| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.96, | |
| "eval_steps": 500, | |
| "global_step": 186, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.016, | |
| "grad_norm": 17.265566540871276, | |
| "learning_rate": 5.263157894736843e-07, | |
| "loss": 1.3849, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 20.089852093670576, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 1.4064, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 6.589490853652612, | |
| "learning_rate": 1.5789473684210526e-06, | |
| "loss": 1.0402, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 5.14296612105976, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 0.895, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 8.124533688517559, | |
| "learning_rate": 2.631578947368421e-06, | |
| "loss": 1.0815, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 8.84827182866774, | |
| "learning_rate": 3.157894736842105e-06, | |
| "loss": 1.2908, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 8.203756927127115, | |
| "learning_rate": 3.6842105263157896e-06, | |
| "loss": 1.2459, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 7.57722281069691, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 0.9707, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 7.102462941011033, | |
| "learning_rate": 4.736842105263158e-06, | |
| "loss": 1.0695, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 3.382564005385294, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 0.8253, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 2.883997467358394, | |
| "learning_rate": 5.789473684210527e-06, | |
| "loss": 0.7806, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 3.48767265499384, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 0.9426, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 3.6801016439021033, | |
| "learning_rate": 6.842105263157896e-06, | |
| "loss": 0.9984, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 5.111818314033425, | |
| "learning_rate": 7.368421052631579e-06, | |
| "loss": 1.0925, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 3.00809107659031, | |
| "learning_rate": 7.894736842105265e-06, | |
| "loss": 0.8113, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 2.5810942766550613, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 0.8985, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 2.239234193084543, | |
| "learning_rate": 8.947368421052632e-06, | |
| "loss": 0.801, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 1.977889876770481, | |
| "learning_rate": 9.473684210526315e-06, | |
| "loss": 0.8024, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 2.5321288733806107, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9323, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 2.3084642254292005, | |
| "learning_rate": 9.999115304121459e-06, | |
| "loss": 0.8174, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 2.275895453495919, | |
| "learning_rate": 9.996461529560553e-06, | |
| "loss": 0.7172, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 2.390705215087479, | |
| "learning_rate": 9.992039615430648e-06, | |
| "loss": 0.7238, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 1.9421657131294014, | |
| "learning_rate": 9.985851126551428e-06, | |
| "loss": 0.8107, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 2.058372329799834, | |
| "learning_rate": 9.977898252895133e-06, | |
| "loss": 0.7229, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.9006752617160494, | |
| "learning_rate": 9.968183808811586e-06, | |
| "loss": 0.6931, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 2.054145992306095, | |
| "learning_rate": 9.95671123203224e-06, | |
| "loss": 0.746, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 1.9468770299243592, | |
| "learning_rate": 9.943484582453653e-06, | |
| "loss": 0.6654, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.814496686840529, | |
| "learning_rate": 9.928508540700775e-06, | |
| "loss": 0.6921, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 1.802486279814975, | |
| "learning_rate": 9.91178840647057e-06, | |
| "loss": 0.7111, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 1.9478710077072412, | |
| "learning_rate": 9.893330096656576e-06, | |
| "loss": 0.6362, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.496, | |
| "grad_norm": 1.8318214451416757, | |
| "learning_rate": 9.873140143255035e-06, | |
| "loss": 0.7653, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 1.8452709208358775, | |
| "learning_rate": 9.851225691053382e-06, | |
| "loss": 0.6662, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 1.983049889607633, | |
| "learning_rate": 9.827594495101824e-06, | |
| "loss": 0.7745, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 1.8290774383219428, | |
| "learning_rate": 9.802254917969033e-06, | |
| "loss": 0.7405, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 1.732618942001559, | |
| "learning_rate": 9.775215926782788e-06, | |
| "loss": 0.705, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.8003122188796479, | |
| "learning_rate": 9.746487090056712e-06, | |
| "loss": 0.6401, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.592, | |
| "grad_norm": 2.0123189922467315, | |
| "learning_rate": 9.71607857430419e-06, | |
| "loss": 0.7768, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 1.9124994424711974, | |
| "learning_rate": 9.68400114044064e-06, | |
| "loss": 0.801, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 1.6874581302472245, | |
| "learning_rate": 9.650266139975474e-06, | |
| "loss": 0.7914, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 1.8036614396987305, | |
| "learning_rate": 9.614885510995047e-06, | |
| "loss": 0.6431, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.656, | |
| "grad_norm": 1.7792173564341356, | |
| "learning_rate": 9.577871773938013e-06, | |
| "loss": 0.6685, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 2.106904023154908, | |
| "learning_rate": 9.539238027164618e-06, | |
| "loss": 0.7304, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.688, | |
| "grad_norm": 1.9512092168832307, | |
| "learning_rate": 9.498997942321484e-06, | |
| "loss": 0.7133, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 1.8794356205862166, | |
| "learning_rate": 9.457165759503492e-06, | |
| "loss": 0.7122, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.9449756616113063, | |
| "learning_rate": 9.413756282214538e-06, | |
| "loss": 0.5962, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 1.749613648866588, | |
| "learning_rate": 9.368784872128877e-06, | |
| "loss": 0.6571, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.752, | |
| "grad_norm": 1.9668503728809823, | |
| "learning_rate": 9.322267443654974e-06, | |
| "loss": 0.7153, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 2.0756580484041245, | |
| "learning_rate": 9.274220458303727e-06, | |
| "loss": 0.6427, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.784, | |
| "grad_norm": 1.7320764922054182, | |
| "learning_rate": 9.224660918863104e-06, | |
| "loss": 0.7456, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 1.7692368287021367, | |
| "learning_rate": 9.173606363381218e-06, | |
| "loss": 0.729, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 1.9599425283213727, | |
| "learning_rate": 9.121074858959997e-06, | |
| "loss": 0.6243, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 1.6116899916566545, | |
| "learning_rate": 9.067084995361623e-06, | |
| "loss": 0.6753, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.848, | |
| "grad_norm": 1.9862542412605975, | |
| "learning_rate": 9.011655878430018e-06, | |
| "loss": 0.656, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 1.8386860819765696, | |
| "learning_rate": 8.954807123329703e-06, | |
| "loss": 0.7196, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 1.6272214888273704, | |
| "learning_rate": 8.896558847604414e-06, | |
| "loss": 0.6435, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 1.7446187430114655, | |
| "learning_rate": 8.836931664057935e-06, | |
| "loss": 0.7281, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 1.7164804652605405, | |
| "learning_rate": 8.775946673459682e-06, | |
| "loss": 0.6243, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 1.5894836987127223, | |
| "learning_rate": 8.713625457077585e-06, | |
| "loss": 0.648, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.944, | |
| "grad_norm": 1.6190441689783222, | |
| "learning_rate": 8.64999006904096e-06, | |
| "loss": 0.6313, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.7601660204750595, | |
| "learning_rate": 8.585063028536015e-06, | |
| "loss": 0.6445, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.976, | |
| "grad_norm": 1.6814846694055416, | |
| "learning_rate": 8.518867311836808e-06, | |
| "loss": 0.6175, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 1.554320926318259, | |
| "learning_rate": 8.451426344174433e-06, | |
| "loss": 0.6353, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.554320926318259, | |
| "learning_rate": 8.382763991447344e-06, | |
| "loss": 0.6396, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.016, | |
| "grad_norm": 2.514616810040317, | |
| "learning_rate": 8.312904551775731e-06, | |
| "loss": 0.589, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.032, | |
| "grad_norm": 1.5914266097871264, | |
| "learning_rate": 8.241872746902934e-06, | |
| "loss": 0.642, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.048, | |
| "grad_norm": 1.6715859096702153, | |
| "learning_rate": 8.16969371344696e-06, | |
| "loss": 0.5919, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.064, | |
| "grad_norm": 1.7221785188395358, | |
| "learning_rate": 8.096392994005177e-06, | |
| "loss": 0.5219, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 1.523934181331874, | |
| "learning_rate": 8.021996528115335e-06, | |
| "loss": 0.4975, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.096, | |
| "grad_norm": 1.7251860309038605, | |
| "learning_rate": 7.946530643076138e-06, | |
| "loss": 0.4045, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.112, | |
| "grad_norm": 2.1387296754306453, | |
| "learning_rate": 7.870022044630569e-06, | |
| "loss": 0.5663, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1280000000000001, | |
| "grad_norm": 1.5678644951260237, | |
| "learning_rate": 7.792497807515317e-06, | |
| "loss": 0.5071, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.144, | |
| "grad_norm": 2.1069613377109864, | |
| "learning_rate": 7.713985365879607e-06, | |
| "loss": 0.4374, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 1.7700256012070563, | |
| "learning_rate": 7.63451250357685e-06, | |
| "loss": 0.5385, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.176, | |
| "grad_norm": 1.4683115653202277, | |
| "learning_rate": 7.55410734433254e-06, | |
| "loss": 0.5088, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.192, | |
| "grad_norm": 3.2315487431671173, | |
| "learning_rate": 7.472798341791877e-06, | |
| "loss": 0.394, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.208, | |
| "grad_norm": 1.7315310300769962, | |
| "learning_rate": 7.390614269450633e-06, | |
| "loss": 0.5133, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.224, | |
| "grad_norm": 1.7186427379174711, | |
| "learning_rate": 7.3075842104728445e-06, | |
| "loss": 0.4349, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 1.718919488729367, | |
| "learning_rate": 7.223737547398898e-06, | |
| "loss": 0.4948, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.256, | |
| "grad_norm": 1.7710317999113074, | |
| "learning_rate": 7.139103951747694e-06, | |
| "loss": 0.4855, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.272, | |
| "grad_norm": 1.7053255613043878, | |
| "learning_rate": 7.053713373516538e-06, | |
| "loss": 0.4599, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.288, | |
| "grad_norm": 1.6455437889543758, | |
| "learning_rate": 6.9675960305824785e-06, | |
| "loss": 0.5597, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.304, | |
| "grad_norm": 1.928985400009411, | |
| "learning_rate": 6.880782398008862e-06, | |
| "loss": 0.4486, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 1.557246422551428, | |
| "learning_rate": 6.7933031972608644e-06, | |
| "loss": 0.5173, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.336, | |
| "grad_norm": 1.6393519103212826, | |
| "learning_rate": 6.70518938533383e-06, | |
| "loss": 0.4891, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3519999999999999, | |
| "grad_norm": 1.5087677558718193, | |
| "learning_rate": 6.61647214379826e-06, | |
| "loss": 0.5729, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.3679999999999999, | |
| "grad_norm": 1.607244932229975, | |
| "learning_rate": 6.527182867765333e-06, | |
| "loss": 0.4997, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.384, | |
| "grad_norm": 1.6372031149188508, | |
| "learning_rate": 6.437353154776848e-06, | |
| "loss": 0.4983, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 1.763840958323373, | |
| "learning_rate": 6.3470147936235485e-06, | |
| "loss": 0.5028, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.416, | |
| "grad_norm": 1.7861540798275108, | |
| "learning_rate": 6.256199753095745e-06, | |
| "loss": 0.5406, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.432, | |
| "grad_norm": 1.499957779372258, | |
| "learning_rate": 6.164940170670266e-06, | |
| "loss": 0.4305, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.448, | |
| "grad_norm": 1.995733104339069, | |
| "learning_rate": 6.073268341137694e-06, | |
| "loss": 0.4544, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.464, | |
| "grad_norm": 1.9151523868805838, | |
| "learning_rate": 5.98121670517393e-06, | |
| "loss": 0.5066, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 1.4756706730929385, | |
| "learning_rate": 5.8888178378601565e-06, | |
| "loss": 0.5019, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.496, | |
| "grad_norm": 1.649862272540868, | |
| "learning_rate": 5.796104437155213e-06, | |
| "loss": 0.4756, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.512, | |
| "grad_norm": 1.7212809492557795, | |
| "learning_rate": 5.703109312324493e-06, | |
| "loss": 0.4896, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.528, | |
| "grad_norm": 1.9194800100352978, | |
| "learning_rate": 5.609865372329461e-06, | |
| "loss": 0.5845, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.544, | |
| "grad_norm": 1.8731612795860104, | |
| "learning_rate": 5.516405614181883e-06, | |
| "loss": 0.5868, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 1.4387957716763344, | |
| "learning_rate": 5.4227631112668955e-06, | |
| "loss": 0.592, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.576, | |
| "grad_norm": 2.2104402624908706, | |
| "learning_rate": 5.328971001639054e-06, | |
| "loss": 0.5143, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.592, | |
| "grad_norm": 1.4292270032965346, | |
| "learning_rate": 5.235062476295488e-06, | |
| "loss": 0.5155, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.608, | |
| "grad_norm": 1.7550654702916557, | |
| "learning_rate": 5.141070767430331e-06, | |
| "loss": 0.4888, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.624, | |
| "grad_norm": 1.5018754865504773, | |
| "learning_rate": 5.047029136674563e-06, | |
| "loss": 0.5572, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6400000000000001, | |
| "grad_norm": 1.51385812350827, | |
| "learning_rate": 4.95297086332544e-06, | |
| "loss": 0.4697, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6560000000000001, | |
| "grad_norm": 2.038637843864613, | |
| "learning_rate": 4.858929232569671e-06, | |
| "loss": 0.4527, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6720000000000002, | |
| "grad_norm": 1.520271301988124, | |
| "learning_rate": 4.7649375237045135e-06, | |
| "loss": 0.4727, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.688, | |
| "grad_norm": 1.5283877837212068, | |
| "learning_rate": 4.671028998360947e-06, | |
| "loss": 0.5009, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.704, | |
| "grad_norm": 1.7732975300862988, | |
| "learning_rate": 4.5772368887331044e-06, | |
| "loss": 0.5175, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 1.7652466043543638, | |
| "learning_rate": 4.483594385818119e-06, | |
| "loss": 0.3725, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.736, | |
| "grad_norm": 2.338369077116064, | |
| "learning_rate": 4.39013462767054e-06, | |
| "loss": 0.6059, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.752, | |
| "grad_norm": 1.693230322723448, | |
| "learning_rate": 4.29689068767551e-06, | |
| "loss": 0.4357, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.768, | |
| "grad_norm": 1.827758943805404, | |
| "learning_rate": 4.203895562844789e-06, | |
| "loss": 0.4563, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.784, | |
| "grad_norm": 1.7904258585266672, | |
| "learning_rate": 4.111182162139844e-06, | |
| "loss": 0.4377, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 1.5612150330770178, | |
| "learning_rate": 4.018783294826071e-06, | |
| "loss": 0.4275, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.8159999999999998, | |
| "grad_norm": 1.523037615179662, | |
| "learning_rate": 3.926731658862307e-06, | |
| "loss": 0.5228, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.8319999999999999, | |
| "grad_norm": 1.5255447437439678, | |
| "learning_rate": 3.8350598293297345e-06, | |
| "loss": 0.5258, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8479999999999999, | |
| "grad_norm": 1.5829211736970477, | |
| "learning_rate": 3.7438002469042567e-06, | |
| "loss": 0.4063, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.8639999999999999, | |
| "grad_norm": 1.5773032081061509, | |
| "learning_rate": 3.652985206376455e-06, | |
| "loss": 0.4821, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.4805521281076799, | |
| "learning_rate": 3.5626468452231534e-06, | |
| "loss": 0.4939, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.896, | |
| "grad_norm": 1.5554298604401637, | |
| "learning_rate": 3.472817132234669e-06, | |
| "loss": 0.5338, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.912, | |
| "grad_norm": 1.5494033362764545, | |
| "learning_rate": 3.3835278562017405e-06, | |
| "loss": 0.4957, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.928, | |
| "grad_norm": 1.5831323686883927, | |
| "learning_rate": 3.29481061466617e-06, | |
| "loss": 0.4285, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.944, | |
| "grad_norm": 1.5898749396215215, | |
| "learning_rate": 3.2066968027391377e-06, | |
| "loss": 0.457, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 1.5157195677831812, | |
| "learning_rate": 3.119217601991139e-06, | |
| "loss": 0.4218, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.976, | |
| "grad_norm": 1.5731407093139658, | |
| "learning_rate": 3.032403969417523e-06, | |
| "loss": 0.3875, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.992, | |
| "grad_norm": 1.7232422836213686, | |
| "learning_rate": 2.946286626483463e-06, | |
| "loss": 0.4412, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.393109486523402, | |
| "learning_rate": 2.8608960482523058e-06, | |
| "loss": 0.4625, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 2.257370296808367, | |
| "learning_rate": 2.776262452601104e-06, | |
| "loss": 0.3664, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.032, | |
| "grad_norm": 2.0798503170345866, | |
| "learning_rate": 2.6924157895271563e-06, | |
| "loss": 0.37, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 1.6371039756658181, | |
| "learning_rate": 2.6093857305493666e-06, | |
| "loss": 0.322, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 1.5223327796553272, | |
| "learning_rate": 2.5272016582081236e-06, | |
| "loss": 0.3695, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.6380378858305447, | |
| "learning_rate": 2.445892655667462e-06, | |
| "loss": 0.3229, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.096, | |
| "grad_norm": 3.120605376278745, | |
| "learning_rate": 2.365487496423152e-06, | |
| "loss": 0.3354, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 1.8530996062038518, | |
| "learning_rate": 2.2860146341203936e-06, | |
| "loss": 0.3207, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.128, | |
| "grad_norm": 2.079838959175905, | |
| "learning_rate": 2.207502192484685e-06, | |
| "loss": 0.2488, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.144, | |
| "grad_norm": 1.8087558940405277, | |
| "learning_rate": 2.1299779553694323e-06, | |
| "loss": 0.4157, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 1.8451844625035758, | |
| "learning_rate": 2.053469356923865e-06, | |
| "loss": 0.4506, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 1.6518452502805208, | |
| "learning_rate": 1.9780034718846653e-06, | |
| "loss": 0.3653, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.192, | |
| "grad_norm": 1.663937556819872, | |
| "learning_rate": 1.9036070059948253e-06, | |
| "loss": 0.3349, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 1.5820059418011216, | |
| "learning_rate": 1.8303062865530407e-06, | |
| "loss": 0.3425, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.224, | |
| "grad_norm": 1.576843734833355, | |
| "learning_rate": 1.7581272530970666e-06, | |
| "loss": 0.3322, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 1.7275837235833509, | |
| "learning_rate": 1.6870954482242707e-06, | |
| "loss": 0.3241, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 1.726720454583051, | |
| "learning_rate": 1.6172360085526567e-06, | |
| "loss": 0.2696, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.2720000000000002, | |
| "grad_norm": 2.010215233237989, | |
| "learning_rate": 1.54857365582557e-06, | |
| "loss": 0.3464, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.288, | |
| "grad_norm": 1.8848106151516786, | |
| "learning_rate": 1.4811326881631937e-06, | |
| "loss": 0.3499, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 1.5388938200969897, | |
| "learning_rate": 1.4149369714639856e-06, | |
| "loss": 0.3419, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 1.5044322184419996, | |
| "learning_rate": 1.3500099309590397e-06, | |
| "loss": 0.4196, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.336, | |
| "grad_norm": 1.6274762480553728, | |
| "learning_rate": 1.2863745429224145e-06, | |
| "loss": 0.3369, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 1.4496643724972158, | |
| "learning_rate": 1.22405332654032e-06, | |
| "loss": 0.4233, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 1.943638242632571, | |
| "learning_rate": 1.1630683359420653e-06, | |
| "loss": 0.358, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.384, | |
| "grad_norm": 1.933655917159831, | |
| "learning_rate": 1.103441152395588e-06, | |
| "loss": 0.2888, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 2.031085917073374, | |
| "learning_rate": 1.045192876670298e-06, | |
| "loss": 0.2591, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.416, | |
| "grad_norm": 1.5293961459532648, | |
| "learning_rate": 9.883441215699824e-07, | |
| "loss": 0.3791, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 1.588731036259167, | |
| "learning_rate": 9.329150046383773e-07, | |
| "loss": 0.331, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 1.4858621936000498, | |
| "learning_rate": 8.789251410400024e-07, | |
| "loss": 0.2818, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.464, | |
| "grad_norm": 1.7426467948455326, | |
| "learning_rate": 8.263936366187825e-07, | |
| "loss": 0.3005, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 1.7460841045163584, | |
| "learning_rate": 7.753390811368972e-07, | |
| "loss": 0.3003, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 1.5976363728033662, | |
| "learning_rate": 7.257795416962754e-07, | |
| "loss": 0.3347, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.512, | |
| "grad_norm": 1.885887380938252, | |
| "learning_rate": 6.777325563450282e-07, | |
| "loss": 0.4411, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.528, | |
| "grad_norm": 1.6121877256346284, | |
| "learning_rate": 6.312151278711237e-07, | |
| "loss": 0.3053, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 1.6916631587022424, | |
| "learning_rate": 5.862437177854629e-07, | |
| "loss": 0.3424, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 1.5105490449456516, | |
| "learning_rate": 5.428342404965076e-07, | |
| "loss": 0.3474, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.576, | |
| "grad_norm": 1.5897890361860791, | |
| "learning_rate": 5.010020576785174e-07, | |
| "loss": 0.3146, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 1.7421226681944129, | |
| "learning_rate": 4.607619728353818e-07, | |
| "loss": 0.3134, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.608, | |
| "grad_norm": 1.6114545605409272, | |
| "learning_rate": 4.221282260619891e-07, | |
| "loss": 0.3337, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 1.6581271061315566, | |
| "learning_rate": 3.851144890049535e-07, | |
| "loss": 0.4023, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 1.6669298907958805, | |
| "learning_rate": 3.497338600245254e-07, | |
| "loss": 0.3021, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.656, | |
| "grad_norm": 1.5082405058478559, | |
| "learning_rate": 3.159988595593616e-07, | |
| "loss": 0.3485, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.672, | |
| "grad_norm": 1.6867142867470823, | |
| "learning_rate": 2.839214256958106e-07, | |
| "loss": 0.3852, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 1.4197906212673126, | |
| "learning_rate": 2.5351290994328703e-07, | |
| "loss": 0.3677, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.7039999999999997, | |
| "grad_norm": 1.6371560986781053, | |
| "learning_rate": 2.2478407321721295e-07, | |
| "loss": 0.2545, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 1.513927276530402, | |
| "learning_rate": 1.9774508203096843e-07, | |
| "loss": 0.2907, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 1.6505772046236602, | |
| "learning_rate": 1.7240550489817652e-07, | |
| "loss": 0.3116, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 1.6467061689919404, | |
| "learning_rate": 1.4877430894662037e-07, | |
| "loss": 0.2597, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.768, | |
| "grad_norm": 1.3910601508999687, | |
| "learning_rate": 1.268598567449647e-07, | |
| "loss": 0.4666, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 1.5158376103770992, | |
| "learning_rate": 1.0666990334342708e-07, | |
| "loss": 0.3047, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 1.4078568916572975, | |
| "learning_rate": 8.821159352943142e-08, | |
| "loss": 0.363, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 1.5414756510402199, | |
| "learning_rate": 7.149145929922607e-08, | |
| "loss": 0.3118, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 1.5590409985711402, | |
| "learning_rate": 5.651541754634726e-08, | |
| "loss": 0.2766, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.848, | |
| "grad_norm": 1.4929985095640963, | |
| "learning_rate": 4.328876796776071e-08, | |
| "loss": 0.3311, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.864, | |
| "grad_norm": 1.5070078137197234, | |
| "learning_rate": 3.181619118841517e-08, | |
| "loss": 0.3399, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.4238651288325814, | |
| "learning_rate": 2.210174710486679e-08, | |
| "loss": 0.4328, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.896, | |
| "grad_norm": 1.4194361676055525, | |
| "learning_rate": 1.4148873448573408e-08, | |
| "loss": 0.3449, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.912, | |
| "grad_norm": 1.5984952095126612, | |
| "learning_rate": 7.96038456935322e-09, | |
| "loss": 0.3814, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 1.4302652833406175, | |
| "learning_rate": 3.538470439448105e-09, | |
| "loss": 0.3605, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 1.4596899305481994, | |
| "learning_rate": 8.846958785418969e-10, | |
| "loss": 0.2759, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 1.4920781635835423, | |
| "learning_rate": 0.0, | |
| "loss": 0.2919, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "step": 186, | |
| "total_flos": 21670257623040.0, | |
| "train_loss": 0.5450129303079779, | |
| "train_runtime": 5234.2515, | |
| "train_samples_per_second": 0.573, | |
| "train_steps_per_second": 0.036 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 186, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 21670257623040.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |