157, | |
| "grad_norm": 47.88185009368047, | |
| "learning_rate": 1.1576023884836472e-09, | |
| "logits/chosen": -0.9520083665847778, | |
| "logits/rejected": -0.9862750172615051, | |
| "logps/chosen": -1.2228929996490479, | |
| "logps/rejected": -1.7746740579605103, | |
| "loss": 1.9683, | |
| "rewards/accuracies": 0.675000011920929, | |
| "rewards/chosen": -1.2228929996490479, | |
| "rewards/margins": 0.5517812967300415, | |
| "rewards/rejected": -1.7746740579605103, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.98403559277676, | |
| "grad_norm": 49.47414271366926, | |
| "learning_rate": 3.940736506780395e-10, | |
| "logits/chosen": -0.8756373524665833, | |
| "logits/rejected": -0.9029091000556946, | |
| "logps/chosen": -1.1716878414154053, | |
| "logps/rejected": -1.6480519771575928, | |
| "loss": 2.017, | |
| "rewards/accuracies": 0.668749988079071, | |
| "rewards/chosen": -1.1716878414154053, | |
| "rewards/margins": 0.4763643741607666, | |
| "rewards/rejected": -1.6480519771575928, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.9945040565297043, | |
| "grad_norm": 20.978298819504356, | |
| "learning_rate": 3.2175747716822744e-11, | |
| "logits/chosen": -0.9227266311645508, | |
| "logits/rejected": -0.9477631449699402, | |
| "logps/chosen": -1.2197462320327759, | |
| "logps/rejected": -1.715574026107788, | |
| "loss": 1.9795, | |
| "rewards/accuracies": 0.5874999761581421, | |
| "rewards/chosen": -1.2197462320327759, | |
| "rewards/margins": 0.495827853679657, | |
| "rewards/rejected": -1.715574026107788, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.998691442030882, | |
| "step": 477, | |
| "total_flos": 0.0, | |
| "train_loss": 2.0152689970764226, | |
| "train_runtime": 14277.4481, | |
| "train_samples_per_second": 4.282, | |
| "train_steps_per_second": 0.033 | |
| } | |
| ], | |
| "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 | |
| } | |