Open Aristomd opened 2 years ago
wujcan
commented
2 years ago 我通过readme给的amazon_book数据集给出的设置aug_type=1,reg=10-4 ,embed_size=64,n_layers=3,ssl_reg=0.5,ssl_ratio=0.1,ssl_temp=0.2,运行了代码，其他地方没有变，早停在110代，60代得到最好的结果recall@20结果0.0142 ndcg@20结果0.0119是我哪里没有设置正确嘛，希望您有时间给回复一下，谢谢
请问reg=10-4还是1e-4?
Aristomd
commented
2 years ago 感谢回复，我设置的是1e-4，和readme的一致，上面打错了
Aristomd
commented
2 years ago test_batch_size=128 num_thread=8 start_testing_epoch=0 proj_path=/home/aristold/babyorange/SGL-main/
SGL's hyperparameters: seed=2021 aug_type=1 reg=1e-4 embed_size=64 n_layers=3 ssl_reg=0.5 ssl_ratio=0.1 ssl_temp=0.2 ssl_mode=both_side lr=0.0001 learner=adam adj_type=pre epochs=1000 batch_size=2048 num_negatives=1 init_method=xavier_uniform stddev=0.01 verbose=1 stop_cnt=50 pretrain=0 save_flag=0
2021-12-05 12:56:26.889: metrics: Precision@20 Recall@20 NDCG@20 MAP@20 MRR@20
2021-12-05 12:57:04.866: 0.00013678 0.00026791 0.00022005 0.00050896 0.00050821
2021-12-05 13:04:51.081: [iter 1 : loss : 7.0245 = 0.6931 + 6.3314 + 0.0000, time: 459.652283]
2021-12-05 13:05:16.374: epoch 1: 0.00132791 0.00282034 0.00238781 0.00469158 0.00500780
2021-12-05 13:05:16.374: Find a better model.
2021-12-05 13:13:01.724: [iter 2 : loss : 7.0210 = 0.6931 + 6.3279 + 0.0000, time: 458.736247]
2021-12-05 13:13:16.557: epoch 2: 0.00187315 0.00385211 0.00329129 0.00667352 0.00715973
2021-12-05 13:13:16.557: Find a better model.
2021-12-05 13:20:58.046: [iter 3 : loss : 7.0205 = 0.6931 + 6.3274 + 0.0000, time: 455.126777]
2021-12-05 13:21:34.795: epoch 3: 0.00222081 0.00438590 0.00382693 0.00792056 0.00847678
2021-12-05 13:21:34.795: Find a better model.
2021-12-05 13:29:12.667: [iter 4 : loss : 7.0201 = 0.6931 + 6.3270 + 0.0000, time: 451.619277]
2021-12-05 13:29:27.637: epoch 4: 0.00229775 0.00467891 0.00403542 0.00838881 0.00895285
2021-12-05 13:29:27.638: Find a better model.
2021-12-05 13:37:11.258: [iter 5 : loss : 7.0199 = 0.6931 + 6.3268 + 0.0000, time: 457.360852]
2021-12-05 13:37:47.902: epoch 5: 0.00253238 0.00520699 0.00446910 0.00922883 0.00982355
2021-12-05 13:37:47.902: Find a better model.
2021-12-05 13:45:23.563: [iter 6 : loss : 7.0197 = 0.6931 + 6.3266 + 0.0000, time: 449.423811]
2021-12-05 13:45:38.397: epoch 6: 0.00257038 0.00527902 0.00452059 0.00951717 0.00997269
2021-12-05 13:45:38.397: Find a better model.
2021-12-05 13:53:19.403: [iter 7 : loss : 7.0196 = 0.6931 + 6.3265 + 0.0000, time: 454.740660]
2021-12-05 13:53:44.434: epoch 7: 0.00269007 0.00543981 0.00474599 0.00999233 0.01064107
2021-12-05 13:53:44.434: Find a better model.
2021-12-05 14:01:23.863: [iter 8 : loss : 7.0195 = 0.6931 + 6.3264 + 0.0000, time: 453.224072]
2021-12-05 14:01:38.872: epoch 8: 0.00277556 0.00568055 0.00490730 0.01042775 0.01099857
2021-12-05 14:01:38.872: Find a better model.
2021-12-05 14:09:21.684: [iter 9 : loss : 7.0196 = 0.6931 + 6.3265 + 0.0000, time: 456.622367]
2021-12-05 14:09:46.395: epoch 9: 0.00296934 0.00613387 0.00525277 0.01105893 0.01171634
2021-12-05 14:09:46.395: Find a better model.
2021-12-05 14:17:28.172: [iter 10 : loss : 7.0196 = 0.6931 + 6.3265 + 0.0000, time: 455.539485]
2021-12-05 14:17:43.180: epoch 10: 0.00306148 0.00639025 0.00557024 0.01173296 0.01250693
2021-12-05 14:17:43.180: Find a better model.
2021-12-05 14:25:25.439: [iter 11 : loss : 7.0196 = 0.6931 + 6.3265 + 0.0000, time: 456.066938]
2021-12-05 14:25:40.434: epoch 11: 0.00305293 0.00635612 0.00548394 0.01150883 0.01232856
2021-12-05 14:33:23.921: [iter 12 : loss : 7.0195 = 0.6931 + 6.3265 + 0.0000, time: 457.302565]
2021-12-05 14:33:38.757: epoch 12: 0.00312322 0.00657155 0.00559645 0.01163304 0.01237344
2021-12-05 14:33:38.757: Find a better model.
2021-12-05 14:41:17.438: [iter 13 : loss : 7.0194 = 0.6930 + 6.3264 + 0.0000, time: 452.503283]
2021-12-05 14:41:32.544: epoch 13: 0.00319922 0.00675679 0.00584879 0.01259855 0.01332147
2021-12-05 14:41:32.544: Find a better model.
2021-12-05 14:49:11.965: [iter 14 : loss : 7.0195 = 0.6930 + 6.3265 + 0.0000, time: 453.141603]
2021-12-05 14:49:27.231: epoch 14: 0.00326952 0.00685065 0.00581834 0.01220620 0.01290802
2021-12-05 14:49:27.231: Find a better model.
2021-12-05 14:57:02.176: [iter 15 : loss : 7.0194 = 0.6930 + 6.3264 + 0.0000, time: 448.785232]
2021-12-05 14:57:27.019: epoch 15: 0.00335215 0.00703984 0.00592311 0.01225964 0.01300933
2021-12-05 14:57:27.019: Find a better model.
2021-12-05 15:05:01.884: [iter 16 : loss : 7.0194 = 0.6930 + 6.3264 + 0.0000, time: 448.622375]
2021-12-05 15:05:16.885: epoch 16: 0.00342340 0.00734873 0.00616704 0.01288376 0.01369006
2021-12-05 15:05:16.885: Find a better model.
2021-12-05 15:12:57.675: [iter 17 : loss : 7.0195 = 0.6930 + 6.3265 + 0.0000, time: 454.619975]
2021-12-05 15:13:34.364: epoch 17: 0.00354689 0.00756746 0.00643243 0.01377840 0.01449539
2021-12-05 15:13:34.364: Find a better model.
2021-12-05 15:21:14.132: [iter 18 : loss : 7.0195 = 0.6930 + 6.3265 + 0.0000, time: 453.513441]
2021-12-05 15:21:29.185: epoch 18: 0.00359628 0.00770576 0.00652727 0.01354942 0.01437996
2021-12-05 15:21:29.185: Find a better model.
2021-12-05 15:29:09.401: [iter 19 : loss : 7.0194 = 0.6930 + 6.3265 + 0.0000, time: 453.997295]
2021-12-05 15:29:24.430: epoch 19: 0.00375682 0.00804430 0.00691976 0.01461282 0.01553059
2021-12-05 15:29:24.431: Find a better model.
2021-12-05 15:37:03.208: [iter 20 : loss : 7.0195 = 0.6930 + 6.3265 + 0.0000, time: 452.558098]
2021-12-05 15:37:18.240: epoch 20: 0.00378626 0.00818399 0.00706324 0.01473585 0.01580802
2021-12-05 15:37:18.240: Find a better model.
2021-12-05 15:44:58.621: [iter 21 : loss : 7.0195 = 0.6930 + 6.3265 + 0.0000, time: 454.119236]
2021-12-05 15:45:23.750: epoch 21: 0.00382521 0.00824080 0.00706217 0.01497195 0.01580744
2021-12-05 15:45:23.750: Find a better model.
2021-12-05 15:53:01.210: [iter 22 : loss : 7.0195 = 0.6930 + 6.3265 + 0.0000, time: 451.139155]
2021-12-05 15:53:16.148: epoch 22: 0.00383471 0.00833742 0.00716185 0.01526928 0.01616655
2021-12-05 15:53:16.148: Find a better model.
2021-12-05 16:00:55.259: [iter 23 : loss : 7.0196 = 0.6930 + 6.3266 + 0.0000, time: 452.750479]
2021-12-05 16:01:10.433: epoch 23: 0.00387175 0.00839106 0.00715334 0.01490544 0.01578999
2021-12-05 16:01:10.433: Find a better model.
2021-12-05 16:08:54.032: [iter 24 : loss : 7.0195 = 0.6929 + 6.3266 + 0.0000, time: 457.055172]
2021-12-05 16:09:19.121: epoch 24: 0.00407598 0.00893420 0.00758035 0.01575238 0.01672077
2021-12-05 16:09:19.121: Find a better model.
2021-12-05 16:17:01.348: [iter 25 : loss : 7.0195 = 0.6929 + 6.3266 + 0.0000, time: 455.974876]
2021-12-05 16:17:16.424: epoch 25: 0.00411398 0.00893893 0.00763608 0.01591949 0.01692522
2021-12-05 16:17:16.424: Find a better model.
2021-12-05 16:24:58.167: [iter 26 : loss : 7.0195 = 0.6929 + 6.3266 + 0.0000, time: 455.492454]
2021-12-05 16:25:13.235: epoch 26: 0.00413678 0.00892622 0.00776005 0.01651694 0.01748559
2021-12-05 16:32:54.083: [iter 27 : loss : 7.0196 = 0.6929 + 6.3266 + 0.0000, time: 454.570099]
2021-12-05 16:33:09.025: epoch 27: 0.00424222 0.00918396 0.00798166 0.01695715 0.01801449
2021-12-05 16:33:09.025: Find a better model.
2021-12-05 16:40:49.110: [iter 28 : loss : 7.0196 = 0.6929 + 6.3267 + 0.0000, time: 453.716040]
2021-12-05 16:41:14.395: epoch 28: 0.00425267 0.00932975 0.00796766 0.01674358 0.01773234
2021-12-05 16:41:14.395: Find a better model.
2021-12-05 16:48:54.686: [iter 29 : loss : 7.0195 = 0.6929 + 6.3266 + 0.0000, time: 453.964447]
2021-12-05 16:49:09.699: epoch 29: 0.00434861 0.00948785 0.00815338 0.01704596 0.01811074
2021-12-05 16:49:09.700: Find a better model.
2021-12-05 16:56:46.985: [iter 30 : loss : 7.0194 = 0.6929 + 6.3265 + 0.0000, time: 450.972741]
2021-12-05 16:57:01.847: epoch 30: 0.00431537 0.00946431 0.00816453 0.01714017 0.01814647
2021-12-05 17:04:35.941: [iter 31 : loss : 7.0195 = 0.6929 + 6.3267 + 0.0000, time: 447.754553]
2021-12-05 17:04:50.177: epoch 31: 0.00439611 0.00957129 0.00821710 0.01716623 0.01835509
2021-12-05 17:04:50.177: Find a better model.
2021-12-05 17:12:26.226: [iter 32 : loss : 7.0195 = 0.6929 + 6.3266 + 0.0000, time: 449.752640]
2021-12-05 17:13:02.829: epoch 32: 0.00451770 0.00986130 0.00844209 0.01772446 0.01880301
2021-12-05 17:13:02.830: Find a better model.
2021-12-05 17:20:39.239: [iter 33 : loss : 7.0196 = 0.6928 + 6.3268 + 0.0000, time: 450.046607]
2021-12-05 17:20:54.388: epoch 33: 0.00450440 0.00990695 0.00839239 0.01751892 0.01852660
2021-12-05 17:20:54.388: Find a better model.
2021-12-05 17:28:33.896: [iter 34 : loss : 7.0197 = 0.6928 + 6.3268 + 0.0000, time: 453.101452]
2021-12-05 17:28:48.953: epoch 34: 0.00459179 0.01013120 0.00862572 0.01787323 0.01912097
2021-12-05 17:28:48.954: Find a better model.
2021-12-05 17:36:31.329: [iter 35 : loss : 7.0195 = 0.6928 + 6.3267 + 0.0000, time: 456.046334]
2021-12-05 17:36:46.446: epoch 35: 0.00471528 0.01036876 0.00886577 0.01836253 0.01963428
2021-12-05 17:36:46.446: Find a better model.
2021-12-05 17:44:28.458: [iter 36 : loss : 7.0195 = 0.6928 + 6.3267 + 0.0000, time: 455.688246]
2021-12-05 17:44:43.783: epoch 36: 0.00480267 0.01068195 0.00900871 0.01836460 0.01953793
2021-12-05 17:44:43.784: Find a better model.
2021-12-05 17:52:26.747: [iter 37 : loss : 7.0197 = 0.6928 + 6.3269 + 0.0000, time: 456.654295]
2021-12-05 17:52:41.807: epoch 37: 0.00491950 0.01092858 0.00929541 0.01919170 0.02049464
2021-12-05 17:52:41.807: Find a better model.
2021-12-05 18:00:24.097: [iter 38 : loss : 7.0196 = 0.6928 + 6.3268 + 0.0000, time: 455.973511]
2021-12-05 18:00:39.055: epoch 38: 0.00487486 0.01075151 0.00909962 0.01863023 0.01994301
2021-12-05 18:08:24.371: [iter 39 : loss : 7.0196 = 0.6928 + 6.3268 + 0.0000, time: 458.954588]
2021-12-05 18:08:39.426: epoch 39: 0.00501163 0.01123516 0.00939996 0.01923214 0.02044516
2021-12-05 18:08:39.426: Find a better model.
2021-12-05 18:16:24.307: [iter 40 : loss : 7.0194 = 0.6928 + 6.3267 + 0.0000, time: 458.563171]
2021-12-05 18:16:39.232: epoch 40: 0.00505246 0.01121261 0.00943427 0.01916034 0.02048936
2021-12-05 18:24:24.226: [iter 41 : loss : 7.0196 = 0.6927 + 6.3268 + 0.0000, time: 458.713924]
2021-12-05 18:24:39.565: epoch 41: 0.00508856 0.01137946 0.00959330 0.01958583 0.02102214
2021-12-05 18:24:39.565: Find a better model.
2021-12-05 18:32:24.592: [iter 42 : loss : 7.0196 = 0.6927 + 6.3269 + 0.0000, time: 458.704556]
2021-12-05 18:32:39.510: epoch 42: 0.00514459 0.01145933 0.00964090 0.01964627 0.02096677
2021-12-05 18:32:39.510: Find a better model.
2021-12-05 18:40:21.757: [iter 43 : loss : 7.0195 = 0.6927 + 6.3268 + 0.0000, time: 455.934913]
2021-12-05 18:40:36.959: epoch 43: 0.00513034 0.01152364 0.00967478 0.01972248 0.02099582
2021-12-05 18:40:36.959: Find a better model.
2021-12-05 18:48:19.893: [iter 44 : loss : 7.0196 = 0.6927 + 6.3269 + 0.0000, time: 456.655233]
2021-12-05 18:48:34.958: epoch 44: 0.00528609 0.01198155 0.00998714 0.02031140 0.02172996
2021-12-05 18:48:34.959: Find a better model.
2021-12-05 18:56:19.939: [iter 45 : loss : 7.0195 = 0.6927 + 6.3269 + 0.0000, time: 458.758062]
2021-12-05 18:56:34.752: epoch 45: 0.00537157 0.01207599 0.01022530 0.02087627 0.02232079
2021-12-05 18:56:34.752: Find a better model.
2021-12-05 19:04:16.833: [iter 46 : loss : 7.0196 = 0.6927 + 6.3269 + 0.0000, time: 455.773155]
2021-12-05 19:04:31.774: epoch 46: 0.00546939 0.01225831 0.01035970 0.02101806 0.02238646
2021-12-05 19:04:31.774: Find a better model.
2021-12-05 19:12:15.660: [iter 47 : loss : 7.0194 = 0.6926 + 6.3268 + 0.0000, time: 457.651506]
2021-12-05 19:12:40.167: epoch 47: 0.00543900 0.01224850 0.01034556 0.02105800 0.02258649
2021-12-05 19:20:25.024: [iter 48 : loss : 7.0194 = 0.6926 + 6.3268 + 0.0000, time: 458.570422]
2021-12-05 19:20:40.175: epoch 48: 0.00552543 0.01236637 0.01044683 0.02104991 0.02270224
2021-12-05 19:20:40.175: Find a better model.
2021-12-05 19:28:25.734: [iter 49 : loss : 7.0196 = 0.6926 + 6.3270 + 0.0000, time: 459.329278]
2021-12-05 19:28:40.640: epoch 49: 0.00572772 0.01302486 0.01089921 0.02180821 0.02349286
2021-12-05 19:28:40.641: Find a better model.
2021-12-05 19:36:23.583: [iter 50 : loss : 7.0195 = 0.6926 + 6.3269 + 0.0000, time: 456.725160]
2021-12-05 19:36:38.569: epoch 50: 0.00568024 0.01297300 0.01087460 0.02201787 0.02356261
2021-12-05 19:44:30.146: [iter 51 : loss : 7.0195 = 0.6926 + 6.3270 + 0.0000, time: 465.336065]
2021-12-05 19:44:58.155: epoch 51: 0.00568594 0.01305893 0.01086565 0.02155350 0.02312946
2021-12-05 19:44:58.155: Find a better model.
2021-12-05 19:52:33.744: [iter 52 : loss : 7.0194 = 0.6925 + 6.3269 + 0.0000, time: 448.905582]
2021-12-05 19:52:45.650: epoch 52: 0.00568023 0.01292337 0.01095453 0.02213627 0.02382617
2021-12-05 20:00:13.222: [iter 53 : loss : 7.0195 = 0.6925 + 6.3270 + 0.0000, time: 441.661253]
2021-12-05 20:00:25.115: epoch 53: 0.00579326 0.01316035 0.01112485 0.02245538 0.02427685
2021-12-05 20:00:25.115: Find a better model.
2021-12-05 20:07:54.590: [iter 54 : loss : 7.0196 = 0.6925 + 6.3271 + 0.0000, time: 443.529709]
2021-12-05 20:08:06.533: epoch 54: 0.00584548 0.01336121 0.01133410 0.02296268 0.02460670
2021-12-05 20:08:06.533: Find a better model.
2021-12-05 20:15:33.593: [iter 55 : loss : 7.0195 = 0.6925 + 6.3270 + 0.0000, time: 441.160195]
2021-12-05 20:15:45.520: epoch 55: 0.00593097 0.01356471 0.01146245 0.02286122 0.02457688
2021-12-05 20:15:45.520: Find a better model.
2021-12-05 20:23:12.847: [iter 56 : loss : 7.0193 = 0.6924 + 6.3269 + 0.0000, time: 441.422473]
2021-12-05 20:23:34.712: epoch 56: 0.00586070 0.01351548 0.01143924 0.02328570 0.02491733
2021-12-05 20:31:01.886: [iter 57 : loss : 7.0193 = 0.6924 + 6.3269 + 0.0000, time: 441.232511]
2021-12-05 20:31:13.817: epoch 57: 0.00599461 0.01372719 0.01162932 0.02334848 0.02511566
2021-12-05 20:31:13.817: Find a better model.
2021-12-05 20:38:44.462: [iter 58 : loss : 7.0195 = 0.6924 + 6.3271 + 0.0000, time: 444.701456]
2021-12-05 20:38:56.392: epoch 58: 0.00615606 0.01417093 0.01183738 0.02337406 0.02531281
2021-12-05 20:38:56.392: Find a better model.
2021-12-05 20:46:23.693: [iter 59 : loss : 7.0194 = 0.6924 + 6.3271 + 0.0000, time: 441.340626]
2021-12-05 20:46:35.562: epoch 59: 0.00612568 0.01416575 0.01182421 0.02330048 0.02510505
2021-12-05 20:54:12.593: [iter 60 : loss : 7.0194 = 0.6923 + 6.3271 + 0.0000, time: 451.073853]
2021-12-05 20:54:24.560: epoch 60: 0.00622825 0.01424114 0.01186671 0.02320280 0.02496259
2021-12-05 20:54:24.560: Find a better model.
2021-12-05 21:01:51.734: [iter 61 : loss : 7.0195 = 0.6923 + 6.3272 + 0.0000, time: 441.191380]
2021-12-05 21:02:03.683: epoch 61: 0.00617980 0.01409679 0.01173235 0.02282025 0.02471266
2021-12-05 21:09:31.463: [iter 62 : loss : 7.0194 = 0.6923 + 6.3271 + 0.0000, time: 441.811908]
2021-12-05 21:09:52.976: epoch 62: 0.00612377 0.01408938 0.01151136 0.02200436 0.02361411
2021-12-05 21:17:22.170: [iter 63 : loss : 7.0195 = 0.6922 + 6.3272 + 0.0000, time: 443.134542]
2021-12-05 21:17:44.306: epoch 63: 0.00619786 0.01419991 0.01173703 0.02261085 0.02430903
2021-12-05 21:25:11.202: [iter 64 : loss : 7.0194 = 0.6922 + 6.3272 + 0.0000, time: 440.842889]
2021-12-05 21:25:32.814: epoch 64: 0.00611712 0.01407418 0.01146277 0.02203372 0.02360403
2021-12-05 21:32:58.960: [iter 65 : loss : 7.0194 = 0.6922 + 6.3272 + 0.0000, time: 440.115290]
2021-12-05 21:33:10.897: epoch 65: 0.00603355 0.01387637 0.01125738 0.02148712 0.02299993
2021-12-05 21:40:40.168: [iter 66 : loss : 7.0193 = 0.6921 + 6.3271 + 0.0000, time: 443.268090]
2021-12-05 21:41:02.573: epoch 66: 0.00586828 0.01368447 0.01087456 0.02029089 0.02171366
2021-12-05 21:48:30.055: [iter 67 : loss : 7.0192 = 0.6921 + 6.3271 + 0.0000, time: 441.445210]
2021-12-05 21:48:41.892: epoch 67: 0.00573437 0.01326444 0.01061559 0.01990361 0.02135481
2021-12-05 21:56:10.222: [iter 68 : loss : 7.0193 = 0.6920 + 6.3273 + 0.0000, time: 442.316750]
2021-12-05 21:56:32.085: epoch 68: 0.00562325 0.01304924 0.01029860 0.01917918 0.02051006
2021-12-05 22:03:57.293: [iter 69 : loss : 7.0192 = 0.6920 + 6.3272 + 0.0000, time: 439.257188]
2021-12-05 22:04:09.212: epoch 69: 0.00531459 0.01235647 0.00967542 0.01764608 0.01886429
2021-12-05 22:11:34.722: [iter 70 : loss : 7.0191 = 0.6919 + 6.3272 + 0.0000, time: 439.536289]
2021-12-05 22:11:56.763: epoch 70: 0.00511419 0.01196421 0.00919309 0.01640973 0.01758342
2021-12-05 22:19:23.020: [iter 71 : loss : 7.0190 = 0.6919 + 6.3271 + 0.0000, time: 440.296077]
2021-12-05 22:19:34.959: epoch 71: 0.00470577 0.01113341 0.00841814 0.01478392 0.01580842
2021-12-05 22:27:04.263: [iter 72 : loss : 7.0191 = 0.6918 + 6.3273 + 0.0000, time: 443.324044]
2021-12-05 22:27:16.191: epoch 72: 0.00421087 0.01001751 0.00754225 0.01311838 0.01401578
2021-12-05 22:34:42.503: [iter 73 : loss : 7.0191 = 0.6918 + 6.3273 + 0.0000, time: 440.306792]
2021-12-05 22:34:54.420: epoch 73: 0.00373497 0.00884819 0.00659261 0.01130044 0.01212232
2021-12-05 22:42:30.959: [iter 74 : loss : 7.0191 = 0.6917 + 6.3274 + 0.0000, time: 450.574821]
2021-12-05 22:42:52.644: epoch 74: 0.00321631 0.00764889 0.00566053 0.00961025 0.01017764
2021-12-05 22:50:21.163: [iter 75 : loss : 7.0189 = 0.6916 + 6.3273 + 0.0000, time: 442.596739]
2021-12-05 22:50:33.141: epoch 75: 0.00279360 0.00671280 0.00500979 0.00842991 0.00899778
2021-12-05 22:58:02.485: [iter 76 : loss : 7.0188 = 0.6915 + 6.3273 + 0.0000, time: 443.461540]
2021-12-05 22:58:14.344: epoch 76: 0.00239274 0.00573203 0.00424882 0.00700419 0.00750803
2021-12-05 23:05:43.297: [iter 77 : loss : 7.0188 = 0.6914 + 6.3274 + 0.0000, time: 443.060531]
2021-12-05 23:06:04.974: epoch 77: 0.00210492 0.00508974 0.00377772 0.00624910 0.00669846
2021-12-05 23:13:32.538: [iter 78 : loss : 7.0187 = 0.6912 + 6.3274 + 0.0000, time: 441.680732]
2021-12-05 23:13:44.488: epoch 78: 0.00190734 0.00460823 0.00339322 0.00548345 0.00587591
2021-12-05 23:21:21.065: [iter 79 : loss : 7.0185 = 0.6911 + 6.3274 + 0.0000, time: 450.707899]
2021-12-05 23:21:53.432: epoch 79: 0.00175726 0.00420767 0.00312522 0.00506147 0.00546334
2021-12-05 23:29:20.681: [iter 80 : loss : 7.0185 = 0.6909 + 6.3275 + 0.0000, time: 441.362761]
2021-12-05 23:29:32.617: epoch 80: 0.00159673 0.00391464 0.00287082 0.00452690 0.00494191
2021-12-05 23:37:05.764: [iter 81 : loss : 7.0183 = 0.6908 + 6.3276 + 0.0000, time: 447.218974]
2021-12-05 23:37:17.738: epoch 81: 0.00153024 0.00371485 0.00272662 0.00428170 0.00464344
2021-12-05 23:44:44.576: [iter 82 : loss : 7.0182 = 0.6906 + 6.3276 + 0.0000, time: 440.936631]
2021-12-05 23:45:06.373: epoch 82: 0.00143145 0.00355802 0.00256393 0.00393142 0.00424311
2021-12-05 23:52:34.688: [iter 83 : loss : 7.0182 = 0.6904 + 6.3277 + 0.0000, time: 442.357977]
2021-12-05 23:52:46.616: epoch 83: 0.00138015 0.00339230 0.00247185 0.00379365 0.00404782
2021-12-06 00:00:14.692: [iter 84 : loss : 7.0179 = 0.6902 + 6.3277 + 0.0000, time: 442.146669]
2021-12-06 00:00:26.668: epoch 84: 0.00132696 0.00329187 0.00237669 0.00344063 0.00369553
2021-12-06 00:07:51.545: [iter 85 : loss : 7.0178 = 0.6901 + 6.3277 + 0.0000, time: 438.951850]
2021-12-06 00:08:03.550: epoch 85: 0.00124147 0.00312916 0.00224692 0.00316420 0.00341764
2021-12-06 00:15:36.862: [iter 86 : loss : 7.0176 = 0.6899 + 6.3277 + 0.0000, time: 447.411109]
2021-12-06 00:15:58.409: epoch 86: 0.00117593 0.00295947 0.00214289 0.00298925 0.00322400
2021-12-06 00:23:24.209: [iter 87 : loss : 7.0175 = 0.6896 + 6.3278 + 0.0000, time: 439.878778]
2021-12-06 00:23:45.738: epoch 87: 0.00115028 0.00289370 0.00213160 0.00292208 0.00315706
2021-12-06 00:31:13.826: [iter 88 : loss : 7.0174 = 0.6894 + 6.3280 + 0.0000, time: 442.146061]
2021-12-06 00:31:36.553: epoch 88: 0.00116073 0.00294334 0.00216012 0.00291061 0.00316347
2021-12-06 00:39:05.904: [iter 89 : loss : 7.0172 = 0.6892 + 6.3281 + 0.0000, time: 443.322508]
2021-12-06 00:39:38.419: epoch 89: 0.00117498 0.00296385 0.00222362 0.00297300 0.00323986
2021-12-06 00:47:05.030: [iter 90 : loss : 7.0169 = 0.6888 + 6.3281 + 0.0000, time: 440.618102]
2021-12-06 00:47:16.956: epoch 90: 0.00118638 0.00302097 0.00229672 0.00307072 0.00330008
2021-12-06 00:54:46.051: [iter 91 : loss : 7.0166 = 0.6884 + 6.3282 + 0.0000, time: 443.098826]
2021-12-06 00:55:18.420: epoch 91: 0.00121772 0.00304963 0.00237123 0.00313851 0.00340587
2021-12-06 01:02:44.423: [iter 92 : loss : 7.0162 = 0.6880 + 6.3282 + 0.0000, time: 440.005979]
2021-12-06 01:02:56.336: epoch 92: 0.00129086 0.00323482 0.00254144 0.00337641 0.00366420
2021-12-06 01:10:25.089: [iter 93 : loss : 7.0159 = 0.6874 + 6.3284 + 0.0000, time: 442.745171]
2021-12-06 01:10:37.032: epoch 93: 0.00136495 0.00343193 0.00268123 0.00352431 0.00380374
2021-12-06 01:18:03.424: [iter 94 : loss : 7.0155 = 0.6868 + 6.3287 + 0.0000, time: 440.367657]
2021-12-06 01:18:15.339: epoch 94: 0.00142860 0.00355960 0.00278953 0.00366216 0.00395296
2021-12-06 01:25:41.883: [iter 95 : loss : 7.0148 = 0.6860 + 6.3288 + 0.0000, time: 440.544437]
2021-12-06 01:25:53.803: epoch 95: 0.00151884 0.00380152 0.00292057 0.00375638 0.00402883
2021-12-06 01:33:19.575: [iter 96 : loss : 7.0140 = 0.6850 + 6.3290 + 0.0000, time: 439.715856]
2021-12-06 01:33:52.141: epoch 96: 0.00160717 0.00400704 0.00307655 0.00398812 0.00428237
2021-12-06 01:41:17.433: [iter 97 : loss : 7.0133 = 0.6840 + 6.3293 + 0.0000, time: 439.209569]
2021-12-06 01:41:29.360: epoch 97: 0.00168696 0.00421445 0.00324417 0.00425457 0.00459564
2021-12-06 01:48:57.331: [iter 98 : loss : 7.0122 = 0.6827 + 6.3295 + 0.0000, time: 441.973793]
2021-12-06 01:49:29.842: epoch 98: 0.00177530 0.00440754 0.00340438 0.00448283 0.00485716
2021-12-06 01:56:55.100: [iter 99 : loss : 7.0111 = 0.6813 + 6.3298 + 0.0000, time: 439.236126]
2021-12-06 01:57:06.979: epoch 99: 0.00186364 0.00462252 0.00355487 0.00468096 0.00505868
2021-12-06 02:04:32.768: [iter 100 : loss : 7.0096 = 0.6795 + 6.3300 + 0.0000, time: 439.741058]
2021-12-06 02:05:05.075: epoch 100: 0.00197858 0.00490771 0.00377314 0.00502548 0.00546285
2021-12-06 02:12:30.123: [iter 101 : loss : 7.0081 = 0.6776 + 6.3303 + 0.0001, time: 439.015037]
2021-12-06 02:12:42.003: epoch 101: 0.00208876 0.00514782 0.00398300 0.00539391 0.00587074
2021-12-06 02:20:07.962: [iter 102 : loss : 7.0063 = 0.6754 + 6.3308 + 0.0001, time: 439.990550]
2021-12-06 02:20:40.339: epoch 102: 0.00222365 0.00546260 0.00422980 0.00580349 0.00629844
2021-12-06 02:28:05.059: [iter 103 : loss : 7.0042 = 0.6731 + 6.3311 + 0.0001, time: 438.730864]
2021-12-06 02:28:17.023: epoch 103: 0.00238703 0.00587428 0.00453482 0.00627779 0.00685429
2021-12-06 02:35:45.254: [iter 104 : loss : 7.0019 = 0.6703 + 6.3315 + 0.0001, time: 442.267055]
2021-12-06 02:36:17.552: epoch 104: 0.00255041 0.00625586 0.00481888 0.00667220 0.00732710
2021-12-06 02:43:39.883: [iter 105 : loss : 6.9993 = 0.6673 + 6.3320 + 0.0001, time: 436.373894]
2021-12-06 02:43:51.777: epoch 105: 0.00275939 0.00675896 0.00516100 0.00711571 0.00785054
2021-12-06 02:51:19.667: [iter 106 : loss : 6.9967 = 0.6641 + 6.3325 + 0.0001, time: 441.943623]
2021-12-06 02:51:51.895: epoch 106: 0.00294082 0.00724663 0.00550646 0.00768337 0.00844135
2021-12-06 02:59:19.544: [iter 107 : loss : 6.9934 = 0.6604 + 6.3329 + 0.0001, time: 441.667479]
2021-12-06 02:59:31.446: epoch 107: 0.00312510 0.00772439 0.00584424 0.00819701 0.00898288
2021-12-06 03:06:58.468: [iter 108 : loss : 6.9898 = 0.6563 + 6.3334 + 0.0001, time: 441.028302]
2021-12-06 03:07:20.352: epoch 108: 0.00333502 0.00825289 0.00628256 0.00897335 0.00987228
2021-12-06 03:14:46.370: [iter 109 : loss : 6.9862 = 0.6521 + 6.3340 + 0.0001, time: 440.083576]
2021-12-06 03:14:58.362: epoch 109: 0.00356395 0.00884222 0.00670888 0.00962812 0.01058583
2021-12-06 03:22:27.668: [iter 110 : loss : 6.9818 = 0.6472 + 6.3345 + 0.0002, time: 443.393281]
2021-12-06 03:22:39.589: epoch 110: 0.00384037 0.00951267 0.00722792 0.01046697 0.01156572
2021-12-06 03:22:39.589: Early stopping is trigger at epoch: 110
2021-12-06 03:22:39.589: best_result@epoch 60:
2021-12-06 03:22:39.589: 0.0062 0.0142 0.0119
这是训练日志，麻烦您看看是哪里出问题了，非常感谢
zwb29
commented
2 years ago lr=0.001试一下
Aristomd
commented
2 years ago 好的感谢，我试一试
我通过readme给的amazon_book数据集给出的设置aug_type=1,reg=10-4 ,embed_size=64,n_layers=3,ssl_reg=0.5,ssl_ratio=0.1,ssl_temp=0.2,运行了代码，其他地方没有变，早停在110代，60代得到最好的结果recall@20结果0.0142 ndcg@20结果0.0119是我哪里没有设置正确嘛，希望您有时间给回复一下，谢谢