| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.976, | |
| "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": 7.007598400115967, | |
| "learning_rate": 5.263157894736843e-07, | |
| "loss": 1.2811, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.032, | |
| "grad_norm": 6.761180400848389, | |
| "learning_rate": 1.0526315789473685e-06, | |
| "loss": 1.2323, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.048, | |
| "grad_norm": 6.8008036613464355, | |
| "learning_rate": 1.5789473684210526e-06, | |
| "loss": 1.2607, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 6.8452467918396, | |
| "learning_rate": 2.105263157894737e-06, | |
| "loss": 1.3123, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 6.010447978973389, | |
| "learning_rate": 2.631578947368421e-06, | |
| "loss": 1.2593, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.096, | |
| "grad_norm": 4.768431663513184, | |
| "learning_rate": 3.157894736842105e-06, | |
| "loss": 1.1822, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.112, | |
| "grad_norm": 2.9961299896240234, | |
| "learning_rate": 3.6842105263157896e-06, | |
| "loss": 1.1504, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 2.724243640899658, | |
| "learning_rate": 4.210526315789474e-06, | |
| "loss": 1.1726, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.144, | |
| "grad_norm": 2.480252265930176, | |
| "learning_rate": 4.736842105263158e-06, | |
| "loss": 1.1204, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 4.808835983276367, | |
| "learning_rate": 5.263157894736842e-06, | |
| "loss": 1.1539, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.176, | |
| "grad_norm": 5.101093292236328, | |
| "learning_rate": 5.789473684210527e-06, | |
| "loss": 1.1375, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 4.860002517700195, | |
| "learning_rate": 6.31578947368421e-06, | |
| "loss": 1.1462, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.208, | |
| "grad_norm": 4.055907249450684, | |
| "learning_rate": 6.842105263157896e-06, | |
| "loss": 1.1125, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.224, | |
| "grad_norm": 3.9394595623016357, | |
| "learning_rate": 7.368421052631579e-06, | |
| "loss": 1.0942, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 3.135388135910034, | |
| "learning_rate": 7.894736842105265e-06, | |
| "loss": 1.0433, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 2.2031311988830566, | |
| "learning_rate": 8.421052631578948e-06, | |
| "loss": 1.049, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.272, | |
| "grad_norm": 2.0445187091827393, | |
| "learning_rate": 8.947368421052632e-06, | |
| "loss": 1.0181, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.288, | |
| "grad_norm": 2.0096278190612793, | |
| "learning_rate": 9.473684210526315e-06, | |
| "loss": 0.9997, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.304, | |
| "grad_norm": 1.7502564191818237, | |
| "learning_rate": 1e-05, | |
| "loss": 0.9879, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 1.5390841960906982, | |
| "learning_rate": 9.999115304121459e-06, | |
| "loss": 1.0304, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.336, | |
| "grad_norm": 1.2779120206832886, | |
| "learning_rate": 9.996461529560553e-06, | |
| "loss": 0.9906, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.352, | |
| "grad_norm": 1.2320711612701416, | |
| "learning_rate": 9.992039615430648e-06, | |
| "loss": 1.011, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.368, | |
| "grad_norm": 1.182990312576294, | |
| "learning_rate": 9.985851126551428e-06, | |
| "loss": 0.9907, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 1.1079776287078857, | |
| "learning_rate": 9.977898252895133e-06, | |
| "loss": 0.9597, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.1073062419891357, | |
| "learning_rate": 9.968183808811586e-06, | |
| "loss": 0.9848, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.416, | |
| "grad_norm": 1.2341408729553223, | |
| "learning_rate": 9.95671123203224e-06, | |
| "loss": 0.9474, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.432, | |
| "grad_norm": 1.1053662300109863, | |
| "learning_rate": 9.943484582453653e-06, | |
| "loss": 0.9498, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.9707959890365601, | |
| "learning_rate": 9.928508540700775e-06, | |
| "loss": 0.9398, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.464, | |
| "grad_norm": 0.8282796144485474, | |
| "learning_rate": 9.91178840647057e-06, | |
| "loss": 0.9456, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 0.8281641602516174, | |
| "learning_rate": 9.893330096656576e-06, | |
| "loss": 0.9409, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.496, | |
| "grad_norm": 0.9116457104682922, | |
| "learning_rate": 9.873140143255035e-06, | |
| "loss": 0.9358, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.8681783080101013, | |
| "learning_rate": 9.851225691053382e-06, | |
| "loss": 0.9528, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.528, | |
| "grad_norm": 0.7963255643844604, | |
| "learning_rate": 9.827594495101824e-06, | |
| "loss": 0.9102, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.544, | |
| "grad_norm": 0.8326982855796814, | |
| "learning_rate": 9.802254917969033e-06, | |
| "loss": 0.9342, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 0.9402235746383667, | |
| "learning_rate": 9.775215926782788e-06, | |
| "loss": 0.9357, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.7626885771751404, | |
| "learning_rate": 9.746487090056712e-06, | |
| "loss": 0.8703, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.592, | |
| "grad_norm": 0.649657130241394, | |
| "learning_rate": 9.71607857430419e-06, | |
| "loss": 0.903, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.608, | |
| "grad_norm": 0.941702663898468, | |
| "learning_rate": 9.68400114044064e-06, | |
| "loss": 0.9288, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.624, | |
| "grad_norm": 0.8211132884025574, | |
| "learning_rate": 9.650266139975474e-06, | |
| "loss": 0.9291, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.6539824604988098, | |
| "learning_rate": 9.614885510995047e-06, | |
| "loss": 0.9348, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.656, | |
| "grad_norm": 0.6680752038955688, | |
| "learning_rate": 9.577871773938013e-06, | |
| "loss": 0.8988, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.672, | |
| "grad_norm": 0.8227656483650208, | |
| "learning_rate": 9.539238027164618e-06, | |
| "loss": 0.8974, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.688, | |
| "grad_norm": 0.9751904606819153, | |
| "learning_rate": 9.498997942321484e-06, | |
| "loss": 0.9121, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.6582592725753784, | |
| "learning_rate": 9.457165759503492e-06, | |
| "loss": 0.8954, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 1.0337086915969849, | |
| "learning_rate": 9.413756282214538e-06, | |
| "loss": 0.9307, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.736, | |
| "grad_norm": 0.8674771189689636, | |
| "learning_rate": 9.368784872128877e-06, | |
| "loss": 0.9164, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.752, | |
| "grad_norm": 0.6859424114227295, | |
| "learning_rate": 9.322267443654974e-06, | |
| "loss": 0.9151, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.827805757522583, | |
| "learning_rate": 9.274220458303727e-06, | |
| "loss": 0.896, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.784, | |
| "grad_norm": 0.701875627040863, | |
| "learning_rate": 9.224660918863104e-06, | |
| "loss": 0.9154, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 0.6655284762382507, | |
| "learning_rate": 9.173606363381218e-06, | |
| "loss": 0.9351, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.816, | |
| "grad_norm": 0.6565393209457397, | |
| "learning_rate": 9.121074858959997e-06, | |
| "loss": 0.921, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.7235894799232483, | |
| "learning_rate": 9.067084995361623e-06, | |
| "loss": 0.9249, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.848, | |
| "grad_norm": 0.6363807916641235, | |
| "learning_rate": 9.011655878430018e-06, | |
| "loss": 0.9208, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.864, | |
| "grad_norm": 0.723604142665863, | |
| "learning_rate": 8.954807123329703e-06, | |
| "loss": 0.9103, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 0.557287335395813, | |
| "learning_rate": 8.896558847604414e-06, | |
| "loss": 0.9309, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.6184971332550049, | |
| "learning_rate": 8.836931664057935e-06, | |
| "loss": 0.9204, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.912, | |
| "grad_norm": 0.5848482251167297, | |
| "learning_rate": 8.775946673459682e-06, | |
| "loss": 0.8756, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.928, | |
| "grad_norm": 0.745296061038971, | |
| "learning_rate": 8.713625457077585e-06, | |
| "loss": 0.8895, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.944, | |
| "grad_norm": 0.7137336730957031, | |
| "learning_rate": 8.64999006904096e-06, | |
| "loss": 0.8999, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6531467437744141, | |
| "learning_rate": 8.585063028536015e-06, | |
| "loss": 0.8874, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.976, | |
| "grad_norm": 0.8325201272964478, | |
| "learning_rate": 8.518867311836808e-06, | |
| "loss": 0.9085, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.992, | |
| "grad_norm": 0.6783756017684937, | |
| "learning_rate": 8.451426344174433e-06, | |
| "loss": 0.9178, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 1.008, | |
| "grad_norm": 0.9748940467834473, | |
| "learning_rate": 8.382763991447344e-06, | |
| "loss": 1.1564, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.6522616744041443, | |
| "learning_rate": 8.312904551775731e-06, | |
| "loss": 1.0139, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 0.7602720856666565, | |
| "learning_rate": 8.241872746902934e-06, | |
| "loss": 0.8157, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.056, | |
| "grad_norm": 0.6934854388237, | |
| "learning_rate": 8.16969371344696e-06, | |
| "loss": 0.7933, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 1.072, | |
| "grad_norm": 0.771884024143219, | |
| "learning_rate": 8.096392994005177e-06, | |
| "loss": 0.9664, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.6898618936538696, | |
| "learning_rate": 8.021996528115335e-06, | |
| "loss": 0.8397, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 1.104, | |
| "grad_norm": 0.7334851026535034, | |
| "learning_rate": 7.946530643076138e-06, | |
| "loss": 0.8189, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.6383283734321594, | |
| "learning_rate": 7.870022044630569e-06, | |
| "loss": 0.8748, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.1360000000000001, | |
| "grad_norm": 0.7002072334289551, | |
| "learning_rate": 7.792497807515317e-06, | |
| "loss": 0.911, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.858782947063446, | |
| "learning_rate": 7.713985365879607e-06, | |
| "loss": 0.9038, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 1.168, | |
| "grad_norm": 0.5368704795837402, | |
| "learning_rate": 7.63451250357685e-06, | |
| "loss": 0.7338, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 1.184, | |
| "grad_norm": 0.9001109004020691, | |
| "learning_rate": 7.55410734433254e-06, | |
| "loss": 0.9501, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.7221422791481018, | |
| "learning_rate": 7.472798341791877e-06, | |
| "loss": 0.8639, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.6477048397064209, | |
| "learning_rate": 7.390614269450633e-06, | |
| "loss": 0.9042, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 1.232, | |
| "grad_norm": 0.8133475184440613, | |
| "learning_rate": 7.3075842104728445e-06, | |
| "loss": 0.7982, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 1.248, | |
| "grad_norm": 0.639371395111084, | |
| "learning_rate": 7.223737547398898e-06, | |
| "loss": 0.8959, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.264, | |
| "grad_norm": 0.5538738369941711, | |
| "learning_rate": 7.139103951747694e-06, | |
| "loss": 0.771, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.698175311088562, | |
| "learning_rate": 7.053713373516538e-06, | |
| "loss": 0.8581, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.296, | |
| "grad_norm": 0.5397263169288635, | |
| "learning_rate": 6.9675960305824785e-06, | |
| "loss": 0.8207, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.312, | |
| "grad_norm": 0.7133552432060242, | |
| "learning_rate": 6.880782398008862e-06, | |
| "loss": 0.8967, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.328, | |
| "grad_norm": 0.5522973537445068, | |
| "learning_rate": 6.7933031972608644e-06, | |
| "loss": 0.8124, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.5570671558380127, | |
| "learning_rate": 6.70518938533383e-06, | |
| "loss": 0.8149, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.3599999999999999, | |
| "grad_norm": 0.6610273122787476, | |
| "learning_rate": 6.61647214379826e-06, | |
| "loss": 0.933, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.376, | |
| "grad_norm": 0.6104078888893127, | |
| "learning_rate": 6.527182867765333e-06, | |
| "loss": 0.825, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.392, | |
| "grad_norm": 0.5505048632621765, | |
| "learning_rate": 6.437353154776848e-06, | |
| "loss": 0.8886, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.6055363416671753, | |
| "learning_rate": 6.3470147936235485e-06, | |
| "loss": 0.7722, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.424, | |
| "grad_norm": 0.5329720973968506, | |
| "learning_rate": 6.256199753095745e-06, | |
| "loss": 0.7361, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 0.6371309161186218, | |
| "learning_rate": 6.164940170670266e-06, | |
| "loss": 0.8923, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.456, | |
| "grad_norm": 0.611217737197876, | |
| "learning_rate": 6.073268341137694e-06, | |
| "loss": 0.8711, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.5608077645301819, | |
| "learning_rate": 5.98121670517393e-06, | |
| "loss": 0.884, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.488, | |
| "grad_norm": 0.5341782569885254, | |
| "learning_rate": 5.8888178378601565e-06, | |
| "loss": 0.8121, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.504, | |
| "grad_norm": 0.4952053129673004, | |
| "learning_rate": 5.796104437155213e-06, | |
| "loss": 0.7203, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 0.5193338394165039, | |
| "learning_rate": 5.703109312324493e-06, | |
| "loss": 0.9225, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.5817763209342957, | |
| "learning_rate": 5.609865372329461e-06, | |
| "loss": 0.8223, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.552, | |
| "grad_norm": 0.5604681968688965, | |
| "learning_rate": 5.516405614181883e-06, | |
| "loss": 0.8242, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.568, | |
| "grad_norm": 0.5329235196113586, | |
| "learning_rate": 5.4227631112668955e-06, | |
| "loss": 0.8194, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.584, | |
| "grad_norm": 0.64899742603302, | |
| "learning_rate": 5.328971001639054e-06, | |
| "loss": 0.8811, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5119101405143738, | |
| "learning_rate": 5.235062476295488e-06, | |
| "loss": 0.813, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.616, | |
| "grad_norm": 0.5360473990440369, | |
| "learning_rate": 5.141070767430331e-06, | |
| "loss": 0.8624, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.6320000000000001, | |
| "grad_norm": 0.5789046883583069, | |
| "learning_rate": 5.047029136674563e-06, | |
| "loss": 0.8966, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.6480000000000001, | |
| "grad_norm": 0.5769147276878357, | |
| "learning_rate": 4.95297086332544e-06, | |
| "loss": 0.7797, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.6945145726203918, | |
| "learning_rate": 4.858929232569671e-06, | |
| "loss": 0.8263, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.6800000000000002, | |
| "grad_norm": 0.6338273286819458, | |
| "learning_rate": 4.7649375237045135e-06, | |
| "loss": 0.8809, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.696, | |
| "grad_norm": 0.5993066430091858, | |
| "learning_rate": 4.671028998360947e-06, | |
| "loss": 0.8502, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.712, | |
| "grad_norm": 0.6613332629203796, | |
| "learning_rate": 4.5772368887331044e-06, | |
| "loss": 0.827, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.6187359690666199, | |
| "learning_rate": 4.483594385818119e-06, | |
| "loss": 0.8734, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.744, | |
| "grad_norm": 0.505432665348053, | |
| "learning_rate": 4.39013462767054e-06, | |
| "loss": 0.8396, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.762208104133606, | |
| "learning_rate": 4.29689068767551e-06, | |
| "loss": 0.8963, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.776, | |
| "grad_norm": 0.5531466603279114, | |
| "learning_rate": 4.203895562844789e-06, | |
| "loss": 0.8408, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.49315544962882996, | |
| "learning_rate": 4.111182162139844e-06, | |
| "loss": 0.7991, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.808, | |
| "grad_norm": 0.596223771572113, | |
| "learning_rate": 4.018783294826071e-06, | |
| "loss": 0.8497, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.8239999999999998, | |
| "grad_norm": 0.5061976313591003, | |
| "learning_rate": 3.926731658862307e-06, | |
| "loss": 0.807, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.8399999999999999, | |
| "grad_norm": 0.4961169362068176, | |
| "learning_rate": 3.8350598293297345e-06, | |
| "loss": 0.7939, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.5992616415023804, | |
| "learning_rate": 3.7438002469042567e-06, | |
| "loss": 0.8904, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.8719999999999999, | |
| "grad_norm": 0.45855075120925903, | |
| "learning_rate": 3.652985206376455e-06, | |
| "loss": 0.6929, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.888, | |
| "grad_norm": 0.5494182705879211, | |
| "learning_rate": 3.5626468452231534e-06, | |
| "loss": 0.8963, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.904, | |
| "grad_norm": 0.4761560559272766, | |
| "learning_rate": 3.472817132234669e-06, | |
| "loss": 0.8742, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.47549012303352356, | |
| "learning_rate": 3.3835278562017405e-06, | |
| "loss": 0.7919, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.936, | |
| "grad_norm": 0.6070746183395386, | |
| "learning_rate": 3.29481061466617e-06, | |
| "loss": 0.8288, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.952, | |
| "grad_norm": 0.5322284698486328, | |
| "learning_rate": 3.2066968027391377e-06, | |
| "loss": 0.857, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.968, | |
| "grad_norm": 0.49000993371009827, | |
| "learning_rate": 3.119217601991139e-06, | |
| "loss": 0.8831, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.466543048620224, | |
| "learning_rate": 3.032403969417523e-06, | |
| "loss": 0.8452, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.912174642086029, | |
| "learning_rate": 2.946286626483463e-06, | |
| "loss": 1.2448, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.016, | |
| "grad_norm": 0.5235552191734314, | |
| "learning_rate": 2.8608960482523058e-06, | |
| "loss": 0.7986, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 2.032, | |
| "grad_norm": 0.46861377358436584, | |
| "learning_rate": 2.776262452601104e-06, | |
| "loss": 0.7906, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.4755309522151947, | |
| "learning_rate": 2.6924157895271563e-06, | |
| "loss": 0.8128, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 2.064, | |
| "grad_norm": 0.5102059841156006, | |
| "learning_rate": 2.6093857305493666e-06, | |
| "loss": 0.8217, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 0.5348610877990723, | |
| "learning_rate": 2.5272016582081236e-06, | |
| "loss": 0.7677, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.096, | |
| "grad_norm": 0.5097714066505432, | |
| "learning_rate": 2.445892655667462e-06, | |
| "loss": 0.8047, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.49139249324798584, | |
| "learning_rate": 2.365487496423152e-06, | |
| "loss": 0.8069, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 2.128, | |
| "grad_norm": 0.4548731744289398, | |
| "learning_rate": 2.2860146341203936e-06, | |
| "loss": 0.7955, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 2.144, | |
| "grad_norm": 0.481595516204834, | |
| "learning_rate": 2.207502192484685e-06, | |
| "loss": 0.8239, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 0.49964967370033264, | |
| "learning_rate": 2.1299779553694323e-06, | |
| "loss": 0.8155, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.4396260976791382, | |
| "learning_rate": 2.053469356923865e-06, | |
| "loss": 0.8109, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 2.192, | |
| "grad_norm": 0.46519672870635986, | |
| "learning_rate": 1.9780034718846653e-06, | |
| "loss": 0.8485, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 2.208, | |
| "grad_norm": 0.4533243179321289, | |
| "learning_rate": 1.9036070059948253e-06, | |
| "loss": 0.7487, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 2.224, | |
| "grad_norm": 0.47146841883659363, | |
| "learning_rate": 1.8303062865530407e-06, | |
| "loss": 0.8066, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.43573132157325745, | |
| "learning_rate": 1.7581272530970666e-06, | |
| "loss": 0.779, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.2560000000000002, | |
| "grad_norm": 0.44934266805648804, | |
| "learning_rate": 1.6870954482242707e-06, | |
| "loss": 0.7981, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 2.2720000000000002, | |
| "grad_norm": 0.44404736161231995, | |
| "learning_rate": 1.6172360085526567e-06, | |
| "loss": 0.7759, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 2.288, | |
| "grad_norm": 0.43881285190582275, | |
| "learning_rate": 1.54857365582557e-06, | |
| "loss": 0.7576, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.5115951895713806, | |
| "learning_rate": 1.4811326881631937e-06, | |
| "loss": 0.805, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 2.32, | |
| "grad_norm": 0.4446362257003784, | |
| "learning_rate": 1.4149369714639856e-06, | |
| "loss": 0.7901, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 2.336, | |
| "grad_norm": 0.4514508843421936, | |
| "learning_rate": 1.3500099309590397e-06, | |
| "loss": 0.7918, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 2.352, | |
| "grad_norm": 0.42255961894989014, | |
| "learning_rate": 1.2863745429224145e-06, | |
| "loss": 0.8227, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.47350555658340454, | |
| "learning_rate": 1.22405332654032e-06, | |
| "loss": 0.7966, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 2.384, | |
| "grad_norm": 0.45856159925460815, | |
| "learning_rate": 1.1630683359420653e-06, | |
| "loss": 0.7853, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 0.42117229104042053, | |
| "learning_rate": 1.103441152395588e-06, | |
| "loss": 0.7966, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.416, | |
| "grad_norm": 0.4378995895385742, | |
| "learning_rate": 1.045192876670298e-06, | |
| "loss": 0.7746, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.42116639018058777, | |
| "learning_rate": 9.883441215699824e-07, | |
| "loss": 0.8016, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 2.448, | |
| "grad_norm": 0.44868284463882446, | |
| "learning_rate": 9.329150046383773e-07, | |
| "loss": 0.8247, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 2.464, | |
| "grad_norm": 0.4263235330581665, | |
| "learning_rate": 8.789251410400024e-07, | |
| "loss": 0.8123, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 2.48, | |
| "grad_norm": 0.43173035979270935, | |
| "learning_rate": 8.263936366187825e-07, | |
| "loss": 0.7802, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4168744683265686, | |
| "learning_rate": 7.753390811368972e-07, | |
| "loss": 0.8144, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.512, | |
| "grad_norm": 0.4433804750442505, | |
| "learning_rate": 7.257795416962754e-07, | |
| "loss": 0.8041, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.528, | |
| "grad_norm": 0.42650994658470154, | |
| "learning_rate": 6.777325563450282e-07, | |
| "loss": 0.8293, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.544, | |
| "grad_norm": 0.3940533697605133, | |
| "learning_rate": 6.312151278711237e-07, | |
| "loss": 0.7941, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.42621371150016785, | |
| "learning_rate": 5.862437177854629e-07, | |
| "loss": 0.7833, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.576, | |
| "grad_norm": 0.37541621923446655, | |
| "learning_rate": 5.428342404965076e-07, | |
| "loss": 0.8054, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.592, | |
| "grad_norm": 0.37826448678970337, | |
| "learning_rate": 5.010020576785174e-07, | |
| "loss": 0.7977, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.608, | |
| "grad_norm": 0.3738263249397278, | |
| "learning_rate": 4.607619728353818e-07, | |
| "loss": 0.7742, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.4060584306716919, | |
| "learning_rate": 4.221282260619891e-07, | |
| "loss": 0.7976, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.64, | |
| "grad_norm": 0.38774457573890686, | |
| "learning_rate": 3.851144890049535e-07, | |
| "loss": 0.869, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.656, | |
| "grad_norm": 0.4467254877090454, | |
| "learning_rate": 3.497338600245254e-07, | |
| "loss": 0.838, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.672, | |
| "grad_norm": 0.38972336053848267, | |
| "learning_rate": 3.159988595593616e-07, | |
| "loss": 0.8194, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.40794649720191956, | |
| "learning_rate": 2.839214256958106e-07, | |
| "loss": 0.7788, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.7039999999999997, | |
| "grad_norm": 0.4033019542694092, | |
| "learning_rate": 2.5351290994328703e-07, | |
| "loss": 0.785, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.7199999999999998, | |
| "grad_norm": 0.4318539500236511, | |
| "learning_rate": 2.2478407321721295e-07, | |
| "loss": 0.8027, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.7359999999999998, | |
| "grad_norm": 0.4013233482837677, | |
| "learning_rate": 1.9774508203096843e-07, | |
| "loss": 0.7901, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.42489951848983765, | |
| "learning_rate": 1.7240550489817652e-07, | |
| "loss": 0.8184, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.768, | |
| "grad_norm": 0.3855217397212982, | |
| "learning_rate": 1.4877430894662037e-07, | |
| "loss": 0.8117, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.784, | |
| "grad_norm": 0.3926891088485718, | |
| "learning_rate": 1.268598567449647e-07, | |
| "loss": 0.7767, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.8, | |
| "grad_norm": 0.385903924703598, | |
| "learning_rate": 1.0666990334342708e-07, | |
| "loss": 0.7666, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.38875097036361694, | |
| "learning_rate": 8.821159352943142e-08, | |
| "loss": 0.816, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.832, | |
| "grad_norm": 0.424413800239563, | |
| "learning_rate": 7.149145929922607e-08, | |
| "loss": 0.7514, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.848, | |
| "grad_norm": 0.4490892291069031, | |
| "learning_rate": 5.651541754634726e-08, | |
| "loss": 0.7848, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.864, | |
| "grad_norm": 0.44026678800582886, | |
| "learning_rate": 4.328876796776071e-08, | |
| "loss": 0.8048, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.4055703282356262, | |
| "learning_rate": 3.181619118841517e-08, | |
| "loss": 0.8017, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.896, | |
| "grad_norm": 0.40867382287979126, | |
| "learning_rate": 2.210174710486679e-08, | |
| "loss": 0.8113, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.912, | |
| "grad_norm": 0.4212925136089325, | |
| "learning_rate": 1.4148873448573408e-08, | |
| "loss": 0.7914, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.928, | |
| "grad_norm": 0.4007076919078827, | |
| "learning_rate": 7.96038456935322e-09, | |
| "loss": 0.7951, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3707263171672821, | |
| "learning_rate": 3.538470439448105e-09, | |
| "loss": 0.7816, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 0.3968121409416199, | |
| "learning_rate": 8.846958785418969e-10, | |
| "loss": 0.812, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "grad_norm": 0.418857604265213, | |
| "learning_rate": 0.0, | |
| "loss": 0.773, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.976, | |
| "step": 186, | |
| "total_flos": 185463521411072.0, | |
| "train_loss": 0.8840919864434068, | |
| "train_runtime": 10559.3674, | |
| "train_samples_per_second": 1.705, | |
| "train_steps_per_second": 0.018 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "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": 185463521411072.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |