-
-
Save stas-sl/8b9a6f046735ae006da6634321fe3832 to your computer and use it in GitHub Desktop.
This file has been truncated, but you can view the full file.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| I0401 12:44:49.957870 6134 solver.cpp:280] Solving mixed_lstm | |
| I0401 12:44:49.957882 6134 solver.cpp:281] Learning Rate Policy: fixed | |
| I0401 12:44:50.306246 6134 solver.cpp:229] Iteration 0, loss = 13.7773 | |
| I0401 12:44:50.306291 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0 | |
| I0401 12:44:50.306309 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.181818 | |
| I0401 12:44:50.306321 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.0222222 | |
| I0401 12:44:50.306339 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 4.39462 (* 0.3 = 1.31839 loss) | |
| I0401 12:44:50.306352 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 4.19871 (* 0.3 = 1.25961 loss) | |
| I0401 12:44:50.306365 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0 | |
| I0401 12:44:50.306397 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0 | |
| I0401 12:44:50.306411 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0 | |
| I0401 12:44:50.306423 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 4.37531 (* 0.3 = 1.31259 loss) | |
| I0401 12:44:50.306437 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 4.36735 (* 0.3 = 1.3102 loss) | |
| I0401 12:44:50.306449 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0222222 | |
| I0401 12:44:50.306462 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.00568182 | |
| I0401 12:44:50.306473 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.0666667 | |
| I0401 12:44:50.306486 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 4.33032 (* 1 = 4.33032 loss) | |
| I0401 12:44:50.306500 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 4.24617 (* 1 = 4.24617 loss) | |
| I0401 12:44:50.306511 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 12:44:50.306522 6134 solver.cpp:245] Train net output #16: total_confidence = 1.19457e-35 | |
| I0401 12:44:50.306540 6134 sgd_solver.cpp:106] Iteration 0, lr = 0.01 | |
| I0401 12:44:50.324127 6134 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 36.1709 > 30) by scale factor 0.829395 | |
| I0401 12:44:50.601414 6134 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 37.4927 > 30) by scale factor 0.800156 | |
| I0401 12:44:50.861389 6134 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 31.9811 > 30) by scale factor 0.938054 | |
| I0401 12:44:51.120638 6134 sgd_solver.cpp:92] Gradient clipping: scaling down gradients (L2 norm 32.4394 > 30) by scale factor 0.924802 | |
| I0401 12:46:58.867997 6134 solver.cpp:229] Iteration 500, loss = 8.63114 | |
| I0401 12:46:58.868300 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0 | |
| I0401 12:46:58.868319 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0401 12:46:58.868332 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.0416667 | |
| I0401 12:46:58.868347 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 4.25996 (* 0.3 = 1.27799 loss) | |
| I0401 12:46:58.868362 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.38435 (* 0.3 = 0.415306 loss) | |
| I0401 12:46:58.868374 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0 | |
| I0401 12:46:58.868386 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.727273 | |
| I0401 12:46:58.868398 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.104167 | |
| I0401 12:46:58.868412 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 4.10623 (* 0.3 = 1.23187 loss) | |
| I0401 12:46:58.868430 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.36563 (* 0.3 = 0.40969 loss) | |
| I0401 12:46:58.868451 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0416667 | |
| I0401 12:46:58.868465 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0401 12:46:58.868477 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.0833333 | |
| I0401 12:46:58.868490 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.71374 (* 1 = 3.71374 loss) | |
| I0401 12:46:58.868505 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.0663 (* 1 = 1.0663 loss) | |
| I0401 12:46:58.868516 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 12:46:58.868527 6134 solver.cpp:245] Train net output #16: total_confidence = 7.97661e-07 | |
| I0401 12:46:58.868540 6134 sgd_solver.cpp:106] Iteration 500, lr = 0.01 | |
| I0401 12:49:07.302662 6134 solver.cpp:229] Iteration 1000, loss = 7.86443 | |
| I0401 12:49:07.302811 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.116279 | |
| I0401 12:49:07.302831 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 12:49:07.302845 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.209302 | |
| I0401 12:49:07.302860 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.45219 (* 0.3 = 1.03566 loss) | |
| I0401 12:49:07.302875 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.16467 (* 0.3 = 0.3494 loss) | |
| I0401 12:49:07.302886 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0465116 | |
| I0401 12:49:07.302899 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 12:49:07.302911 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.186047 | |
| I0401 12:49:07.302924 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.46772 (* 0.3 = 1.04032 loss) | |
| I0401 12:49:07.302937 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.05089 (* 0.3 = 0.315267 loss) | |
| I0401 12:49:07.302949 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0232558 | |
| I0401 12:49:07.302961 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0401 12:49:07.302973 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.255814 | |
| I0401 12:49:07.302986 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.30487 (* 1 = 3.30487 loss) | |
| I0401 12:49:07.302999 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.890591 (* 1 = 0.890591 loss) | |
| I0401 12:49:07.303011 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 12:49:07.303022 6134 solver.cpp:245] Train net output #16: total_confidence = 4.59384e-05 | |
| I0401 12:49:07.303037 6134 sgd_solver.cpp:106] Iteration 1000, lr = 0.01 | |
| I0401 12:51:15.845552 6134 solver.cpp:229] Iteration 1500, loss = 7.64125 | |
| I0401 12:51:15.845660 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0454545 | |
| I0401 12:51:15.845680 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0401 12:51:15.845692 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.136364 | |
| I0401 12:51:15.845707 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.51853 (* 0.3 = 1.05556 loss) | |
| I0401 12:51:15.845722 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.23506 (* 0.3 = 0.370518 loss) | |
| I0401 12:51:15.845734 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0681818 | |
| I0401 12:51:15.845746 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 12:51:15.845758 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.272727 | |
| I0401 12:51:15.845772 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.42376 (* 0.3 = 1.02713 loss) | |
| I0401 12:51:15.845785 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.01984 (* 0.3 = 0.305953 loss) | |
| I0401 12:51:15.845796 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0227273 | |
| I0401 12:51:15.845808 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0401 12:51:15.845819 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.181818 | |
| I0401 12:51:15.845832 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.35834 (* 1 = 3.35834 loss) | |
| I0401 12:51:15.845846 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.97117 (* 1 = 0.97117 loss) | |
| I0401 12:51:15.845857 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 12:51:15.845868 6134 solver.cpp:245] Train net output #16: total_confidence = 3.35111e-05 | |
| I0401 12:51:15.845880 6134 sgd_solver.cpp:106] Iteration 1500, lr = 0.01 | |
| I0401 12:53:24.300691 6134 solver.cpp:229] Iteration 2000, loss = 7.50205 | |
| I0401 12:53:24.300822 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.131579 | |
| I0401 12:53:24.300843 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 12:53:24.300856 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.210526 | |
| I0401 12:53:24.300871 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.87501 (* 0.3 = 1.1625 loss) | |
| I0401 12:53:24.300885 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.1176 (* 0.3 = 0.33528 loss) | |
| I0401 12:53:24.300897 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.105263 | |
| I0401 12:53:24.300910 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 12:53:24.300921 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.157895 | |
| I0401 12:53:24.300935 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.80084 (* 0.3 = 1.14025 loss) | |
| I0401 12:53:24.300957 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.06115 (* 0.3 = 0.318346 loss) | |
| I0401 12:53:24.300973 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0 | |
| I0401 12:53:24.300984 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0401 12:53:24.300997 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.131579 | |
| I0401 12:53:24.301009 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.74166 (* 1 = 3.74166 loss) | |
| I0401 12:53:24.301023 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.04449 (* 1 = 1.04449 loss) | |
| I0401 12:53:24.301033 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 12:53:24.301062 6134 solver.cpp:245] Train net output #16: total_confidence = 4.22573e-06 | |
| I0401 12:53:24.301077 6134 sgd_solver.cpp:106] Iteration 2000, lr = 0.01 | |
| I0401 12:55:32.811437 6134 solver.cpp:229] Iteration 2500, loss = 7.40309 | |
| I0401 12:55:32.811549 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0232558 | |
| I0401 12:55:32.811568 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0401 12:55:32.811581 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.162791 | |
| I0401 12:55:32.811596 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.79671 (* 0.3 = 1.13901 loss) | |
| I0401 12:55:32.811611 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.2318 (* 0.3 = 0.36954 loss) | |
| I0401 12:55:32.811624 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0232558 | |
| I0401 12:55:32.811635 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0401 12:55:32.811647 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.162791 | |
| I0401 12:55:32.811661 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.65853 (* 0.3 = 1.09756 loss) | |
| I0401 12:55:32.811674 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.13254 (* 0.3 = 0.339763 loss) | |
| I0401 12:55:32.811686 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0465116 | |
| I0401 12:55:32.811697 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.732955 | |
| I0401 12:55:32.811709 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.116279 | |
| I0401 12:55:32.811722 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.67876 (* 1 = 3.67876 loss) | |
| I0401 12:55:32.811736 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.14378 (* 1 = 1.14378 loss) | |
| I0401 12:55:32.811748 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 12:55:32.811759 6134 solver.cpp:245] Train net output #16: total_confidence = 1.95148e-06 | |
| I0401 12:55:32.811771 6134 sgd_solver.cpp:106] Iteration 2500, lr = 0.01 | |
| I0401 12:57:41.305498 6134 solver.cpp:229] Iteration 3000, loss = 7.29351 | |
| I0401 12:57:41.305788 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0392157 | |
| I0401 12:57:41.305809 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.721591 | |
| I0401 12:57:41.305822 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.137255 | |
| I0401 12:57:41.305838 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.58241 (* 0.3 = 1.07472 loss) | |
| I0401 12:57:41.305852 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.17041 (* 0.3 = 0.351124 loss) | |
| I0401 12:57:41.305865 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0588235 | |
| I0401 12:57:41.305876 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.727273 | |
| I0401 12:57:41.305888 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.196078 | |
| I0401 12:57:41.305902 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.67634 (* 0.3 = 1.1029 loss) | |
| I0401 12:57:41.305915 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.17258 (* 0.3 = 0.351773 loss) | |
| I0401 12:57:41.305927 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0980392 | |
| I0401 12:57:41.305939 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0401 12:57:41.305950 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.117647 | |
| I0401 12:57:41.305964 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.56638 (* 1 = 3.56638 loss) | |
| I0401 12:57:41.305979 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.09412 (* 1 = 1.09412 loss) | |
| I0401 12:57:41.305996 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 12:57:41.306018 6134 solver.cpp:245] Train net output #16: total_confidence = 3.28912e-06 | |
| I0401 12:57:41.306041 6134 sgd_solver.cpp:106] Iteration 3000, lr = 0.01 | |
| I0401 12:59:49.774981 6134 solver.cpp:229] Iteration 3500, loss = 7.21689 | |
| I0401 12:59:49.775131 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0555556 | |
| I0401 12:59:49.775152 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.710227 | |
| I0401 12:59:49.775166 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.166667 | |
| I0401 12:59:49.775182 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.71542 (* 0.3 = 1.11463 loss) | |
| I0401 12:59:49.775195 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.23676 (* 0.3 = 0.371028 loss) | |
| I0401 12:59:49.775207 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.037037 | |
| I0401 12:59:49.775219 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.704545 | |
| I0401 12:59:49.775231 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.148148 | |
| I0401 12:59:49.775245 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.86867 (* 0.3 = 1.1606 loss) | |
| I0401 12:59:49.775259 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.25572 (* 0.3 = 0.376717 loss) | |
| I0401 12:59:49.775270 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.111111 | |
| I0401 12:59:49.775284 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.721591 | |
| I0401 12:59:49.775295 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.203704 | |
| I0401 12:59:49.775308 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.67225 (* 1 = 3.67225 loss) | |
| I0401 12:59:49.775321 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.20688 (* 1 = 1.20688 loss) | |
| I0401 12:59:49.775333 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 12:59:49.775344 6134 solver.cpp:245] Train net output #16: total_confidence = 7.09011e-07 | |
| I0401 12:59:49.775357 6134 sgd_solver.cpp:106] Iteration 3500, lr = 0.01 | |
| I0401 13:01:58.159101 6134 solver.cpp:229] Iteration 4000, loss = 7.10611 | |
| I0401 13:01:58.159226 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0227273 | |
| I0401 13:01:58.159246 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0401 13:01:58.159260 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.136364 | |
| I0401 13:01:58.159274 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 4.09134 (* 0.3 = 1.2274 loss) | |
| I0401 13:01:58.159288 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.13627 (* 0.3 = 0.340881 loss) | |
| I0401 13:01:58.159301 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0909091 | |
| I0401 13:01:58.159312 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 13:01:58.159324 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.227273 | |
| I0401 13:01:58.159337 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.97358 (* 0.3 = 1.19207 loss) | |
| I0401 13:01:58.159350 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.11581 (* 0.3 = 0.334743 loss) | |
| I0401 13:01:58.159363 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0454545 | |
| I0401 13:01:58.159374 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0401 13:01:58.159385 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.181818 | |
| I0401 13:01:58.159399 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 4.14542 (* 1 = 4.14542 loss) | |
| I0401 13:01:58.159411 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.17096 (* 1 = 1.17096 loss) | |
| I0401 13:01:58.159423 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:01:58.159435 6134 solver.cpp:245] Train net output #16: total_confidence = 6.01958e-05 | |
| I0401 13:01:58.159446 6134 sgd_solver.cpp:106] Iteration 4000, lr = 0.01 | |
| I0401 13:04:06.616226 6134 solver.cpp:229] Iteration 4500, loss = 7.11091 | |
| I0401 13:04:06.616375 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0243902 | |
| I0401 13:04:06.616396 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 13:04:06.616410 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.195122 | |
| I0401 13:04:06.616425 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.45376 (* 0.3 = 1.03613 loss) | |
| I0401 13:04:06.616439 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.974194 (* 0.3 = 0.292258 loss) | |
| I0401 13:04:06.616451 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0 | |
| I0401 13:04:06.616463 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 13:04:06.616475 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.0731707 | |
| I0401 13:04:06.616489 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.68507 (* 0.3 = 1.10552 loss) | |
| I0401 13:04:06.616503 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.995368 (* 0.3 = 0.29861 loss) | |
| I0401 13:04:06.616515 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0487805 | |
| I0401 13:04:06.616530 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0401 13:04:06.616542 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.146341 | |
| I0401 13:04:06.616555 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.43503 (* 1 = 3.43503 loss) | |
| I0401 13:04:06.616569 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.976691 (* 1 = 0.976691 loss) | |
| I0401 13:04:06.616581 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:04:06.616592 6134 solver.cpp:245] Train net output #16: total_confidence = 5.86373e-05 | |
| I0401 13:04:06.616605 6134 sgd_solver.cpp:106] Iteration 4500, lr = 0.01 | |
| I0401 13:06:14.928297 6134 solver.cpp:338] Iteration 5000, Testing net (#0) | |
| I0401 13:06:44.721635 6134 solver.cpp:393] Test loss: 6.48153 | |
| I0401 13:06:44.721679 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.105584 | |
| I0401 13:06:44.721695 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.779636 | |
| I0401 13:06:44.721707 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.304764 | |
| I0401 13:06:44.721724 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 3.51898 (* 0.3 = 1.05569 loss) | |
| I0401 13:06:44.721737 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.888278 (* 0.3 = 0.266483 loss) | |
| I0401 13:06:44.721750 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.140229 | |
| I0401 13:06:44.721761 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.787954 | |
| I0401 13:06:44.721772 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.318096 | |
| I0401 13:06:44.721786 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 3.57113 (* 0.3 = 1.07134 loss) | |
| I0401 13:06:44.721801 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.901538 (* 0.3 = 0.270461 loss) | |
| I0401 13:06:44.721812 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.143378 | |
| I0401 13:06:44.721823 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.787227 | |
| I0401 13:06:44.721834 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.350022 | |
| I0401 13:06:44.721848 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 3.04125 (* 1 = 3.04125 loss) | |
| I0401 13:06:44.721860 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.776293 (* 1 = 0.776293 loss) | |
| I0401 13:06:44.721873 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.001 | |
| I0401 13:06:44.721884 6134 solver.cpp:406] Test net output #16: total_confidence = 0.000799455 | |
| I0401 13:06:44.871942 6134 solver.cpp:229] Iteration 5000, loss = 7.08775 | |
| I0401 13:06:44.871978 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0238095 | |
| I0401 13:06:44.871994 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0401 13:06:44.872006 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.119048 | |
| I0401 13:06:44.872022 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.36235 (* 0.3 = 1.0087 loss) | |
| I0401 13:06:44.872036 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.04216 (* 0.3 = 0.312649 loss) | |
| I0401 13:06:44.872051 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0714286 | |
| I0401 13:06:44.872064 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 13:06:44.872076 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.119048 | |
| I0401 13:06:44.872088 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.40962 (* 0.3 = 1.02288 loss) | |
| I0401 13:06:44.872102 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.01223 (* 0.3 = 0.30367 loss) | |
| I0401 13:06:44.872114 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0714286 | |
| I0401 13:06:44.872125 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0401 13:06:44.872138 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.142857 | |
| I0401 13:06:44.872151 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.14789 (* 1 = 3.14789 loss) | |
| I0401 13:06:44.872165 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.89757 (* 1 = 0.89757 loss) | |
| I0401 13:06:44.872176 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:06:44.872187 6134 solver.cpp:245] Train net output #16: total_confidence = 1.3146e-05 | |
| I0401 13:06:44.872200 6134 sgd_solver.cpp:106] Iteration 5000, lr = 0.01 | |
| I0401 13:08:53.338459 6134 solver.cpp:229] Iteration 5500, loss = 6.98019 | |
| I0401 13:08:53.338604 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0697674 | |
| I0401 13:08:53.338625 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 13:08:53.338639 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.0930233 | |
| I0401 13:08:53.338654 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.43841 (* 0.3 = 1.03152 loss) | |
| I0401 13:08:53.338668 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.97645 (* 0.3 = 0.292935 loss) | |
| I0401 13:08:53.338680 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0232558 | |
| I0401 13:08:53.338692 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0401 13:08:53.338704 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.162791 | |
| I0401 13:08:53.338717 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.63452 (* 0.3 = 1.09036 loss) | |
| I0401 13:08:53.338732 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.00852 (* 0.3 = 0.302557 loss) | |
| I0401 13:08:53.338742 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0232558 | |
| I0401 13:08:53.338754 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0401 13:08:53.338765 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.116279 | |
| I0401 13:08:53.338779 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.42606 (* 1 = 3.42606 loss) | |
| I0401 13:08:53.338793 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.96259 (* 1 = 0.96259 loss) | |
| I0401 13:08:53.338804 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:08:53.338815 6134 solver.cpp:245] Train net output #16: total_confidence = 1.13981e-05 | |
| I0401 13:08:53.338827 6134 sgd_solver.cpp:106] Iteration 5500, lr = 0.01 | |
| I0401 13:11:01.703521 6134 solver.cpp:229] Iteration 6000, loss = 6.95745 | |
| I0401 13:11:01.703629 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0769231 | |
| I0401 13:11:01.703649 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.721591 | |
| I0401 13:11:01.703662 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.115385 | |
| I0401 13:11:01.703677 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 4.11588 (* 0.3 = 1.23477 loss) | |
| I0401 13:11:01.703692 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.30326 (* 0.3 = 0.390977 loss) | |
| I0401 13:11:01.703704 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0576923 | |
| I0401 13:11:01.703716 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.721591 | |
| I0401 13:11:01.703728 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.134615 | |
| I0401 13:11:01.703742 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.74357 (* 0.3 = 1.12307 loss) | |
| I0401 13:11:01.703754 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.16968 (* 0.3 = 0.350905 loss) | |
| I0401 13:11:01.703766 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0384615 | |
| I0401 13:11:01.703778 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.715909 | |
| I0401 13:11:01.703789 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.134615 | |
| I0401 13:11:01.703804 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.69385 (* 1 = 3.69385 loss) | |
| I0401 13:11:01.703817 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.13357 (* 1 = 1.13357 loss) | |
| I0401 13:11:01.703829 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:11:01.703840 6134 solver.cpp:245] Train net output #16: total_confidence = 1.18051e-05 | |
| I0401 13:11:01.703851 6134 sgd_solver.cpp:106] Iteration 6000, lr = 0.01 | |
| I0401 13:13:10.045214 6134 solver.cpp:229] Iteration 6500, loss = 6.90748 | |
| I0401 13:13:10.045346 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0 | |
| I0401 13:13:10.045366 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 13:13:10.045378 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.176471 | |
| I0401 13:13:10.045395 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.62205 (* 0.3 = 1.08662 loss) | |
| I0401 13:13:10.045410 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.946883 (* 0.3 = 0.284065 loss) | |
| I0401 13:13:10.045423 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0 | |
| I0401 13:13:10.045434 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0401 13:13:10.045446 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.0588235 | |
| I0401 13:13:10.045460 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.7926 (* 0.3 = 1.13778 loss) | |
| I0401 13:13:10.045474 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.03574 (* 0.3 = 0.310723 loss) | |
| I0401 13:13:10.045486 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0294118 | |
| I0401 13:13:10.045497 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0401 13:13:10.045509 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.147059 | |
| I0401 13:13:10.045526 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.58827 (* 1 = 3.58827 loss) | |
| I0401 13:13:10.045539 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.99052 (* 1 = 0.99052 loss) | |
| I0401 13:13:10.045552 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:13:10.045562 6134 solver.cpp:245] Train net output #16: total_confidence = 1.84045e-06 | |
| I0401 13:13:10.045575 6134 sgd_solver.cpp:106] Iteration 6500, lr = 0.01 | |
| I0401 13:15:18.489142 6134 solver.cpp:229] Iteration 7000, loss = 6.88976 | |
| I0401 13:15:18.489253 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.15 | |
| I0401 13:15:18.489272 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 13:15:18.489284 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.325 | |
| I0401 13:15:18.489300 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.99769 (* 0.3 = 0.899307 loss) | |
| I0401 13:15:18.489315 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.795148 (* 0.3 = 0.238545 loss) | |
| I0401 13:15:18.489326 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.075 | |
| I0401 13:15:18.489338 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 13:15:18.489351 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.225 | |
| I0401 13:15:18.489363 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.07115 (* 0.3 = 0.921344 loss) | |
| I0401 13:15:18.489377 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.895987 (* 0.3 = 0.268796 loss) | |
| I0401 13:15:18.489388 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.1 | |
| I0401 13:15:18.489400 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0401 13:15:18.489413 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.3 | |
| I0401 13:15:18.489425 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.88599 (* 1 = 2.88599 loss) | |
| I0401 13:15:18.489439 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.772585 (* 1 = 0.772585 loss) | |
| I0401 13:15:18.489450 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:15:18.489462 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000124423 | |
| I0401 13:15:18.489475 6134 sgd_solver.cpp:106] Iteration 7000, lr = 0.01 | |
| I0401 13:17:27.034993 6134 solver.cpp:229] Iteration 7500, loss = 6.87051 | |
| I0401 13:17:27.035276 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.02 | |
| I0401 13:17:27.035298 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.721591 | |
| I0401 13:17:27.035310 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.12 | |
| I0401 13:17:27.035327 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.43534 (* 0.3 = 1.0306 loss) | |
| I0401 13:17:27.035341 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.02762 (* 0.3 = 0.308286 loss) | |
| I0401 13:17:27.035353 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.04 | |
| I0401 13:17:27.035365 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.721591 | |
| I0401 13:17:27.035377 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.16 | |
| I0401 13:17:27.035390 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.46977 (* 0.3 = 1.04093 loss) | |
| I0401 13:17:27.035403 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.06672 (* 0.3 = 0.320016 loss) | |
| I0401 13:17:27.035415 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.08 | |
| I0401 13:17:27.035428 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0401 13:17:27.035439 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.2 | |
| I0401 13:17:27.035451 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.23281 (* 1 = 3.23281 loss) | |
| I0401 13:17:27.035465 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.955508 (* 1 = 0.955508 loss) | |
| I0401 13:17:27.035477 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:17:27.035488 6134 solver.cpp:245] Train net output #16: total_confidence = 4.60563e-06 | |
| I0401 13:17:27.035501 6134 sgd_solver.cpp:106] Iteration 7500, lr = 0.01 | |
| I0401 13:19:35.514016 6134 solver.cpp:229] Iteration 8000, loss = 6.85358 | |
| I0401 13:19:35.514122 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0612245 | |
| I0401 13:19:35.514140 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0401 13:19:35.514153 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.22449 | |
| I0401 13:19:35.514168 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.61572 (* 0.3 = 1.08471 loss) | |
| I0401 13:19:35.514183 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.06568 (* 0.3 = 0.319705 loss) | |
| I0401 13:19:35.514195 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0612245 | |
| I0401 13:19:35.514207 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.732955 | |
| I0401 13:19:35.514219 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.244898 | |
| I0401 13:19:35.514231 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.67513 (* 0.3 = 1.10254 loss) | |
| I0401 13:19:35.514245 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.09881 (* 0.3 = 0.329642 loss) | |
| I0401 13:19:35.514257 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0612245 | |
| I0401 13:19:35.514268 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0401 13:19:35.514281 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.183673 | |
| I0401 13:19:35.514293 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.6218 (* 1 = 3.6218 loss) | |
| I0401 13:19:35.514307 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.05419 (* 1 = 1.05419 loss) | |
| I0401 13:19:35.514318 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:19:35.514330 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000183274 | |
| I0401 13:19:35.514341 6134 sgd_solver.cpp:106] Iteration 8000, lr = 0.01 | |
| I0401 13:21:44.015879 6134 solver.cpp:229] Iteration 8500, loss = 6.81192 | |
| I0401 13:21:44.016007 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0869565 | |
| I0401 13:21:44.016032 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0401 13:21:44.016046 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.326087 | |
| I0401 13:21:44.016062 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.00757 (* 0.3 = 0.902271 loss) | |
| I0401 13:21:44.016075 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.927711 (* 0.3 = 0.278313 loss) | |
| I0401 13:21:44.016091 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.108696 | |
| I0401 13:21:44.016103 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0401 13:21:44.016115 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.26087 | |
| I0401 13:21:44.016129 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.13177 (* 0.3 = 0.93953 loss) | |
| I0401 13:21:44.016142 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.985022 (* 0.3 = 0.295506 loss) | |
| I0401 13:21:44.016155 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0652174 | |
| I0401 13:21:44.016167 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0401 13:21:44.016178 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.26087 | |
| I0401 13:21:44.016192 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.03434 (* 1 = 3.03434 loss) | |
| I0401 13:21:44.016206 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.925309 (* 1 = 0.925309 loss) | |
| I0401 13:21:44.016217 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:21:44.016228 6134 solver.cpp:245] Train net output #16: total_confidence = 9.67964e-06 | |
| I0401 13:21:44.016240 6134 sgd_solver.cpp:106] Iteration 8500, lr = 0.01 | |
| I0401 13:23:52.505467 6134 solver.cpp:229] Iteration 9000, loss = 6.75599 | |
| I0401 13:23:52.505576 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.133333 | |
| I0401 13:23:52.505595 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 13:23:52.505609 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.311111 | |
| I0401 13:23:52.505625 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.99297 (* 0.3 = 0.89789 loss) | |
| I0401 13:23:52.505638 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.91502 (* 0.3 = 0.274506 loss) | |
| I0401 13:23:52.505650 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.111111 | |
| I0401 13:23:52.505663 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 13:23:52.505674 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.244444 | |
| I0401 13:23:52.505687 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.06223 (* 0.3 = 0.91867 loss) | |
| I0401 13:23:52.505702 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.896422 (* 0.3 = 0.268927 loss) | |
| I0401 13:23:52.505713 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.133333 | |
| I0401 13:23:52.505724 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.772727 | |
| I0401 13:23:52.505736 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.377778 | |
| I0401 13:23:52.505749 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.9204 (* 1 = 2.9204 loss) | |
| I0401 13:23:52.505764 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.861425 (* 1 = 0.861425 loss) | |
| I0401 13:23:52.505774 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:23:52.505786 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000434245 | |
| I0401 13:23:52.505797 6134 sgd_solver.cpp:106] Iteration 9000, lr = 0.01 | |
| I0401 13:26:01.022151 6134 solver.cpp:229] Iteration 9500, loss = 6.76842 | |
| I0401 13:26:01.022465 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0930233 | |
| I0401 13:26:01.022487 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 13:26:01.022500 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.209302 | |
| I0401 13:26:01.022518 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 4.00866 (* 0.3 = 1.2026 loss) | |
| I0401 13:26:01.022533 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.15429 (* 0.3 = 0.346286 loss) | |
| I0401 13:26:01.022547 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.116279 | |
| I0401 13:26:01.022558 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0401 13:26:01.022569 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.27907 | |
| I0401 13:26:01.022583 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.83016 (* 0.3 = 1.14905 loss) | |
| I0401 13:26:01.022598 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.05824 (* 0.3 = 0.317472 loss) | |
| I0401 13:26:01.022610 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0697674 | |
| I0401 13:26:01.022621 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0401 13:26:01.022634 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.255814 | |
| I0401 13:26:01.022647 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.83008 (* 1 = 3.83008 loss) | |
| I0401 13:26:01.022660 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.04035 (* 1 = 1.04035 loss) | |
| I0401 13:26:01.022672 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:26:01.022683 6134 solver.cpp:245] Train net output #16: total_confidence = 4.44319e-05 | |
| I0401 13:26:01.022696 6134 sgd_solver.cpp:106] Iteration 9500, lr = 0.01 | |
| I0401 13:28:09.377058 6134 solver.cpp:338] Iteration 10000, Testing net (#0) | |
| I0401 13:28:39.116127 6134 solver.cpp:393] Test loss: 6.09995 | |
| I0401 13:28:39.116170 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.0993023 | |
| I0401 13:28:39.116186 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.777682 | |
| I0401 13:28:39.116199 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.281737 | |
| I0401 13:28:39.116214 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 3.10383 (* 0.3 = 0.931149 loss) | |
| I0401 13:28:39.116228 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.782006 (* 0.3 = 0.234602 loss) | |
| I0401 13:28:39.116240 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.105045 | |
| I0401 13:28:39.116252 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.779181 | |
| I0401 13:28:39.116263 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.299468 | |
| I0401 13:28:39.116277 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 3.32187 (* 0.3 = 0.99656 loss) | |
| I0401 13:28:39.116291 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.833804 (* 0.3 = 0.250141 loss) | |
| I0401 13:28:39.116302 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.101468 | |
| I0401 13:28:39.116314 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.775591 | |
| I0401 13:28:39.116325 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.293905 | |
| I0401 13:28:39.116339 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 2.92605 (* 1 = 2.92605 loss) | |
| I0401 13:28:39.116353 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.761456 (* 1 = 0.761456 loss) | |
| I0401 13:28:39.116364 6134 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0401 13:28:39.116375 6134 solver.cpp:406] Test net output #16: total_confidence = 0.000425581 | |
| I0401 13:28:39.267325 6134 solver.cpp:229] Iteration 10000, loss = 6.7372 | |
| I0401 13:28:39.267362 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0416667 | |
| I0401 13:28:39.267379 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0401 13:28:39.267392 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.229167 | |
| I0401 13:28:39.267406 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.41505 (* 0.3 = 1.02451 loss) | |
| I0401 13:28:39.267421 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.01102 (* 0.3 = 0.303306 loss) | |
| I0401 13:28:39.267432 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.104167 | |
| I0401 13:28:39.267446 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.744318 | |
| I0401 13:28:39.267457 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.25 | |
| I0401 13:28:39.267470 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.28085 (* 0.3 = 0.984255 loss) | |
| I0401 13:28:39.267483 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.978654 (* 0.3 = 0.293596 loss) | |
| I0401 13:28:39.267496 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.104167 | |
| I0401 13:28:39.267508 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.75 | |
| I0401 13:28:39.267519 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.25 | |
| I0401 13:28:39.267532 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.23297 (* 1 = 3.23297 loss) | |
| I0401 13:28:39.267546 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.941816 (* 1 = 0.941816 loss) | |
| I0401 13:28:39.267557 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:28:39.267570 6134 solver.cpp:245] Train net output #16: total_confidence = 3.92164e-06 | |
| I0401 13:28:39.267582 6134 sgd_solver.cpp:106] Iteration 10000, lr = 0.01 | |
| I0401 13:30:47.659692 6134 solver.cpp:229] Iteration 10500, loss = 6.69848 | |
| I0401 13:30:47.659816 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0930233 | |
| I0401 13:30:47.659837 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 13:30:47.659849 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.162791 | |
| I0401 13:30:47.659864 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.16975 (* 0.3 = 0.950926 loss) | |
| I0401 13:30:47.659879 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.879164 (* 0.3 = 0.263749 loss) | |
| I0401 13:30:47.659891 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.116279 | |
| I0401 13:30:47.659904 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 13:30:47.659914 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.302326 | |
| I0401 13:30:47.659929 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.92261 (* 0.3 = 0.876783 loss) | |
| I0401 13:30:47.659943 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.814878 (* 0.3 = 0.244463 loss) | |
| I0401 13:30:47.659955 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.162791 | |
| I0401 13:30:47.659966 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0401 13:30:47.659978 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.27907 | |
| I0401 13:30:47.659991 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.02839 (* 1 = 3.02839 loss) | |
| I0401 13:30:47.660006 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.793209 (* 1 = 0.793209 loss) | |
| I0401 13:30:47.660017 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:30:47.660028 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000151129 | |
| I0401 13:30:47.660040 6134 sgd_solver.cpp:106] Iteration 10500, lr = 0.01 | |
| I0401 13:32:56.169900 6134 solver.cpp:229] Iteration 11000, loss = 6.6783 | |
| I0401 13:32:56.170035 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.130435 | |
| I0401 13:32:56.170056 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 13:32:56.170069 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.26087 | |
| I0401 13:32:56.170084 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.92371 (* 0.3 = 0.877112 loss) | |
| I0401 13:32:56.170099 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.843243 (* 0.3 = 0.252973 loss) | |
| I0401 13:32:56.170111 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.152174 | |
| I0401 13:32:56.170123 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 13:32:56.170135 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.217391 | |
| I0401 13:32:56.170147 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.04203 (* 0.3 = 0.912609 loss) | |
| I0401 13:32:56.170161 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.856727 (* 0.3 = 0.257018 loss) | |
| I0401 13:32:56.170173 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.130435 | |
| I0401 13:32:56.170184 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.772727 | |
| I0401 13:32:56.170195 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.282609 | |
| I0401 13:32:56.170209 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.91161 (* 1 = 2.91161 loss) | |
| I0401 13:32:56.170223 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.804483 (* 1 = 0.804483 loss) | |
| I0401 13:32:56.170234 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:32:56.170246 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000167535 | |
| I0401 13:32:56.170258 6134 sgd_solver.cpp:106] Iteration 11000, lr = 0.01 | |
| I0401 13:35:04.630028 6134 solver.cpp:229] Iteration 11500, loss = 6.64544 | |
| I0401 13:35:04.630148 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0434783 | |
| I0401 13:35:04.630168 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0401 13:35:04.630182 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.26087 | |
| I0401 13:35:04.630198 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.03007 (* 0.3 = 0.90902 loss) | |
| I0401 13:35:04.630211 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.894923 (* 0.3 = 0.268477 loss) | |
| I0401 13:35:04.630223 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0434783 | |
| I0401 13:35:04.630235 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.744318 | |
| I0401 13:35:04.630247 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.195652 | |
| I0401 13:35:04.630262 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.04383 (* 0.3 = 0.913149 loss) | |
| I0401 13:35:04.630276 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.886182 (* 0.3 = 0.265855 loss) | |
| I0401 13:35:04.630288 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.108696 | |
| I0401 13:35:04.630300 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.75 | |
| I0401 13:35:04.630311 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.195652 | |
| I0401 13:35:04.630326 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.94977 (* 1 = 2.94977 loss) | |
| I0401 13:35:04.630339 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.886716 (* 1 = 0.886716 loss) | |
| I0401 13:35:04.630352 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:35:04.630362 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000104385 | |
| I0401 13:35:04.630374 6134 sgd_solver.cpp:106] Iteration 11500, lr = 0.01 | |
| I0401 13:37:13.119850 6134 solver.cpp:229] Iteration 12000, loss = 6.67639 | |
| I0401 13:37:13.120581 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.191489 | |
| I0401 13:37:13.120604 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 13:37:13.120615 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.276596 | |
| I0401 13:37:13.120631 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.1482 (* 0.3 = 0.94446 loss) | |
| I0401 13:37:13.120646 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.897649 (* 0.3 = 0.269295 loss) | |
| I0401 13:37:13.120659 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.170213 | |
| I0401 13:37:13.120671 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 13:37:13.120682 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.340426 | |
| I0401 13:37:13.120697 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.15494 (* 0.3 = 0.946483 loss) | |
| I0401 13:37:13.120709 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.967485 (* 0.3 = 0.290245 loss) | |
| I0401 13:37:13.120721 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.12766 | |
| I0401 13:37:13.120733 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0401 13:37:13.120744 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.255319 | |
| I0401 13:37:13.120759 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.22467 (* 1 = 3.22467 loss) | |
| I0401 13:37:13.120772 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.922504 (* 1 = 0.922504 loss) | |
| I0401 13:37:13.120784 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:37:13.120795 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000141357 | |
| I0401 13:37:13.120808 6134 sgd_solver.cpp:106] Iteration 12000, lr = 0.01 | |
| I0401 13:39:21.547664 6134 solver.cpp:229] Iteration 12500, loss = 6.61706 | |
| I0401 13:39:21.547793 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.121951 | |
| I0401 13:39:21.547813 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 13:39:21.547826 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.317073 | |
| I0401 13:39:21.547842 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.83492 (* 0.3 = 0.850476 loss) | |
| I0401 13:39:21.547855 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.799583 (* 0.3 = 0.239875 loss) | |
| I0401 13:39:21.547868 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.097561 | |
| I0401 13:39:21.547880 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 13:39:21.547893 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.317073 | |
| I0401 13:39:21.547905 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.89554 (* 0.3 = 0.868661 loss) | |
| I0401 13:39:21.547919 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.804542 (* 0.3 = 0.241363 loss) | |
| I0401 13:39:21.547930 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.195122 | |
| I0401 13:39:21.547942 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0401 13:39:21.547953 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.317073 | |
| I0401 13:39:21.547967 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.76444 (* 1 = 2.76444 loss) | |
| I0401 13:39:21.547981 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.72786 (* 1 = 0.72786 loss) | |
| I0401 13:39:21.547992 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:39:21.548003 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000145038 | |
| I0401 13:39:21.548015 6134 sgd_solver.cpp:106] Iteration 12500, lr = 0.01 | |
| I0401 13:41:30.160312 6134 solver.cpp:229] Iteration 13000, loss = 6.66031 | |
| I0401 13:41:30.160455 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.142857 | |
| I0401 13:41:30.160485 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 13:41:30.160509 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.261905 | |
| I0401 13:41:30.160537 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.06373 (* 0.3 = 0.919118 loss) | |
| I0401 13:41:30.160552 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.88797 (* 0.3 = 0.266391 loss) | |
| I0401 13:41:30.160565 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.119048 | |
| I0401 13:41:30.160578 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0401 13:41:30.160588 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.357143 | |
| I0401 13:41:30.160603 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.07338 (* 0.3 = 0.922014 loss) | |
| I0401 13:41:30.160615 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.0168 (* 0.3 = 0.30504 loss) | |
| I0401 13:41:30.160627 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.142857 | |
| I0401 13:41:30.160640 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.778409 | |
| I0401 13:41:30.160650 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.285714 | |
| I0401 13:41:30.160665 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.90885 (* 1 = 2.90885 loss) | |
| I0401 13:41:30.160677 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.844784 (* 1 = 0.844784 loss) | |
| I0401 13:41:30.160689 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:41:30.160701 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000100537 | |
| I0401 13:41:30.160712 6134 sgd_solver.cpp:106] Iteration 13000, lr = 0.01 | |
| I0401 13:43:38.648793 6134 solver.cpp:229] Iteration 13500, loss = 6.61744 | |
| I0401 13:43:38.648906 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.133333 | |
| I0401 13:43:38.648937 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0401 13:43:38.648962 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.244444 | |
| I0401 13:43:38.648989 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.14507 (* 0.3 = 0.943521 loss) | |
| I0401 13:43:38.649019 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.971647 (* 0.3 = 0.291494 loss) | |
| I0401 13:43:38.649058 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.111111 | |
| I0401 13:43:38.649085 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0401 13:43:38.649109 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.288889 | |
| I0401 13:43:38.649135 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.13077 (* 0.3 = 0.939232 loss) | |
| I0401 13:43:38.649161 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.929451 (* 0.3 = 0.278835 loss) | |
| I0401 13:43:38.649184 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.133333 | |
| I0401 13:43:38.649206 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0401 13:43:38.649227 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.266667 | |
| I0401 13:43:38.649253 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.08032 (* 1 = 3.08032 loss) | |
| I0401 13:43:38.649279 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.878164 (* 1 = 0.878164 loss) | |
| I0401 13:43:38.649301 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:43:38.649322 6134 solver.cpp:245] Train net output #16: total_confidence = 6.79612e-05 | |
| I0401 13:43:38.649343 6134 sgd_solver.cpp:106] Iteration 13500, lr = 0.01 | |
| I0401 13:45:47.230180 6134 solver.cpp:229] Iteration 14000, loss = 6.54827 | |
| I0401 13:45:47.230305 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0 | |
| I0401 13:45:47.230324 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0401 13:45:47.230337 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.170213 | |
| I0401 13:45:47.230352 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.40451 (* 0.3 = 1.02135 loss) | |
| I0401 13:45:47.230367 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.01213 (* 0.3 = 0.303639 loss) | |
| I0401 13:45:47.230379 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0 | |
| I0401 13:45:47.230391 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.721591 | |
| I0401 13:45:47.230402 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.191489 | |
| I0401 13:45:47.230417 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.35998 (* 0.3 = 1.00799 loss) | |
| I0401 13:45:47.230429 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.00391 (* 0.3 = 0.301174 loss) | |
| I0401 13:45:47.230440 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0212766 | |
| I0401 13:45:47.230453 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.732955 | |
| I0401 13:45:47.230463 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.0851064 | |
| I0401 13:45:47.230478 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.30757 (* 1 = 3.30757 loss) | |
| I0401 13:45:47.230490 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.00783 (* 1 = 1.00783 loss) | |
| I0401 13:45:47.230501 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:45:47.230512 6134 solver.cpp:245] Train net output #16: total_confidence = 1.37611e-06 | |
| I0401 13:45:47.230527 6134 sgd_solver.cpp:106] Iteration 14000, lr = 0.01 | |
| I0401 13:47:55.755676 6134 solver.cpp:229] Iteration 14500, loss = 6.51078 | |
| I0401 13:47:55.755910 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0731707 | |
| I0401 13:47:55.755929 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0401 13:47:55.755941 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.243902 | |
| I0401 13:47:55.755957 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.13267 (* 0.3 = 0.939802 loss) | |
| I0401 13:47:55.755971 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.01623 (* 0.3 = 0.304868 loss) | |
| I0401 13:47:55.755983 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0243902 | |
| I0401 13:47:55.755995 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0401 13:47:55.756007 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.292683 | |
| I0401 13:47:55.756021 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.01015 (* 0.3 = 0.903046 loss) | |
| I0401 13:47:55.756034 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.950414 (* 0.3 = 0.285124 loss) | |
| I0401 13:47:55.756047 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0731707 | |
| I0401 13:47:55.756058 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.75 | |
| I0401 13:47:55.756069 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.341463 | |
| I0401 13:47:55.756083 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.9942 (* 1 = 2.9942 loss) | |
| I0401 13:47:55.756096 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.914748 (* 1 = 0.914748 loss) | |
| I0401 13:47:55.756111 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:47:55.756122 6134 solver.cpp:245] Train net output #16: total_confidence = 3.24563e-05 | |
| I0401 13:47:55.756135 6134 sgd_solver.cpp:106] Iteration 14500, lr = 0.01 | |
| I0401 13:50:04.162964 6134 solver.cpp:338] Iteration 15000, Testing net (#0) | |
| I0401 13:50:33.878303 6134 solver.cpp:393] Test loss: 6.42501 | |
| I0401 13:50:33.878360 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.0867452 | |
| I0401 13:50:33.878376 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.774682 | |
| I0401 13:50:33.878388 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.249524 | |
| I0401 13:50:33.878404 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 3.43181 (* 0.3 = 1.02954 loss) | |
| I0401 13:50:33.878418 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.864127 (* 0.3 = 0.259238 loss) | |
| I0401 13:50:33.878430 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.103944 | |
| I0401 13:50:33.878442 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.777591 | |
| I0401 13:50:33.878453 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.289435 | |
| I0401 13:50:33.878466 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 3.24657 (* 0.3 = 0.973971 loss) | |
| I0401 13:50:33.878479 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.829063 (* 0.3 = 0.248719 loss) | |
| I0401 13:50:33.878491 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.110642 | |
| I0401 13:50:33.878502 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.778818 | |
| I0401 13:50:33.878515 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.311218 | |
| I0401 13:50:33.878530 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 3.10868 (* 1 = 3.10868 loss) | |
| I0401 13:50:33.878545 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.804862 (* 1 = 0.804862 loss) | |
| I0401 13:50:33.878556 6134 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0401 13:50:33.878567 6134 solver.cpp:406] Test net output #16: total_confidence = 0.000205893 | |
| I0401 13:50:34.030284 6134 solver.cpp:229] Iteration 15000, loss = 6.48884 | |
| I0401 13:50:34.030345 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0851064 | |
| I0401 13:50:34.030364 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 13:50:34.030376 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.170213 | |
| I0401 13:50:34.030392 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.36107 (* 0.3 = 1.00832 loss) | |
| I0401 13:50:34.030407 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.943656 (* 0.3 = 0.283097 loss) | |
| I0401 13:50:34.030419 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.106383 | |
| I0401 13:50:34.030431 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0401 13:50:34.030443 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.191489 | |
| I0401 13:50:34.030457 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.29119 (* 0.3 = 0.987357 loss) | |
| I0401 13:50:34.030470 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.985205 (* 0.3 = 0.295561 loss) | |
| I0401 13:50:34.030483 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.106383 | |
| I0401 13:50:34.030494 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0401 13:50:34.030509 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.212766 | |
| I0401 13:50:34.030524 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.18615 (* 1 = 3.18615 loss) | |
| I0401 13:50:34.030537 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.917527 (* 1 = 0.917527 loss) | |
| I0401 13:50:34.030550 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:50:34.030561 6134 solver.cpp:245] Train net output #16: total_confidence = 7.491e-06 | |
| I0401 13:50:34.030573 6134 sgd_solver.cpp:106] Iteration 15000, lr = 0.01 | |
| I0401 13:52:42.377859 6134 solver.cpp:229] Iteration 15500, loss = 6.48074 | |
| I0401 13:52:42.377984 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.12766 | |
| I0401 13:52:42.378003 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0401 13:52:42.378016 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.234043 | |
| I0401 13:52:42.378032 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.04257 (* 0.3 = 0.91277 loss) | |
| I0401 13:52:42.378046 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.931176 (* 0.3 = 0.279353 loss) | |
| I0401 13:52:42.378058 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.170213 | |
| I0401 13:52:42.378072 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 13:52:42.378082 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.319149 | |
| I0401 13:52:42.378098 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.01987 (* 0.3 = 0.905961 loss) | |
| I0401 13:52:42.378113 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.884767 (* 0.3 = 0.26543 loss) | |
| I0401 13:52:42.378125 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.148936 | |
| I0401 13:52:42.378136 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0401 13:52:42.378149 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.382979 | |
| I0401 13:52:42.378161 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.87586 (* 1 = 2.87586 loss) | |
| I0401 13:52:42.378175 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.878418 (* 1 = 0.878418 loss) | |
| I0401 13:52:42.378186 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:52:42.378198 6134 solver.cpp:245] Train net output #16: total_confidence = 6.54283e-06 | |
| I0401 13:52:42.378211 6134 sgd_solver.cpp:106] Iteration 15500, lr = 0.01 | |
| I0401 13:54:50.725260 6134 solver.cpp:229] Iteration 16000, loss = 6.47345 | |
| I0401 13:54:50.725365 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.139535 | |
| I0401 13:54:50.725384 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 13:54:50.725396 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.27907 | |
| I0401 13:54:50.725412 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.99559 (* 0.3 = 0.898678 loss) | |
| I0401 13:54:50.725426 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.785119 (* 0.3 = 0.235536 loss) | |
| I0401 13:54:50.725438 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0697674 | |
| I0401 13:54:50.725450 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 13:54:50.725462 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.255814 | |
| I0401 13:54:50.725476 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.02296 (* 0.3 = 0.906888 loss) | |
| I0401 13:54:50.725491 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.797284 (* 0.3 = 0.239185 loss) | |
| I0401 13:54:50.725502 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.209302 | |
| I0401 13:54:50.725514 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0401 13:54:50.725529 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.418605 | |
| I0401 13:54:50.725543 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.82141 (* 1 = 2.82141 loss) | |
| I0401 13:54:50.725556 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.774697 (* 1 = 0.774697 loss) | |
| I0401 13:54:50.725569 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:54:50.725580 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000154658 | |
| I0401 13:54:50.725592 6134 sgd_solver.cpp:106] Iteration 16000, lr = 0.01 | |
| I0401 13:56:59.262763 6134 solver.cpp:229] Iteration 16500, loss = 6.43571 | |
| I0401 13:56:59.263056 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0784314 | |
| I0401 13:56:59.263077 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0401 13:56:59.263088 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.352941 | |
| I0401 13:56:59.263104 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.01052 (* 0.3 = 0.903156 loss) | |
| I0401 13:56:59.263118 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.946859 (* 0.3 = 0.284058 loss) | |
| I0401 13:56:59.263131 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0980392 | |
| I0401 13:56:59.263144 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.727273 | |
| I0401 13:56:59.263155 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.235294 | |
| I0401 13:56:59.263169 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.14041 (* 0.3 = 0.942122 loss) | |
| I0401 13:56:59.263182 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.988132 (* 0.3 = 0.29644 loss) | |
| I0401 13:56:59.263195 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.137255 | |
| I0401 13:56:59.263206 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.75 | |
| I0401 13:56:59.263217 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.392157 | |
| I0401 13:56:59.263231 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.86422 (* 1 = 2.86422 loss) | |
| I0401 13:56:59.263245 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.904078 (* 1 = 0.904078 loss) | |
| I0401 13:56:59.263257 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:56:59.263268 6134 solver.cpp:245] Train net output #16: total_confidence = 8.94474e-05 | |
| I0401 13:56:59.263280 6134 sgd_solver.cpp:106] Iteration 16500, lr = 0.01 | |
| I0401 13:59:07.818303 6134 solver.cpp:229] Iteration 17000, loss = 6.3994 | |
| I0401 13:59:07.818411 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.217391 | |
| I0401 13:59:07.818429 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 13:59:07.818441 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.326087 | |
| I0401 13:59:07.818457 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.91126 (* 0.3 = 0.873378 loss) | |
| I0401 13:59:07.818471 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.842408 (* 0.3 = 0.252723 loss) | |
| I0401 13:59:07.818485 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.173913 | |
| I0401 13:59:07.818496 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 13:59:07.818507 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.304348 | |
| I0401 13:59:07.818524 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.97308 (* 0.3 = 0.891924 loss) | |
| I0401 13:59:07.818537 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.895719 (* 0.3 = 0.268716 loss) | |
| I0401 13:59:07.818549 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.130435 | |
| I0401 13:59:07.818562 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0401 13:59:07.818573 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.347826 | |
| I0401 13:59:07.818586 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.86259 (* 1 = 2.86259 loss) | |
| I0401 13:59:07.818600 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.855918 (* 1 = 0.855918 loss) | |
| I0401 13:59:07.818611 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 13:59:07.818624 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000222226 | |
| I0401 13:59:07.818634 6134 sgd_solver.cpp:106] Iteration 17000, lr = 0.01 | |
| I0401 14:01:16.261456 6134 solver.cpp:229] Iteration 17500, loss = 6.38739 | |
| I0401 14:01:16.261582 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.08 | |
| I0401 14:01:16.261601 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0401 14:01:16.261615 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.24 | |
| I0401 14:01:16.261629 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.94549 (* 0.3 = 0.883646 loss) | |
| I0401 14:01:16.261644 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.873793 (* 0.3 = 0.262138 loss) | |
| I0401 14:01:16.261656 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.08 | |
| I0401 14:01:16.261668 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0401 14:01:16.261679 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.22 | |
| I0401 14:01:16.261693 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.13115 (* 0.3 = 0.939346 loss) | |
| I0401 14:01:16.261706 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.919798 (* 0.3 = 0.275939 loss) | |
| I0401 14:01:16.261718 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.16 | |
| I0401 14:01:16.261729 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0401 14:01:16.261741 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.22 | |
| I0401 14:01:16.261754 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.91146 (* 1 = 2.91146 loss) | |
| I0401 14:01:16.261768 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.880542 (* 1 = 0.880542 loss) | |
| I0401 14:01:16.261780 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:01:16.261790 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000202682 | |
| I0401 14:01:16.261802 6134 sgd_solver.cpp:106] Iteration 17500, lr = 0.01 | |
| I0401 14:03:24.519723 6134 solver.cpp:229] Iteration 18000, loss = 6.40169 | |
| I0401 14:03:24.519819 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0454545 | |
| I0401 14:03:24.519837 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0401 14:03:24.519850 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.181818 | |
| I0401 14:03:24.519865 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.09581 (* 0.3 = 0.928743 loss) | |
| I0401 14:03:24.519879 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.02641 (* 0.3 = 0.307922 loss) | |
| I0401 14:03:24.519891 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0454545 | |
| I0401 14:03:24.519903 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.721591 | |
| I0401 14:03:24.519915 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.295455 | |
| I0401 14:03:24.519928 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.03341 (* 0.3 = 0.910022 loss) | |
| I0401 14:03:24.519945 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.944012 (* 0.3 = 0.283204 loss) | |
| I0401 14:03:24.519958 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0909091 | |
| I0401 14:03:24.519970 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.75 | |
| I0401 14:03:24.519982 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.318182 | |
| I0401 14:03:24.519996 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.06774 (* 1 = 3.06774 loss) | |
| I0401 14:03:24.520010 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.924301 (* 1 = 0.924301 loss) | |
| I0401 14:03:24.520022 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:03:24.520035 6134 solver.cpp:245] Train net output #16: total_confidence = 1.76834e-05 | |
| I0401 14:03:24.520048 6134 sgd_solver.cpp:106] Iteration 18000, lr = 0.01 | |
| I0401 14:05:32.841389 6134 solver.cpp:229] Iteration 18500, loss = 6.32362 | |
| I0401 14:05:32.841513 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0217391 | |
| I0401 14:05:32.841533 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0401 14:05:32.841545 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.130435 | |
| I0401 14:05:32.841562 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.66005 (* 0.3 = 1.09801 loss) | |
| I0401 14:05:32.841575 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.02577 (* 0.3 = 0.307731 loss) | |
| I0401 14:05:32.841588 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0652174 | |
| I0401 14:05:32.841599 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0401 14:05:32.841611 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.195652 | |
| I0401 14:05:32.841624 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.53725 (* 0.3 = 1.06117 loss) | |
| I0401 14:05:32.841639 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.01876 (* 0.3 = 0.305628 loss) | |
| I0401 14:05:32.841650 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.0869565 | |
| I0401 14:05:32.841661 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.75 | |
| I0401 14:05:32.841675 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.217391 | |
| I0401 14:05:32.841687 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.4097 (* 1 = 3.4097 loss) | |
| I0401 14:05:32.841701 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.968536 (* 1 = 0.968536 loss) | |
| I0401 14:05:32.841712 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:05:32.841724 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000249993 | |
| I0401 14:05:32.841735 6134 sgd_solver.cpp:106] Iteration 18500, lr = 0.01 | |
| I0401 14:07:41.178032 6134 solver.cpp:229] Iteration 19000, loss = 6.28594 | |
| I0401 14:07:41.178264 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0681818 | |
| I0401 14:07:41.178284 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 14:07:41.178297 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.227273 | |
| I0401 14:07:41.178313 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.14856 (* 0.3 = 0.944568 loss) | |
| I0401 14:07:41.178328 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.889782 (* 0.3 = 0.266934 loss) | |
| I0401 14:07:41.178339 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0909091 | |
| I0401 14:07:41.178351 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0401 14:07:41.178364 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.136364 | |
| I0401 14:07:41.178377 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.14876 (* 0.3 = 0.944627 loss) | |
| I0401 14:07:41.178391 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.911489 (* 0.3 = 0.273447 loss) | |
| I0401 14:07:41.178403 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.159091 | |
| I0401 14:07:41.178414 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0401 14:07:41.178426 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.272727 | |
| I0401 14:07:41.178439 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.92702 (* 1 = 2.92702 loss) | |
| I0401 14:07:41.178453 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.824432 (* 1 = 0.824432 loss) | |
| I0401 14:07:41.178465 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:07:41.178477 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000255109 | |
| I0401 14:07:41.178490 6134 sgd_solver.cpp:106] Iteration 19000, lr = 0.01 | |
| I0401 14:09:49.365310 6134 solver.cpp:229] Iteration 19500, loss = 6.30001 | |
| I0401 14:09:49.365448 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0512821 | |
| I0401 14:09:49.365470 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 14:09:49.365483 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.179487 | |
| I0401 14:09:49.365499 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.02695 (* 0.3 = 0.908085 loss) | |
| I0401 14:09:49.365514 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.957549 (* 0.3 = 0.287265 loss) | |
| I0401 14:09:49.365530 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0769231 | |
| I0401 14:09:49.365542 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 14:09:49.365555 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.25641 | |
| I0401 14:09:49.365574 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.98019 (* 0.3 = 0.894057 loss) | |
| I0401 14:09:49.365592 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.97887 (* 0.3 = 0.293661 loss) | |
| I0401 14:09:49.365604 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.128205 | |
| I0401 14:09:49.365617 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0401 14:09:49.365628 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.307692 | |
| I0401 14:09:49.365641 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.03064 (* 1 = 3.03064 loss) | |
| I0401 14:09:49.365655 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.95672 (* 1 = 0.95672 loss) | |
| I0401 14:09:49.365667 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:09:49.365679 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0004928 | |
| I0401 14:09:49.365690 6134 sgd_solver.cpp:106] Iteration 19500, lr = 0.01 | |
| I0401 14:11:57.720243 6134 solver.cpp:338] Iteration 20000, Testing net (#0) | |
| I0401 14:12:27.450562 6134 solver.cpp:393] Test loss: 5.67653 | |
| I0401 14:12:27.450604 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.0984261 | |
| I0401 14:12:27.450621 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.778317 | |
| I0401 14:12:27.450634 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.316652 | |
| I0401 14:12:27.450649 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.89744 (* 0.3 = 0.869232 loss) | |
| I0401 14:12:27.450664 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.736694 (* 0.3 = 0.221008 loss) | |
| I0401 14:12:27.450675 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.101628 | |
| I0401 14:12:27.450687 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.780136 | |
| I0401 14:12:27.450698 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.330421 | |
| I0401 14:12:27.450712 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 3.02979 (* 0.3 = 0.908937 loss) | |
| I0401 14:12:27.450726 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.761713 (* 0.3 = 0.228514 loss) | |
| I0401 14:12:27.450737 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.14364 | |
| I0401 14:12:27.450749 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.788409 | |
| I0401 14:12:27.450760 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.400337 | |
| I0401 14:12:27.450773 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 2.73885 (* 1 = 2.73885 loss) | |
| I0401 14:12:27.450786 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.709991 (* 1 = 0.709991 loss) | |
| I0401 14:12:27.450798 6134 solver.cpp:406] Test net output #15: total_accuracy = 0 | |
| I0401 14:12:27.450809 6134 solver.cpp:406] Test net output #16: total_confidence = 0.00033956 | |
| I0401 14:12:27.601985 6134 solver.cpp:229] Iteration 20000, loss = 6.24553 | |
| I0401 14:12:27.602025 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.230769 | |
| I0401 14:12:27.602043 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 14:12:27.602056 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.410256 | |
| I0401 14:12:27.602071 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.83642 (* 0.3 = 0.850927 loss) | |
| I0401 14:12:27.602084 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.866708 (* 0.3 = 0.260013 loss) | |
| I0401 14:12:27.602097 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.128205 | |
| I0401 14:12:27.602108 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 14:12:27.602120 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.461538 | |
| I0401 14:12:27.602133 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.66897 (* 0.3 = 0.80069 loss) | |
| I0401 14:12:27.602147 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.788197 (* 0.3 = 0.236459 loss) | |
| I0401 14:12:27.602159 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.282051 | |
| I0401 14:12:27.602171 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.8125 | |
| I0401 14:12:27.602182 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.487179 | |
| I0401 14:12:27.602196 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.51113 (* 1 = 2.51113 loss) | |
| I0401 14:12:27.602210 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.688626 (* 1 = 0.688626 loss) | |
| I0401 14:12:27.602221 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:12:27.602232 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000148568 | |
| I0401 14:12:27.602246 6134 sgd_solver.cpp:106] Iteration 20000, lr = 0.01 | |
| I0401 14:14:35.983271 6134 solver.cpp:229] Iteration 20500, loss = 6.25179 | |
| I0401 14:14:35.983397 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.139535 | |
| I0401 14:14:35.983417 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 14:14:35.983429 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.255814 | |
| I0401 14:14:35.983444 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.09803 (* 0.3 = 0.929408 loss) | |
| I0401 14:14:35.983459 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.896102 (* 0.3 = 0.268831 loss) | |
| I0401 14:14:35.983471 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.139535 | |
| I0401 14:14:35.983484 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 14:14:35.983495 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.232558 | |
| I0401 14:14:35.983510 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.01971 (* 0.3 = 0.905913 loss) | |
| I0401 14:14:35.983530 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.848249 (* 0.3 = 0.254475 loss) | |
| I0401 14:14:35.983551 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.116279 | |
| I0401 14:14:35.983575 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.772727 | |
| I0401 14:14:35.983599 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.395349 | |
| I0401 14:14:35.983623 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.84029 (* 1 = 2.84029 loss) | |
| I0401 14:14:35.983639 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.800029 (* 1 = 0.800029 loss) | |
| I0401 14:14:35.983650 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:14:35.983661 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000774003 | |
| I0401 14:14:35.983675 6134 sgd_solver.cpp:106] Iteration 20500, lr = 0.01 | |
| I0401 14:16:44.109031 6134 solver.cpp:229] Iteration 21000, loss = 6.20462 | |
| I0401 14:16:44.109333 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.097561 | |
| I0401 14:16:44.109354 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 14:16:44.109367 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.243902 | |
| I0401 14:16:44.109383 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.95243 (* 0.3 = 0.885729 loss) | |
| I0401 14:16:44.109397 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.805067 (* 0.3 = 0.24152 loss) | |
| I0401 14:16:44.109410 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.121951 | |
| I0401 14:16:44.109423 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 14:16:44.109434 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.243902 | |
| I0401 14:16:44.109448 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.95422 (* 0.3 = 0.886265 loss) | |
| I0401 14:16:44.109470 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.754996 (* 0.3 = 0.226499 loss) | |
| I0401 14:16:44.109488 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.195122 | |
| I0401 14:16:44.109500 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0401 14:16:44.109511 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.341463 | |
| I0401 14:16:44.109527 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.78921 (* 1 = 2.78921 loss) | |
| I0401 14:16:44.109541 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.746161 (* 1 = 0.746161 loss) | |
| I0401 14:16:44.109554 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:16:44.109565 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000348426 | |
| I0401 14:16:44.109577 6134 sgd_solver.cpp:106] Iteration 21000, lr = 0.01 | |
| I0401 14:18:52.370321 6134 solver.cpp:229] Iteration 21500, loss = 6.19981 | |
| I0401 14:18:52.370440 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.08 | |
| I0401 14:18:52.370460 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0401 14:18:52.370473 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.32 | |
| I0401 14:18:52.370489 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.16668 (* 0.3 = 0.950004 loss) | |
| I0401 14:18:52.370503 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.942075 (* 0.3 = 0.282622 loss) | |
| I0401 14:18:52.370515 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.06 | |
| I0401 14:18:52.370528 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.732955 | |
| I0401 14:18:52.370540 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.24 | |
| I0401 14:18:52.370553 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.09341 (* 0.3 = 0.928023 loss) | |
| I0401 14:18:52.370568 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.901903 (* 0.3 = 0.270571 loss) | |
| I0401 14:18:52.370579 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.06 | |
| I0401 14:18:52.370591 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.727273 | |
| I0401 14:18:52.370602 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.38 | |
| I0401 14:18:52.370615 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.9485 (* 1 = 2.9485 loss) | |
| I0401 14:18:52.370630 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.888137 (* 1 = 0.888137 loss) | |
| I0401 14:18:52.370641 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:18:52.370652 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000840563 | |
| I0401 14:18:52.370663 6134 sgd_solver.cpp:106] Iteration 21500, lr = 0.01 | |
| I0401 14:21:00.918875 6134 solver.cpp:229] Iteration 22000, loss = 6.19478 | |
| I0401 14:21:00.919011 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0851064 | |
| I0401 14:21:00.919033 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 14:21:00.919044 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.234043 | |
| I0401 14:21:00.919060 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.16361 (* 0.3 = 0.949083 loss) | |
| I0401 14:21:00.919075 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.87757 (* 0.3 = 0.263271 loss) | |
| I0401 14:21:00.919087 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0851064 | |
| I0401 14:21:00.919100 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 14:21:00.919111 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.212766 | |
| I0401 14:21:00.919124 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.23011 (* 0.3 = 0.969033 loss) | |
| I0401 14:21:00.919138 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.892032 (* 0.3 = 0.26761 loss) | |
| I0401 14:21:00.919150 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.148936 | |
| I0401 14:21:00.919162 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.772727 | |
| I0401 14:21:00.919173 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.276596 | |
| I0401 14:21:00.919186 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.08438 (* 1 = 3.08438 loss) | |
| I0401 14:21:00.919200 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.840522 (* 1 = 0.840522 loss) | |
| I0401 14:21:00.919211 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:21:00.919222 6134 solver.cpp:245] Train net output #16: total_confidence = 2.36222e-05 | |
| I0401 14:21:00.919234 6134 sgd_solver.cpp:106] Iteration 22000, lr = 0.01 | |
| I0401 14:23:09.300401 6134 solver.cpp:229] Iteration 22500, loss = 6.09174 | |
| I0401 14:23:09.300518 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0833333 | |
| I0401 14:23:09.300537 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0401 14:23:09.300549 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.354167 | |
| I0401 14:23:09.300565 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.19377 (* 0.3 = 0.95813 loss) | |
| I0401 14:23:09.300580 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.941009 (* 0.3 = 0.282303 loss) | |
| I0401 14:23:09.300592 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.145833 | |
| I0401 14:23:09.300604 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 14:23:09.300616 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.270833 | |
| I0401 14:23:09.300628 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.17031 (* 0.3 = 0.951094 loss) | |
| I0401 14:23:09.300642 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.936717 (* 0.3 = 0.281015 loss) | |
| I0401 14:23:09.300654 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.104167 | |
| I0401 14:23:09.300667 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0401 14:23:09.300678 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.375 | |
| I0401 14:23:09.300691 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.91858 (* 1 = 2.91858 loss) | |
| I0401 14:23:09.300705 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.866064 (* 1 = 0.866064 loss) | |
| I0401 14:23:09.300717 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:23:09.300729 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00235747 | |
| I0401 14:23:09.300740 6134 sgd_solver.cpp:106] Iteration 22500, lr = 0.01 | |
| I0401 14:25:17.690780 6134 solver.cpp:229] Iteration 23000, loss = 6.07299 | |
| I0401 14:25:17.690929 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.173077 | |
| I0401 14:25:17.690950 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0401 14:25:17.690963 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.288462 | |
| I0401 14:25:17.690979 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.1844 (* 0.3 = 0.95532 loss) | |
| I0401 14:25:17.690994 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.03574 (* 0.3 = 0.310723 loss) | |
| I0401 14:25:17.691006 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0769231 | |
| I0401 14:25:17.691020 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.727273 | |
| I0401 14:25:17.691031 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.346154 | |
| I0401 14:25:17.691045 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.02355 (* 0.3 = 0.907065 loss) | |
| I0401 14:25:17.691059 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.939014 (* 0.3 = 0.281704 loss) | |
| I0401 14:25:17.691071 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.153846 | |
| I0401 14:25:17.691082 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.75 | |
| I0401 14:25:17.691094 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.288462 | |
| I0401 14:25:17.691108 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.01952 (* 1 = 3.01952 loss) | |
| I0401 14:25:17.691123 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.940993 (* 1 = 0.940993 loss) | |
| I0401 14:25:17.691133 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:25:17.691145 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000598684 | |
| I0401 14:25:17.691157 6134 sgd_solver.cpp:106] Iteration 23000, lr = 0.01 | |
| I0401 14:27:25.961655 6134 solver.cpp:229] Iteration 23500, loss = 6.12305 | |
| I0401 14:27:25.961863 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0666667 | |
| I0401 14:27:25.961880 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0401 14:27:25.961894 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.244444 | |
| I0401 14:27:25.961908 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.20304 (* 0.3 = 0.960912 loss) | |
| I0401 14:27:25.961922 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.945829 (* 0.3 = 0.283749 loss) | |
| I0401 14:27:25.961935 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.133333 | |
| I0401 14:27:25.961947 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 14:27:25.961958 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.355556 | |
| I0401 14:27:25.961972 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.18468 (* 0.3 = 0.955406 loss) | |
| I0401 14:27:25.961985 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.903248 (* 0.3 = 0.270975 loss) | |
| I0401 14:27:25.962000 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.155556 | |
| I0401 14:27:25.962013 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.772727 | |
| I0401 14:27:25.962024 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.244444 | |
| I0401 14:27:25.962038 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.02773 (* 1 = 3.02773 loss) | |
| I0401 14:27:25.962051 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.859375 (* 1 = 0.859375 loss) | |
| I0401 14:27:25.962062 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:27:25.962074 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000242582 | |
| I0401 14:27:25.962086 6134 sgd_solver.cpp:106] Iteration 23500, lr = 0.01 | |
| I0401 14:29:34.544431 6134 solver.cpp:229] Iteration 24000, loss = 6.06155 | |
| I0401 14:29:34.544560 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.115385 | |
| I0401 14:29:34.544580 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0401 14:29:34.544592 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.384615 | |
| I0401 14:29:34.544607 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.06231 (* 0.3 = 0.918694 loss) | |
| I0401 14:29:34.544622 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.966566 (* 0.3 = 0.28997 loss) | |
| I0401 14:29:34.544634 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.153846 | |
| I0401 14:29:34.544646 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0401 14:29:34.544658 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.423077 | |
| I0401 14:29:34.544672 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.83202 (* 0.3 = 0.849606 loss) | |
| I0401 14:29:34.544687 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.897475 (* 0.3 = 0.269243 loss) | |
| I0401 14:29:34.544698 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.25 | |
| I0401 14:29:34.544710 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0401 14:29:34.544721 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.423077 | |
| I0401 14:29:34.544735 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.55844 (* 1 = 2.55844 loss) | |
| I0401 14:29:34.544749 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.847198 (* 1 = 0.847198 loss) | |
| I0401 14:29:34.544760 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:29:34.544772 6134 solver.cpp:245] Train net output #16: total_confidence = 8.74741e-05 | |
| I0401 14:29:34.544785 6134 sgd_solver.cpp:106] Iteration 24000, lr = 0.01 | |
| I0401 14:31:43.142940 6134 solver.cpp:229] Iteration 24500, loss = 6.01358 | |
| I0401 14:31:43.143038 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.145833 | |
| I0401 14:31:43.143057 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 14:31:43.143070 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.354167 | |
| I0401 14:31:43.143085 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.88606 (* 0.3 = 0.865819 loss) | |
| I0401 14:31:43.143102 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.874057 (* 0.3 = 0.262217 loss) | |
| I0401 14:31:43.143115 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.125 | |
| I0401 14:31:43.143127 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 14:31:43.143139 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.395833 | |
| I0401 14:31:43.143153 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.87254 (* 0.3 = 0.861763 loss) | |
| I0401 14:31:43.143167 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.86401 (* 0.3 = 0.259203 loss) | |
| I0401 14:31:43.143179 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.145833 | |
| I0401 14:31:43.143190 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0401 14:31:43.143203 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.395833 | |
| I0401 14:31:43.143216 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.70178 (* 1 = 2.70178 loss) | |
| I0401 14:31:43.143229 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.765352 (* 1 = 0.765352 loss) | |
| I0401 14:31:43.143241 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:31:43.143252 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00135932 | |
| I0401 14:31:43.143265 6134 sgd_solver.cpp:106] Iteration 24500, lr = 0.01 | |
| I0401 14:33:51.404153 6134 solver.cpp:338] Iteration 25000, Testing net (#0) | |
| I0401 14:34:20.895989 6134 solver.cpp:393] Test loss: 5.27374 | |
| I0401 14:34:20.896034 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.145434 | |
| I0401 14:34:20.896049 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.789273 | |
| I0401 14:34:20.896061 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.380587 | |
| I0401 14:34:20.896076 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.8704 (* 0.3 = 0.861121 loss) | |
| I0401 14:34:20.896091 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.721657 (* 0.3 = 0.216497 loss) | |
| I0401 14:34:20.896103 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.15224 | |
| I0401 14:34:20.896116 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.790545 | |
| I0401 14:34:20.896126 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.402902 | |
| I0401 14:34:20.896139 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 2.79291 (* 0.3 = 0.837874 loss) | |
| I0401 14:34:20.896152 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.709173 (* 0.3 = 0.212752 loss) | |
| I0401 14:34:20.896164 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.246248 | |
| I0401 14:34:20.896176 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.808045 | |
| I0401 14:34:20.896188 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.502916 | |
| I0401 14:34:20.896200 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 2.48413 (* 1 = 2.48413 loss) | |
| I0401 14:34:20.896214 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.661377 (* 1 = 0.661377 loss) | |
| I0401 14:34:20.896225 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.001 | |
| I0401 14:34:20.896236 6134 solver.cpp:406] Test net output #16: total_confidence = 0.00168814 | |
| I0401 14:34:21.046411 6134 solver.cpp:229] Iteration 25000, loss = 6.04105 | |
| I0401 14:34:21.046453 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.190476 | |
| I0401 14:34:21.046470 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 14:34:21.046483 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.261905 | |
| I0401 14:34:21.046497 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.94503 (* 0.3 = 0.883511 loss) | |
| I0401 14:34:21.046512 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.824608 (* 0.3 = 0.247382 loss) | |
| I0401 14:34:21.046525 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.166667 | |
| I0401 14:34:21.046536 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 14:34:21.046548 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.285714 | |
| I0401 14:34:21.046561 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.94304 (* 0.3 = 0.882913 loss) | |
| I0401 14:34:21.046576 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.844135 (* 0.3 = 0.25324 loss) | |
| I0401 14:34:21.046587 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.166667 | |
| I0401 14:34:21.046599 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0401 14:34:21.046614 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.404762 | |
| I0401 14:34:21.046628 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.71409 (* 1 = 2.71409 loss) | |
| I0401 14:34:21.046643 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.838025 (* 1 = 0.838025 loss) | |
| I0401 14:34:21.046654 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:34:21.046665 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000377854 | |
| I0401 14:34:21.046677 6134 sgd_solver.cpp:106] Iteration 25000, lr = 0.01 | |
| I0401 14:36:29.520390 6134 solver.cpp:229] Iteration 25500, loss = 5.95791 | |
| I0401 14:36:29.520671 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.155556 | |
| I0401 14:36:29.520691 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 14:36:29.520704 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.422222 | |
| I0401 14:36:29.520720 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.82395 (* 0.3 = 0.847185 loss) | |
| I0401 14:36:29.520735 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.826006 (* 0.3 = 0.247802 loss) | |
| I0401 14:36:29.520746 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.222222 | |
| I0401 14:36:29.520758 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 14:36:29.520771 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.4 | |
| I0401 14:36:29.520783 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.74795 (* 0.3 = 0.824386 loss) | |
| I0401 14:36:29.520797 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.809584 (* 0.3 = 0.242875 loss) | |
| I0401 14:36:29.520809 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.288889 | |
| I0401 14:36:29.520822 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.806818 | |
| I0401 14:36:29.520833 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.577778 | |
| I0401 14:36:29.520846 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.45834 (* 1 = 2.45834 loss) | |
| I0401 14:36:29.520859 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.724035 (* 1 = 0.724035 loss) | |
| I0401 14:36:29.520871 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:36:29.520882 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000420386 | |
| I0401 14:36:29.520895 6134 sgd_solver.cpp:106] Iteration 25500, lr = 0.01 | |
| I0401 14:38:37.940440 6134 solver.cpp:229] Iteration 26000, loss = 5.91521 | |
| I0401 14:38:37.940551 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.16 | |
| I0401 14:38:37.940572 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 14:38:37.940584 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.38 | |
| I0401 14:38:37.940599 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.86079 (* 0.3 = 0.858236 loss) | |
| I0401 14:38:37.940614 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.860673 (* 0.3 = 0.258202 loss) | |
| I0401 14:38:37.940626 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.18 | |
| I0401 14:38:37.940639 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 14:38:37.940649 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.42 | |
| I0401 14:38:37.940664 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.83694 (* 0.3 = 0.851081 loss) | |
| I0401 14:38:37.940676 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.881951 (* 0.3 = 0.264585 loss) | |
| I0401 14:38:37.940688 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.16 | |
| I0401 14:38:37.940701 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0401 14:38:37.940711 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.42 | |
| I0401 14:38:37.940724 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.64177 (* 1 = 2.64177 loss) | |
| I0401 14:38:37.940738 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.78255 (* 1 = 0.78255 loss) | |
| I0401 14:38:37.940749 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:38:37.940760 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000234915 | |
| I0401 14:38:37.940773 6134 sgd_solver.cpp:106] Iteration 26000, lr = 0.01 | |
| I0401 14:40:46.329681 6134 solver.cpp:229] Iteration 26500, loss = 5.91922 | |
| I0401 14:40:46.329821 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.142857 | |
| I0401 14:40:46.329849 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 14:40:46.329862 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.357143 | |
| I0401 14:40:46.329879 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.06536 (* 0.3 = 0.919609 loss) | |
| I0401 14:40:46.329892 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.92049 (* 0.3 = 0.276147 loss) | |
| I0401 14:40:46.329907 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.190476 | |
| I0401 14:40:46.329931 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0401 14:40:46.329951 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.285714 | |
| I0401 14:40:46.329964 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.98299 (* 0.3 = 0.894898 loss) | |
| I0401 14:40:46.329979 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.877862 (* 0.3 = 0.263358 loss) | |
| I0401 14:40:46.329991 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.190476 | |
| I0401 14:40:46.330003 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0401 14:40:46.330014 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.452381 | |
| I0401 14:40:46.330029 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.67546 (* 1 = 2.67546 loss) | |
| I0401 14:40:46.330042 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.849048 (* 1 = 0.849048 loss) | |
| I0401 14:40:46.330054 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:40:46.330065 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000160137 | |
| I0401 14:40:46.330080 6134 sgd_solver.cpp:106] Iteration 26500, lr = 0.01 | |
| I0401 14:42:54.801506 6134 solver.cpp:229] Iteration 27000, loss = 5.86701 | |
| I0401 14:42:54.801605 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.163265 | |
| I0401 14:42:54.801625 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 14:42:54.801636 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.265306 | |
| I0401 14:42:54.801652 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.30597 (* 0.3 = 0.99179 loss) | |
| I0401 14:42:54.801667 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.976866 (* 0.3 = 0.29306 loss) | |
| I0401 14:42:54.801679 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.142857 | |
| I0401 14:42:54.801692 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 14:42:54.801703 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.326531 | |
| I0401 14:42:54.801720 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.0116 (* 0.3 = 0.903481 loss) | |
| I0401 14:42:54.801743 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.912434 (* 0.3 = 0.27373 loss) | |
| I0401 14:42:54.801756 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.122449 | |
| I0401 14:42:54.801769 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0401 14:42:54.801779 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.326531 | |
| I0401 14:42:54.801794 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.78729 (* 1 = 2.78729 loss) | |
| I0401 14:42:54.801807 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.85407 (* 1 = 0.85407 loss) | |
| I0401 14:42:54.801820 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:42:54.801831 6134 solver.cpp:245] Train net output #16: total_confidence = 4.12092e-05 | |
| I0401 14:42:54.801842 6134 sgd_solver.cpp:106] Iteration 27000, lr = 0.01 | |
| I0401 14:45:03.017666 6134 solver.cpp:229] Iteration 27500, loss = 5.82302 | |
| I0401 14:45:03.017801 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.133333 | |
| I0401 14:45:03.017822 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 14:45:03.017833 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.266667 | |
| I0401 14:45:03.017849 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.9734 (* 0.3 = 0.892021 loss) | |
| I0401 14:45:03.017863 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.898534 (* 0.3 = 0.26956 loss) | |
| I0401 14:45:03.017875 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.133333 | |
| I0401 14:45:03.017889 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 14:45:03.017900 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.288889 | |
| I0401 14:45:03.017913 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.98951 (* 0.3 = 0.896852 loss) | |
| I0401 14:45:03.017927 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.892923 (* 0.3 = 0.267877 loss) | |
| I0401 14:45:03.017942 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.133333 | |
| I0401 14:45:03.017966 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.755682 | |
| I0401 14:45:03.017987 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.444444 | |
| I0401 14:45:03.018013 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.61492 (* 1 = 2.61492 loss) | |
| I0401 14:45:03.018041 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.77947 (* 1 = 0.77947 loss) | |
| I0401 14:45:03.018057 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:45:03.018069 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000228397 | |
| I0401 14:45:03.018081 6134 sgd_solver.cpp:106] Iteration 27500, lr = 0.01 | |
| I0401 14:47:11.444795 6134 solver.cpp:229] Iteration 28000, loss = 5.78782 | |
| I0401 14:47:11.445057 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.219512 | |
| I0401 14:47:11.445082 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 14:47:11.445096 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.390244 | |
| I0401 14:47:11.445111 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.62532 (* 0.3 = 0.787596 loss) | |
| I0401 14:47:11.445125 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.711521 (* 0.3 = 0.213456 loss) | |
| I0401 14:47:11.445137 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.170732 | |
| I0401 14:47:11.445153 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 14:47:11.445165 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.365854 | |
| I0401 14:47:11.445179 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.63961 (* 0.3 = 0.791882 loss) | |
| I0401 14:47:11.445194 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.741352 (* 0.3 = 0.222406 loss) | |
| I0401 14:47:11.445205 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.317073 | |
| I0401 14:47:11.445219 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0401 14:47:11.445240 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.609756 | |
| I0401 14:47:11.445256 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.45643 (* 1 = 2.45643 loss) | |
| I0401 14:47:11.445271 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.670944 (* 1 = 0.670944 loss) | |
| I0401 14:47:11.445282 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:47:11.445294 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00229379 | |
| I0401 14:47:11.445307 6134 sgd_solver.cpp:106] Iteration 28000, lr = 0.01 | |
| I0401 14:49:19.765692 6134 solver.cpp:229] Iteration 28500, loss = 5.77501 | |
| I0401 14:49:19.765820 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.222222 | |
| I0401 14:49:19.765839 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 14:49:19.765851 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.333333 | |
| I0401 14:49:19.765867 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.98837 (* 0.3 = 0.89651 loss) | |
| I0401 14:49:19.765882 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.88875 (* 0.3 = 0.266625 loss) | |
| I0401 14:49:19.765895 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.222222 | |
| I0401 14:49:19.765907 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 14:49:19.765918 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.444444 | |
| I0401 14:49:19.765931 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.65748 (* 0.3 = 0.797245 loss) | |
| I0401 14:49:19.765945 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.809635 (* 0.3 = 0.24289 loss) | |
| I0401 14:49:19.765957 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.311111 | |
| I0401 14:49:19.765969 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.806818 | |
| I0401 14:49:19.765980 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.511111 | |
| I0401 14:49:19.765995 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.50959 (* 1 = 2.50959 loss) | |
| I0401 14:49:19.766007 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.741936 (* 1 = 0.741936 loss) | |
| I0401 14:49:19.766019 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:49:19.766031 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000215916 | |
| I0401 14:49:19.766043 6134 sgd_solver.cpp:106] Iteration 28500, lr = 0.01 | |
| I0401 14:51:28.217803 6134 solver.cpp:229] Iteration 29000, loss = 5.77616 | |
| I0401 14:51:28.217921 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0909091 | |
| I0401 14:51:28.217943 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.715909 | |
| I0401 14:51:28.217957 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.363636 | |
| I0401 14:51:28.217972 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.98137 (* 0.3 = 0.894411 loss) | |
| I0401 14:51:28.217986 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.966119 (* 0.3 = 0.289836 loss) | |
| I0401 14:51:28.217998 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.127273 | |
| I0401 14:51:28.218010 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.727273 | |
| I0401 14:51:28.218022 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.327273 | |
| I0401 14:51:28.218035 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.95539 (* 0.3 = 0.886617 loss) | |
| I0401 14:51:28.218050 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.947604 (* 0.3 = 0.284281 loss) | |
| I0401 14:51:28.218061 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.181818 | |
| I0401 14:51:28.218075 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0401 14:51:28.218087 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.509091 | |
| I0401 14:51:28.218101 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.63932 (* 1 = 2.63932 loss) | |
| I0401 14:51:28.218114 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.858855 (* 1 = 0.858855 loss) | |
| I0401 14:51:28.218127 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:51:28.218137 6134 solver.cpp:245] Train net output #16: total_confidence = 5.6095e-05 | |
| I0401 14:51:28.218149 6134 sgd_solver.cpp:106] Iteration 29000, lr = 0.01 | |
| I0401 14:53:36.570986 6134 solver.cpp:229] Iteration 29500, loss = 5.68008 | |
| I0401 14:53:36.571118 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.153846 | |
| I0401 14:53:36.571140 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 14:53:36.571151 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.410256 | |
| I0401 14:53:36.571167 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.82276 (* 0.3 = 0.846828 loss) | |
| I0401 14:53:36.571182 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.825591 (* 0.3 = 0.247677 loss) | |
| I0401 14:53:36.571193 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.179487 | |
| I0401 14:53:36.571207 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0401 14:53:36.571218 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.487179 | |
| I0401 14:53:36.571231 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.93953 (* 0.3 = 0.881858 loss) | |
| I0401 14:53:36.571244 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.849905 (* 0.3 = 0.254972 loss) | |
| I0401 14:53:36.571257 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.282051 | |
| I0401 14:53:36.571269 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0401 14:53:36.571280 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.512821 | |
| I0401 14:53:36.571293 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.35164 (* 1 = 2.35164 loss) | |
| I0401 14:53:36.571306 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.727965 (* 1 = 0.727965 loss) | |
| I0401 14:53:36.571318 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:53:36.571331 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00139026 | |
| I0401 14:53:36.571342 6134 sgd_solver.cpp:106] Iteration 29500, lr = 0.01 | |
| I0401 14:55:44.894884 6134 solver.cpp:338] Iteration 30000, Testing net (#0) | |
| I0401 14:56:14.640405 6134 solver.cpp:393] Test loss: 5.12796 | |
| I0401 14:56:14.640450 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.164273 | |
| I0401 14:56:14.640465 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.790363 | |
| I0401 14:56:14.640477 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.398009 | |
| I0401 14:56:14.640492 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.77347 (* 0.3 = 0.832041 loss) | |
| I0401 14:56:14.640507 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.739214 (* 0.3 = 0.221764 loss) | |
| I0401 14:56:14.640522 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.201178 | |
| I0401 14:56:14.640535 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.80091 | |
| I0401 14:56:14.640547 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.488425 | |
| I0401 14:56:14.640560 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 2.62329 (* 0.3 = 0.786988 loss) | |
| I0401 14:56:14.640574 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.687587 (* 0.3 = 0.206276 loss) | |
| I0401 14:56:14.640585 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.276553 | |
| I0401 14:56:14.640597 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.811319 | |
| I0401 14:56:14.640609 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.576795 | |
| I0401 14:56:14.640621 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 2.4214 (* 1 = 2.4214 loss) | |
| I0401 14:56:14.640635 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.659487 (* 1 = 0.659487 loss) | |
| I0401 14:56:14.640646 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.002 | |
| I0401 14:56:14.640657 6134 solver.cpp:406] Test net output #16: total_confidence = 0.00351599 | |
| I0401 14:56:14.791204 6134 solver.cpp:229] Iteration 30000, loss = 5.64372 | |
| I0401 14:56:14.791244 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0697674 | |
| I0401 14:56:14.791260 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 14:56:14.791272 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.232558 | |
| I0401 14:56:14.791287 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.93291 (* 0.3 = 0.879872 loss) | |
| I0401 14:56:14.791301 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.803378 (* 0.3 = 0.241014 loss) | |
| I0401 14:56:14.791314 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.139535 | |
| I0401 14:56:14.791327 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 14:56:14.791339 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.372093 | |
| I0401 14:56:14.791352 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.73877 (* 0.3 = 0.821631 loss) | |
| I0401 14:56:14.791366 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.80316 (* 0.3 = 0.240948 loss) | |
| I0401 14:56:14.791378 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.186047 | |
| I0401 14:56:14.791390 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0401 14:56:14.791401 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.44186 | |
| I0401 14:56:14.791415 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.52431 (* 1 = 2.52431 loss) | |
| I0401 14:56:14.791429 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.688534 (* 1 = 0.688534 loss) | |
| I0401 14:56:14.791440 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:56:14.791451 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00207242 | |
| I0401 14:56:14.791465 6134 sgd_solver.cpp:106] Iteration 30000, lr = 0.01 | |
| I0401 14:58:23.067124 6134 solver.cpp:229] Iteration 30500, loss = 5.61595 | |
| I0401 14:58:23.067430 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.142857 | |
| I0401 14:58:23.067451 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 14:58:23.067463 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.404762 | |
| I0401 14:58:23.067478 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.89773 (* 0.3 = 0.869321 loss) | |
| I0401 14:58:23.067492 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.775126 (* 0.3 = 0.232538 loss) | |
| I0401 14:58:23.067505 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.166667 | |
| I0401 14:58:23.067520 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 14:58:23.067533 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.428571 | |
| I0401 14:58:23.067546 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.93739 (* 0.3 = 0.881217 loss) | |
| I0401 14:58:23.067561 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.769859 (* 0.3 = 0.230958 loss) | |
| I0401 14:58:23.067574 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.214286 | |
| I0401 14:58:23.067584 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0401 14:58:23.067595 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.428571 | |
| I0401 14:58:23.067610 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.65376 (* 1 = 2.65376 loss) | |
| I0401 14:58:23.067622 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.71221 (* 1 = 0.71221 loss) | |
| I0401 14:58:23.067634 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 14:58:23.067646 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00202828 | |
| I0401 14:58:23.067657 6134 sgd_solver.cpp:106] Iteration 30500, lr = 0.01 | |
| I0401 15:00:31.577500 6134 solver.cpp:229] Iteration 31000, loss = 5.63256 | |
| I0401 15:00:31.577646 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0740741 | |
| I0401 15:00:31.577666 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.715909 | |
| I0401 15:00:31.577679 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.277778 | |
| I0401 15:00:31.577695 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.0718 (* 0.3 = 0.92154 loss) | |
| I0401 15:00:31.577710 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.976732 (* 0.3 = 0.29302 loss) | |
| I0401 15:00:31.577723 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.12963 | |
| I0401 15:00:31.577735 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.732955 | |
| I0401 15:00:31.577746 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.240741 | |
| I0401 15:00:31.577761 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.16537 (* 0.3 = 0.949611 loss) | |
| I0401 15:00:31.577775 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.995386 (* 0.3 = 0.298616 loss) | |
| I0401 15:00:31.577786 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.12963 | |
| I0401 15:00:31.577798 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.727273 | |
| I0401 15:00:31.577811 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.296296 | |
| I0401 15:00:31.577823 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.03443 (* 1 = 3.03443 loss) | |
| I0401 15:00:31.577837 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.966408 (* 1 = 0.966408 loss) | |
| I0401 15:00:31.577849 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:00:31.577860 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000221629 | |
| I0401 15:00:31.577872 6134 sgd_solver.cpp:106] Iteration 31000, lr = 0.01 | |
| I0401 15:02:40.168670 6134 solver.cpp:229] Iteration 31500, loss = 5.60251 | |
| I0401 15:02:40.168787 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.116279 | |
| I0401 15:02:40.168805 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 15:02:40.168817 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.232558 | |
| I0401 15:02:40.168833 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.08718 (* 0.3 = 0.926154 loss) | |
| I0401 15:02:40.168848 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.836157 (* 0.3 = 0.250847 loss) | |
| I0401 15:02:40.168860 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.116279 | |
| I0401 15:02:40.168872 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 15:02:40.168884 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.348837 | |
| I0401 15:02:40.168897 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.90159 (* 0.3 = 0.870477 loss) | |
| I0401 15:02:40.168911 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.832883 (* 0.3 = 0.249865 loss) | |
| I0401 15:02:40.168925 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.186047 | |
| I0401 15:02:40.168936 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0401 15:02:40.168947 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.465116 | |
| I0401 15:02:40.168962 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.615 (* 1 = 2.615 loss) | |
| I0401 15:02:40.168974 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.729809 (* 1 = 0.729809 loss) | |
| I0401 15:02:40.168987 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:02:40.168998 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000354832 | |
| I0401 15:02:40.169009 6134 sgd_solver.cpp:106] Iteration 31500, lr = 0.01 | |
| I0401 15:04:48.541823 6134 solver.cpp:229] Iteration 32000, loss = 5.61025 | |
| I0401 15:04:48.541962 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.170213 | |
| I0401 15:04:48.541982 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 15:04:48.541995 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.446809 | |
| I0401 15:04:48.542009 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.70011 (* 0.3 = 0.810033 loss) | |
| I0401 15:04:48.542024 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.868042 (* 0.3 = 0.260413 loss) | |
| I0401 15:04:48.542037 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.170213 | |
| I0401 15:04:48.542048 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0401 15:04:48.542060 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.446809 | |
| I0401 15:04:48.542073 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.64557 (* 0.3 = 0.793671 loss) | |
| I0401 15:04:48.542088 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.838326 (* 0.3 = 0.251498 loss) | |
| I0401 15:04:48.542099 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.255319 | |
| I0401 15:04:48.542111 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.767045 | |
| I0401 15:04:48.542122 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.553191 | |
| I0401 15:04:48.542136 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.31058 (* 1 = 2.31058 loss) | |
| I0401 15:04:48.542150 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.722816 (* 1 = 0.722816 loss) | |
| I0401 15:04:48.542162 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:04:48.542173 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000453888 | |
| I0401 15:04:48.542186 6134 sgd_solver.cpp:106] Iteration 32000, lr = 0.01 | |
| I0401 15:06:56.997766 6134 solver.cpp:229] Iteration 32500, loss = 5.52099 | |
| I0401 15:06:56.998049 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.037037 | |
| I0401 15:06:56.998072 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.704545 | |
| I0401 15:06:56.998086 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.240741 | |
| I0401 15:06:56.998103 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.24351 (* 0.3 = 0.973052 loss) | |
| I0401 15:06:56.998118 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.02368 (* 0.3 = 0.307105 loss) | |
| I0401 15:06:56.998131 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.12963 | |
| I0401 15:06:56.998143 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.732955 | |
| I0401 15:06:56.998155 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.240741 | |
| I0401 15:06:56.998168 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.16876 (* 0.3 = 0.950628 loss) | |
| I0401 15:06:56.998183 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.999521 (* 0.3 = 0.299856 loss) | |
| I0401 15:06:56.998195 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.148148 | |
| I0401 15:06:56.998208 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0401 15:06:56.998219 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.407407 | |
| I0401 15:06:56.998232 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.08366 (* 1 = 3.08366 loss) | |
| I0401 15:06:56.998246 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.967447 (* 1 = 0.967447 loss) | |
| I0401 15:06:56.998258 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:06:56.998270 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000333085 | |
| I0401 15:06:56.998282 6134 sgd_solver.cpp:106] Iteration 32500, lr = 0.01 | |
| I0401 15:09:05.352882 6134 solver.cpp:229] Iteration 33000, loss = 5.48267 | |
| I0401 15:09:05.353021 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.133333 | |
| I0401 15:09:05.353042 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 15:09:05.353055 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.355556 | |
| I0401 15:09:05.353071 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.0237 (* 0.3 = 0.90711 loss) | |
| I0401 15:09:05.353088 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.811496 (* 0.3 = 0.243449 loss) | |
| I0401 15:09:05.353101 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.133333 | |
| I0401 15:09:05.353113 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 15:09:05.353126 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.333333 | |
| I0401 15:09:05.353150 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.84536 (* 0.3 = 0.853608 loss) | |
| I0401 15:09:05.353168 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.766473 (* 0.3 = 0.229942 loss) | |
| I0401 15:09:05.353188 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.311111 | |
| I0401 15:09:05.353211 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0401 15:09:05.353231 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.444444 | |
| I0401 15:09:05.353257 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.62487 (* 1 = 2.62487 loss) | |
| I0401 15:09:05.353283 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.702 (* 1 = 0.702 loss) | |
| I0401 15:09:05.353303 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:09:05.353317 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00352263 | |
| I0401 15:09:05.353328 6134 sgd_solver.cpp:106] Iteration 33000, lr = 0.01 | |
| I0401 15:11:13.795368 6134 solver.cpp:229] Iteration 33500, loss = 5.45655 | |
| I0401 15:11:13.795480 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.195652 | |
| I0401 15:11:13.795500 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 15:11:13.795513 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.369565 | |
| I0401 15:11:13.795531 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.67764 (* 0.3 = 0.803291 loss) | |
| I0401 15:11:13.795547 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.808648 (* 0.3 = 0.242594 loss) | |
| I0401 15:11:13.795560 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.173913 | |
| I0401 15:11:13.795572 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 15:11:13.795584 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.478261 | |
| I0401 15:11:13.795598 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.47011 (* 0.3 = 0.741033 loss) | |
| I0401 15:11:13.795611 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.738019 (* 0.3 = 0.221406 loss) | |
| I0401 15:11:13.795624 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.369565 | |
| I0401 15:11:13.795634 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0401 15:11:13.795646 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.543478 | |
| I0401 15:11:13.795660 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.15767 (* 1 = 2.15767 loss) | |
| I0401 15:11:13.795672 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.632344 (* 1 = 0.632344 loss) | |
| I0401 15:11:13.795685 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:11:13.795696 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00165655 | |
| I0401 15:11:13.795708 6134 sgd_solver.cpp:106] Iteration 33500, lr = 0.01 | |
| I0401 15:13:22.152452 6134 solver.cpp:229] Iteration 34000, loss = 5.43822 | |
| I0401 15:13:22.152590 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.12766 | |
| I0401 15:13:22.152611 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0401 15:13:22.152623 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.276596 | |
| I0401 15:13:22.152639 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.56039 (* 0.3 = 1.06812 loss) | |
| I0401 15:13:22.152654 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.07914 (* 0.3 = 0.323743 loss) | |
| I0401 15:13:22.152667 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.106383 | |
| I0401 15:13:22.152679 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0401 15:13:22.152691 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.255319 | |
| I0401 15:13:22.152704 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.45259 (* 0.3 = 1.03578 loss) | |
| I0401 15:13:22.152719 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.0574 (* 0.3 = 0.317221 loss) | |
| I0401 15:13:22.152730 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.234043 | |
| I0401 15:13:22.152742 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.761364 | |
| I0401 15:13:22.152753 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.340426 | |
| I0401 15:13:22.152767 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.0247 (* 1 = 3.0247 loss) | |
| I0401 15:13:22.152781 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.932011 (* 1 = 0.932011 loss) | |
| I0401 15:13:22.152792 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:13:22.152803 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000740599 | |
| I0401 15:13:22.152815 6134 sgd_solver.cpp:106] Iteration 34000, lr = 0.01 | |
| I0401 15:15:30.702716 6134 solver.cpp:229] Iteration 34500, loss = 5.40896 | |
| I0401 15:15:30.702824 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.232558 | |
| I0401 15:15:30.702843 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 15:15:30.702855 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.418605 | |
| I0401 15:15:30.702872 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.62253 (* 0.3 = 0.78676 loss) | |
| I0401 15:15:30.702886 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.704151 (* 0.3 = 0.211245 loss) | |
| I0401 15:15:30.702898 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.186047 | |
| I0401 15:15:30.702911 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0401 15:15:30.702924 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.465116 | |
| I0401 15:15:30.702936 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.58708 (* 0.3 = 0.776124 loss) | |
| I0401 15:15:30.702950 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.699844 (* 0.3 = 0.209953 loss) | |
| I0401 15:15:30.702970 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.465116 | |
| I0401 15:15:30.702993 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0401 15:15:30.703017 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.674419 | |
| I0401 15:15:30.703047 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.83125 (* 1 = 1.83125 loss) | |
| I0401 15:15:30.703063 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.512182 (* 1 = 0.512182 loss) | |
| I0401 15:15:30.703074 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 15:15:30.703086 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0120649 | |
| I0401 15:15:30.703099 6134 sgd_solver.cpp:106] Iteration 34500, lr = 0.01 | |
| I0401 15:17:38.755453 6134 solver.cpp:338] Iteration 35000, Testing net (#0) | |
| I0401 15:18:08.495537 6134 solver.cpp:393] Test loss: 4.61024 | |
| I0401 15:18:08.495594 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.214214 | |
| I0401 15:18:08.495609 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.802909 | |
| I0401 15:18:08.495622 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.480776 | |
| I0401 15:18:08.495640 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.56555 (* 0.3 = 0.769665 loss) | |
| I0401 15:18:08.495654 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.676454 (* 0.3 = 0.202936 loss) | |
| I0401 15:18:08.495666 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.239841 | |
| I0401 15:18:08.495677 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.810637 | |
| I0401 15:18:08.495689 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.529415 | |
| I0401 15:18:08.495702 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 2.50448 (* 0.3 = 0.751343 loss) | |
| I0401 15:18:08.495717 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.642538 (* 0.3 = 0.192762 loss) | |
| I0401 15:18:08.495728 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.384719 | |
| I0401 15:18:08.495739 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.840047 | |
| I0401 15:18:08.495750 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.635139 | |
| I0401 15:18:08.495764 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 2.12172 (* 1 = 2.12172 loss) | |
| I0401 15:18:08.495776 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.571811 (* 1 = 0.571811 loss) | |
| I0401 15:18:08.495789 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.002 | |
| I0401 15:18:08.495800 6134 solver.cpp:406] Test net output #16: total_confidence = 0.00690701 | |
| I0401 15:18:08.646796 6134 solver.cpp:229] Iteration 35000, loss = 5.34427 | |
| I0401 15:18:08.646841 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.186047 | |
| I0401 15:18:08.646857 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 15:18:08.646869 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.395349 | |
| I0401 15:18:08.646884 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.82389 (* 0.3 = 0.847168 loss) | |
| I0401 15:18:08.646898 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.75579 (* 0.3 = 0.226737 loss) | |
| I0401 15:18:08.646910 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.116279 | |
| I0401 15:18:08.646922 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 15:18:08.646934 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.372093 | |
| I0401 15:18:08.646947 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.73159 (* 0.3 = 0.819476 loss) | |
| I0401 15:18:08.646961 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.764398 (* 0.3 = 0.229319 loss) | |
| I0401 15:18:08.646973 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.348837 | |
| I0401 15:18:08.646986 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.829545 | |
| I0401 15:18:08.646997 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.604651 | |
| I0401 15:18:08.647011 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.26298 (* 1 = 2.26298 loss) | |
| I0401 15:18:08.647024 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.622973 (* 1 = 0.622973 loss) | |
| I0401 15:18:08.647037 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:18:08.647050 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00236579 | |
| I0401 15:18:08.647063 6134 sgd_solver.cpp:106] Iteration 35000, lr = 0.01 | |
| I0401 15:20:17.113476 6134 solver.cpp:229] Iteration 35500, loss = 5.34331 | |
| I0401 15:20:17.113610 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0401 15:20:17.113631 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 15:20:17.113643 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.452381 | |
| I0401 15:20:17.113659 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.65528 (* 0.3 = 0.796584 loss) | |
| I0401 15:20:17.113673 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.722775 (* 0.3 = 0.216833 loss) | |
| I0401 15:20:17.113685 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.357143 | |
| I0401 15:20:17.113698 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0401 15:20:17.113709 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.52381 | |
| I0401 15:20:17.113723 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.38596 (* 0.3 = 0.715787 loss) | |
| I0401 15:20:17.113737 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.668123 (* 0.3 = 0.200437 loss) | |
| I0401 15:20:17.113749 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.380952 | |
| I0401 15:20:17.113761 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.829545 | |
| I0401 15:20:17.113772 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.619048 | |
| I0401 15:20:17.113786 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.02598 (* 1 = 2.02598 loss) | |
| I0401 15:20:17.113801 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.550661 (* 1 = 0.550661 loss) | |
| I0401 15:20:17.113812 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:20:17.113823 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00913007 | |
| I0401 15:20:17.113837 6134 sgd_solver.cpp:106] Iteration 35500, lr = 0.01 | |
| I0401 15:22:25.735653 6134 solver.cpp:229] Iteration 36000, loss = 5.33662 | |
| I0401 15:22:25.735754 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.181818 | |
| I0401 15:22:25.735772 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 15:22:25.735785 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.363636 | |
| I0401 15:22:25.735800 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.75668 (* 0.3 = 0.827003 loss) | |
| I0401 15:22:25.735816 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.878282 (* 0.3 = 0.263484 loss) | |
| I0401 15:22:25.735828 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.25 | |
| I0401 15:22:25.735841 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 15:22:25.735852 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.477273 | |
| I0401 15:22:25.735865 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.68576 (* 0.3 = 0.805729 loss) | |
| I0401 15:22:25.735879 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.899154 (* 0.3 = 0.269746 loss) | |
| I0401 15:22:25.735895 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.295455 | |
| I0401 15:22:25.735908 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0401 15:22:25.735918 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.431818 | |
| I0401 15:22:25.735932 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.53234 (* 1 = 2.53234 loss) | |
| I0401 15:22:25.735946 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.774421 (* 1 = 0.774421 loss) | |
| I0401 15:22:25.735959 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:22:25.735970 6134 solver.cpp:245] Train net output #16: total_confidence = 0.000978123 | |
| I0401 15:22:25.735981 6134 sgd_solver.cpp:106] Iteration 36000, lr = 0.01 | |
| I0401 15:24:33.980589 6134 solver.cpp:229] Iteration 36500, loss = 5.24859 | |
| I0401 15:24:33.980729 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.229167 | |
| I0401 15:24:33.980751 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 15:24:33.980764 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.520833 | |
| I0401 15:24:33.980779 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.73328 (* 0.3 = 0.819985 loss) | |
| I0401 15:24:33.980794 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.810041 (* 0.3 = 0.243012 loss) | |
| I0401 15:24:33.980806 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.208333 | |
| I0401 15:24:33.980818 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 15:24:33.980830 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.520833 | |
| I0401 15:24:33.980844 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.64413 (* 0.3 = 0.793238 loss) | |
| I0401 15:24:33.980857 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.776303 (* 0.3 = 0.232891 loss) | |
| I0401 15:24:33.980870 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.375 | |
| I0401 15:24:33.980881 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0401 15:24:33.980893 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.604167 | |
| I0401 15:24:33.980906 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.10785 (* 1 = 2.10785 loss) | |
| I0401 15:24:33.980921 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.63085 (* 1 = 0.63085 loss) | |
| I0401 15:24:33.980932 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:24:33.980943 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00441509 | |
| I0401 15:24:33.980955 6134 sgd_solver.cpp:106] Iteration 36500, lr = 0.01 | |
| I0401 15:26:42.378340 6134 solver.cpp:229] Iteration 37000, loss = 5.24584 | |
| I0401 15:26:42.378602 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0731707 | |
| I0401 15:26:42.378623 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0401 15:26:42.378635 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.292683 | |
| I0401 15:26:42.378651 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.93944 (* 0.3 = 0.881832 loss) | |
| I0401 15:26:42.378665 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.92282 (* 0.3 = 0.276846 loss) | |
| I0401 15:26:42.378677 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.219512 | |
| I0401 15:26:42.378690 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0401 15:26:42.378701 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.390244 | |
| I0401 15:26:42.378715 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.7267 (* 0.3 = 0.81801 loss) | |
| I0401 15:26:42.378729 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.933107 (* 0.3 = 0.279932 loss) | |
| I0401 15:26:42.378741 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.292683 | |
| I0401 15:26:42.378753 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.772727 | |
| I0401 15:26:42.378764 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.512195 | |
| I0401 15:26:42.378777 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.29913 (* 1 = 2.29913 loss) | |
| I0401 15:26:42.378792 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.753465 (* 1 = 0.753465 loss) | |
| I0401 15:26:42.378803 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:26:42.378814 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00611665 | |
| I0401 15:26:42.378826 6134 sgd_solver.cpp:106] Iteration 37000, lr = 0.01 | |
| I0401 15:28:50.807435 6134 solver.cpp:229] Iteration 37500, loss = 5.26914 | |
| I0401 15:28:50.807585 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.204082 | |
| I0401 15:28:50.807605 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 15:28:50.807617 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.408163 | |
| I0401 15:28:50.807633 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.73905 (* 0.3 = 0.821714 loss) | |
| I0401 15:28:50.807648 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.788707 (* 0.3 = 0.236612 loss) | |
| I0401 15:28:50.807660 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.183673 | |
| I0401 15:28:50.807672 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 15:28:50.807684 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.367347 | |
| I0401 15:28:50.807698 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.67966 (* 0.3 = 0.803897 loss) | |
| I0401 15:28:50.807711 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.768196 (* 0.3 = 0.230459 loss) | |
| I0401 15:28:50.807723 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.285714 | |
| I0401 15:28:50.807735 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0401 15:28:50.807746 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.469388 | |
| I0401 15:28:50.807760 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.32343 (* 1 = 2.32343 loss) | |
| I0401 15:28:50.807773 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.66458 (* 1 = 0.66458 loss) | |
| I0401 15:28:50.807785 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:28:50.807797 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0207813 | |
| I0401 15:28:50.807808 6134 sgd_solver.cpp:106] Iteration 37500, lr = 0.01 | |
| I0401 15:30:59.318486 6134 solver.cpp:229] Iteration 38000, loss = 5.17791 | |
| I0401 15:30:59.318586 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.125 | |
| I0401 15:30:59.318604 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.715909 | |
| I0401 15:30:59.318617 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.214286 | |
| I0401 15:30:59.318634 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.05678 (* 0.3 = 0.917034 loss) | |
| I0401 15:30:59.318647 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.02917 (* 0.3 = 0.308751 loss) | |
| I0401 15:30:59.318660 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.160714 | |
| I0401 15:30:59.318672 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.732955 | |
| I0401 15:30:59.318684 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.303571 | |
| I0401 15:30:59.318698 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.05211 (* 0.3 = 0.915632 loss) | |
| I0401 15:30:59.318711 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.997909 (* 0.3 = 0.299373 loss) | |
| I0401 15:30:59.318723 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.196429 | |
| I0401 15:30:59.318735 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0401 15:30:59.318747 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.410714 | |
| I0401 15:30:59.318760 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.70404 (* 1 = 2.70404 loss) | |
| I0401 15:30:59.318773 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.899963 (* 1 = 0.899963 loss) | |
| I0401 15:30:59.318785 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:30:59.318797 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00771179 | |
| I0401 15:30:59.318809 6134 sgd_solver.cpp:106] Iteration 38000, lr = 0.01 | |
| I0401 15:33:07.742674 6134 solver.cpp:229] Iteration 38500, loss = 5.18309 | |
| I0401 15:33:07.742846 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.348837 | |
| I0401 15:33:07.742867 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 15:33:07.742880 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.581395 | |
| I0401 15:33:07.742897 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.26158 (* 0.3 = 0.678473 loss) | |
| I0401 15:33:07.742911 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.616077 (* 0.3 = 0.184823 loss) | |
| I0401 15:33:07.742924 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.302326 | |
| I0401 15:33:07.742936 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0401 15:33:07.742947 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.581395 | |
| I0401 15:33:07.742961 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.29252 (* 0.3 = 0.687756 loss) | |
| I0401 15:33:07.742974 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.652144 (* 0.3 = 0.195643 loss) | |
| I0401 15:33:07.742986 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.395349 | |
| I0401 15:33:07.742998 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.8125 | |
| I0401 15:33:07.743010 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.651163 | |
| I0401 15:33:07.743026 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.70267 (* 1 = 1.70267 loss) | |
| I0401 15:33:07.743039 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.534672 (* 1 = 0.534672 loss) | |
| I0401 15:33:07.743052 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:33:07.743062 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0199046 | |
| I0401 15:33:07.743074 6134 sgd_solver.cpp:106] Iteration 38500, lr = 0.01 | |
| I0401 15:35:16.297148 6134 solver.cpp:229] Iteration 39000, loss = 5.10351 | |
| I0401 15:35:16.297265 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.159091 | |
| I0401 15:35:16.297283 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 15:35:16.297297 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.477273 | |
| I0401 15:35:16.297312 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.58408 (* 0.3 = 0.775225 loss) | |
| I0401 15:35:16.297327 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.727002 (* 0.3 = 0.218101 loss) | |
| I0401 15:35:16.297338 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.25 | |
| I0401 15:35:16.297350 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 15:35:16.297363 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.545455 | |
| I0401 15:35:16.297376 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.31459 (* 0.3 = 0.694377 loss) | |
| I0401 15:35:16.297390 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.67903 (* 0.3 = 0.203709 loss) | |
| I0401 15:35:16.297410 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0401 15:35:16.297427 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0401 15:35:16.297440 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.727273 | |
| I0401 15:35:16.297454 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.87596 (* 1 = 1.87596 loss) | |
| I0401 15:35:16.297467 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.562416 (* 1 = 0.562416 loss) | |
| I0401 15:35:16.297479 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:35:16.297492 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0144186 | |
| I0401 15:35:16.297503 6134 sgd_solver.cpp:106] Iteration 39000, lr = 0.01 | |
| I0401 15:37:24.773236 6134 solver.cpp:229] Iteration 39500, loss = 5.1535 | |
| I0401 15:37:24.773521 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.272727 | |
| I0401 15:37:24.773541 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 15:37:24.773555 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.431818 | |
| I0401 15:37:24.773571 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.50711 (* 0.3 = 0.752133 loss) | |
| I0401 15:37:24.773584 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.681222 (* 0.3 = 0.204367 loss) | |
| I0401 15:37:24.773597 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.318182 | |
| I0401 15:37:24.773610 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0401 15:37:24.773622 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.522727 | |
| I0401 15:37:24.773635 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.2415 (* 0.3 = 0.672449 loss) | |
| I0401 15:37:24.773649 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.640859 (* 0.3 = 0.192258 loss) | |
| I0401 15:37:24.773661 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.431818 | |
| I0401 15:37:24.773674 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0401 15:37:24.773684 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.636364 | |
| I0401 15:37:24.773699 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.83194 (* 1 = 1.83194 loss) | |
| I0401 15:37:24.773712 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.499599 (* 1 = 0.499599 loss) | |
| I0401 15:37:24.773725 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:37:24.773736 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00350031 | |
| I0401 15:37:24.773746 6134 sgd_solver.cpp:106] Iteration 39500, lr = 0.01 | |
| I0401 15:39:33.031548 6134 solver.cpp:338] Iteration 40000, Testing net (#0) | |
| I0401 15:40:02.492645 6134 solver.cpp:393] Test loss: 4.46918 | |
| I0401 15:40:02.492694 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.19482 | |
| I0401 15:40:02.492710 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.796 | |
| I0401 15:40:02.492722 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.468429 | |
| I0401 15:40:02.492738 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.63986 (* 0.3 = 0.791959 loss) | |
| I0401 15:40:02.492753 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.71024 (* 0.3 = 0.213072 loss) | |
| I0401 15:40:02.492763 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.278834 | |
| I0401 15:40:02.492775 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.816637 | |
| I0401 15:40:02.492787 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.593573 | |
| I0401 15:40:02.492800 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 2.35326 (* 0.3 = 0.705978 loss) | |
| I0401 15:40:02.492815 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.624441 (* 0.3 = 0.187332 loss) | |
| I0401 15:40:02.492825 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.416387 | |
| I0401 15:40:02.492838 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.837956 | |
| I0401 15:40:02.492849 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.693108 | |
| I0401 15:40:02.492863 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 2.0022 (* 1 = 2.0022 loss) | |
| I0401 15:40:02.492877 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.56864 (* 1 = 0.56864 loss) | |
| I0401 15:40:02.492888 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.007 | |
| I0401 15:40:02.492907 6134 solver.cpp:406] Test net output #16: total_confidence = 0.014633 | |
| I0401 15:40:02.647099 6134 solver.cpp:229] Iteration 40000, loss = 5.04233 | |
| I0401 15:40:02.647184 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.145833 | |
| I0401 15:40:02.647217 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 15:40:02.647239 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.3125 | |
| I0401 15:40:02.647267 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.02751 (* 0.3 = 0.908252 loss) | |
| I0401 15:40:02.647300 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.889922 (* 0.3 = 0.266977 loss) | |
| I0401 15:40:02.647322 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.229167 | |
| I0401 15:40:02.647346 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 15:40:02.647366 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.4375 | |
| I0401 15:40:02.647392 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.89548 (* 0.3 = 0.868643 loss) | |
| I0401 15:40:02.647416 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.884563 (* 0.3 = 0.265369 loss) | |
| I0401 15:40:02.647438 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.354167 | |
| I0401 15:40:02.647459 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.8125 | |
| I0401 15:40:02.647478 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.541667 | |
| I0401 15:40:02.647503 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.37596 (* 1 = 2.37596 loss) | |
| I0401 15:40:02.647527 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.741458 (* 1 = 0.741458 loss) | |
| I0401 15:40:02.647548 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 15:40:02.647578 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00797599 | |
| I0401 15:40:02.647600 6134 sgd_solver.cpp:106] Iteration 40000, lr = 0.01 | |
| I0401 15:42:10.793766 6134 solver.cpp:229] Iteration 40500, loss = 5.06271 | |
| I0401 15:42:10.793892 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.275 | |
| I0401 15:42:10.793913 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 15:42:10.793926 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55 | |
| I0401 15:42:10.793941 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.5271 (* 0.3 = 0.758131 loss) | |
| I0401 15:42:10.793956 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.692224 (* 0.3 = 0.207667 loss) | |
| I0401 15:42:10.793968 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.275 | |
| I0401 15:42:10.793979 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 15:42:10.793992 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.4 | |
| I0401 15:42:10.794005 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.43301 (* 0.3 = 0.729902 loss) | |
| I0401 15:42:10.794018 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.719813 (* 0.3 = 0.215944 loss) | |
| I0401 15:42:10.794030 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.4 | |
| I0401 15:42:10.794042 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0401 15:42:10.794054 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.575 | |
| I0401 15:42:10.794070 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.13082 (* 1 = 2.13082 loss) | |
| I0401 15:42:10.794085 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.591507 (* 1 = 0.591507 loss) | |
| I0401 15:42:10.794096 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 15:42:10.794107 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00773209 | |
| I0401 15:42:10.794119 6134 sgd_solver.cpp:106] Iteration 40500, lr = 0.01 | |
| I0401 15:44:19.348440 6134 solver.cpp:229] Iteration 41000, loss = 4.98494 | |
| I0401 15:44:19.348590 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.172414 | |
| I0401 15:44:19.348611 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0401 15:44:19.348623 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.344828 | |
| I0401 15:44:19.348639 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.80358 (* 0.3 = 0.841074 loss) | |
| I0401 15:44:19.348654 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.946067 (* 0.3 = 0.28382 loss) | |
| I0401 15:44:19.348666 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.189655 | |
| I0401 15:44:19.348678 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.727273 | |
| I0401 15:44:19.348690 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.396552 | |
| I0401 15:44:19.348703 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.7367 (* 0.3 = 0.821009 loss) | |
| I0401 15:44:19.348717 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.932362 (* 0.3 = 0.279709 loss) | |
| I0401 15:44:19.348729 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.241379 | |
| I0401 15:44:19.348742 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.744318 | |
| I0401 15:44:19.348752 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.517241 | |
| I0401 15:44:19.348765 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.35806 (* 1 = 2.35806 loss) | |
| I0401 15:44:19.348779 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.80588 (* 1 = 0.80588 loss) | |
| I0401 15:44:19.348791 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:44:19.348803 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00203591 | |
| I0401 15:44:19.348815 6134 sgd_solver.cpp:106] Iteration 41000, lr = 0.01 | |
| I0401 15:46:27.807608 6134 solver.cpp:229] Iteration 41500, loss = 4.96139 | |
| I0401 15:46:27.807837 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.24 | |
| I0401 15:46:27.807854 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 15:46:27.807868 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.48 | |
| I0401 15:46:27.807883 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.53109 (* 0.3 = 0.759326 loss) | |
| I0401 15:46:27.807896 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.764471 (* 0.3 = 0.229341 loss) | |
| I0401 15:46:27.807909 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.36 | |
| I0401 15:46:27.807921 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 15:46:27.807932 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6 | |
| I0401 15:46:27.807946 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.34846 (* 0.3 = 0.704539 loss) | |
| I0401 15:46:27.807968 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.700915 (* 0.3 = 0.210275 loss) | |
| I0401 15:46:27.807993 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.42 | |
| I0401 15:46:27.808018 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0401 15:46:27.808039 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.72 | |
| I0401 15:46:27.808055 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.84665 (* 1 = 1.84665 loss) | |
| I0401 15:46:27.808069 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.553904 (* 1 = 0.553904 loss) | |
| I0401 15:46:27.808084 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:46:27.808095 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00925162 | |
| I0401 15:46:27.808107 6134 sgd_solver.cpp:106] Iteration 41500, lr = 0.01 | |
| I0401 15:48:36.375327 6134 solver.cpp:229] Iteration 42000, loss = 4.97625 | |
| I0401 15:48:36.375473 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.125 | |
| I0401 15:48:36.375494 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 15:48:36.375505 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.3 | |
| I0401 15:48:36.375524 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.09904 (* 0.3 = 0.929713 loss) | |
| I0401 15:48:36.375538 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.857041 (* 0.3 = 0.257112 loss) | |
| I0401 15:48:36.375551 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.2 | |
| I0401 15:48:36.375563 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0401 15:48:36.375574 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.425 | |
| I0401 15:48:36.375588 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.97335 (* 0.3 = 0.892006 loss) | |
| I0401 15:48:36.375602 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.887957 (* 0.3 = 0.266387 loss) | |
| I0401 15:48:36.375614 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.25 | |
| I0401 15:48:36.375633 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.778409 | |
| I0401 15:48:36.375644 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.6 | |
| I0401 15:48:36.375658 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.42271 (* 1 = 2.42271 loss) | |
| I0401 15:48:36.375671 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.745218 (* 1 = 0.745218 loss) | |
| I0401 15:48:36.375684 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:48:36.375694 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00355502 | |
| I0401 15:48:36.375706 6134 sgd_solver.cpp:106] Iteration 42000, lr = 0.01 | |
| I0401 15:50:44.551810 6134 solver.cpp:229] Iteration 42500, loss = 4.95348 | |
| I0401 15:50:44.551941 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.229167 | |
| I0401 15:50:44.551960 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 15:50:44.551973 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.416667 | |
| I0401 15:50:44.551988 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.01484 (* 0.3 = 0.904451 loss) | |
| I0401 15:50:44.552002 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.936293 (* 0.3 = 0.280888 loss) | |
| I0401 15:50:44.552016 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.229167 | |
| I0401 15:50:44.552027 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 15:50:44.552038 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.416667 | |
| I0401 15:50:44.552052 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.70974 (* 0.3 = 0.812923 loss) | |
| I0401 15:50:44.552065 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.829557 (* 0.3 = 0.248867 loss) | |
| I0401 15:50:44.552078 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.458333 | |
| I0401 15:50:44.552088 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0401 15:50:44.552100 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.604167 | |
| I0401 15:50:44.552114 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.32564 (* 1 = 2.32564 loss) | |
| I0401 15:50:44.552127 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.727543 (* 1 = 0.727543 loss) | |
| I0401 15:50:44.552139 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:50:44.552150 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00220908 | |
| I0401 15:50:44.552162 6134 sgd_solver.cpp:106] Iteration 42500, lr = 0.01 | |
| I0401 15:52:53.023571 6134 solver.cpp:229] Iteration 43000, loss = 4.91357 | |
| I0401 15:52:53.023792 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.173913 | |
| I0401 15:52:53.023813 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 15:52:53.023826 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.413043 | |
| I0401 15:52:53.023843 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.46927 (* 0.3 = 1.04078 loss) | |
| I0401 15:52:53.023857 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.0098 (* 0.3 = 0.30294 loss) | |
| I0401 15:52:53.023869 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.173913 | |
| I0401 15:52:53.023881 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 15:52:53.023893 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.456522 | |
| I0401 15:52:53.023907 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.51392 (* 0.3 = 1.05418 loss) | |
| I0401 15:52:53.023921 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.991949 (* 0.3 = 0.297585 loss) | |
| I0401 15:52:53.023933 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.23913 | |
| I0401 15:52:53.023946 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.789773 | |
| I0401 15:52:53.023957 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.478261 | |
| I0401 15:52:53.023972 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.9518 (* 1 = 2.9518 loss) | |
| I0401 15:52:53.023985 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.840349 (* 1 = 0.840349 loss) | |
| I0401 15:52:53.023998 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:52:53.024009 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00199914 | |
| I0401 15:52:53.024022 6134 sgd_solver.cpp:106] Iteration 43000, lr = 0.01 | |
| I0401 15:55:01.492164 6134 solver.cpp:229] Iteration 43500, loss = 4.92539 | |
| I0401 15:55:01.492283 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0784314 | |
| I0401 15:55:01.492303 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.715909 | |
| I0401 15:55:01.492316 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.254902 | |
| I0401 15:55:01.492331 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.23261 (* 0.3 = 0.969782 loss) | |
| I0401 15:55:01.492346 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.02998 (* 0.3 = 0.308995 loss) | |
| I0401 15:55:01.492357 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0784314 | |
| I0401 15:55:01.492370 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.715909 | |
| I0401 15:55:01.492382 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.352941 | |
| I0401 15:55:01.492395 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.06792 (* 0.3 = 0.920377 loss) | |
| I0401 15:55:01.492409 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.968393 (* 0.3 = 0.290518 loss) | |
| I0401 15:55:01.492421 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.137255 | |
| I0401 15:55:01.492432 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.738636 | |
| I0401 15:55:01.492450 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.45098 | |
| I0401 15:55:01.492477 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.53561 (* 1 = 2.53561 loss) | |
| I0401 15:55:01.492502 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.816407 (* 1 = 0.816407 loss) | |
| I0401 15:55:01.492529 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:55:01.492550 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0105986 | |
| I0401 15:55:01.492564 6134 sgd_solver.cpp:106] Iteration 43500, lr = 0.01 | |
| I0401 15:57:09.668195 6134 solver.cpp:229] Iteration 44000, loss = 4.86729 | |
| I0401 15:57:09.668462 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.25 | |
| I0401 15:57:09.668481 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 15:57:09.668494 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.522727 | |
| I0401 15:57:09.668509 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.63116 (* 0.3 = 0.789347 loss) | |
| I0401 15:57:09.668526 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.781721 (* 0.3 = 0.234516 loss) | |
| I0401 15:57:09.668540 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.318182 | |
| I0401 15:57:09.668551 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 15:57:09.668563 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.522727 | |
| I0401 15:57:09.668576 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.34666 (* 0.3 = 0.703998 loss) | |
| I0401 15:57:09.668591 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.726103 (* 0.3 = 0.217831 loss) | |
| I0401 15:57:09.668602 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.295455 | |
| I0401 15:57:09.668613 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0401 15:57:09.668625 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.704545 | |
| I0401 15:57:09.668638 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.92371 (* 1 = 1.92371 loss) | |
| I0401 15:57:09.668653 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.592157 (* 1 = 0.592157 loss) | |
| I0401 15:57:09.668664 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:57:09.668675 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0185316 | |
| I0401 15:57:09.668687 6134 sgd_solver.cpp:106] Iteration 44000, lr = 0.01 | |
| I0401 15:59:17.910464 6134 solver.cpp:229] Iteration 44500, loss = 4.79339 | |
| I0401 15:59:17.910581 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.282609 | |
| I0401 15:59:17.910610 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 15:59:17.910631 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.543478 | |
| I0401 15:59:17.910647 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.42638 (* 0.3 = 0.727915 loss) | |
| I0401 15:59:17.910662 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.802901 (* 0.3 = 0.24087 loss) | |
| I0401 15:59:17.910675 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.304348 | |
| I0401 15:59:17.910686 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0401 15:59:17.910698 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.543478 | |
| I0401 15:59:17.910712 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.2567 (* 0.3 = 0.67701 loss) | |
| I0401 15:59:17.910725 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.715648 (* 0.3 = 0.214694 loss) | |
| I0401 15:59:17.910737 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.521739 | |
| I0401 15:59:17.910749 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0401 15:59:17.910760 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.76087 | |
| I0401 15:59:17.910774 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.59357 (* 1 = 1.59357 loss) | |
| I0401 15:59:17.910789 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.501617 (* 1 = 0.501617 loss) | |
| I0401 15:59:17.910800 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 15:59:17.910811 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0562163 | |
| I0401 15:59:17.910823 6134 sgd_solver.cpp:106] Iteration 44500, lr = 0.01 | |
| I0401 16:01:26.292325 6134 solver.cpp:338] Iteration 45000, Testing net (#0) | |
| I0401 16:01:56.068552 6134 solver.cpp:393] Test loss: 4.59382 | |
| I0401 16:01:56.068599 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.216705 | |
| I0401 16:01:56.068616 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.804772 | |
| I0401 16:01:56.068629 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.50792 | |
| I0401 16:01:56.068645 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.59144 (* 0.3 = 0.777431 loss) | |
| I0401 16:01:56.068658 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.664045 (* 0.3 = 0.199214 loss) | |
| I0401 16:01:56.068671 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.262835 | |
| I0401 16:01:56.068682 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.816636 | |
| I0401 16:01:56.068694 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.596615 | |
| I0401 16:01:56.068707 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 2.42824 (* 0.3 = 0.728472 loss) | |
| I0401 16:01:56.068720 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.616192 (* 0.3 = 0.184858 loss) | |
| I0401 16:01:56.068732 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.403984 | |
| I0401 16:01:56.068743 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.851365 | |
| I0401 16:01:56.068754 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.693662 | |
| I0401 16:01:56.068768 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 2.15867 (* 1 = 2.15867 loss) | |
| I0401 16:01:56.068780 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.545178 (* 1 = 0.545178 loss) | |
| I0401 16:01:56.068792 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.022 | |
| I0401 16:01:56.068804 6134 solver.cpp:406] Test net output #16: total_confidence = 0.0515882 | |
| I0401 16:01:56.220038 6134 solver.cpp:229] Iteration 45000, loss = 4.79855 | |
| I0401 16:01:56.220088 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.228571 | |
| I0401 16:01:56.220105 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 16:01:56.220118 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.457143 | |
| I0401 16:01:56.220132 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.72478 (* 0.3 = 0.817435 loss) | |
| I0401 16:01:56.220151 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.730862 (* 0.3 = 0.219259 loss) | |
| I0401 16:01:56.220165 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.314286 | |
| I0401 16:01:56.220176 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0401 16:01:56.220187 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6 | |
| I0401 16:01:56.220201 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.30472 (* 0.3 = 0.691417 loss) | |
| I0401 16:01:56.220216 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.610633 (* 0.3 = 0.18319 loss) | |
| I0401 16:01:56.220228 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.542857 | |
| I0401 16:01:56.220240 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 16:01:56.220252 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.742857 | |
| I0401 16:01:56.220265 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.60836 (* 1 = 1.60836 loss) | |
| I0401 16:01:56.220279 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.454924 (* 1 = 0.454924 loss) | |
| I0401 16:01:56.220291 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:01:56.220304 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0109512 | |
| I0401 16:01:56.220315 6134 sgd_solver.cpp:106] Iteration 45000, lr = 0.01 | |
| I0401 16:04:04.515130 6134 solver.cpp:229] Iteration 45500, loss = 4.77291 | |
| I0401 16:04:04.515256 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.219512 | |
| I0401 16:04:04.515276 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 16:04:04.515288 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.414634 | |
| I0401 16:04:04.515303 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.9329 (* 0.3 = 0.879871 loss) | |
| I0401 16:04:04.515318 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.793017 (* 0.3 = 0.237905 loss) | |
| I0401 16:04:04.515331 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.243902 | |
| I0401 16:04:04.515342 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 16:04:04.515354 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.365854 | |
| I0401 16:04:04.515368 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.88022 (* 0.3 = 0.864066 loss) | |
| I0401 16:04:04.515382 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.791643 (* 0.3 = 0.237493 loss) | |
| I0401 16:04:04.515393 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.243902 | |
| I0401 16:04:04.515405 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0401 16:04:04.515418 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.536585 | |
| I0401 16:04:04.515430 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.48113 (* 1 = 2.48113 loss) | |
| I0401 16:04:04.515444 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.665521 (* 1 = 0.665521 loss) | |
| I0401 16:04:04.515455 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:04:04.515467 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00580688 | |
| I0401 16:04:04.515480 6134 sgd_solver.cpp:106] Iteration 45500, lr = 0.01 | |
| I0401 16:06:12.864107 6134 solver.cpp:229] Iteration 46000, loss = 4.71955 | |
| I0401 16:06:12.864351 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.217391 | |
| I0401 16:06:12.864382 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 16:06:12.864397 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.347826 | |
| I0401 16:06:12.864413 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.79121 (* 0.3 = 0.837362 loss) | |
| I0401 16:06:12.864426 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.824111 (* 0.3 = 0.247233 loss) | |
| I0401 16:06:12.864439 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.152174 | |
| I0401 16:06:12.864451 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0401 16:06:12.864462 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.478261 | |
| I0401 16:06:12.864476 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.57867 (* 0.3 = 0.773601 loss) | |
| I0401 16:06:12.864490 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.759582 (* 0.3 = 0.227875 loss) | |
| I0401 16:06:12.864501 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.478261 | |
| I0401 16:06:12.864513 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0401 16:06:12.864526 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.630435 | |
| I0401 16:06:12.864549 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.72135 (* 1 = 1.72135 loss) | |
| I0401 16:06:12.864575 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.539923 (* 1 = 0.539923 loss) | |
| I0401 16:06:12.864598 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 16:06:12.864622 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0284555 | |
| I0401 16:06:12.864646 6134 sgd_solver.cpp:106] Iteration 46000, lr = 0.01 | |
| I0401 16:08:21.296548 6134 solver.cpp:229] Iteration 46500, loss = 4.70397 | |
| I0401 16:08:21.296720 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.23913 | |
| I0401 16:08:21.296741 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 16:08:21.296752 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.478261 | |
| I0401 16:08:21.296768 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.64089 (* 0.3 = 0.792267 loss) | |
| I0401 16:08:21.296782 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.780271 (* 0.3 = 0.234081 loss) | |
| I0401 16:08:21.296794 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.26087 | |
| I0401 16:08:21.296807 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 16:08:21.296818 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.608696 | |
| I0401 16:08:21.296831 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.36738 (* 0.3 = 0.710213 loss) | |
| I0401 16:08:21.296845 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.692546 (* 0.3 = 0.207764 loss) | |
| I0401 16:08:21.296857 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.434783 | |
| I0401 16:08:21.296869 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0401 16:08:21.296880 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.673913 | |
| I0401 16:08:21.296895 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.77823 (* 1 = 1.77823 loss) | |
| I0401 16:08:21.296907 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.530523 (* 1 = 0.530523 loss) | |
| I0401 16:08:21.296919 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:08:21.296931 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0176603 | |
| I0401 16:08:21.296943 6134 sgd_solver.cpp:106] Iteration 46500, lr = 0.01 | |
| I0401 16:10:29.719319 6134 solver.cpp:229] Iteration 47000, loss = 4.73169 | |
| I0401 16:10:29.719446 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.215686 | |
| I0401 16:10:29.719467 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 16:10:29.719480 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.392157 | |
| I0401 16:10:29.719494 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.86857 (* 0.3 = 0.860572 loss) | |
| I0401 16:10:29.719509 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.85453 (* 0.3 = 0.256359 loss) | |
| I0401 16:10:29.719524 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.235294 | |
| I0401 16:10:29.719537 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 16:10:29.719549 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.529412 | |
| I0401 16:10:29.719563 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.70036 (* 0.3 = 0.810107 loss) | |
| I0401 16:10:29.719578 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.811096 (* 0.3 = 0.243329 loss) | |
| I0401 16:10:29.719589 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.431373 | |
| I0401 16:10:29.719601 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0401 16:10:29.719614 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.72549 | |
| I0401 16:10:29.719627 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.9902 (* 1 = 1.9902 loss) | |
| I0401 16:10:29.719640 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.595502 (* 1 = 0.595502 loss) | |
| I0401 16:10:29.719652 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:10:29.719663 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0140925 | |
| I0401 16:10:29.719676 6134 sgd_solver.cpp:106] Iteration 47000, lr = 0.01 | |
| I0401 16:12:38.122110 6134 solver.cpp:229] Iteration 47500, loss = 4.66086 | |
| I0401 16:12:38.122277 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.136364 | |
| I0401 16:12:38.122298 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 16:12:38.122310 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.477273 | |
| I0401 16:12:38.122328 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.66197 (* 0.3 = 0.798593 loss) | |
| I0401 16:12:38.122342 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.764738 (* 0.3 = 0.229421 loss) | |
| I0401 16:12:38.122354 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.318182 | |
| I0401 16:12:38.122366 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 16:12:38.122378 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.431818 | |
| I0401 16:12:38.122392 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.53969 (* 0.3 = 0.761907 loss) | |
| I0401 16:12:38.122406 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.715424 (* 0.3 = 0.214627 loss) | |
| I0401 16:12:38.122417 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0401 16:12:38.122429 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0401 16:12:38.122442 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.75 | |
| I0401 16:12:38.122454 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.71302 (* 1 = 1.71302 loss) | |
| I0401 16:12:38.122468 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.493616 (* 1 = 0.493616 loss) | |
| I0401 16:12:38.122480 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:12:38.122493 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0199906 | |
| I0401 16:12:38.122504 6134 sgd_solver.cpp:106] Iteration 47500, lr = 0.01 | |
| I0401 16:14:46.673321 6134 solver.cpp:229] Iteration 48000, loss = 4.6379 | |
| I0401 16:14:46.673424 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0833333 | |
| I0401 16:14:46.673444 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0401 16:14:46.673455 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.291667 | |
| I0401 16:14:46.673471 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.99153 (* 0.3 = 0.89746 loss) | |
| I0401 16:14:46.673485 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.856671 (* 0.3 = 0.257001 loss) | |
| I0401 16:14:46.673497 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.1875 | |
| I0401 16:14:46.673509 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 16:14:46.673521 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.354167 | |
| I0401 16:14:46.673535 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.90828 (* 0.3 = 0.872485 loss) | |
| I0401 16:14:46.673549 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.843683 (* 0.3 = 0.253105 loss) | |
| I0401 16:14:46.673562 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.416667 | |
| I0401 16:14:46.673573 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.829545 | |
| I0401 16:14:46.673584 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.708333 | |
| I0401 16:14:46.673600 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.05135 (* 1 = 2.05135 loss) | |
| I0401 16:14:46.673626 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.589644 (* 1 = 0.589644 loss) | |
| I0401 16:14:46.673647 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:14:46.673667 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00581666 | |
| I0401 16:14:46.673687 6134 sgd_solver.cpp:106] Iteration 48000, lr = 0.01 | |
| I0401 16:16:55.041704 6134 solver.cpp:229] Iteration 48500, loss = 4.65587 | |
| I0401 16:16:55.042021 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.191489 | |
| I0401 16:16:55.042042 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 16:16:55.042054 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.361702 | |
| I0401 16:16:55.042070 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.69053 (* 0.3 = 0.807159 loss) | |
| I0401 16:16:55.042084 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.780274 (* 0.3 = 0.234082 loss) | |
| I0401 16:16:55.042098 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.255319 | |
| I0401 16:16:55.042109 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 16:16:55.042121 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.531915 | |
| I0401 16:16:55.042135 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.37794 (* 0.3 = 0.713382 loss) | |
| I0401 16:16:55.042148 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.704275 (* 0.3 = 0.211282 loss) | |
| I0401 16:16:55.042160 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.446809 | |
| I0401 16:16:55.042172 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0401 16:16:55.042183 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.702128 | |
| I0401 16:16:55.042197 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.80822 (* 1 = 1.80822 loss) | |
| I0401 16:16:55.042210 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.534649 (* 1 = 0.534649 loss) | |
| I0401 16:16:55.042222 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:16:55.042233 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0086921 | |
| I0401 16:16:55.042246 6134 sgd_solver.cpp:106] Iteration 48500, lr = 0.01 | |
| I0401 16:19:03.709995 6134 solver.cpp:229] Iteration 49000, loss = 4.5895 | |
| I0401 16:19:03.710124 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.192308 | |
| I0401 16:19:03.710145 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 16:19:03.710158 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.288462 | |
| I0401 16:19:03.710175 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.90711 (* 0.3 = 0.872134 loss) | |
| I0401 16:19:03.710188 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.902999 (* 0.3 = 0.2709 loss) | |
| I0401 16:19:03.710201 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.192308 | |
| I0401 16:19:03.710213 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 16:19:03.710224 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.423077 | |
| I0401 16:19:03.710237 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.91047 (* 0.3 = 0.87314 loss) | |
| I0401 16:19:03.710252 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.900256 (* 0.3 = 0.270077 loss) | |
| I0401 16:19:03.710263 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.346154 | |
| I0401 16:19:03.710275 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0401 16:19:03.710286 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.480769 | |
| I0401 16:19:03.710300 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.21203 (* 1 = 2.21203 loss) | |
| I0401 16:19:03.710314 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.700363 (* 1 = 0.700363 loss) | |
| I0401 16:19:03.710326 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:19:03.710337 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0189673 | |
| I0401 16:19:03.710350 6134 sgd_solver.cpp:106] Iteration 49000, lr = 0.01 | |
| I0401 16:21:12.245713 6134 solver.cpp:229] Iteration 49500, loss = 4.55382 | |
| I0401 16:21:12.245863 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.170213 | |
| I0401 16:21:12.245890 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 16:21:12.245903 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.425532 | |
| I0401 16:21:12.245919 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.64517 (* 0.3 = 0.793552 loss) | |
| I0401 16:21:12.245934 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.757379 (* 0.3 = 0.227214 loss) | |
| I0401 16:21:12.245946 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.234043 | |
| I0401 16:21:12.245959 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0401 16:21:12.245970 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.489362 | |
| I0401 16:21:12.245992 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.45101 (* 0.3 = 0.735302 loss) | |
| I0401 16:21:12.246009 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.717443 (* 0.3 = 0.215233 loss) | |
| I0401 16:21:12.246021 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.425532 | |
| I0401 16:21:12.246033 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.829545 | |
| I0401 16:21:12.246045 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.765957 | |
| I0401 16:21:12.246058 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.809 (* 1 = 1.809 loss) | |
| I0401 16:21:12.246075 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.569607 (* 1 = 0.569607 loss) | |
| I0401 16:21:12.246088 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:21:12.246098 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00414086 | |
| I0401 16:21:12.246111 6134 sgd_solver.cpp:106] Iteration 49500, lr = 0.01 | |
| I0401 16:23:20.560545 6134 solver.cpp:456] Snapshotting to binary proto file /mnt/snapshots/mixed_lstm9_bn_iter_50000.caffemodel | |
| I0401 16:23:20.912770 6134 sgd_solver.cpp:273] Snapshotting solver state to binary proto file /mnt/snapshots/mixed_lstm9_bn_iter_50000.solverstate | |
| I0401 16:23:21.073299 6134 solver.cpp:338] Iteration 50000, Testing net (#0) | |
| I0401 16:23:50.878435 6134 solver.cpp:393] Test loss: 3.82045 | |
| I0401 16:23:50.878528 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.264661 | |
| I0401 16:23:50.878546 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.815727 | |
| I0401 16:23:50.878559 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.57935 | |
| I0401 16:23:50.878576 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.37403 (* 0.3 = 0.71221 loss) | |
| I0401 16:23:50.878590 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.60565 (* 0.3 = 0.181695 loss) | |
| I0401 16:23:50.878603 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.359141 | |
| I0401 16:23:50.878617 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.839365 | |
| I0401 16:23:50.878628 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.658634 | |
| I0401 16:23:50.878648 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 2.14045 (* 0.3 = 0.642136 loss) | |
| I0401 16:23:50.878676 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.548039 (* 0.3 = 0.164412 loss) | |
| I0401 16:23:50.878702 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.518421 | |
| I0401 16:23:50.878726 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.87523 | |
| I0401 16:23:50.878739 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.778481 | |
| I0401 16:23:50.878753 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.68104 (* 1 = 1.68104 loss) | |
| I0401 16:23:50.878767 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.438956 (* 1 = 0.438956 loss) | |
| I0401 16:23:50.878779 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.047 | |
| I0401 16:23:50.878793 6134 solver.cpp:406] Test net output #16: total_confidence = 0.0641473 | |
| I0401 16:23:51.029693 6134 solver.cpp:229] Iteration 50000, loss = 4.5686 | |
| I0401 16:23:51.029731 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.234043 | |
| I0401 16:23:51.029749 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 16:23:51.029762 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.425532 | |
| I0401 16:23:51.029778 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.04709 (* 0.3 = 0.914127 loss) | |
| I0401 16:23:51.029791 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.840032 (* 0.3 = 0.25201 loss) | |
| I0401 16:23:51.029803 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.234043 | |
| I0401 16:23:51.029815 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0401 16:23:51.029827 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.404255 | |
| I0401 16:23:51.029840 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.8862 (* 0.3 = 0.865859 loss) | |
| I0401 16:23:51.029862 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.7928 (* 0.3 = 0.23784 loss) | |
| I0401 16:23:51.029875 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.404255 | |
| I0401 16:23:51.029887 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0401 16:23:51.029898 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.595745 | |
| I0401 16:23:51.029913 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.5705 (* 1 = 2.5705 loss) | |
| I0401 16:23:51.029927 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.698751 (* 1 = 0.698751 loss) | |
| I0401 16:23:51.029940 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:23:51.029952 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0295253 | |
| I0401 16:23:51.029964 6134 sgd_solver.cpp:106] Iteration 50000, lr = 0.01 | |
| I0401 16:25:59.447907 6134 solver.cpp:229] Iteration 50500, loss = 4.57295 | |
| I0401 16:25:59.448036 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.225 | |
| I0401 16:25:59.448057 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 16:25:59.448071 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.575 | |
| I0401 16:25:59.448086 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.4798 (* 0.3 = 0.743941 loss) | |
| I0401 16:25:59.448101 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.658687 (* 0.3 = 0.197606 loss) | |
| I0401 16:25:59.448113 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.275 | |
| I0401 16:25:59.448125 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0401 16:25:59.448137 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.575 | |
| I0401 16:25:59.448151 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.28023 (* 0.3 = 0.684069 loss) | |
| I0401 16:25:59.448165 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.60541 (* 0.3 = 0.181623 loss) | |
| I0401 16:25:59.448178 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0401 16:25:59.448189 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 16:25:59.448201 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.725 | |
| I0401 16:25:59.448215 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.7394 (* 1 = 1.7394 loss) | |
| I0401 16:25:59.448230 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.484624 (* 1 = 0.484624 loss) | |
| I0401 16:25:59.448241 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:25:59.448252 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0254133 | |
| I0401 16:25:59.448264 6134 sgd_solver.cpp:106] Iteration 50500, lr = 0.01 | |
| I0401 16:28:07.706301 6134 solver.cpp:229] Iteration 51000, loss = 4.49551 | |
| I0401 16:28:07.706666 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.222222 | |
| I0401 16:28:07.706689 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 16:28:07.706702 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.377778 | |
| I0401 16:28:07.706719 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.68479 (* 0.3 = 0.805439 loss) | |
| I0401 16:28:07.706734 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.755654 (* 0.3 = 0.226696 loss) | |
| I0401 16:28:07.706748 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0401 16:28:07.706760 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 16:28:07.706773 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6 | |
| I0401 16:28:07.706787 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.40422 (* 0.3 = 0.721266 loss) | |
| I0401 16:28:07.706801 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.695421 (* 0.3 = 0.208626 loss) | |
| I0401 16:28:07.706815 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.4 | |
| I0401 16:28:07.706826 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0401 16:28:07.706838 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.6 | |
| I0401 16:28:07.706852 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.06257 (* 1 = 2.06257 loss) | |
| I0401 16:28:07.706866 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.56783 (* 1 = 0.56783 loss) | |
| I0401 16:28:07.706879 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:28:07.706892 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0397731 | |
| I0401 16:28:07.706905 6134 sgd_solver.cpp:106] Iteration 51000, lr = 0.01 | |
| I0401 16:30:16.125602 6134 solver.cpp:229] Iteration 51500, loss = 4.5073 | |
| I0401 16:30:16.125699 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.211538 | |
| I0401 16:30:16.125730 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 16:30:16.125756 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.384615 | |
| I0401 16:30:16.125782 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.68613 (* 0.3 = 0.80584 loss) | |
| I0401 16:30:16.125811 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.851546 (* 0.3 = 0.255464 loss) | |
| I0401 16:30:16.125839 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.211538 | |
| I0401 16:30:16.125864 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0401 16:30:16.125887 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.519231 | |
| I0401 16:30:16.125913 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.48493 (* 0.3 = 0.745479 loss) | |
| I0401 16:30:16.125941 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.794302 (* 0.3 = 0.238291 loss) | |
| I0401 16:30:16.125963 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.384615 | |
| I0401 16:30:16.125985 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.806818 | |
| I0401 16:30:16.126008 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.615385 | |
| I0401 16:30:16.126034 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.87839 (* 1 = 1.87839 loss) | |
| I0401 16:30:16.126060 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.595138 (* 1 = 0.595138 loss) | |
| I0401 16:30:16.126087 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:30:16.126109 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0240039 | |
| I0401 16:30:16.126132 6134 sgd_solver.cpp:106] Iteration 51500, lr = 0.01 | |
| I0401 16:32:24.702296 6134 solver.cpp:229] Iteration 52000, loss = 4.43542 | |
| I0401 16:32:24.702404 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.18 | |
| I0401 16:32:24.702425 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0401 16:32:24.702437 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.4 | |
| I0401 16:32:24.702453 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.54994 (* 0.3 = 0.764983 loss) | |
| I0401 16:32:24.702468 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.794915 (* 0.3 = 0.238475 loss) | |
| I0401 16:32:24.702481 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.34 | |
| I0401 16:32:24.702494 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 16:32:24.702507 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.64 | |
| I0401 16:32:24.702520 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.27849 (* 0.3 = 0.683548 loss) | |
| I0401 16:32:24.702535 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.71489 (* 0.3 = 0.214467 loss) | |
| I0401 16:32:24.702548 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.38 | |
| I0401 16:32:24.702559 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0401 16:32:24.702571 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.68 | |
| I0401 16:32:24.702585 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.919 (* 1 = 1.919 loss) | |
| I0401 16:32:24.702600 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.6138 (* 1 = 0.6138 loss) | |
| I0401 16:32:24.702611 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 16:32:24.702623 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0229252 | |
| I0401 16:32:24.702636 6134 sgd_solver.cpp:106] Iteration 52000, lr = 0.01 | |
| I0401 16:34:32.925926 6134 solver.cpp:229] Iteration 52500, loss = 4.45091 | |
| I0401 16:34:32.926025 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26087 | |
| I0401 16:34:32.926044 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 16:34:32.926057 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.521739 | |
| I0401 16:34:32.926074 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.51585 (* 0.3 = 0.754755 loss) | |
| I0401 16:34:32.926089 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.712858 (* 0.3 = 0.213858 loss) | |
| I0401 16:34:32.926101 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.26087 | |
| I0401 16:34:32.926113 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 16:34:32.926126 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.434783 | |
| I0401 16:34:32.926139 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.62332 (* 0.3 = 0.786995 loss) | |
| I0401 16:34:32.926153 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.752708 (* 0.3 = 0.225812 loss) | |
| I0401 16:34:32.926165 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.391304 | |
| I0401 16:34:32.926177 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0401 16:34:32.926189 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.782609 | |
| I0401 16:34:32.926203 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.96735 (* 1 = 1.96735 loss) | |
| I0401 16:34:32.926218 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.57561 (* 1 = 0.57561 loss) | |
| I0401 16:34:32.926229 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:34:32.926241 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00602476 | |
| I0401 16:34:32.926254 6134 sgd_solver.cpp:106] Iteration 52500, lr = 0.01 | |
| I0401 16:36:41.349750 6134 solver.cpp:229] Iteration 53000, loss = 4.34456 | |
| I0401 16:36:41.350090 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.36 | |
| I0401 16:36:41.350111 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 16:36:41.350124 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.58 | |
| I0401 16:36:41.350142 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.2931 (* 0.3 = 0.687931 loss) | |
| I0401 16:36:41.350157 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.69894 (* 0.3 = 0.209682 loss) | |
| I0401 16:36:41.350169 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.38 | |
| I0401 16:36:41.350181 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 16:36:41.350193 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.62 | |
| I0401 16:36:41.350208 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.09828 (* 0.3 = 0.629484 loss) | |
| I0401 16:36:41.350221 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.644801 (* 0.3 = 0.19344 loss) | |
| I0401 16:36:41.350234 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.58 | |
| I0401 16:36:41.350245 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 16:36:41.350258 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.84 | |
| I0401 16:36:41.350272 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.27721 (* 1 = 1.27721 loss) | |
| I0401 16:36:41.350286 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.389369 (* 1 = 0.389369 loss) | |
| I0401 16:36:41.350298 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:36:41.350311 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0415145 | |
| I0401 16:36:41.350323 6134 sgd_solver.cpp:106] Iteration 53000, lr = 0.01 | |
| I0401 16:38:49.682837 6134 solver.cpp:229] Iteration 53500, loss = 4.41207 | |
| I0401 16:38:49.682953 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.184211 | |
| I0401 16:38:49.682974 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 16:38:49.682987 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.552632 | |
| I0401 16:38:49.683003 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.50395 (* 0.3 = 0.751186 loss) | |
| I0401 16:38:49.683018 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.704339 (* 0.3 = 0.211302 loss) | |
| I0401 16:38:49.683032 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 16:38:49.683043 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 16:38:49.683055 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.605263 | |
| I0401 16:38:49.683069 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.21249 (* 0.3 = 0.663748 loss) | |
| I0401 16:38:49.683084 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.612038 (* 0.3 = 0.183612 loss) | |
| I0401 16:38:49.683096 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.473684 | |
| I0401 16:38:49.683109 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 16:38:49.683120 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.657895 | |
| I0401 16:38:49.683135 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.70285 (* 1 = 1.70285 loss) | |
| I0401 16:38:49.683148 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.472532 (* 1 = 0.472532 loss) | |
| I0401 16:38:49.683161 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 16:38:49.683172 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0851777 | |
| I0401 16:38:49.683185 6134 sgd_solver.cpp:106] Iteration 53500, lr = 0.01 | |
| I0401 16:40:58.169673 6134 solver.cpp:229] Iteration 54000, loss = 4.34797 | |
| I0401 16:40:58.169796 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.22449 | |
| I0401 16:40:58.169816 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 16:40:58.169829 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.387755 | |
| I0401 16:40:58.169844 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.85763 (* 0.3 = 0.857289 loss) | |
| I0401 16:40:58.169859 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.833104 (* 0.3 = 0.249931 loss) | |
| I0401 16:40:58.169872 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.22449 | |
| I0401 16:40:58.169884 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 16:40:58.169896 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.428571 | |
| I0401 16:40:58.169911 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.70563 (* 0.3 = 0.81169 loss) | |
| I0401 16:40:58.169925 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.794022 (* 0.3 = 0.238207 loss) | |
| I0401 16:40:58.169939 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.469388 | |
| I0401 16:40:58.169950 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0401 16:40:58.169962 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.591837 | |
| I0401 16:40:58.169976 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.95701 (* 1 = 1.95701 loss) | |
| I0401 16:40:58.169991 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.592322 (* 1 = 0.592322 loss) | |
| I0401 16:40:58.170002 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:40:58.170014 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0491011 | |
| I0401 16:40:58.170027 6134 sgd_solver.cpp:106] Iteration 54000, lr = 0.01 | |
| I0401 16:43:06.618003 6134 solver.cpp:229] Iteration 54500, loss = 4.37137 | |
| I0401 16:43:06.618115 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.175 | |
| I0401 16:43:06.618136 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 16:43:06.618149 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.35 | |
| I0401 16:43:06.618165 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.71262 (* 0.3 = 0.813786 loss) | |
| I0401 16:43:06.618180 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.833035 (* 0.3 = 0.249911 loss) | |
| I0401 16:43:06.618192 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.275 | |
| I0401 16:43:06.618206 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 16:43:06.618217 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.45 | |
| I0401 16:43:06.618232 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.59971 (* 0.3 = 0.779912 loss) | |
| I0401 16:43:06.618247 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.787684 (* 0.3 = 0.236305 loss) | |
| I0401 16:43:06.618258 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.4 | |
| I0401 16:43:06.618271 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0401 16:43:06.618283 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.625 | |
| I0401 16:43:06.618297 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.22757 (* 1 = 2.22757 loss) | |
| I0401 16:43:06.618310 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.705487 (* 1 = 0.705487 loss) | |
| I0401 16:43:06.618324 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:43:06.618335 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0287242 | |
| I0401 16:43:06.618347 6134 sgd_solver.cpp:106] Iteration 54500, lr = 0.01 | |
| I0401 16:45:14.969254 6134 solver.cpp:338] Iteration 55000, Testing net (#0) | |
| I0401 16:45:44.751217 6134 solver.cpp:393] Test loss: 3.91808 | |
| I0401 16:45:44.751274 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.215138 | |
| I0401 16:45:44.751291 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.806682 | |
| I0401 16:45:44.751304 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.529206 | |
| I0401 16:45:44.751322 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.6196 (* 0.3 = 0.78588 loss) | |
| I0401 16:45:44.751337 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.657737 (* 0.3 = 0.197321 loss) | |
| I0401 16:45:44.751349 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.339031 | |
| I0401 16:45:44.751361 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.833683 | |
| I0401 16:45:44.751374 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.665517 | |
| I0401 16:45:44.751386 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 2.17212 (* 0.3 = 0.651636 loss) | |
| I0401 16:45:44.751400 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.559888 (* 0.3 = 0.167966 loss) | |
| I0401 16:45:44.751412 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.526206 | |
| I0401 16:45:44.751425 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.876639 | |
| I0401 16:45:44.751435 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.771034 | |
| I0401 16:45:44.751449 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.67162 (* 1 = 1.67162 loss) | |
| I0401 16:45:44.751463 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.443653 (* 1 = 0.443653 loss) | |
| I0401 16:45:44.751476 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.067 | |
| I0401 16:45:44.751487 6134 solver.cpp:406] Test net output #16: total_confidence = 0.0602164 | |
| I0401 16:45:44.902441 6134 solver.cpp:229] Iteration 55000, loss = 4.36464 | |
| I0401 16:45:44.902495 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.156863 | |
| I0401 16:45:44.902513 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0401 16:45:44.902526 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.313726 | |
| I0401 16:45:44.902542 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.885 (* 0.3 = 0.8655 loss) | |
| I0401 16:45:44.902557 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.875208 (* 0.3 = 0.262562 loss) | |
| I0401 16:45:44.902570 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.27451 | |
| I0401 16:45:44.902582 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0401 16:45:44.902595 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.372549 | |
| I0401 16:45:44.902608 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.52831 (* 0.3 = 0.758493 loss) | |
| I0401 16:45:44.902622 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.77721 (* 0.3 = 0.233163 loss) | |
| I0401 16:45:44.902634 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.294118 | |
| I0401 16:45:44.902647 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.784091 | |
| I0401 16:45:44.902658 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.568627 | |
| I0401 16:45:44.902673 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.27669 (* 1 = 2.27669 loss) | |
| I0401 16:45:44.902686 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.693314 (* 1 = 0.693314 loss) | |
| I0401 16:45:44.902698 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:45:44.902710 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0137687 | |
| I0401 16:45:44.902724 6134 sgd_solver.cpp:106] Iteration 55000, lr = 0.01 | |
| I0401 16:47:53.618798 6134 solver.cpp:229] Iteration 55500, loss = 4.32605 | |
| I0401 16:47:53.619098 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.2 | |
| I0401 16:47:53.619118 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 16:47:53.619132 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.488889 | |
| I0401 16:47:53.619148 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.50041 (* 0.3 = 0.750122 loss) | |
| I0401 16:47:53.619163 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.762967 (* 0.3 = 0.22889 loss) | |
| I0401 16:47:53.619175 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.266667 | |
| I0401 16:47:53.619189 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 16:47:53.619199 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.555556 | |
| I0401 16:47:53.619213 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.37501 (* 0.3 = 0.712504 loss) | |
| I0401 16:47:53.619227 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.685706 (* 0.3 = 0.205712 loss) | |
| I0401 16:47:53.619240 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.488889 | |
| I0401 16:47:53.619251 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0401 16:47:53.619263 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.755556 | |
| I0401 16:47:53.619277 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.72326 (* 1 = 1.72326 loss) | |
| I0401 16:47:53.619290 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.494902 (* 1 = 0.494902 loss) | |
| I0401 16:47:53.619302 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 16:47:53.619314 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0539322 | |
| I0401 16:47:53.619328 6134 sgd_solver.cpp:106] Iteration 55500, lr = 0.01 | |
| I0401 16:50:02.152067 6134 solver.cpp:229] Iteration 56000, loss = 4.30134 | |
| I0401 16:50:02.152204 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.309524 | |
| I0401 16:50:02.152225 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 16:50:02.152237 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.619048 | |
| I0401 16:50:02.152254 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.32915 (* 0.3 = 0.698746 loss) | |
| I0401 16:50:02.152268 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.65617 (* 0.3 = 0.196851 loss) | |
| I0401 16:50:02.152281 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.357143 | |
| I0401 16:50:02.152294 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0401 16:50:02.152307 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0401 16:50:02.152320 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.0951 (* 0.3 = 0.628531 loss) | |
| I0401 16:50:02.152334 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.625023 (* 0.3 = 0.187507 loss) | |
| I0401 16:50:02.152348 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.619048 | |
| I0401 16:50:02.152359 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0401 16:50:02.152370 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.880952 | |
| I0401 16:50:02.152384 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.09931 (* 1 = 1.09931 loss) | |
| I0401 16:50:02.152400 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.331912 (* 1 = 0.331912 loss) | |
| I0401 16:50:02.152411 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:50:02.152423 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0484895 | |
| I0401 16:50:02.152436 6134 sgd_solver.cpp:106] Iteration 56000, lr = 0.01 | |
| I0401 16:52:10.333323 6134 solver.cpp:229] Iteration 56500, loss = 4.26406 | |
| I0401 16:52:10.333453 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.195652 | |
| I0401 16:52:10.333473 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 16:52:10.333487 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0401 16:52:10.333503 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.46087 (* 0.3 = 0.738262 loss) | |
| I0401 16:52:10.333519 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.719671 (* 0.3 = 0.215901 loss) | |
| I0401 16:52:10.333534 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.347826 | |
| I0401 16:52:10.333545 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0401 16:52:10.333557 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.630435 | |
| I0401 16:52:10.333571 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.21972 (* 0.3 = 0.665917 loss) | |
| I0401 16:52:10.333585 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.613126 (* 0.3 = 0.183938 loss) | |
| I0401 16:52:10.333598 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.434783 | |
| I0401 16:52:10.333609 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0401 16:52:10.333622 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.673913 | |
| I0401 16:52:10.333636 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.56724 (* 1 = 1.56724 loss) | |
| I0401 16:52:10.333649 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.446201 (* 1 = 0.446201 loss) | |
| I0401 16:52:10.333662 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 16:52:10.333673 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0205402 | |
| I0401 16:52:10.333685 6134 sgd_solver.cpp:106] Iteration 56500, lr = 0.01 | |
| I0401 16:54:19.323324 6134 solver.cpp:229] Iteration 57000, loss = 4.20321 | |
| I0401 16:54:19.323434 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.309524 | |
| I0401 16:54:19.323456 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 16:54:19.323468 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.547619 | |
| I0401 16:54:19.323483 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.2966 (* 0.3 = 0.688979 loss) | |
| I0401 16:54:19.323498 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.733264 (* 0.3 = 0.219979 loss) | |
| I0401 16:54:19.323511 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.309524 | |
| I0401 16:54:19.323528 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 16:54:19.323539 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.547619 | |
| I0401 16:54:19.323554 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.55352 (* 0.3 = 0.766055 loss) | |
| I0401 16:54:19.323568 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.816663 (* 0.3 = 0.244999 loss) | |
| I0401 16:54:19.323580 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.571429 | |
| I0401 16:54:19.323592 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0401 16:54:19.323604 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.738095 | |
| I0401 16:54:19.323618 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.87746 (* 1 = 1.87746 loss) | |
| I0401 16:54:19.323632 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.761483 (* 1 = 0.761483 loss) | |
| I0401 16:54:19.323644 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:54:19.323657 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0912121 | |
| I0401 16:54:19.323668 6134 sgd_solver.cpp:106] Iteration 57000, lr = 0.01 | |
| I0401 16:56:27.846416 6134 solver.cpp:229] Iteration 57500, loss = 4.24011 | |
| I0401 16:56:27.846686 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.339623 | |
| I0401 16:56:27.846707 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 16:56:27.846720 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.603774 | |
| I0401 16:56:27.846735 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.15414 (* 0.3 = 0.646243 loss) | |
| I0401 16:56:27.846750 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.69076 (* 0.3 = 0.207228 loss) | |
| I0401 16:56:27.846763 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.339623 | |
| I0401 16:56:27.846776 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 16:56:27.846788 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.698113 | |
| I0401 16:56:27.846802 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.07317 (* 0.3 = 0.621952 loss) | |
| I0401 16:56:27.846817 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.648807 (* 0.3 = 0.194642 loss) | |
| I0401 16:56:27.846828 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.54717 | |
| I0401 16:56:27.846842 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 16:56:27.846853 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.811321 | |
| I0401 16:56:27.846866 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.42307 (* 1 = 1.42307 loss) | |
| I0401 16:56:27.846880 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.441425 (* 1 = 0.441425 loss) | |
| I0401 16:56:27.846894 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:56:27.846904 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0121024 | |
| I0401 16:56:27.846916 6134 sgd_solver.cpp:106] Iteration 57500, lr = 0.01 | |
| I0401 16:58:36.323050 6134 solver.cpp:229] Iteration 58000, loss = 4.18723 | |
| I0401 16:58:36.323153 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.133333 | |
| I0401 16:58:36.323173 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 16:58:36.323185 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.266667 | |
| I0401 16:58:36.323201 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.20292 (* 0.3 = 0.960876 loss) | |
| I0401 16:58:36.323216 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.873513 (* 0.3 = 0.262054 loss) | |
| I0401 16:58:36.323230 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.177778 | |
| I0401 16:58:36.323242 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0401 16:58:36.323254 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.4 | |
| I0401 16:58:36.323268 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.9468 (* 0.3 = 0.88404 loss) | |
| I0401 16:58:36.323282 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.822487 (* 0.3 = 0.246746 loss) | |
| I0401 16:58:36.323295 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.288889 | |
| I0401 16:58:36.323307 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.806818 | |
| I0401 16:58:36.323319 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.488889 | |
| I0401 16:58:36.323333 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.3939 (* 1 = 2.3939 loss) | |
| I0401 16:58:36.323348 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.668756 (* 1 = 0.668756 loss) | |
| I0401 16:58:36.323359 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 16:58:36.323371 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0112969 | |
| I0401 16:58:36.323384 6134 sgd_solver.cpp:106] Iteration 58000, lr = 0.01 | |
| I0401 17:00:44.732712 6134 solver.cpp:229] Iteration 58500, loss = 4.1727 | |
| I0401 17:00:44.732842 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.289474 | |
| I0401 17:00:44.732862 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 17:00:44.732875 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.605263 | |
| I0401 17:00:44.732892 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.32132 (* 0.3 = 0.696395 loss) | |
| I0401 17:00:44.732908 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.668407 (* 0.3 = 0.200522 loss) | |
| I0401 17:00:44.732920 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.315789 | |
| I0401 17:00:44.732933 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 17:00:44.732945 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.631579 | |
| I0401 17:00:44.732959 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.12901 (* 0.3 = 0.638704 loss) | |
| I0401 17:00:44.732974 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.662232 (* 0.3 = 0.19867 loss) | |
| I0401 17:00:44.732985 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.552632 | |
| I0401 17:00:44.732998 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 17:00:44.733009 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.789474 | |
| I0401 17:00:44.733023 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.3822 (* 1 = 1.3822 loss) | |
| I0401 17:00:44.733037 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.466671 (* 1 = 0.466671 loss) | |
| I0401 17:00:44.733062 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 17:00:44.733078 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0950895 | |
| I0401 17:00:44.733089 6134 sgd_solver.cpp:106] Iteration 58500, lr = 0.01 | |
| I0401 17:02:53.124496 6134 solver.cpp:229] Iteration 59000, loss = 4.20952 | |
| I0401 17:02:53.124603 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.214286 | |
| I0401 17:02:53.124624 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 17:02:53.124635 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.380952 | |
| I0401 17:02:53.124651 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.61615 (* 0.3 = 0.784845 loss) | |
| I0401 17:02:53.124666 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.760364 (* 0.3 = 0.228109 loss) | |
| I0401 17:02:53.124680 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.380952 | |
| I0401 17:02:53.124693 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 17:02:53.124706 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.47619 | |
| I0401 17:02:53.124719 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.4222 (* 0.3 = 0.726661 loss) | |
| I0401 17:02:53.124733 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.686662 (* 0.3 = 0.205998 loss) | |
| I0401 17:02:53.124747 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.452381 | |
| I0401 17:02:53.124758 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0401 17:02:53.124769 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.785714 | |
| I0401 17:02:53.124784 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.60866 (* 1 = 1.60866 loss) | |
| I0401 17:02:53.124799 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.475694 (* 1 = 0.475694 loss) | |
| I0401 17:02:53.124810 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:02:53.124822 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0790929 | |
| I0401 17:02:53.124835 6134 sgd_solver.cpp:106] Iteration 59000, lr = 0.01 | |
| I0401 17:05:01.393643 6134 solver.cpp:229] Iteration 59500, loss = 4.1575 | |
| I0401 17:05:01.393805 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.215686 | |
| I0401 17:05:01.393826 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0401 17:05:01.393838 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.470588 | |
| I0401 17:05:01.393854 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.77614 (* 0.3 = 0.832843 loss) | |
| I0401 17:05:01.393869 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.89338 (* 0.3 = 0.268014 loss) | |
| I0401 17:05:01.393882 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.294118 | |
| I0401 17:05:01.393896 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 17:05:01.393908 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.54902 | |
| I0401 17:05:01.393923 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.60781 (* 0.3 = 0.782344 loss) | |
| I0401 17:05:01.393936 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.867039 (* 0.3 = 0.260112 loss) | |
| I0401 17:05:01.393949 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.470588 | |
| I0401 17:05:01.393960 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.8125 | |
| I0401 17:05:01.393972 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.627451 | |
| I0401 17:05:01.393986 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.02737 (* 1 = 2.02737 loss) | |
| I0401 17:05:01.394001 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.699985 (* 1 = 0.699985 loss) | |
| I0401 17:05:01.394012 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:05:01.394024 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0253565 | |
| I0401 17:05:01.394037 6134 sgd_solver.cpp:106] Iteration 59500, lr = 0.01 | |
| I0401 17:07:09.801879 6134 solver.cpp:338] Iteration 60000, Testing net (#0) | |
| I0401 17:07:39.379954 6134 solver.cpp:393] Test loss: 3.61448 | |
| I0401 17:07:39.379998 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.288406 | |
| I0401 17:07:39.380015 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.816364 | |
| I0401 17:07:39.380028 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.603254 | |
| I0401 17:07:39.380043 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.27332 (* 0.3 = 0.681995 loss) | |
| I0401 17:07:39.380058 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.612319 (* 0.3 = 0.183696 loss) | |
| I0401 17:07:39.380070 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.402706 | |
| I0401 17:07:39.380082 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.84091 | |
| I0401 17:07:39.380095 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.708801 | |
| I0401 17:07:39.380107 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.95105 (* 0.3 = 0.585316 loss) | |
| I0401 17:07:39.380121 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.533538 (* 0.3 = 0.160061 loss) | |
| I0401 17:07:39.380133 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.557871 | |
| I0401 17:07:39.380146 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.881139 | |
| I0401 17:07:39.380156 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.798977 | |
| I0401 17:07:39.380170 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.57314 (* 1 = 1.57314 loss) | |
| I0401 17:07:39.380183 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.430265 (* 1 = 0.430265 loss) | |
| I0401 17:07:39.380195 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.074 | |
| I0401 17:07:39.380206 6134 solver.cpp:406] Test net output #16: total_confidence = 0.0991732 | |
| I0401 17:07:39.531630 6134 solver.cpp:229] Iteration 60000, loss = 4.18178 | |
| I0401 17:07:39.531669 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.0465116 | |
| I0401 17:07:39.531687 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0401 17:07:39.531698 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.348837 | |
| I0401 17:07:39.531713 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.14844 (* 0.3 = 0.944533 loss) | |
| I0401 17:07:39.531728 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.869117 (* 0.3 = 0.260735 loss) | |
| I0401 17:07:39.531740 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.0697674 | |
| I0401 17:07:39.531752 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0401 17:07:39.531764 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.232558 | |
| I0401 17:07:39.531779 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.496 (* 0.3 = 1.0488 loss) | |
| I0401 17:07:39.531792 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.990287 (* 0.3 = 0.297086 loss) | |
| I0401 17:07:39.531807 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.209302 | |
| I0401 17:07:39.531821 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0401 17:07:39.531832 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.44186 | |
| I0401 17:07:39.531847 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.74698 (* 1 = 2.74698 loss) | |
| I0401 17:07:39.531860 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.712457 (* 1 = 0.712457 loss) | |
| I0401 17:07:39.531872 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:07:39.531884 6134 solver.cpp:245] Train net output #16: total_confidence = 0.00846092 | |
| I0401 17:07:39.531896 6134 sgd_solver.cpp:106] Iteration 60000, lr = 0.01 | |
| I0401 17:09:47.780514 6134 solver.cpp:229] Iteration 60500, loss = 4.20862 | |
| I0401 17:09:47.780704 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.234043 | |
| I0401 17:09:47.780725 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 17:09:47.780738 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.595745 | |
| I0401 17:09:47.780755 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.40981 (* 0.3 = 0.722942 loss) | |
| I0401 17:09:47.780769 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.694382 (* 0.3 = 0.208315 loss) | |
| I0401 17:09:47.780782 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.319149 | |
| I0401 17:09:47.780796 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 17:09:47.780807 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.617021 | |
| I0401 17:09:47.780822 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.18545 (* 0.3 = 0.655634 loss) | |
| I0401 17:09:47.780835 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.621871 (* 0.3 = 0.186561 loss) | |
| I0401 17:09:47.780848 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.617021 | |
| I0401 17:09:47.780859 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 17:09:47.780871 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.744681 | |
| I0401 17:09:47.780885 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.36393 (* 1 = 1.36393 loss) | |
| I0401 17:09:47.780900 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.396573 (* 1 = 0.396573 loss) | |
| I0401 17:09:47.780911 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 17:09:47.780925 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0421525 | |
| I0401 17:09:47.780936 6134 sgd_solver.cpp:106] Iteration 60500, lr = 0.01 | |
| I0401 17:11:56.402694 6134 solver.cpp:229] Iteration 61000, loss = 4.04744 | |
| I0401 17:11:56.402799 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.22 | |
| I0401 17:11:56.402820 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 17:11:56.402833 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.44 | |
| I0401 17:11:56.402848 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.53101 (* 0.3 = 0.759302 loss) | |
| I0401 17:11:56.402863 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.747477 (* 0.3 = 0.224243 loss) | |
| I0401 17:11:56.402876 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.32 | |
| I0401 17:11:56.402889 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 17:11:56.402901 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.68 | |
| I0401 17:11:56.402915 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.22921 (* 0.3 = 0.668763 loss) | |
| I0401 17:11:56.402930 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.68374 (* 0.3 = 0.205122 loss) | |
| I0401 17:11:56.402941 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.58 | |
| I0401 17:11:56.402953 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0401 17:11:56.402966 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.68 | |
| I0401 17:11:56.402979 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.63217 (* 1 = 1.63217 loss) | |
| I0401 17:11:56.402993 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.506888 (* 1 = 0.506888 loss) | |
| I0401 17:11:56.403005 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 17:11:56.403017 6134 solver.cpp:245] Train net output #16: total_confidence = 0.136028 | |
| I0401 17:11:56.403029 6134 sgd_solver.cpp:106] Iteration 61000, lr = 0.01 | |
| I0401 17:14:04.776231 6134 solver.cpp:229] Iteration 61500, loss = 4.08256 | |
| I0401 17:14:04.776365 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0401 17:14:04.776386 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 17:14:04.776398 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.595238 | |
| I0401 17:14:04.776415 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.3161 (* 0.3 = 0.694831 loss) | |
| I0401 17:14:04.776429 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.665416 (* 0.3 = 0.199625 loss) | |
| I0401 17:14:04.776443 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.404762 | |
| I0401 17:14:04.776455 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0401 17:14:04.776468 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.714286 | |
| I0401 17:14:04.776481 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.02993 (* 0.3 = 0.60898 loss) | |
| I0401 17:14:04.776495 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.611924 (* 0.3 = 0.183577 loss) | |
| I0401 17:14:04.776507 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.738095 | |
| I0401 17:14:04.776522 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0401 17:14:04.776535 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.857143 | |
| I0401 17:14:04.776549 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.14992 (* 1 = 1.14992 loss) | |
| I0401 17:14:04.776563 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.350391 (* 1 = 0.350391 loss) | |
| I0401 17:14:04.776576 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 17:14:04.776587 6134 solver.cpp:245] Train net output #16: total_confidence = 0.147976 | |
| I0401 17:14:04.776599 6134 sgd_solver.cpp:106] Iteration 61500, lr = 0.01 | |
| I0401 17:16:13.252694 6134 solver.cpp:229] Iteration 62000, loss = 4.08776 | |
| I0401 17:16:13.252964 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.276596 | |
| I0401 17:16:13.252985 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 17:16:13.252997 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.531915 | |
| I0401 17:16:13.253013 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.53043 (* 0.3 = 0.759129 loss) | |
| I0401 17:16:13.253028 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.742455 (* 0.3 = 0.222737 loss) | |
| I0401 17:16:13.253041 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.255319 | |
| I0401 17:16:13.253072 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 17:16:13.253084 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.595745 | |
| I0401 17:16:13.253098 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.35816 (* 0.3 = 0.707449 loss) | |
| I0401 17:16:13.253113 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.679641 (* 0.3 = 0.203892 loss) | |
| I0401 17:16:13.253129 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.531915 | |
| I0401 17:16:13.253141 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0401 17:16:13.253154 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.765957 | |
| I0401 17:16:13.253167 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.60305 (* 1 = 1.60305 loss) | |
| I0401 17:16:13.253181 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.450345 (* 1 = 0.450345 loss) | |
| I0401 17:16:13.253193 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:16:13.253206 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0546284 | |
| I0401 17:16:13.253218 6134 sgd_solver.cpp:106] Iteration 62000, lr = 0.01 | |
| I0401 17:18:21.609097 6134 solver.cpp:229] Iteration 62500, loss = 4.03367 | |
| I0401 17:18:21.609254 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.238095 | |
| I0401 17:18:21.609287 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 17:18:21.609309 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.595238 | |
| I0401 17:18:21.609338 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.28999 (* 0.3 = 0.686997 loss) | |
| I0401 17:18:21.609367 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.745375 (* 0.3 = 0.223612 loss) | |
| I0401 17:18:21.609390 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 17:18:21.609411 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0401 17:18:21.609432 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.738095 | |
| I0401 17:18:21.609459 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.65312 (* 0.3 = 0.495936 loss) | |
| I0401 17:18:21.609486 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.491943 (* 0.3 = 0.147583 loss) | |
| I0401 17:18:21.609508 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.595238 | |
| I0401 17:18:21.609532 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0401 17:18:21.609555 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.785714 | |
| I0401 17:18:21.609580 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.17949 (* 1 = 1.17949 loss) | |
| I0401 17:18:21.609604 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.346597 (* 1 = 0.346597 loss) | |
| I0401 17:18:21.609625 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 17:18:21.609649 6134 solver.cpp:245] Train net output #16: total_confidence = 0.12039 | |
| I0401 17:18:21.609670 6134 sgd_solver.cpp:106] Iteration 62500, lr = 0.01 | |
| I0401 17:20:30.122526 6134 solver.cpp:229] Iteration 63000, loss = 4.09879 | |
| I0401 17:20:30.122658 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.238095 | |
| I0401 17:20:30.122678 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 17:20:30.122691 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.47619 | |
| I0401 17:20:30.122709 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.70128 (* 0.3 = 0.810383 loss) | |
| I0401 17:20:30.122723 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.744399 (* 0.3 = 0.22332 loss) | |
| I0401 17:20:30.122735 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.261905 | |
| I0401 17:20:30.122748 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 17:20:30.122761 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.452381 | |
| I0401 17:20:30.122774 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.80185 (* 0.3 = 0.840555 loss) | |
| I0401 17:20:30.122788 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.774722 (* 0.3 = 0.232417 loss) | |
| I0401 17:20:30.122800 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.404762 | |
| I0401 17:20:30.122812 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0401 17:20:30.122824 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.666667 | |
| I0401 17:20:30.122838 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.03589 (* 1 = 2.03589 loss) | |
| I0401 17:20:30.122853 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.51274 (* 1 = 0.51274 loss) | |
| I0401 17:20:30.122864 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:20:30.122876 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0243124 | |
| I0401 17:20:30.122889 6134 sgd_solver.cpp:106] Iteration 63000, lr = 0.01 | |
| I0401 17:22:38.729537 6134 solver.cpp:229] Iteration 63500, loss = 4.00095 | |
| I0401 17:22:38.729651 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.317073 | |
| I0401 17:22:38.729671 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 17:22:38.729683 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.512195 | |
| I0401 17:22:38.729699 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.64808 (* 0.3 = 0.794424 loss) | |
| I0401 17:22:38.729715 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.750546 (* 0.3 = 0.225164 loss) | |
| I0401 17:22:38.729730 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.243902 | |
| I0401 17:22:38.729743 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0401 17:22:38.729755 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.487805 | |
| I0401 17:22:38.729769 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.43638 (* 0.3 = 0.730914 loss) | |
| I0401 17:22:38.729784 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.687589 (* 0.3 = 0.206277 loss) | |
| I0401 17:22:38.729795 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.536585 | |
| I0401 17:22:38.729809 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 17:22:38.729821 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.780488 | |
| I0401 17:22:38.729835 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.70698 (* 1 = 1.70698 loss) | |
| I0401 17:22:38.729849 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.45668 (* 1 = 0.45668 loss) | |
| I0401 17:22:38.729861 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:22:38.729873 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0248108 | |
| I0401 17:22:38.729885 6134 sgd_solver.cpp:106] Iteration 63500, lr = 0.01 | |
| I0401 17:24:47.200341 6134 solver.cpp:229] Iteration 64000, loss = 4.05785 | |
| I0401 17:24:47.200456 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.236364 | |
| I0401 17:24:47.200477 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 17:24:47.200490 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.363636 | |
| I0401 17:24:47.200507 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.26893 (* 0.3 = 0.980678 loss) | |
| I0401 17:24:47.200525 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.05363 (* 0.3 = 0.31609 loss) | |
| I0401 17:24:47.200538 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.254545 | |
| I0401 17:24:47.200551 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0401 17:24:47.200562 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.4 | |
| I0401 17:24:47.200577 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.21192 (* 0.3 = 0.963575 loss) | |
| I0401 17:24:47.200590 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.02529 (* 0.3 = 0.307587 loss) | |
| I0401 17:24:47.200603 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.381818 | |
| I0401 17:24:47.200615 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.801136 | |
| I0401 17:24:47.200628 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.490909 | |
| I0401 17:24:47.200641 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.80135 (* 1 = 2.80135 loss) | |
| I0401 17:24:47.200655 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.909448 (* 1 = 0.909448 loss) | |
| I0401 17:24:47.200667 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:24:47.200680 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0529742 | |
| I0401 17:24:47.200692 6134 sgd_solver.cpp:106] Iteration 64000, lr = 0.01 | |
| I0401 17:26:55.690649 6134 solver.cpp:229] Iteration 64500, loss = 3.97861 | |
| I0401 17:26:55.690960 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.2 | |
| I0401 17:26:55.690980 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 17:26:55.690994 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.555556 | |
| I0401 17:26:55.691009 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.59943 (* 0.3 = 0.779828 loss) | |
| I0401 17:26:55.691027 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.714718 (* 0.3 = 0.214416 loss) | |
| I0401 17:26:55.691051 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.266667 | |
| I0401 17:26:55.691071 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 17:26:55.691083 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.644444 | |
| I0401 17:26:55.691097 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.23082 (* 0.3 = 0.669245 loss) | |
| I0401 17:26:55.691112 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.620719 (* 0.3 = 0.186216 loss) | |
| I0401 17:26:55.691124 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.533333 | |
| I0401 17:26:55.691136 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 17:26:55.691148 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8 | |
| I0401 17:26:55.691161 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.53231 (* 1 = 1.53231 loss) | |
| I0401 17:26:55.691175 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.440889 (* 1 = 0.440889 loss) | |
| I0401 17:26:55.691187 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:26:55.691200 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0974831 | |
| I0401 17:26:55.691211 6134 sgd_solver.cpp:106] Iteration 64500, lr = 0.01 | |
| I0401 17:29:03.952927 6134 solver.cpp:338] Iteration 65000, Testing net (#0) | |
| I0401 17:29:33.548275 6134 solver.cpp:393] Test loss: 3.28882 | |
| I0401 17:29:33.548327 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.34678 | |
| I0401 17:29:33.548344 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.814 | |
| I0401 17:29:33.548357 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.634625 | |
| I0401 17:29:33.548372 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.16158 (* 0.3 = 0.648474 loss) | |
| I0401 17:29:33.548388 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.643679 (* 0.3 = 0.193104 loss) | |
| I0401 17:29:33.548400 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.46816 | |
| I0401 17:29:33.548413 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.851048 | |
| I0401 17:29:33.548424 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.75709 | |
| I0401 17:29:33.548437 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.77256 (* 0.3 = 0.531767 loss) | |
| I0401 17:29:33.548452 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.508139 (* 0.3 = 0.152442 loss) | |
| I0401 17:29:33.548465 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.609433 | |
| I0401 17:29:33.548475 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.893821 | |
| I0401 17:29:33.548487 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.825553 | |
| I0401 17:29:33.548501 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.37826 (* 1 = 1.37826 loss) | |
| I0401 17:29:33.548514 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.38477 (* 1 = 0.38477 loss) | |
| I0401 17:29:33.548530 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.119 | |
| I0401 17:29:33.548542 6134 solver.cpp:406] Test net output #16: total_confidence = 0.0936669 | |
| I0401 17:29:33.700016 6134 solver.cpp:229] Iteration 65000, loss = 3.96614 | |
| I0401 17:29:33.700058 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.254902 | |
| I0401 17:29:33.700075 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 17:29:33.700091 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.490196 | |
| I0401 17:29:33.700106 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.24574 (* 0.3 = 0.673721 loss) | |
| I0401 17:29:33.700121 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.745885 (* 0.3 = 0.223765 loss) | |
| I0401 17:29:33.700134 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.372549 | |
| I0401 17:29:33.700146 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 17:29:33.700158 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.686275 | |
| I0401 17:29:33.700172 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.98942 (* 0.3 = 0.596827 loss) | |
| I0401 17:29:33.700186 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.650467 (* 0.3 = 0.19514 loss) | |
| I0401 17:29:33.700198 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.607843 | |
| I0401 17:29:33.700211 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0401 17:29:33.700222 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.862745 | |
| I0401 17:29:33.700237 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.06405 (* 1 = 1.06405 loss) | |
| I0401 17:29:33.700250 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.318267 (* 1 = 0.318267 loss) | |
| I0401 17:29:33.700263 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 17:29:33.700274 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0813158 | |
| I0401 17:29:33.700287 6134 sgd_solver.cpp:106] Iteration 65000, lr = 0.01 | |
| I0401 17:31:41.990008 6134 solver.cpp:229] Iteration 65500, loss = 4.0379 | |
| I0401 17:31:41.990145 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.341463 | |
| I0401 17:31:41.990166 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 17:31:41.990180 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.512195 | |
| I0401 17:31:41.990195 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.88506 (* 0.3 = 0.865518 loss) | |
| I0401 17:31:41.990211 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.766231 (* 0.3 = 0.229869 loss) | |
| I0401 17:31:41.990223 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.414634 | |
| I0401 17:31:41.990236 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 17:31:41.990247 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.585366 | |
| I0401 17:31:41.990262 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.26447 (* 0.3 = 0.679341 loss) | |
| I0401 17:31:41.990277 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.61447 (* 0.3 = 0.184341 loss) | |
| I0401 17:31:41.990288 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.634146 | |
| I0401 17:31:41.990300 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 17:31:41.990313 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.780488 | |
| I0401 17:31:41.990326 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.06672 (* 1 = 2.06672 loss) | |
| I0401 17:31:41.990340 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.550776 (* 1 = 0.550776 loss) | |
| I0401 17:31:41.990352 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 17:31:41.990365 6134 solver.cpp:245] Train net output #16: total_confidence = 0.108971 | |
| I0401 17:31:41.990376 6134 sgd_solver.cpp:106] Iteration 65500, lr = 0.01 | |
| I0401 17:33:50.185969 6134 solver.cpp:229] Iteration 66000, loss = 3.93847 | |
| I0401 17:33:50.186103 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0401 17:33:50.186122 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 17:33:50.186136 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.571429 | |
| I0401 17:33:50.186151 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.34294 (* 0.3 = 0.702882 loss) | |
| I0401 17:33:50.186167 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.704074 (* 0.3 = 0.211222 loss) | |
| I0401 17:33:50.186178 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.306122 | |
| I0401 17:33:50.186192 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0401 17:33:50.186203 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.77551 | |
| I0401 17:33:50.186218 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.92261 (* 0.3 = 0.576784 loss) | |
| I0401 17:33:50.186231 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.608189 (* 0.3 = 0.182457 loss) | |
| I0401 17:33:50.186244 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.591837 | |
| I0401 17:33:50.186256 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0401 17:33:50.186269 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.816327 | |
| I0401 17:33:50.186282 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.2872 (* 1 = 1.2872 loss) | |
| I0401 17:33:50.186296 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.386263 (* 1 = 0.386263 loss) | |
| I0401 17:33:50.186310 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:33:50.186321 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0770456 | |
| I0401 17:33:50.186332 6134 sgd_solver.cpp:106] Iteration 66000, lr = 0.01 | |
| I0401 17:35:58.407938 6134 solver.cpp:229] Iteration 66500, loss = 3.90261 | |
| I0401 17:35:58.408046 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26087 | |
| I0401 17:35:58.408067 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 17:35:58.408080 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.456522 | |
| I0401 17:35:58.408097 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.54622 (* 0.3 = 0.763867 loss) | |
| I0401 17:35:58.408112 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.7207 (* 0.3 = 0.21621 loss) | |
| I0401 17:35:58.408124 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.282609 | |
| I0401 17:35:58.408138 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 17:35:58.408149 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.521739 | |
| I0401 17:35:58.408164 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.22801 (* 0.3 = 0.668404 loss) | |
| I0401 17:35:58.408177 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.614371 (* 0.3 = 0.184311 loss) | |
| I0401 17:35:58.408190 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0401 17:35:58.408202 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 17:35:58.408213 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.804348 | |
| I0401 17:35:58.408227 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.49572 (* 1 = 1.49572 loss) | |
| I0401 17:35:58.408241 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.424602 (* 1 = 0.424602 loss) | |
| I0401 17:35:58.408253 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 17:35:58.408265 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0762414 | |
| I0401 17:35:58.408277 6134 sgd_solver.cpp:106] Iteration 66500, lr = 0.01 | |
| I0401 17:38:06.808480 6134 solver.cpp:229] Iteration 67000, loss = 3.88239 | |
| I0401 17:38:06.808804 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.375 | |
| I0401 17:38:06.808825 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0401 17:38:06.808838 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.53125 | |
| I0401 17:38:06.808854 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.10152 (* 0.3 = 0.630456 loss) | |
| I0401 17:38:06.808869 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.567548 (* 0.3 = 0.170264 loss) | |
| I0401 17:38:06.808882 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 17:38:06.808894 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 17:38:06.808907 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.71875 | |
| I0401 17:38:06.808919 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.80599 (* 0.3 = 0.541796 loss) | |
| I0401 17:38:06.808933 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.532591 (* 0.3 = 0.159777 loss) | |
| I0401 17:38:06.808946 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.65625 | |
| I0401 17:38:06.808959 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 17:38:06.808970 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.90625 | |
| I0401 17:38:06.808984 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.892781 (* 1 = 0.892781 loss) | |
| I0401 17:38:06.808998 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.25169 (* 1 = 0.25169 loss) | |
| I0401 17:38:06.809010 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 17:38:06.809022 6134 solver.cpp:245] Train net output #16: total_confidence = 0.205258 | |
| I0401 17:38:06.809034 6134 sgd_solver.cpp:106] Iteration 67000, lr = 0.01 | |
| I0401 17:40:15.463862 6134 solver.cpp:229] Iteration 67500, loss = 3.94169 | |
| I0401 17:40:15.463968 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.452381 | |
| I0401 17:40:15.463986 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 17:40:15.463999 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.571429 | |
| I0401 17:40:15.464015 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.2634 (* 0.3 = 0.679019 loss) | |
| I0401 17:40:15.464030 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.719392 (* 0.3 = 0.215818 loss) | |
| I0401 17:40:15.464042 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0401 17:40:15.464056 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 17:40:15.464071 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.690476 | |
| I0401 17:40:15.464098 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.95889 (* 0.3 = 0.587667 loss) | |
| I0401 17:40:15.464119 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.621226 (* 0.3 = 0.186368 loss) | |
| I0401 17:40:15.464131 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.595238 | |
| I0401 17:40:15.464144 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0401 17:40:15.464156 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.690476 | |
| I0401 17:40:15.464170 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.49385 (* 1 = 1.49385 loss) | |
| I0401 17:40:15.464184 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.465433 (* 1 = 0.465433 loss) | |
| I0401 17:40:15.464197 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 17:40:15.464210 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0699506 | |
| I0401 17:40:15.464221 6134 sgd_solver.cpp:106] Iteration 67500, lr = 0.01 | |
| I0401 17:42:23.990182 6134 solver.cpp:229] Iteration 68000, loss = 3.90662 | |
| I0401 17:42:23.990309 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.276596 | |
| I0401 17:42:23.990329 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 17:42:23.990341 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.510638 | |
| I0401 17:42:23.990356 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.52124 (* 0.3 = 0.756373 loss) | |
| I0401 17:42:23.990371 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.72541 (* 0.3 = 0.217623 loss) | |
| I0401 17:42:23.990384 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.382979 | |
| I0401 17:42:23.990397 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 17:42:23.990408 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.489362 | |
| I0401 17:42:23.990422 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.20122 (* 0.3 = 0.660366 loss) | |
| I0401 17:42:23.990437 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.619357 (* 0.3 = 0.185807 loss) | |
| I0401 17:42:23.990449 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.595745 | |
| I0401 17:42:23.990461 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0401 17:42:23.990473 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.787234 | |
| I0401 17:42:23.990488 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.36657 (* 1 = 1.36657 loss) | |
| I0401 17:42:23.990501 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.39679 (* 1 = 0.39679 loss) | |
| I0401 17:42:23.990514 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:42:23.990528 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0332478 | |
| I0401 17:42:23.990541 6134 sgd_solver.cpp:106] Iteration 68000, lr = 0.01 | |
| I0401 17:44:32.273198 6134 solver.cpp:229] Iteration 68500, loss = 3.84175 | |
| I0401 17:44:32.273357 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0401 17:44:32.273378 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 17:44:32.273392 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.489796 | |
| I0401 17:44:32.273408 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.45252 (* 0.3 = 0.735757 loss) | |
| I0401 17:44:32.273423 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.729678 (* 0.3 = 0.218903 loss) | |
| I0401 17:44:32.273435 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.408163 | |
| I0401 17:44:32.273448 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 17:44:32.273461 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.714286 | |
| I0401 17:44:32.273475 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.04819 (* 0.3 = 0.614457 loss) | |
| I0401 17:44:32.273489 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.624055 (* 0.3 = 0.187217 loss) | |
| I0401 17:44:32.273502 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.489796 | |
| I0401 17:44:32.273514 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0401 17:44:32.273530 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.816327 | |
| I0401 17:44:32.273543 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.52999 (* 1 = 1.52999 loss) | |
| I0401 17:44:32.273557 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.441626 (* 1 = 0.441626 loss) | |
| I0401 17:44:32.273571 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 17:44:32.273582 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0919086 | |
| I0401 17:44:32.273596 6134 sgd_solver.cpp:106] Iteration 68500, lr = 0.01 | |
| I0401 17:46:40.903487 6134 solver.cpp:229] Iteration 69000, loss = 3.8406 | |
| I0401 17:46:40.903774 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.226415 | |
| I0401 17:46:40.903795 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 17:46:40.903807 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.415094 | |
| I0401 17:46:40.903823 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.77535 (* 0.3 = 0.832604 loss) | |
| I0401 17:46:40.903838 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.893631 (* 0.3 = 0.268089 loss) | |
| I0401 17:46:40.903851 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.377358 | |
| I0401 17:46:40.903863 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 17:46:40.903875 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.584906 | |
| I0401 17:46:40.903888 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.22162 (* 0.3 = 0.666486 loss) | |
| I0401 17:46:40.903903 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.748036 (* 0.3 = 0.224411 loss) | |
| I0401 17:46:40.903914 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.396226 | |
| I0401 17:46:40.903928 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0401 17:46:40.903939 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.698113 | |
| I0401 17:46:40.903952 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.82232 (* 1 = 1.82232 loss) | |
| I0401 17:46:40.903966 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.63058 (* 1 = 0.63058 loss) | |
| I0401 17:46:40.903978 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:46:40.903990 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0795323 | |
| I0401 17:46:40.904003 6134 sgd_solver.cpp:106] Iteration 69000, lr = 0.01 | |
| I0401 17:48:49.080871 6134 solver.cpp:229] Iteration 69500, loss = 3.82833 | |
| I0401 17:48:49.080971 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.282609 | |
| I0401 17:48:49.080989 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 17:48:49.081002 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0401 17:48:49.081022 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.33711 (* 0.3 = 0.701132 loss) | |
| I0401 17:48:49.081037 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.677185 (* 0.3 = 0.203155 loss) | |
| I0401 17:48:49.081049 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.347826 | |
| I0401 17:48:49.081063 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0401 17:48:49.081074 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.630435 | |
| I0401 17:48:49.081089 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.0729 (* 0.3 = 0.62187 loss) | |
| I0401 17:48:49.081115 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.578448 (* 0.3 = 0.173534 loss) | |
| I0401 17:48:49.081130 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.565217 | |
| I0401 17:48:49.081142 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 17:48:49.081154 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.847826 | |
| I0401 17:48:49.081168 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.32064 (* 1 = 1.32064 loss) | |
| I0401 17:48:49.081182 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.417528 (* 1 = 0.417528 loss) | |
| I0401 17:48:49.081194 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 17:48:49.081207 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0512663 | |
| I0401 17:48:49.081219 6134 sgd_solver.cpp:106] Iteration 69500, lr = 0.01 | |
| I0401 17:50:57.428182 6134 solver.cpp:338] Iteration 70000, Testing net (#0) | |
| I0401 17:51:27.176236 6134 solver.cpp:393] Test loss: 3.1736 | |
| I0401 17:51:27.176282 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.328981 | |
| I0401 17:51:27.176301 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.827637 | |
| I0401 17:51:27.176312 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.63043 | |
| I0401 17:51:27.176328 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.21178 (* 0.3 = 0.663533 loss) | |
| I0401 17:51:27.176342 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.579677 (* 0.3 = 0.173903 loss) | |
| I0401 17:51:27.176354 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.473664 | |
| I0401 17:51:27.176367 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.861503 | |
| I0401 17:51:27.176378 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.74374 | |
| I0401 17:51:27.176391 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.80672 (* 0.3 = 0.542016 loss) | |
| I0401 17:51:27.176405 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.479979 (* 0.3 = 0.143994 loss) | |
| I0401 17:51:27.176417 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.653201 | |
| I0401 17:51:27.176429 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.902275 | |
| I0401 17:51:27.176440 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.836883 | |
| I0401 17:51:27.176453 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.28923 (* 1 = 1.28923 loss) | |
| I0401 17:51:27.176467 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.360926 (* 1 = 0.360926 loss) | |
| I0401 17:51:27.176478 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.15 | |
| I0401 17:51:27.176491 6134 solver.cpp:406] Test net output #16: total_confidence = 0.13014 | |
| I0401 17:51:27.327000 6134 solver.cpp:229] Iteration 70000, loss = 3.8372 | |
| I0401 17:51:27.327040 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.210526 | |
| I0401 17:51:27.327059 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 17:51:27.327072 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.368421 | |
| I0401 17:51:27.327087 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.04736 (* 0.3 = 0.914209 loss) | |
| I0401 17:51:27.327102 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.744872 (* 0.3 = 0.223462 loss) | |
| I0401 17:51:27.327114 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.289474 | |
| I0401 17:51:27.327126 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0401 17:51:27.327137 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.605263 | |
| I0401 17:51:27.327152 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.37881 (* 0.3 = 0.713643 loss) | |
| I0401 17:51:27.327165 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.575177 (* 0.3 = 0.172553 loss) | |
| I0401 17:51:27.327178 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.421053 | |
| I0401 17:51:27.327189 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0401 17:51:27.327200 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.684211 | |
| I0401 17:51:27.327214 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.78649 (* 1 = 1.78649 loss) | |
| I0401 17:51:27.327227 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.447483 (* 1 = 0.447483 loss) | |
| I0401 17:51:27.327239 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:51:27.327251 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0254203 | |
| I0401 17:51:27.327263 6134 sgd_solver.cpp:106] Iteration 70000, lr = 0.01 | |
| I0401 17:53:35.587855 6134 solver.cpp:229] Iteration 70500, loss = 3.77914 | |
| I0401 17:53:35.587985 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.211538 | |
| I0401 17:53:35.588006 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 17:53:35.588018 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.365385 | |
| I0401 17:53:35.588034 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.88647 (* 0.3 = 0.86594 loss) | |
| I0401 17:53:35.588049 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.890258 (* 0.3 = 0.267077 loss) | |
| I0401 17:53:35.588062 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.230769 | |
| I0401 17:53:35.588074 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 17:53:35.588086 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.5 | |
| I0401 17:53:35.588100 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.41529 (* 0.3 = 0.724588 loss) | |
| I0401 17:53:35.588114 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.745194 (* 0.3 = 0.223558 loss) | |
| I0401 17:53:35.588126 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.576923 | |
| I0401 17:53:35.588138 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 17:53:35.588150 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.807692 | |
| I0401 17:53:35.588165 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.41175 (* 1 = 1.41175 loss) | |
| I0401 17:53:35.588178 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.422466 (* 1 = 0.422466 loss) | |
| I0401 17:53:35.588191 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 17:53:35.588203 6134 solver.cpp:245] Train net output #16: total_confidence = 0.175238 | |
| I0401 17:53:35.588215 6134 sgd_solver.cpp:106] Iteration 70500, lr = 0.01 | |
| I0401 17:55:43.908488 6134 solver.cpp:229] Iteration 71000, loss = 3.75153 | |
| I0401 17:55:43.908596 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.166667 | |
| I0401 17:55:43.908617 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0401 17:55:43.908628 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.296296 | |
| I0401 17:55:43.908645 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.44922 (* 0.3 = 1.03477 loss) | |
| I0401 17:55:43.908660 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.08867 (* 0.3 = 0.326601 loss) | |
| I0401 17:55:43.908673 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.203704 | |
| I0401 17:55:43.908685 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.755682 | |
| I0401 17:55:43.908697 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.444444 | |
| I0401 17:55:43.908711 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 3.29326 (* 0.3 = 0.987979 loss) | |
| I0401 17:55:43.908725 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 1.03915 (* 0.3 = 0.311744 loss) | |
| I0401 17:55:43.908737 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.388889 | |
| I0401 17:55:43.908751 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.806818 | |
| I0401 17:55:43.908762 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.574074 | |
| I0401 17:55:43.908776 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 3.33628 (* 1 = 3.33628 loss) | |
| I0401 17:55:43.908789 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 1.03202 (* 1 = 1.03202 loss) | |
| I0401 17:55:43.908802 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 17:55:43.908814 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0631788 | |
| I0401 17:55:43.908826 6134 sgd_solver.cpp:106] Iteration 71000, lr = 0.01 | |
| I0401 17:57:52.105262 6134 solver.cpp:229] Iteration 71500, loss = 3.74939 | |
| I0401 17:57:52.105579 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.25 | |
| I0401 17:57:52.105600 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 17:57:52.105613 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55 | |
| I0401 17:57:52.105630 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.29688 (* 0.3 = 0.689065 loss) | |
| I0401 17:57:52.105645 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.669629 (* 0.3 = 0.200889 loss) | |
| I0401 17:57:52.105657 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.375 | |
| I0401 17:57:52.105670 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 17:57:52.105682 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.575 | |
| I0401 17:57:52.105696 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.23297 (* 0.3 = 0.66989 loss) | |
| I0401 17:57:52.105711 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.599549 (* 0.3 = 0.179865 loss) | |
| I0401 17:57:52.105723 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.425 | |
| I0401 17:57:52.105736 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0401 17:57:52.105747 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.65 | |
| I0401 17:57:52.105762 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.82398 (* 1 = 1.82398 loss) | |
| I0401 17:57:52.105775 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.465368 (* 1 = 0.465368 loss) | |
| I0401 17:57:52.105787 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 17:57:52.105799 6134 solver.cpp:245] Train net output #16: total_confidence = 0.130631 | |
| I0401 17:57:52.105811 6134 sgd_solver.cpp:106] Iteration 71500, lr = 0.01 | |
| I0401 18:00:00.514313 6134 solver.cpp:229] Iteration 72000, loss = 3.74243 | |
| I0401 18:00:00.514421 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.342105 | |
| I0401 18:00:00.514441 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 18:00:00.514453 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0401 18:00:00.514470 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.48317 (* 0.3 = 0.744951 loss) | |
| I0401 18:00:00.514484 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.750185 (* 0.3 = 0.225056 loss) | |
| I0401 18:00:00.514497 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.394737 | |
| I0401 18:00:00.514510 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 18:00:00.514524 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.710526 | |
| I0401 18:00:00.514539 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.98971 (* 0.3 = 0.596914 loss) | |
| I0401 18:00:00.514554 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.666043 (* 0.3 = 0.199813 loss) | |
| I0401 18:00:00.514566 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.552632 | |
| I0401 18:00:00.514578 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 18:00:00.514590 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.763158 | |
| I0401 18:00:00.514605 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.46106 (* 1 = 1.46106 loss) | |
| I0401 18:00:00.514618 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.50215 (* 1 = 0.50215 loss) | |
| I0401 18:00:00.514631 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:00:00.514642 6134 solver.cpp:245] Train net output #16: total_confidence = 0.175871 | |
| I0401 18:00:00.514654 6134 sgd_solver.cpp:106] Iteration 72000, lr = 0.01 | |
| I0401 18:02:08.934254 6134 solver.cpp:229] Iteration 72500, loss = 3.79193 | |
| I0401 18:02:08.934386 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.153846 | |
| I0401 18:02:08.934406 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0401 18:02:08.934418 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.384615 | |
| I0401 18:02:08.934434 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.67971 (* 0.3 = 0.803913 loss) | |
| I0401 18:02:08.934449 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.86045 (* 0.3 = 0.258135 loss) | |
| I0401 18:02:08.934463 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.192308 | |
| I0401 18:02:08.934474 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.727273 | |
| I0401 18:02:08.934486 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.461538 | |
| I0401 18:02:08.934500 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.54134 (* 0.3 = 0.762402 loss) | |
| I0401 18:02:08.934514 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.856744 (* 0.3 = 0.257023 loss) | |
| I0401 18:02:08.934530 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.442308 | |
| I0401 18:02:08.934541 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.8125 | |
| I0401 18:02:08.934554 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.653846 | |
| I0401 18:02:08.934568 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.88182 (* 1 = 1.88182 loss) | |
| I0401 18:02:08.934582 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.628757 (* 1 = 0.628757 loss) | |
| I0401 18:02:08.934594 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 18:02:08.934607 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0114213 | |
| I0401 18:02:08.934619 6134 sgd_solver.cpp:106] Iteration 72500, lr = 0.01 | |
| I0401 18:04:17.212348 6134 solver.cpp:229] Iteration 73000, loss = 3.73631 | |
| I0401 18:04:17.212460 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.326087 | |
| I0401 18:04:17.212481 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 18:04:17.212492 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.630435 | |
| I0401 18:04:17.212508 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.15291 (* 0.3 = 0.645872 loss) | |
| I0401 18:04:17.212527 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.630798 (* 0.3 = 0.189239 loss) | |
| I0401 18:04:17.212539 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.478261 | |
| I0401 18:04:17.212551 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 18:04:17.212563 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.652174 | |
| I0401 18:04:17.212577 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.82347 (* 0.3 = 0.547041 loss) | |
| I0401 18:04:17.212591 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.532243 (* 0.3 = 0.159673 loss) | |
| I0401 18:04:17.212604 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.73913 | |
| I0401 18:04:17.212616 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 18:04:17.212628 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.891304 | |
| I0401 18:04:17.212642 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.966641 (* 1 = 0.966641 loss) | |
| I0401 18:04:17.212656 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.27974 (* 1 = 0.27974 loss) | |
| I0401 18:04:17.212669 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:04:17.212682 6134 solver.cpp:245] Train net output #16: total_confidence = 0.131373 | |
| I0401 18:04:17.212693 6134 sgd_solver.cpp:106] Iteration 73000, lr = 0.01 | |
| I0401 18:06:25.800891 6134 solver.cpp:229] Iteration 73500, loss = 3.6838 | |
| I0401 18:06:25.801257 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.145455 | |
| I0401 18:06:25.801280 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0401 18:06:25.801293 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.509091 | |
| I0401 18:06:25.801311 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.67939 (* 0.3 = 0.803816 loss) | |
| I0401 18:06:25.801326 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.853752 (* 0.3 = 0.256126 loss) | |
| I0401 18:06:25.801337 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.163636 | |
| I0401 18:06:25.801349 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.738636 | |
| I0401 18:06:25.801362 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.4 | |
| I0401 18:06:25.801375 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.82543 (* 0.3 = 0.847631 loss) | |
| I0401 18:06:25.801389 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.897462 (* 0.3 = 0.269239 loss) | |
| I0401 18:06:25.801403 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.381818 | |
| I0401 18:06:25.801414 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.795455 | |
| I0401 18:06:25.801426 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.581818 | |
| I0401 18:06:25.801440 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.32059 (* 1 = 2.32059 loss) | |
| I0401 18:06:25.801455 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.753907 (* 1 = 0.753907 loss) | |
| I0401 18:06:25.801466 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 18:06:25.801478 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0355633 | |
| I0401 18:06:25.801491 6134 sgd_solver.cpp:106] Iteration 73500, lr = 0.01 | |
| I0401 18:08:34.352701 6134 solver.cpp:229] Iteration 74000, loss = 3.67055 | |
| I0401 18:08:34.352814 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.395349 | |
| I0401 18:08:34.352835 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 18:08:34.352849 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.744186 | |
| I0401 18:08:34.352864 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.79029 (* 0.3 = 0.537088 loss) | |
| I0401 18:08:34.352880 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.546258 (* 0.3 = 0.163877 loss) | |
| I0401 18:08:34.352892 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.534884 | |
| I0401 18:08:34.352905 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0401 18:08:34.352916 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.744186 | |
| I0401 18:08:34.352931 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.61499 (* 0.3 = 0.484496 loss) | |
| I0401 18:08:34.352946 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.466189 (* 0.3 = 0.139857 loss) | |
| I0401 18:08:34.352957 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.72093 | |
| I0401 18:08:34.352970 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0401 18:08:34.352982 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.906977 | |
| I0401 18:08:34.352996 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.740577 (* 1 = 0.740577 loss) | |
| I0401 18:08:34.353010 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.259306 (* 1 = 0.259306 loss) | |
| I0401 18:08:34.353023 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:08:34.353034 6134 solver.cpp:245] Train net output #16: total_confidence = 0.131714 | |
| I0401 18:08:34.353062 6134 sgd_solver.cpp:106] Iteration 74000, lr = 0.01 | |
| I0401 18:10:42.635143 6134 solver.cpp:229] Iteration 74500, loss = 3.64782 | |
| I0401 18:10:42.635311 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.365385 | |
| I0401 18:10:42.635334 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 18:10:42.635347 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.557692 | |
| I0401 18:10:42.635363 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.36393 (* 0.3 = 0.709178 loss) | |
| I0401 18:10:42.635378 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.731393 (* 0.3 = 0.219418 loss) | |
| I0401 18:10:42.635390 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.384615 | |
| I0401 18:10:42.635403 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 18:10:42.635416 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.692308 | |
| I0401 18:10:42.635428 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.97748 (* 0.3 = 0.593243 loss) | |
| I0401 18:10:42.635443 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.599985 (* 0.3 = 0.179995 loss) | |
| I0401 18:10:42.635455 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.576923 | |
| I0401 18:10:42.635468 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0401 18:10:42.635479 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.769231 | |
| I0401 18:10:42.635493 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.44214 (* 1 = 1.44214 loss) | |
| I0401 18:10:42.635507 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.4516 (* 1 = 0.4516 loss) | |
| I0401 18:10:42.635522 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 18:10:42.635535 6134 solver.cpp:245] Train net output #16: total_confidence = 0.100938 | |
| I0401 18:10:42.635548 6134 sgd_solver.cpp:106] Iteration 74500, lr = 0.01 | |
| I0401 18:12:50.868577 6134 solver.cpp:338] Iteration 75000, Testing net (#0) | |
| I0401 18:13:20.626327 6134 solver.cpp:393] Test loss: 3.30018 | |
| I0401 18:13:20.626386 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.330592 | |
| I0401 18:13:20.626405 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.828728 | |
| I0401 18:13:20.626417 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.643262 | |
| I0401 18:13:20.626433 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.18397 (* 0.3 = 0.65519 loss) | |
| I0401 18:13:20.626449 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.571765 (* 0.3 = 0.171529 loss) | |
| I0401 18:13:20.626461 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.477485 | |
| I0401 18:13:20.626473 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.866821 | |
| I0401 18:13:20.626485 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.780307 | |
| I0401 18:13:20.626499 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.75956 (* 0.3 = 0.527869 loss) | |
| I0401 18:13:20.626513 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.456135 (* 0.3 = 0.136841 loss) | |
| I0401 18:13:20.626528 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.639665 | |
| I0401 18:13:20.626540 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.908502 | |
| I0401 18:13:20.626552 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.836854 | |
| I0401 18:13:20.626566 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.43436 (* 1 = 1.43436 loss) | |
| I0401 18:13:20.626579 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.374386 (* 1 = 0.374386 loss) | |
| I0401 18:13:20.626591 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.207 | |
| I0401 18:13:20.626603 6134 solver.cpp:406] Test net output #16: total_confidence = 0.248664 | |
| I0401 18:13:20.777840 6134 solver.cpp:229] Iteration 75000, loss = 3.70051 | |
| I0401 18:13:20.777889 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.372093 | |
| I0401 18:13:20.777906 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 18:13:20.777920 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55814 | |
| I0401 18:13:20.777935 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.57136 (* 0.3 = 0.771407 loss) | |
| I0401 18:13:20.777951 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.730975 (* 0.3 = 0.219293 loss) | |
| I0401 18:13:20.777963 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.511628 | |
| I0401 18:13:20.777976 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 18:13:20.777988 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.767442 | |
| I0401 18:13:20.778002 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.8358 (* 0.3 = 0.55074 loss) | |
| I0401 18:13:20.778019 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.521182 (* 0.3 = 0.156355 loss) | |
| I0401 18:13:20.778033 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.674419 | |
| I0401 18:13:20.778045 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 18:13:20.778056 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.883721 | |
| I0401 18:13:20.778072 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.14841 (* 1 = 1.14841 loss) | |
| I0401 18:13:20.778086 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.302187 (* 1 = 0.302187 loss) | |
| I0401 18:13:20.778098 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:13:20.778110 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0954722 | |
| I0401 18:13:20.778123 6134 sgd_solver.cpp:106] Iteration 75000, lr = 0.01 | |
| I0401 18:15:29.253588 6134 solver.cpp:229] Iteration 75500, loss = 3.73708 | |
| I0401 18:15:29.253716 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0401 18:15:29.253737 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 18:15:29.253751 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.595238 | |
| I0401 18:15:29.253767 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.21328 (* 0.3 = 0.663985 loss) | |
| I0401 18:15:29.253782 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.655727 (* 0.3 = 0.196718 loss) | |
| I0401 18:15:29.253794 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0401 18:15:29.253806 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0401 18:15:29.253818 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0401 18:15:29.253832 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.07178 (* 0.3 = 0.621533 loss) | |
| I0401 18:15:29.253847 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.676919 (* 0.3 = 0.203076 loss) | |
| I0401 18:15:29.253859 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.619048 | |
| I0401 18:15:29.253871 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0401 18:15:29.253883 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.857143 | |
| I0401 18:15:29.253896 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.24579 (* 1 = 1.24579 loss) | |
| I0401 18:15:29.253911 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.406359 (* 1 = 0.406359 loss) | |
| I0401 18:15:29.253922 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 18:15:29.253934 6134 solver.cpp:245] Train net output #16: total_confidence = 0.125765 | |
| I0401 18:15:29.253947 6134 sgd_solver.cpp:106] Iteration 75500, lr = 0.01 | |
| I0401 18:17:37.760825 6134 solver.cpp:229] Iteration 76000, loss = 3.73529 | |
| I0401 18:17:37.761139 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.297872 | |
| I0401 18:17:37.761159 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 18:17:37.761173 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.510638 | |
| I0401 18:17:37.761189 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.36241 (* 0.3 = 0.708723 loss) | |
| I0401 18:17:37.761204 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.671083 (* 0.3 = 0.201325 loss) | |
| I0401 18:17:37.761216 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.425532 | |
| I0401 18:17:37.761229 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 18:17:37.761241 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.617021 | |
| I0401 18:17:37.761255 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.30858 (* 0.3 = 0.692574 loss) | |
| I0401 18:17:37.761270 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.647261 (* 0.3 = 0.194178 loss) | |
| I0401 18:17:37.761281 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.595745 | |
| I0401 18:17:37.761293 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 18:17:37.761306 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.829787 | |
| I0401 18:17:37.761319 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.34348 (* 1 = 1.34348 loss) | |
| I0401 18:17:37.761333 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.375593 (* 1 = 0.375593 loss) | |
| I0401 18:17:37.761345 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 18:17:37.761358 6134 solver.cpp:245] Train net output #16: total_confidence = 0.110559 | |
| I0401 18:17:37.761370 6134 sgd_solver.cpp:106] Iteration 76000, lr = 0.01 | |
| I0401 18:19:46.244030 6134 solver.cpp:229] Iteration 76500, loss = 3.62966 | |
| I0401 18:19:46.244135 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.282609 | |
| I0401 18:19:46.244155 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 18:19:46.244168 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.543478 | |
| I0401 18:19:46.244185 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.55073 (* 0.3 = 0.765218 loss) | |
| I0401 18:19:46.244200 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.736422 (* 0.3 = 0.220927 loss) | |
| I0401 18:19:46.244212 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.478261 | |
| I0401 18:19:46.244225 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 18:19:46.244237 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.630435 | |
| I0401 18:19:46.244251 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.09183 (* 0.3 = 0.627548 loss) | |
| I0401 18:19:46.244266 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.646853 (* 0.3 = 0.194056 loss) | |
| I0401 18:19:46.244278 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.586957 | |
| I0401 18:19:46.244290 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 18:19:46.244302 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.826087 | |
| I0401 18:19:46.244316 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.66813 (* 1 = 1.66813 loss) | |
| I0401 18:19:46.244330 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.450424 (* 1 = 0.450424 loss) | |
| I0401 18:19:46.244343 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:19:46.244355 6134 solver.cpp:245] Train net output #16: total_confidence = 0.188193 | |
| I0401 18:19:46.244367 6134 sgd_solver.cpp:106] Iteration 76500, lr = 0.01 | |
| I0401 18:21:55.018409 6134 solver.cpp:229] Iteration 77000, loss = 3.72351 | |
| I0401 18:21:55.018537 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.12963 | |
| I0401 18:21:55.018556 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.732955 | |
| I0401 18:21:55.018570 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.37037 | |
| I0401 18:21:55.018586 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.88418 (* 0.3 = 0.865253 loss) | |
| I0401 18:21:55.018601 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.921656 (* 0.3 = 0.276497 loss) | |
| I0401 18:21:55.018613 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.222222 | |
| I0401 18:21:55.018626 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0401 18:21:55.018638 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.462963 | |
| I0401 18:21:55.018651 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.60298 (* 0.3 = 0.780895 loss) | |
| I0401 18:21:55.018666 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.827465 (* 0.3 = 0.248239 loss) | |
| I0401 18:21:55.018678 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.425926 | |
| I0401 18:21:55.018690 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0401 18:21:55.018702 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.611111 | |
| I0401 18:21:55.018717 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.12539 (* 1 = 2.12539 loss) | |
| I0401 18:21:55.018730 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.677844 (* 1 = 0.677844 loss) | |
| I0401 18:21:55.018743 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 18:21:55.018754 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0360006 | |
| I0401 18:21:55.018766 6134 sgd_solver.cpp:106] Iteration 77000, lr = 0.01 | |
| I0401 18:24:03.829697 6134 solver.cpp:229] Iteration 77500, loss = 3.66458 | |
| I0401 18:24:03.829824 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.22 | |
| I0401 18:24:03.829845 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 18:24:03.829859 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.42 | |
| I0401 18:24:03.829875 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.803 (* 0.3 = 0.840901 loss) | |
| I0401 18:24:03.829890 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.885845 (* 0.3 = 0.265754 loss) | |
| I0401 18:24:03.829903 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.28 | |
| I0401 18:24:03.829916 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 18:24:03.829928 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.52 | |
| I0401 18:24:03.829942 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.60076 (* 0.3 = 0.780228 loss) | |
| I0401 18:24:03.829957 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.819843 (* 0.3 = 0.245953 loss) | |
| I0401 18:24:03.829969 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.54 | |
| I0401 18:24:03.829982 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0401 18:24:03.829993 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.78 | |
| I0401 18:24:03.830008 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.70471 (* 1 = 1.70471 loss) | |
| I0401 18:24:03.830023 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.529584 (* 1 = 0.529584 loss) | |
| I0401 18:24:03.830034 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:24:03.830046 6134 solver.cpp:245] Train net output #16: total_confidence = 0.171903 | |
| I0401 18:24:03.830059 6134 sgd_solver.cpp:106] Iteration 77500, lr = 0.01 | |
| I0401 18:26:12.902709 6134 solver.cpp:229] Iteration 78000, loss = 3.71421 | |
| I0401 18:26:12.903139 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.195652 | |
| I0401 18:26:12.903161 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 18:26:12.903174 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.543478 | |
| I0401 18:26:12.903192 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.58681 (* 0.3 = 0.776044 loss) | |
| I0401 18:26:12.903206 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.774121 (* 0.3 = 0.232236 loss) | |
| I0401 18:26:12.903220 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 18:26:12.903234 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 18:26:12.903245 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.586957 | |
| I0401 18:26:12.903259 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.08786 (* 0.3 = 0.626357 loss) | |
| I0401 18:26:12.903275 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.650636 (* 0.3 = 0.195191 loss) | |
| I0401 18:26:12.903286 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.717391 | |
| I0401 18:26:12.903300 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0401 18:26:12.903311 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.847826 | |
| I0401 18:26:12.903326 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.14852 (* 1 = 1.14852 loss) | |
| I0401 18:26:12.903340 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.361035 (* 1 = 0.361035 loss) | |
| I0401 18:26:12.903352 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:26:12.903364 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0785482 | |
| I0401 18:26:12.903378 6134 sgd_solver.cpp:106] Iteration 78000, lr = 0.01 | |
| I0401 18:28:21.532691 6134 solver.cpp:229] Iteration 78500, loss = 3.65651 | |
| I0401 18:28:21.532809 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.18 | |
| I0401 18:28:21.532830 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 18:28:21.532843 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.48 | |
| I0401 18:28:21.532860 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.76868 (* 0.3 = 0.830604 loss) | |
| I0401 18:28:21.532874 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.806897 (* 0.3 = 0.242069 loss) | |
| I0401 18:28:21.532887 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.42 | |
| I0401 18:28:21.532899 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 18:28:21.532912 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.58 | |
| I0401 18:28:21.532925 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.43004 (* 0.3 = 0.729012 loss) | |
| I0401 18:28:21.532939 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.696357 (* 0.3 = 0.208907 loss) | |
| I0401 18:28:21.532953 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.7 | |
| I0401 18:28:21.532963 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0401 18:28:21.532975 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.78 | |
| I0401 18:28:21.532989 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.36938 (* 1 = 1.36938 loss) | |
| I0401 18:28:21.533004 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.40521 (* 1 = 0.40521 loss) | |
| I0401 18:28:21.533016 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:28:21.533028 6134 solver.cpp:245] Train net output #16: total_confidence = 0.149489 | |
| I0401 18:28:21.533041 6134 sgd_solver.cpp:106] Iteration 78500, lr = 0.01 | |
| I0401 18:30:30.260901 6134 solver.cpp:229] Iteration 79000, loss = 3.59306 | |
| I0401 18:30:30.261044 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.181818 | |
| I0401 18:30:30.261065 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 18:30:30.261077 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.340909 | |
| I0401 18:30:30.261102 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.1264 (* 0.3 = 0.937921 loss) | |
| I0401 18:30:30.261117 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.850531 (* 0.3 = 0.255159 loss) | |
| I0401 18:30:30.261129 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.181818 | |
| I0401 18:30:30.261142 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0401 18:30:30.261154 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.386364 | |
| I0401 18:30:30.261168 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.84522 (* 0.3 = 0.853567 loss) | |
| I0401 18:30:30.261183 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.78377 (* 0.3 = 0.235131 loss) | |
| I0401 18:30:30.261194 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.295455 | |
| I0401 18:30:30.261207 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.823864 | |
| I0401 18:30:30.261219 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.522727 | |
| I0401 18:30:30.261234 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.23049 (* 1 = 2.23049 loss) | |
| I0401 18:30:30.261247 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.590498 (* 1 = 0.590498 loss) | |
| I0401 18:30:30.261260 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 18:30:30.261271 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0241516 | |
| I0401 18:30:30.261284 6134 sgd_solver.cpp:106] Iteration 79000, lr = 0.01 | |
| I0401 18:32:39.091327 6134 solver.cpp:229] Iteration 79500, loss = 3.63374 | |
| I0401 18:32:39.091442 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.183673 | |
| I0401 18:32:39.091462 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.761364 | |
| I0401 18:32:39.091475 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.469388 | |
| I0401 18:32:39.091491 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.03702 (* 0.3 = 0.911107 loss) | |
| I0401 18:32:39.091506 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.901503 (* 0.3 = 0.270451 loss) | |
| I0401 18:32:39.091521 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.326531 | |
| I0401 18:32:39.091536 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0401 18:32:39.091547 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.591837 | |
| I0401 18:32:39.091562 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.55482 (* 0.3 = 0.766447 loss) | |
| I0401 18:32:39.091577 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.75138 (* 0.3 = 0.225414 loss) | |
| I0401 18:32:39.091588 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.44898 | |
| I0401 18:32:39.091600 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.840909 | |
| I0401 18:32:39.091612 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.714286 | |
| I0401 18:32:39.091626 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.06939 (* 1 = 2.06939 loss) | |
| I0401 18:32:39.091640 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.612282 (* 1 = 0.612282 loss) | |
| I0401 18:32:39.091653 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 18:32:39.091665 6134 solver.cpp:245] Train net output #16: total_confidence = 0.153413 | |
| I0401 18:32:39.091677 6134 sgd_solver.cpp:106] Iteration 79500, lr = 0.01 | |
| I0401 18:34:47.871453 6134 solver.cpp:338] Iteration 80000, Testing net (#0) | |
| I0401 18:35:17.778465 6134 solver.cpp:393] Test loss: 2.9809 | |
| I0401 18:35:17.778513 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.365732 | |
| I0401 18:35:17.778532 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.83582 | |
| I0401 18:35:17.778545 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.653551 | |
| I0401 18:35:17.778561 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.07837 (* 0.3 = 0.623512 loss) | |
| I0401 18:35:17.778575 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.556518 (* 0.3 = 0.166956 loss) | |
| I0401 18:35:17.778589 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.516335 | |
| I0401 18:35:17.778600 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.870775 | |
| I0401 18:35:17.778612 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.78435 | |
| I0401 18:35:17.778625 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.6623 (* 0.3 = 0.498691 loss) | |
| I0401 18:35:17.778640 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.453169 (* 0.3 = 0.135951 loss) | |
| I0401 18:35:17.778651 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.675791 | |
| I0401 18:35:17.778663 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.913456 | |
| I0401 18:35:17.778674 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.845124 | |
| I0401 18:35:17.778688 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.22299 (* 1 = 1.22299 loss) | |
| I0401 18:35:17.778702 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.332808 (* 1 = 0.332808 loss) | |
| I0401 18:35:17.778713 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.222 | |
| I0401 18:35:17.778725 6134 solver.cpp:406] Test net output #16: total_confidence = 0.206636 | |
| I0401 18:35:17.930132 6134 solver.cpp:229] Iteration 80000, loss = 3.53752 | |
| I0401 18:35:17.930229 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.270833 | |
| I0401 18:35:17.930248 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 18:35:17.930261 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0401 18:35:17.930277 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.53855 (* 0.3 = 0.761564 loss) | |
| I0401 18:35:17.930292 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.794218 (* 0.3 = 0.238265 loss) | |
| I0401 18:35:17.930305 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0401 18:35:17.930316 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 18:35:17.930327 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.541667 | |
| I0401 18:35:17.930341 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.45116 (* 0.3 = 0.735348 loss) | |
| I0401 18:35:17.930356 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.781652 (* 0.3 = 0.234496 loss) | |
| I0401 18:35:17.930367 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.458333 | |
| I0401 18:35:17.930378 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0401 18:35:17.930390 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.625 | |
| I0401 18:35:17.930403 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.01601 (* 1 = 2.01601 loss) | |
| I0401 18:35:17.930418 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.656107 (* 1 = 0.656107 loss) | |
| I0401 18:35:17.930430 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 18:35:17.930443 6134 solver.cpp:245] Train net output #16: total_confidence = 0.115224 | |
| I0401 18:35:17.930455 6134 sgd_solver.cpp:106] Iteration 80000, lr = 0.01 | |
| I0401 18:37:26.797549 6134 solver.cpp:229] Iteration 80500, loss = 3.53801 | |
| I0401 18:37:26.797909 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.254902 | |
| I0401 18:37:26.797930 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 18:37:26.797943 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54902 | |
| I0401 18:37:26.797960 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.39837 (* 0.3 = 0.719511 loss) | |
| I0401 18:37:26.797983 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.756839 (* 0.3 = 0.227052 loss) | |
| I0401 18:37:26.797996 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0401 18:37:26.798008 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 18:37:26.798020 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.529412 | |
| I0401 18:37:26.798039 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.39304 (* 0.3 = 0.717911 loss) | |
| I0401 18:37:26.798053 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.73354 (* 0.3 = 0.220062 loss) | |
| I0401 18:37:26.798066 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.588235 | |
| I0401 18:37:26.798079 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0401 18:37:26.798090 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.843137 | |
| I0401 18:37:26.798105 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.23427 (* 1 = 1.23427 loss) | |
| I0401 18:37:26.798118 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.422827 (* 1 = 0.422827 loss) | |
| I0401 18:37:26.798130 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 18:37:26.798142 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0818318 | |
| I0401 18:37:26.798154 6134 sgd_solver.cpp:106] Iteration 80500, lr = 0.01 | |
| I0401 18:39:35.673382 6134 solver.cpp:229] Iteration 81000, loss = 3.55559 | |
| I0401 18:39:35.673480 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0401 18:39:35.673499 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 18:39:35.673513 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0401 18:39:35.673529 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.99775 (* 0.3 = 0.599325 loss) | |
| I0401 18:39:35.673544 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.654974 (* 0.3 = 0.196492 loss) | |
| I0401 18:39:35.673557 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488889 | |
| I0401 18:39:35.673570 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0401 18:39:35.673583 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.755556 | |
| I0401 18:39:35.673596 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.76535 (* 0.3 = 0.529604 loss) | |
| I0401 18:39:35.673610 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.584736 (* 0.3 = 0.175421 loss) | |
| I0401 18:39:35.673624 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.777778 | |
| I0401 18:39:35.673635 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0401 18:39:35.673647 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.911111 | |
| I0401 18:39:35.673661 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.781245 (* 1 = 0.781245 loss) | |
| I0401 18:39:35.673676 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.27184 (* 1 = 0.27184 loss) | |
| I0401 18:39:35.673688 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:39:35.673701 6134 solver.cpp:245] Train net output #16: total_confidence = 0.237028 | |
| I0401 18:39:35.673712 6134 sgd_solver.cpp:106] Iteration 81000, lr = 0.01 | |
| I0401 18:41:44.611836 6134 solver.cpp:229] Iteration 81500, loss = 3.57285 | |
| I0401 18:41:44.611976 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.22449 | |
| I0401 18:41:44.612005 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0401 18:41:44.612018 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.469388 | |
| I0401 18:41:44.612035 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.80364 (* 0.3 = 0.841093 loss) | |
| I0401 18:41:44.612049 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.849722 (* 0.3 = 0.254917 loss) | |
| I0401 18:41:44.612061 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.306122 | |
| I0401 18:41:44.612079 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 18:41:44.612092 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.469388 | |
| I0401 18:41:44.612104 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.68281 (* 0.3 = 0.804843 loss) | |
| I0401 18:41:44.612119 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.804082 (* 0.3 = 0.241225 loss) | |
| I0401 18:41:44.612131 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.489796 | |
| I0401 18:41:44.612143 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0401 18:41:44.612155 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.653061 | |
| I0401 18:41:44.612175 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.92368 (* 1 = 1.92368 loss) | |
| I0401 18:41:44.612190 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.563289 (* 1 = 0.563289 loss) | |
| I0401 18:41:44.612201 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:41:44.612212 6134 solver.cpp:245] Train net output #16: total_confidence = 0.122421 | |
| I0401 18:41:44.612233 6134 sgd_solver.cpp:106] Iteration 81500, lr = 0.01 | |
| I0401 18:43:53.438127 6134 solver.cpp:229] Iteration 82000, loss = 3.60812 | |
| I0401 18:43:53.438256 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.255814 | |
| I0401 18:43:53.438290 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 18:43:53.438311 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.488372 | |
| I0401 18:43:53.438338 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.44057 (* 0.3 = 0.732172 loss) | |
| I0401 18:43:53.438352 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.652981 (* 0.3 = 0.195894 loss) | |
| I0401 18:43:53.438365 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.372093 | |
| I0401 18:43:53.438379 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0401 18:43:53.438397 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.55814 | |
| I0401 18:43:53.438411 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.2212 (* 0.3 = 0.666359 loss) | |
| I0401 18:43:53.438426 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.607782 (* 0.3 = 0.182335 loss) | |
| I0401 18:43:53.438439 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.44186 | |
| I0401 18:43:53.438452 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0401 18:43:53.438464 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.674419 | |
| I0401 18:43:53.438478 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.91378 (* 1 = 1.91378 loss) | |
| I0401 18:43:53.438493 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.506762 (* 1 = 0.506762 loss) | |
| I0401 18:43:53.438504 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 18:43:53.438519 6134 solver.cpp:245] Train net output #16: total_confidence = 0.085381 | |
| I0401 18:43:53.438532 6134 sgd_solver.cpp:106] Iteration 82000, lr = 0.01 | |
| I0401 18:46:02.602617 6134 solver.cpp:229] Iteration 82500, loss = 3.50874 | |
| I0401 18:46:02.602818 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.3 | |
| I0401 18:46:02.602838 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 18:46:02.602852 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.45 | |
| I0401 18:46:02.602869 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.30401 (* 0.3 = 0.691203 loss) | |
| I0401 18:46:02.602885 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.636129 (* 0.3 = 0.190839 loss) | |
| I0401 18:46:02.602897 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0401 18:46:02.602910 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 18:46:02.602922 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6 | |
| I0401 18:46:02.602936 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.12135 (* 0.3 = 0.636404 loss) | |
| I0401 18:46:02.602952 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.562041 (* 0.3 = 0.168612 loss) | |
| I0401 18:46:02.602963 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.575 | |
| I0401 18:46:02.602977 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 18:46:02.602988 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.675 | |
| I0401 18:46:02.603003 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.51261 (* 1 = 1.51261 loss) | |
| I0401 18:46:02.603016 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.414931 (* 1 = 0.414931 loss) | |
| I0401 18:46:02.603029 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:46:02.603040 6134 solver.cpp:245] Train net output #16: total_confidence = 0.222587 | |
| I0401 18:46:02.603054 6134 sgd_solver.cpp:106] Iteration 82500, lr = 0.01 | |
| I0401 18:48:11.268380 6134 solver.cpp:229] Iteration 83000, loss = 3.47609 | |
| I0401 18:48:11.268672 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.265306 | |
| I0401 18:48:11.268692 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 18:48:11.268704 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55102 | |
| I0401 18:48:11.268720 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.51786 (* 0.3 = 0.755357 loss) | |
| I0401 18:48:11.268735 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.760446 (* 0.3 = 0.228134 loss) | |
| I0401 18:48:11.268748 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.408163 | |
| I0401 18:48:11.268761 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 18:48:11.268772 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.693878 | |
| I0401 18:48:11.268787 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.88789 (* 0.3 = 0.566366 loss) | |
| I0401 18:48:11.268802 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.600731 (* 0.3 = 0.180219 loss) | |
| I0401 18:48:11.268815 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.693878 | |
| I0401 18:48:11.268826 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 18:48:11.268838 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.816327 | |
| I0401 18:48:11.268853 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.21274 (* 1 = 1.21274 loss) | |
| I0401 18:48:11.268867 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.373533 (* 1 = 0.373533 loss) | |
| I0401 18:48:11.268879 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 18:48:11.268892 6134 solver.cpp:245] Train net output #16: total_confidence = 0.203744 | |
| I0401 18:48:11.268904 6134 sgd_solver.cpp:106] Iteration 83000, lr = 0.01 | |
| I0401 18:50:20.171488 6134 solver.cpp:229] Iteration 83500, loss = 3.55813 | |
| I0401 18:50:20.171703 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.340426 | |
| I0401 18:50:20.171725 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 18:50:20.171739 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.702128 | |
| I0401 18:50:20.171756 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.97 (* 0.3 = 0.590999 loss) | |
| I0401 18:50:20.171771 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.581874 (* 0.3 = 0.174562 loss) | |
| I0401 18:50:20.171783 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.404255 | |
| I0401 18:50:20.171797 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 18:50:20.171809 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.744681 | |
| I0401 18:50:20.171823 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.79114 (* 0.3 = 0.537341 loss) | |
| I0401 18:50:20.171838 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.527763 (* 0.3 = 0.158329 loss) | |
| I0401 18:50:20.171849 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.531915 | |
| I0401 18:50:20.171862 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 18:50:20.171874 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.851064 | |
| I0401 18:50:20.171887 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.34425 (* 1 = 1.34425 loss) | |
| I0401 18:50:20.171910 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.398512 (* 1 = 0.398512 loss) | |
| I0401 18:50:20.171922 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 18:50:20.171934 6134 solver.cpp:245] Train net output #16: total_confidence = 0.144705 | |
| I0401 18:50:20.171947 6134 sgd_solver.cpp:106] Iteration 83500, lr = 0.01 | |
| I0401 18:52:29.128466 6134 solver.cpp:229] Iteration 84000, loss = 3.52494 | |
| I0401 18:52:29.128585 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.288889 | |
| I0401 18:52:29.128607 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 18:52:29.128628 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6 | |
| I0401 18:52:29.128643 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.42899 (* 0.3 = 0.728697 loss) | |
| I0401 18:52:29.128659 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.713524 (* 0.3 = 0.214057 loss) | |
| I0401 18:52:29.128671 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.444444 | |
| I0401 18:52:29.128684 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 18:52:29.128696 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.711111 | |
| I0401 18:52:29.128718 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.06323 (* 0.3 = 0.618968 loss) | |
| I0401 18:52:29.128732 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.66904 (* 0.3 = 0.200712 loss) | |
| I0401 18:52:29.128744 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.533333 | |
| I0401 18:52:29.128758 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0401 18:52:29.128779 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.755556 | |
| I0401 18:52:29.128793 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.48619 (* 1 = 1.48619 loss) | |
| I0401 18:52:29.128808 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.439656 (* 1 = 0.439656 loss) | |
| I0401 18:52:29.128819 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 18:52:29.128832 6134 solver.cpp:245] Train net output #16: total_confidence = 0.262026 | |
| I0401 18:52:29.128844 6134 sgd_solver.cpp:106] Iteration 84000, lr = 0.01 | |
| I0401 18:54:38.077003 6134 solver.cpp:229] Iteration 84500, loss = 3.5228 | |
| I0401 18:54:38.077158 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.372093 | |
| I0401 18:54:38.077178 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0401 18:54:38.077193 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.651163 | |
| I0401 18:54:38.077208 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.00964 (* 0.3 = 0.602892 loss) | |
| I0401 18:54:38.077224 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.559048 (* 0.3 = 0.167714 loss) | |
| I0401 18:54:38.077241 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.511628 | |
| I0401 18:54:38.077255 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 18:54:38.077266 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.627907 | |
| I0401 18:54:38.077280 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.61604 (* 0.3 = 0.484811 loss) | |
| I0401 18:54:38.077294 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.461145 (* 0.3 = 0.138344 loss) | |
| I0401 18:54:38.077306 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.651163 | |
| I0401 18:54:38.077319 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0401 18:54:38.077330 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.813953 | |
| I0401 18:54:38.077344 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.13288 (* 1 = 1.13288 loss) | |
| I0401 18:54:38.077358 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.317953 (* 1 = 0.317953 loss) | |
| I0401 18:54:38.077370 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 18:54:38.077383 6134 solver.cpp:245] Train net output #16: total_confidence = 0.342421 | |
| I0401 18:54:38.077395 6134 sgd_solver.cpp:106] Iteration 84500, lr = 0.01 | |
| I0401 18:56:46.875203 6134 solver.cpp:338] Iteration 85000, Testing net (#0) | |
| I0401 18:57:16.777572 6134 solver.cpp:393] Test loss: 2.9218 | |
| I0401 18:57:16.777627 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.380547 | |
| I0401 18:57:16.777644 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.832592 | |
| I0401 18:57:16.777657 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.694125 | |
| I0401 18:57:16.777673 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.02448 (* 0.3 = 0.607345 loss) | |
| I0401 18:57:16.777688 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.562597 (* 0.3 = 0.168779 loss) | |
| I0401 18:57:16.777700 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.542106 | |
| I0401 18:57:16.777714 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.873322 | |
| I0401 18:57:16.777725 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.807774 | |
| I0401 18:57:16.777740 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.54615 (* 0.3 = 0.463844 loss) | |
| I0401 18:57:16.777753 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.431394 (* 0.3 = 0.129418 loss) | |
| I0401 18:57:16.777765 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.673585 | |
| I0401 18:57:16.777777 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.914138 | |
| I0401 18:57:16.777789 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.854004 | |
| I0401 18:57:16.777803 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.22133 (* 1 = 1.22133 loss) | |
| I0401 18:57:16.777817 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.331086 (* 1 = 0.331086 loss) | |
| I0401 18:57:16.777829 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.253 | |
| I0401 18:57:16.777842 6134 solver.cpp:406] Test net output #16: total_confidence = 0.216195 | |
| I0401 18:57:16.929653 6134 solver.cpp:229] Iteration 85000, loss = 3.47953 | |
| I0401 18:57:16.929806 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.195122 | |
| I0401 18:57:16.929838 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 18:57:16.929852 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.585366 | |
| I0401 18:57:16.929867 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.323 (* 0.3 = 0.6969 loss) | |
| I0401 18:57:16.929890 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.756631 (* 0.3 = 0.226989 loss) | |
| I0401 18:57:16.929903 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.390244 | |
| I0401 18:57:16.929914 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0401 18:57:16.929927 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.682927 | |
| I0401 18:57:16.929940 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.96952 (* 0.3 = 0.590855 loss) | |
| I0401 18:57:16.929961 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.621203 (* 0.3 = 0.186361 loss) | |
| I0401 18:57:16.929973 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.682927 | |
| I0401 18:57:16.929986 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 18:57:16.929997 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.829268 | |
| I0401 18:57:16.930011 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.224 (* 1 = 1.224 loss) | |
| I0401 18:57:16.930025 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.419483 (* 1 = 0.419483 loss) | |
| I0401 18:57:16.930037 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 18:57:16.930049 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0771766 | |
| I0401 18:57:16.930063 6134 sgd_solver.cpp:106] Iteration 85000, lr = 0.01 | |
| I0401 18:59:25.535673 6134 solver.cpp:229] Iteration 85500, loss = 3.47317 | |
| I0401 18:59:25.535785 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.27451 | |
| I0401 18:59:25.535805 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 18:59:25.535818 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54902 | |
| I0401 18:59:25.535835 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.52221 (* 0.3 = 0.756664 loss) | |
| I0401 18:59:25.535850 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.752122 (* 0.3 = 0.225636 loss) | |
| I0401 18:59:25.535862 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0401 18:59:25.535876 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0401 18:59:25.535888 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.647059 | |
| I0401 18:59:25.535902 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.37195 (* 0.3 = 0.711586 loss) | |
| I0401 18:59:25.535917 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.751409 (* 0.3 = 0.225423 loss) | |
| I0401 18:59:25.535929 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.627451 | |
| I0401 18:59:25.535941 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0401 18:59:25.535954 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.862745 | |
| I0401 18:59:25.535969 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.45077 (* 1 = 1.45077 loss) | |
| I0401 18:59:25.535990 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.451532 (* 1 = 0.451532 loss) | |
| I0401 18:59:25.536002 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 18:59:25.536015 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0688208 | |
| I0401 18:59:25.536027 6134 sgd_solver.cpp:106] Iteration 85500, lr = 0.01 | |
| I0401 19:01:34.676713 6134 solver.cpp:229] Iteration 86000, loss = 3.51608 | |
| I0401 19:01:34.676862 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.255319 | |
| I0401 19:01:34.676882 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 19:01:34.676895 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.489362 | |
| I0401 19:01:34.676913 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.37145 (* 0.3 = 0.711435 loss) | |
| I0401 19:01:34.676935 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.683147 (* 0.3 = 0.204944 loss) | |
| I0401 19:01:34.676947 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.425532 | |
| I0401 19:01:34.676960 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 19:01:34.676972 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.659574 | |
| I0401 19:01:34.676986 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.96256 (* 0.3 = 0.588769 loss) | |
| I0401 19:01:34.677000 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.576866 (* 0.3 = 0.17306 loss) | |
| I0401 19:01:34.677013 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.617021 | |
| I0401 19:01:34.677037 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 19:01:34.677053 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.851064 | |
| I0401 19:01:34.677067 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.22737 (* 1 = 1.22737 loss) | |
| I0401 19:01:34.677083 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.384739 (* 1 = 0.384739 loss) | |
| I0401 19:01:34.677094 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 19:01:34.677106 6134 solver.cpp:245] Train net output #16: total_confidence = 0.14583 | |
| I0401 19:01:34.677119 6134 sgd_solver.cpp:106] Iteration 86000, lr = 0.01 | |
| I0401 19:03:43.668409 6134 solver.cpp:229] Iteration 86500, loss = 3.45769 | |
| I0401 19:03:43.668535 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.244898 | |
| I0401 19:03:43.668555 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 19:03:43.668568 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.367347 | |
| I0401 19:03:43.668584 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.75806 (* 0.3 = 0.827417 loss) | |
| I0401 19:03:43.668599 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.802324 (* 0.3 = 0.240697 loss) | |
| I0401 19:03:43.668612 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.306122 | |
| I0401 19:03:43.668625 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 19:03:43.668637 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.530612 | |
| I0401 19:03:43.668651 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.51634 (* 0.3 = 0.754903 loss) | |
| I0401 19:03:43.668665 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.731936 (* 0.3 = 0.219581 loss) | |
| I0401 19:03:43.668678 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.55102 | |
| I0401 19:03:43.668690 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0401 19:03:43.668702 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.734694 | |
| I0401 19:03:43.668716 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.77308 (* 1 = 1.77308 loss) | |
| I0401 19:03:43.668730 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.548514 (* 1 = 0.548514 loss) | |
| I0401 19:03:43.668742 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 19:03:43.668754 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0533114 | |
| I0401 19:03:43.668766 6134 sgd_solver.cpp:106] Iteration 86500, lr = 0.01 | |
| I0401 19:05:52.494949 6134 solver.cpp:229] Iteration 87000, loss = 3.39918 | |
| I0401 19:05:52.495100 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0401 19:05:52.495121 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0401 19:05:52.495134 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.7 | |
| I0401 19:05:52.495151 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.99168 (* 0.3 = 0.597505 loss) | |
| I0401 19:05:52.495167 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.53251 (* 0.3 = 0.159753 loss) | |
| I0401 19:05:52.495188 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0401 19:05:52.495200 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 19:05:52.495213 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0401 19:05:52.495228 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.77321 (* 0.3 = 0.531962 loss) | |
| I0401 19:05:52.495241 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.49162 (* 0.3 = 0.147486 loss) | |
| I0401 19:05:52.495254 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.725 | |
| I0401 19:05:52.495266 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 19:05:52.495278 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.875 | |
| I0401 19:05:52.495292 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.845572 (* 1 = 0.845572 loss) | |
| I0401 19:05:52.495306 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.249115 (* 1 = 0.249115 loss) | |
| I0401 19:05:52.495327 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 19:05:52.495339 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0642938 | |
| I0401 19:05:52.495352 6134 sgd_solver.cpp:106] Iteration 87000, lr = 0.01 | |
| I0401 19:08:01.372947 6134 solver.cpp:229] Iteration 87500, loss = 3.39869 | |
| I0401 19:08:01.373314 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.243902 | |
| I0401 19:08:01.373345 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 19:08:01.373369 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.439024 | |
| I0401 19:08:01.373399 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.75741 (* 0.3 = 0.827224 loss) | |
| I0401 19:08:01.373427 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.762107 (* 0.3 = 0.228632 loss) | |
| I0401 19:08:01.373450 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.390244 | |
| I0401 19:08:01.373471 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 19:08:01.373492 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.634146 | |
| I0401 19:08:01.373522 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.43802 (* 0.3 = 0.731406 loss) | |
| I0401 19:08:01.373551 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.747906 (* 0.3 = 0.224372 loss) | |
| I0401 19:08:01.373572 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.536585 | |
| I0401 19:08:01.373594 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.857955 | |
| I0401 19:08:01.373613 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.707317 | |
| I0401 19:08:01.373639 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.86326 (* 1 = 1.86326 loss) | |
| I0401 19:08:01.373666 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.566846 (* 1 = 0.566846 loss) | |
| I0401 19:08:01.373687 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 19:08:01.373709 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0619094 | |
| I0401 19:08:01.373733 6134 sgd_solver.cpp:106] Iteration 87500, lr = 0.01 | |
| I0401 19:10:10.248014 6134 solver.cpp:229] Iteration 88000, loss = 3.41325 | |
| I0401 19:10:10.248154 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.404762 | |
| I0401 19:10:10.248174 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 19:10:10.248188 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.738095 | |
| I0401 19:10:10.248203 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.9042 (* 0.3 = 0.571259 loss) | |
| I0401 19:10:10.248219 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.548035 (* 0.3 = 0.164411 loss) | |
| I0401 19:10:10.248231 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.380952 | |
| I0401 19:10:10.248253 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 19:10:10.248265 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.738095 | |
| I0401 19:10:10.248280 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.64865 (* 0.3 = 0.494594 loss) | |
| I0401 19:10:10.248293 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.476903 (* 0.3 = 0.143071 loss) | |
| I0401 19:10:10.248306 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.761905 | |
| I0401 19:10:10.248318 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 19:10:10.248330 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.880952 | |
| I0401 19:10:10.248344 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.672744 (* 1 = 0.672744 loss) | |
| I0401 19:10:10.248359 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.199674 (* 1 = 0.199674 loss) | |
| I0401 19:10:10.248371 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 19:10:10.248383 6134 solver.cpp:245] Train net output #16: total_confidence = 0.177925 | |
| I0401 19:10:10.248396 6134 sgd_solver.cpp:106] Iteration 88000, lr = 0.01 | |
| I0401 19:12:19.335609 6134 solver.cpp:229] Iteration 88500, loss = 3.5066 | |
| I0401 19:12:19.335736 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.255319 | |
| I0401 19:12:19.335757 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 19:12:19.335770 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.531915 | |
| I0401 19:12:19.335788 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.48742 (* 0.3 = 0.746226 loss) | |
| I0401 19:12:19.335803 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.713992 (* 0.3 = 0.214198 loss) | |
| I0401 19:12:19.335815 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.361702 | |
| I0401 19:12:19.335827 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0401 19:12:19.335839 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.659574 | |
| I0401 19:12:19.335853 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.34272 (* 0.3 = 0.702816 loss) | |
| I0401 19:12:19.335867 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.66639 (* 0.3 = 0.199917 loss) | |
| I0401 19:12:19.335880 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.489362 | |
| I0401 19:12:19.335892 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0401 19:12:19.335904 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.787234 | |
| I0401 19:12:19.335918 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.9673 (* 1 = 1.9673 loss) | |
| I0401 19:12:19.335932 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.564786 (* 1 = 0.564786 loss) | |
| I0401 19:12:19.335944 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 19:12:19.335957 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0808652 | |
| I0401 19:12:19.335968 6134 sgd_solver.cpp:106] Iteration 88500, lr = 0.01 | |
| I0401 19:14:28.315026 6134 solver.cpp:229] Iteration 89000, loss = 3.38718 | |
| I0401 19:14:28.315158 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.173077 | |
| I0401 19:14:28.315179 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0401 19:14:28.315192 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.423077 | |
| I0401 19:14:28.315209 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.85266 (* 0.3 = 0.855797 loss) | |
| I0401 19:14:28.315224 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.862972 (* 0.3 = 0.258892 loss) | |
| I0401 19:14:28.315237 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.192308 | |
| I0401 19:14:28.315250 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.75 | |
| I0401 19:14:28.315263 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.442308 | |
| I0401 19:14:28.315275 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.80422 (* 0.3 = 0.841268 loss) | |
| I0401 19:14:28.315290 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.876174 (* 0.3 = 0.262852 loss) | |
| I0401 19:14:28.315302 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0401 19:14:28.315315 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0401 19:14:28.315327 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.673077 | |
| I0401 19:14:28.315341 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.93334 (* 1 = 1.93334 loss) | |
| I0401 19:14:28.315356 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.629895 (* 1 = 0.629895 loss) | |
| I0401 19:14:28.315367 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 19:14:28.315379 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0787688 | |
| I0401 19:14:28.315392 6134 sgd_solver.cpp:106] Iteration 89000, lr = 0.01 | |
| I0401 19:16:36.926424 6134 solver.cpp:229] Iteration 89500, loss = 3.41752 | |
| I0401 19:16:36.926695 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.166667 | |
| I0401 19:16:36.926714 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.744318 | |
| I0401 19:16:36.926728 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.37037 | |
| I0401 19:16:36.926744 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.92177 (* 0.3 = 0.876532 loss) | |
| I0401 19:16:36.926759 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.916945 (* 0.3 = 0.275084 loss) | |
| I0401 19:16:36.926772 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.259259 | |
| I0401 19:16:36.926784 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 19:16:36.926797 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.5 | |
| I0401 19:16:36.926811 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.61181 (* 0.3 = 0.783542 loss) | |
| I0401 19:16:36.926826 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.815099 (* 0.3 = 0.24453 loss) | |
| I0401 19:16:36.926837 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.555556 | |
| I0401 19:16:36.926849 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 19:16:36.926862 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.685185 | |
| I0401 19:16:36.926875 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.61962 (* 1 = 1.61962 loss) | |
| I0401 19:16:36.926890 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.509453 (* 1 = 0.509453 loss) | |
| I0401 19:16:36.926903 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 19:16:36.926915 6134 solver.cpp:245] Train net output #16: total_confidence = 0.207926 | |
| I0401 19:16:36.926926 6134 sgd_solver.cpp:106] Iteration 89500, lr = 0.01 | |
| I0401 19:18:45.828851 6134 solver.cpp:338] Iteration 90000, Testing net (#0) | |
| I0401 19:19:15.643659 6134 solver.cpp:393] Test loss: 2.95026 | |
| I0401 19:19:15.643710 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.356947 | |
| I0401 19:19:15.643728 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.831137 | |
| I0401 19:19:15.643740 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.656821 | |
| I0401 19:19:15.643756 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.08718 (* 0.3 = 0.626155 loss) | |
| I0401 19:19:15.643771 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.559833 (* 0.3 = 0.16795 loss) | |
| I0401 19:19:15.643784 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.528757 | |
| I0401 19:19:15.643795 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.87423 | |
| I0401 19:19:15.643807 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.796985 | |
| I0401 19:19:15.643821 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.61335 (* 0.3 = 0.484005 loss) | |
| I0401 19:19:15.643836 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.439288 (* 0.3 = 0.131786 loss) | |
| I0401 19:19:15.643848 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.67437 | |
| I0401 19:19:15.643860 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.916502 | |
| I0401 19:19:15.643872 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.854996 | |
| I0401 19:19:15.643885 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.21449 (* 1 = 1.21449 loss) | |
| I0401 19:19:15.643899 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.325868 (* 1 = 0.325868 loss) | |
| I0401 19:19:15.643911 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.231 | |
| I0401 19:19:15.643923 6134 solver.cpp:406] Test net output #16: total_confidence = 0.192292 | |
| I0401 19:19:15.794682 6134 solver.cpp:229] Iteration 90000, loss = 3.44104 | |
| I0401 19:19:15.794723 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.288889 | |
| I0401 19:19:15.794739 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 19:19:15.794752 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.533333 | |
| I0401 19:19:15.794769 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.4848 (* 0.3 = 0.74544 loss) | |
| I0401 19:19:15.794783 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.715942 (* 0.3 = 0.214783 loss) | |
| I0401 19:19:15.794795 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0401 19:19:15.794808 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 19:19:15.794821 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0401 19:19:15.794833 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.03878 (* 0.3 = 0.611635 loss) | |
| I0401 19:19:15.794847 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.572865 (* 0.3 = 0.171859 loss) | |
| I0401 19:19:15.794859 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.688889 | |
| I0401 19:19:15.794872 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 19:19:15.794884 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.733333 | |
| I0401 19:19:15.794898 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.29847 (* 1 = 1.29847 loss) | |
| I0401 19:19:15.794911 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.367888 (* 1 = 0.367888 loss) | |
| I0401 19:19:15.794924 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 19:19:15.794935 6134 solver.cpp:245] Train net output #16: total_confidence = 0.200964 | |
| I0401 19:19:15.794947 6134 sgd_solver.cpp:106] Iteration 90000, lr = 0.01 | |
| I0401 19:21:25.277760 6134 solver.cpp:229] Iteration 90500, loss = 3.45246 | |
| I0401 19:21:25.277891 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.488889 | |
| I0401 19:21:25.277921 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0401 19:21:25.277945 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.8 | |
| I0401 19:21:25.277974 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.93793 (* 0.3 = 0.581378 loss) | |
| I0401 19:21:25.278002 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.578375 (* 0.3 = 0.173512 loss) | |
| I0401 19:21:25.278024 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.555556 | |
| I0401 19:21:25.278048 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0401 19:21:25.278070 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.866667 | |
| I0401 19:21:25.278096 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.40093 (* 0.3 = 0.420279 loss) | |
| I0401 19:21:25.278121 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.415374 (* 0.3 = 0.124612 loss) | |
| I0401 19:21:25.278142 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.866667 | |
| I0401 19:21:25.278175 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0401 19:21:25.278210 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.955556 | |
| I0401 19:21:25.278237 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.54379 (* 1 = 0.54379 loss) | |
| I0401 19:21:25.278262 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.185083 (* 1 = 0.185083 loss) | |
| I0401 19:21:25.278283 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 19:21:25.278304 6134 solver.cpp:245] Train net output #16: total_confidence = 0.170176 | |
| I0401 19:21:25.278326 6134 sgd_solver.cpp:106] Iteration 90500, lr = 0.01 | |
| I0401 19:23:33.962676 6134 solver.cpp:229] Iteration 91000, loss = 3.39351 | |
| I0401 19:23:33.962788 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.269231 | |
| I0401 19:23:33.962807 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 19:23:33.962821 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.384615 | |
| I0401 19:23:33.962837 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.57195 (* 0.3 = 0.771585 loss) | |
| I0401 19:23:33.962852 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.789182 (* 0.3 = 0.236755 loss) | |
| I0401 19:23:33.962865 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.230769 | |
| I0401 19:23:33.962877 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 19:23:33.962889 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.538462 | |
| I0401 19:23:33.962903 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.49992 (* 0.3 = 0.749976 loss) | |
| I0401 19:23:33.962918 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.750393 (* 0.3 = 0.225118 loss) | |
| I0401 19:23:33.962929 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.480769 | |
| I0401 19:23:33.962942 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.846591 | |
| I0401 19:23:33.962954 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.769231 | |
| I0401 19:23:33.962968 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.47489 (* 1 = 1.47489 loss) | |
| I0401 19:23:33.962982 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.452223 (* 1 = 0.452223 loss) | |
| I0401 19:23:33.962995 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 19:23:33.963006 6134 solver.cpp:245] Train net output #16: total_confidence = 0.166366 | |
| I0401 19:23:33.963019 6134 sgd_solver.cpp:106] Iteration 91000, lr = 0.01 | |
| I0401 19:25:43.331502 6134 solver.cpp:229] Iteration 91500, loss = 3.32188 | |
| I0401 19:25:43.331645 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.255319 | |
| I0401 19:25:43.331665 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 19:25:43.331678 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.404255 | |
| I0401 19:25:43.331694 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.5418 (* 0.3 = 0.762539 loss) | |
| I0401 19:25:43.331709 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.76531 (* 0.3 = 0.229593 loss) | |
| I0401 19:25:43.331722 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.297872 | |
| I0401 19:25:43.331734 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 19:25:43.331746 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.489362 | |
| I0401 19:25:43.331760 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.37549 (* 0.3 = 0.712648 loss) | |
| I0401 19:25:43.331774 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.698922 (* 0.3 = 0.209677 loss) | |
| I0401 19:25:43.331786 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.468085 | |
| I0401 19:25:43.331799 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0401 19:25:43.331810 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.638298 | |
| I0401 19:25:43.331825 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.95186 (* 1 = 1.95186 loss) | |
| I0401 19:25:43.331838 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.556776 (* 1 = 0.556776 loss) | |
| I0401 19:25:43.331851 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 19:25:43.331862 6134 solver.cpp:245] Train net output #16: total_confidence = 0.251275 | |
| I0401 19:25:43.331876 6134 sgd_solver.cpp:106] Iteration 91500, lr = 0.01 | |
| I0401 19:27:52.555338 6134 solver.cpp:229] Iteration 92000, loss = 3.40879 | |
| I0401 19:27:52.555609 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.404762 | |
| I0401 19:27:52.555629 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 19:27:52.555642 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.619048 | |
| I0401 19:27:52.555658 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.04694 (* 0.3 = 0.614083 loss) | |
| I0401 19:27:52.555673 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.588766 (* 0.3 = 0.17663 loss) | |
| I0401 19:27:52.555686 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.380952 | |
| I0401 19:27:52.555698 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 19:27:52.555711 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.642857 | |
| I0401 19:27:52.555724 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.76593 (* 0.3 = 0.529779 loss) | |
| I0401 19:27:52.555739 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.515849 (* 0.3 = 0.154755 loss) | |
| I0401 19:27:52.555752 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0401 19:27:52.555763 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0401 19:27:52.555775 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.857143 | |
| I0401 19:27:52.555789 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.989929 (* 1 = 0.989929 loss) | |
| I0401 19:27:52.555804 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.288039 (* 1 = 0.288039 loss) | |
| I0401 19:27:52.555815 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 19:27:52.555827 6134 solver.cpp:245] Train net output #16: total_confidence = 0.248512 | |
| I0401 19:27:52.555840 6134 sgd_solver.cpp:106] Iteration 92000, lr = 0.01 | |
| I0401 19:30:01.440426 6134 solver.cpp:229] Iteration 92500, loss = 3.39395 | |
| I0401 19:30:01.440588 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.326531 | |
| I0401 19:30:01.440608 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 19:30:01.440623 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.591837 | |
| I0401 19:30:01.440639 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.19735 (* 0.3 = 0.659206 loss) | |
| I0401 19:30:01.440654 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.635497 (* 0.3 = 0.190649 loss) | |
| I0401 19:30:01.440666 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.44898 | |
| I0401 19:30:01.440680 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 19:30:01.440690 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.714286 | |
| I0401 19:30:01.440704 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.85977 (* 0.3 = 0.557931 loss) | |
| I0401 19:30:01.440718 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.547706 (* 0.3 = 0.164312 loss) | |
| I0401 19:30:01.440732 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.612245 | |
| I0401 19:30:01.440743 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 19:30:01.440757 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.795918 | |
| I0401 19:30:01.440770 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.26756 (* 1 = 1.26756 loss) | |
| I0401 19:30:01.440784 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.400658 (* 1 = 0.400658 loss) | |
| I0401 19:30:01.440796 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 19:30:01.440809 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0917591 | |
| I0401 19:30:01.440820 6134 sgd_solver.cpp:106] Iteration 92500, lr = 0.01 | |
| I0401 19:32:10.633386 6134 solver.cpp:229] Iteration 93000, loss = 3.32726 | |
| I0401 19:32:10.633496 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.296296 | |
| I0401 19:32:10.633518 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 19:32:10.633532 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.481481 | |
| I0401 19:32:10.633548 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.61596 (* 0.3 = 0.784787 loss) | |
| I0401 19:32:10.633563 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.810064 (* 0.3 = 0.243019 loss) | |
| I0401 19:32:10.633576 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.277778 | |
| I0401 19:32:10.633589 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 19:32:10.633601 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.611111 | |
| I0401 19:32:10.633615 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.25078 (* 0.3 = 0.675234 loss) | |
| I0401 19:32:10.633630 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.711713 (* 0.3 = 0.213514 loss) | |
| I0401 19:32:10.633641 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.574074 | |
| I0401 19:32:10.633654 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0401 19:32:10.633666 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.703704 | |
| I0401 19:32:10.633679 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.64922 (* 1 = 1.64922 loss) | |
| I0401 19:32:10.633693 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.528839 (* 1 = 0.528839 loss) | |
| I0401 19:32:10.633705 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 19:32:10.633718 6134 solver.cpp:245] Train net output #16: total_confidence = 0.160231 | |
| I0401 19:32:10.633729 6134 sgd_solver.cpp:106] Iteration 93000, lr = 0.01 | |
| I0401 19:34:19.353746 6134 solver.cpp:229] Iteration 93500, loss = 3.30063 | |
| I0401 19:34:19.353880 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.264706 | |
| I0401 19:34:19.353900 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 19:34:19.353914 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.558824 | |
| I0401 19:34:19.353929 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.30305 (* 0.3 = 0.690914 loss) | |
| I0401 19:34:19.353945 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.623035 (* 0.3 = 0.186911 loss) | |
| I0401 19:34:19.353957 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.352941 | |
| I0401 19:34:19.353970 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0401 19:34:19.353981 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.588235 | |
| I0401 19:34:19.353996 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.15483 (* 0.3 = 0.646449 loss) | |
| I0401 19:34:19.354009 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.690331 (* 0.3 = 0.207099 loss) | |
| I0401 19:34:19.354022 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.647059 | |
| I0401 19:34:19.354033 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 19:34:19.354045 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.882353 | |
| I0401 19:34:19.354059 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.31861 (* 1 = 1.31861 loss) | |
| I0401 19:34:19.354074 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.316877 (* 1 = 0.316877 loss) | |
| I0401 19:34:19.354085 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 19:34:19.354097 6134 solver.cpp:245] Train net output #16: total_confidence = 0.235227 | |
| I0401 19:34:19.354110 6134 sgd_solver.cpp:106] Iteration 93500, lr = 0.01 | |
| I0401 19:36:28.144903 6134 solver.cpp:229] Iteration 94000, loss = 3.38031 | |
| I0401 19:36:28.145191 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.32 | |
| I0401 19:36:28.145211 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 19:36:28.145225 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0401 19:36:28.145241 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.68394 (* 0.3 = 0.805182 loss) | |
| I0401 19:36:28.145256 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.819753 (* 0.3 = 0.245926 loss) | |
| I0401 19:36:28.145267 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.44 | |
| I0401 19:36:28.145280 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0401 19:36:28.145292 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.62 | |
| I0401 19:36:28.145306 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.06796 (* 0.3 = 0.620387 loss) | |
| I0401 19:36:28.145320 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.651333 (* 0.3 = 0.1954 loss) | |
| I0401 19:36:28.145333 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.58 | |
| I0401 19:36:28.145346 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0401 19:36:28.145369 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.74 | |
| I0401 19:36:28.145395 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.61652 (* 1 = 1.61652 loss) | |
| I0401 19:36:28.145411 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.530533 (* 1 = 0.530533 loss) | |
| I0401 19:36:28.145424 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 19:36:28.145437 6134 solver.cpp:245] Train net output #16: total_confidence = 0.209337 | |
| I0401 19:36:28.145449 6134 sgd_solver.cpp:106] Iteration 94000, lr = 0.01 | |
| I0401 19:38:37.091224 6134 solver.cpp:229] Iteration 94500, loss = 3.32765 | |
| I0401 19:38:37.091387 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.27907 | |
| I0401 19:38:37.091408 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 19:38:37.091423 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.465116 | |
| I0401 19:38:37.091439 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.4608 (* 0.3 = 0.73824 loss) | |
| I0401 19:38:37.091454 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.713075 (* 0.3 = 0.213922 loss) | |
| I0401 19:38:37.091466 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488372 | |
| I0401 19:38:37.091478 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 19:38:37.091490 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.744186 | |
| I0401 19:38:37.091505 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.64249 (* 0.3 = 0.492746 loss) | |
| I0401 19:38:37.091521 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.485628 (* 0.3 = 0.145689 loss) | |
| I0401 19:38:37.091534 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.860465 | |
| I0401 19:38:37.091547 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0401 19:38:37.091558 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.953488 | |
| I0401 19:38:37.091573 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.663428 (* 1 = 0.663428 loss) | |
| I0401 19:38:37.091588 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.214376 (* 1 = 0.214376 loss) | |
| I0401 19:38:37.091599 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 19:38:37.091611 6134 solver.cpp:245] Train net output #16: total_confidence = 0.187172 | |
| I0401 19:38:37.091624 6134 sgd_solver.cpp:106] Iteration 94500, lr = 0.01 | |
| I0401 19:40:45.465842 6134 solver.cpp:338] Iteration 95000, Testing net (#0) | |
| I0401 19:41:15.223929 6134 solver.cpp:393] Test loss: 2.99736 | |
| I0401 19:41:15.223980 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.405579 | |
| I0401 19:41:15.224009 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.847048 | |
| I0401 19:41:15.224031 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.693052 | |
| I0401 19:41:15.224061 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.00675 (* 0.3 = 0.602024 loss) | |
| I0401 19:41:15.224087 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.531383 (* 0.3 = 0.159415 loss) | |
| I0401 19:41:15.224108 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.53103 | |
| I0401 19:41:15.224131 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.872594 | |
| I0401 19:41:15.224153 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.797492 | |
| I0401 19:41:15.224179 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.63168 (* 0.3 = 0.489505 loss) | |
| I0401 19:41:15.224202 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.446671 (* 0.3 = 0.134001 loss) | |
| I0401 19:41:15.224223 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.681689 | |
| I0401 19:41:15.224246 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.908501 | |
| I0401 19:41:15.224267 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.845496 | |
| I0401 19:41:15.224292 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.25069 (* 1 = 1.25069 loss) | |
| I0401 19:41:15.224316 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.361723 (* 1 = 0.361723 loss) | |
| I0401 19:41:15.224336 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.225 | |
| I0401 19:41:15.224357 6134 solver.cpp:406] Test net output #16: total_confidence = 0.1857 | |
| I0401 19:41:15.376260 6134 solver.cpp:229] Iteration 95000, loss = 3.31022 | |
| I0401 19:41:15.376307 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.434783 | |
| I0401 19:41:15.376335 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0401 19:41:15.376359 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.73913 | |
| I0401 19:41:15.376389 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.92293 (* 0.3 = 0.57688 loss) | |
| I0401 19:41:15.376420 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.542532 (* 0.3 = 0.16276 loss) | |
| I0401 19:41:15.376443 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.521739 | |
| I0401 19:41:15.376466 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0401 19:41:15.376489 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.717391 | |
| I0401 19:41:15.376515 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.76838 (* 0.3 = 0.530513 loss) | |
| I0401 19:41:15.376540 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.479416 (* 0.3 = 0.143825 loss) | |
| I0401 19:41:15.376562 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.717391 | |
| I0401 19:41:15.376585 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 19:41:15.376607 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.891304 | |
| I0401 19:41:15.376633 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.28681 (* 1 = 1.28681 loss) | |
| I0401 19:41:15.376659 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.35494 (* 1 = 0.35494 loss) | |
| I0401 19:41:15.376682 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 19:41:15.376703 6134 solver.cpp:245] Train net output #16: total_confidence = 0.260555 | |
| I0401 19:41:15.376723 6134 sgd_solver.cpp:106] Iteration 95000, lr = 0.01 | |
| I0401 19:43:24.191296 6134 solver.cpp:229] Iteration 95500, loss = 3.34948 | |
| I0401 19:43:24.191436 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.22 | |
| I0401 19:43:24.191455 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 19:43:24.191468 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.52 | |
| I0401 19:43:24.191485 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.62029 (* 0.3 = 0.786087 loss) | |
| I0401 19:43:24.191500 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.818887 (* 0.3 = 0.245666 loss) | |
| I0401 19:43:24.191514 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.24 | |
| I0401 19:43:24.191529 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.767045 | |
| I0401 19:43:24.191541 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.54 | |
| I0401 19:43:24.191555 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.54026 (* 0.3 = 0.762079 loss) | |
| I0401 19:43:24.191570 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.79881 (* 0.3 = 0.239643 loss) | |
| I0401 19:43:24.191582 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.52 | |
| I0401 19:43:24.191594 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 19:43:24.191606 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.76 | |
| I0401 19:43:24.191620 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.68723 (* 1 = 1.68723 loss) | |
| I0401 19:43:24.191634 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.503743 (* 1 = 0.503743 loss) | |
| I0401 19:43:24.191648 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 19:43:24.191659 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0899354 | |
| I0401 19:43:24.191671 6134 sgd_solver.cpp:106] Iteration 95500, lr = 0.01 | |
| I0401 19:45:32.816009 6134 solver.cpp:229] Iteration 96000, loss = 3.30869 | |
| I0401 19:45:32.816153 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.369565 | |
| I0401 19:45:32.816171 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 19:45:32.816184 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.652174 | |
| I0401 19:45:32.816200 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.13395 (* 0.3 = 0.640186 loss) | |
| I0401 19:45:32.816216 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.6269 (* 0.3 = 0.18807 loss) | |
| I0401 19:45:32.816229 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.565217 | |
| I0401 19:45:32.816241 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0401 19:45:32.816253 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.826087 | |
| I0401 19:45:32.816267 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.61264 (* 0.3 = 0.483792 loss) | |
| I0401 19:45:32.816282 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.459855 (* 0.3 = 0.137956 loss) | |
| I0401 19:45:32.816294 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.869565 | |
| I0401 19:45:32.816308 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0401 19:45:32.816318 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.934783 | |
| I0401 19:45:32.816334 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.584369 (* 1 = 0.584369 loss) | |
| I0401 19:45:32.816347 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.16049 (* 1 = 0.16049 loss) | |
| I0401 19:45:32.816359 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 19:45:32.816371 6134 solver.cpp:245] Train net output #16: total_confidence = 0.269621 | |
| I0401 19:45:32.816383 6134 sgd_solver.cpp:106] Iteration 96000, lr = 0.01 | |
| I0401 19:47:41.421155 6134 solver.cpp:229] Iteration 96500, loss = 3.29373 | |
| I0401 19:47:41.421411 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.487805 | |
| I0401 19:47:41.421435 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0401 19:47:41.421449 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.634146 | |
| I0401 19:47:41.421465 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.98688 (* 0.3 = 0.596065 loss) | |
| I0401 19:47:41.421480 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.567554 (* 0.3 = 0.170266 loss) | |
| I0401 19:47:41.421492 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.487805 | |
| I0401 19:47:41.421505 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0401 19:47:41.421519 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.731707 | |
| I0401 19:47:41.421532 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.84783 (* 0.3 = 0.554348 loss) | |
| I0401 19:47:41.421545 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.512161 (* 0.3 = 0.153648 loss) | |
| I0401 19:47:41.421558 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.756098 | |
| I0401 19:47:41.421571 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0401 19:47:41.421582 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.878049 | |
| I0401 19:47:41.421597 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.887716 (* 1 = 0.887716 loss) | |
| I0401 19:47:41.421610 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.21652 (* 1 = 0.21652 loss) | |
| I0401 19:47:41.421623 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0401 19:47:41.421635 6134 solver.cpp:245] Train net output #16: total_confidence = 0.276346 | |
| I0401 19:47:41.421648 6134 sgd_solver.cpp:106] Iteration 96500, lr = 0.01 | |
| I0401 19:49:50.188525 6134 solver.cpp:229] Iteration 97000, loss = 3.29655 | |
| I0401 19:49:50.188647 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.367347 | |
| I0401 19:49:50.188668 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 19:49:50.188680 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.632653 | |
| I0401 19:49:50.188696 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.94394 (* 0.3 = 0.583181 loss) | |
| I0401 19:49:50.188711 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.568922 (* 0.3 = 0.170677 loss) | |
| I0401 19:49:50.188729 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.591837 | |
| I0401 19:49:50.188741 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0401 19:49:50.188753 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.897959 | |
| I0401 19:49:50.188767 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.48106 (* 0.3 = 0.444318 loss) | |
| I0401 19:49:50.188781 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.431203 (* 0.3 = 0.129361 loss) | |
| I0401 19:49:50.188793 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.77551 | |
| I0401 19:49:50.188805 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0401 19:49:50.188817 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.959184 | |
| I0401 19:49:50.188832 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.693493 (* 1 = 0.693493 loss) | |
| I0401 19:49:50.188845 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.216479 (* 1 = 0.216479 loss) | |
| I0401 19:49:50.188858 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 19:49:50.188869 6134 solver.cpp:245] Train net output #16: total_confidence = 0.174353 | |
| I0401 19:49:50.188881 6134 sgd_solver.cpp:106] Iteration 97000, lr = 0.01 | |
| I0401 19:51:58.981158 6134 solver.cpp:229] Iteration 97500, loss = 3.23455 | |
| I0401 19:51:58.981274 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0401 19:51:58.981294 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 19:51:58.981307 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.714286 | |
| I0401 19:51:58.981323 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.11897 (* 0.3 = 0.635691 loss) | |
| I0401 19:51:58.981338 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.652007 (* 0.3 = 0.195602 loss) | |
| I0401 19:51:58.981351 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.547619 | |
| I0401 19:51:58.981364 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0401 19:51:58.981375 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.714286 | |
| I0401 19:51:58.981389 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.67856 (* 0.3 = 0.503569 loss) | |
| I0401 19:51:58.981403 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.499254 (* 0.3 = 0.149776 loss) | |
| I0401 19:51:58.981416 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.738095 | |
| I0401 19:51:58.981428 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0401 19:51:58.981439 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.857143 | |
| I0401 19:51:58.981453 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.897481 (* 1 = 0.897481 loss) | |
| I0401 19:51:58.981467 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.258078 (* 1 = 0.258078 loss) | |
| I0401 19:51:58.981479 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 19:51:58.981492 6134 solver.cpp:245] Train net output #16: total_confidence = 0.213662 | |
| I0401 19:51:58.981503 6134 sgd_solver.cpp:106] Iteration 97500, lr = 0.01 | |
| I0401 19:54:07.649899 6134 solver.cpp:229] Iteration 98000, loss = 3.29608 | |
| I0401 19:54:07.650076 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26087 | |
| I0401 19:54:07.650097 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 19:54:07.650110 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.543478 | |
| I0401 19:54:07.650127 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.34317 (* 0.3 = 0.70295 loss) | |
| I0401 19:54:07.650142 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.650607 (* 0.3 = 0.195182 loss) | |
| I0401 19:54:07.650156 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.434783 | |
| I0401 19:54:07.650168 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 19:54:07.650180 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.673913 | |
| I0401 19:54:07.650194 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.09448 (* 0.3 = 0.628345 loss) | |
| I0401 19:54:07.650208 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.585245 (* 0.3 = 0.175573 loss) | |
| I0401 19:54:07.650221 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.73913 | |
| I0401 19:54:07.650233 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 19:54:07.650245 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.804348 | |
| I0401 19:54:07.650259 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.973422 (* 1 = 0.973422 loss) | |
| I0401 19:54:07.650274 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.288827 (* 1 = 0.288827 loss) | |
| I0401 19:54:07.650286 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 19:54:07.650298 6134 solver.cpp:245] Train net output #16: total_confidence = 0.169364 | |
| I0401 19:54:07.650310 6134 sgd_solver.cpp:106] Iteration 98000, lr = 0.01 | |
| I0401 19:56:16.544996 6134 solver.cpp:229] Iteration 98500, loss = 3.24721 | |
| I0401 19:56:16.545269 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.416667 | |
| I0401 19:56:16.545289 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0401 19:56:16.545301 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6875 | |
| I0401 19:56:16.545316 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.04615 (* 0.3 = 0.613844 loss) | |
| I0401 19:56:16.545332 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.586076 (* 0.3 = 0.175823 loss) | |
| I0401 19:56:16.545344 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.520833 | |
| I0401 19:56:16.545357 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0401 19:56:16.545369 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.729167 | |
| I0401 19:56:16.545382 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.96732 (* 0.3 = 0.590196 loss) | |
| I0401 19:56:16.545397 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.57782 (* 0.3 = 0.173346 loss) | |
| I0401 19:56:16.545409 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.6875 | |
| I0401 19:56:16.545421 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 19:56:16.545433 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8125 | |
| I0401 19:56:16.545447 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.22342 (* 1 = 1.22342 loss) | |
| I0401 19:56:16.545461 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.366115 (* 1 = 0.366115 loss) | |
| I0401 19:56:16.545474 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 19:56:16.545485 6134 solver.cpp:245] Train net output #16: total_confidence = 0.130445 | |
| I0401 19:56:16.545498 6134 sgd_solver.cpp:106] Iteration 98500, lr = 0.01 | |
| I0401 19:58:25.126678 6134 solver.cpp:229] Iteration 99000, loss = 3.23117 | |
| I0401 19:58:25.126816 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.418605 | |
| I0401 19:58:25.126842 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 19:58:25.126857 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.744186 | |
| I0401 19:58:25.126873 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.78278 (* 0.3 = 0.534835 loss) | |
| I0401 19:58:25.126888 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.522065 (* 0.3 = 0.156619 loss) | |
| I0401 19:58:25.126901 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488372 | |
| I0401 19:58:25.126914 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 19:58:25.126925 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.837209 | |
| I0401 19:58:25.126940 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.37965 (* 0.3 = 0.413895 loss) | |
| I0401 19:58:25.126953 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.419511 (* 0.3 = 0.125853 loss) | |
| I0401 19:58:25.126965 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.790698 | |
| I0401 19:58:25.126977 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0401 19:58:25.126989 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.953488 | |
| I0401 19:58:25.127003 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.680987 (* 1 = 0.680987 loss) | |
| I0401 19:58:25.127017 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.180235 (* 1 = 0.180235 loss) | |
| I0401 19:58:25.127029 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 19:58:25.127041 6134 solver.cpp:245] Train net output #16: total_confidence = 0.318496 | |
| I0401 19:58:25.127055 6134 sgd_solver.cpp:106] Iteration 99000, lr = 0.01 | |
| I0401 20:00:33.890247 6134 solver.cpp:229] Iteration 99500, loss = 3.20241 | |
| I0401 20:00:33.890357 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.306122 | |
| I0401 20:00:33.890377 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 20:00:33.890389 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55102 | |
| I0401 20:00:33.890405 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.30074 (* 0.3 = 0.690223 loss) | |
| I0401 20:00:33.890419 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.728714 (* 0.3 = 0.218614 loss) | |
| I0401 20:00:33.890432 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.387755 | |
| I0401 20:00:33.890444 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 20:00:33.890457 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.755102 | |
| I0401 20:00:33.890473 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.94499 (* 0.3 = 0.583498 loss) | |
| I0401 20:00:33.890487 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.623766 (* 0.3 = 0.18713 loss) | |
| I0401 20:00:33.890499 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.571429 | |
| I0401 20:00:33.890512 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0401 20:00:33.890527 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.734694 | |
| I0401 20:00:33.890542 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.82813 (* 1 = 1.82813 loss) | |
| I0401 20:00:33.890555 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.529022 (* 1 = 0.529022 loss) | |
| I0401 20:00:33.890568 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 20:00:33.890579 6134 solver.cpp:245] Train net output #16: total_confidence = 0.323279 | |
| I0401 20:00:33.890591 6134 sgd_solver.cpp:106] Iteration 99500, lr = 0.01 | |
| I0401 20:02:42.481189 6134 solver.cpp:456] Snapshotting to binary proto file /mnt/snapshots/mixed_lstm9_bn_iter_100000.caffemodel | |
| I0401 20:02:42.800722 6134 sgd_solver.cpp:273] Snapshotting solver state to binary proto file /mnt/snapshots/mixed_lstm9_bn_iter_100000.solverstate | |
| I0401 20:02:42.962452 6134 solver.cpp:338] Iteration 100000, Testing net (#0) | |
| I0401 20:03:12.799787 6134 solver.cpp:393] Test loss: 2.76884 | |
| I0401 20:03:12.799892 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.413312 | |
| I0401 20:03:12.799912 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.841093 | |
| I0401 20:03:12.799926 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.69398 | |
| I0401 20:03:12.799940 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.97237 (* 0.3 = 0.591711 loss) | |
| I0401 20:03:12.799955 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.548115 (* 0.3 = 0.164435 loss) | |
| I0401 20:03:12.799968 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.570581 | |
| I0401 20:03:12.799979 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.880412 | |
| I0401 20:03:12.799993 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.822585 | |
| I0401 20:03:12.800005 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.50075 (* 0.3 = 0.450226 loss) | |
| I0401 20:03:12.800019 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.417193 (* 0.3 = 0.125158 loss) | |
| I0401 20:03:12.800030 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.706627 | |
| I0401 20:03:12.800042 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.91791 | |
| I0401 20:03:12.800053 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.85918 | |
| I0401 20:03:12.800067 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.12313 (* 1 = 1.12313 loss) | |
| I0401 20:03:12.800081 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.314179 (* 1 = 0.314179 loss) | |
| I0401 20:03:12.800093 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.275 | |
| I0401 20:03:12.800104 6134 solver.cpp:406] Test net output #16: total_confidence = 0.228995 | |
| I0401 20:03:12.950626 6134 solver.cpp:229] Iteration 100000, loss = 3.18428 | |
| I0401 20:03:12.950670 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.23913 | |
| I0401 20:03:12.950686 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 20:03:12.950700 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.456522 | |
| I0401 20:03:12.950714 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.58922 (* 0.3 = 0.776765 loss) | |
| I0401 20:03:12.950728 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.732481 (* 0.3 = 0.219744 loss) | |
| I0401 20:03:12.950742 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.26087 | |
| I0401 20:03:12.950753 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0401 20:03:12.950764 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.5 | |
| I0401 20:03:12.950778 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.3851 (* 0.3 = 0.71553 loss) | |
| I0401 20:03:12.950793 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.70836 (* 0.3 = 0.212508 loss) | |
| I0401 20:03:12.950805 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.608696 | |
| I0401 20:03:12.950817 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0401 20:03:12.950829 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.782609 | |
| I0401 20:03:12.950842 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.56916 (* 1 = 1.56916 loss) | |
| I0401 20:03:12.950861 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.452885 (* 1 = 0.452885 loss) | |
| I0401 20:03:12.950873 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 20:03:12.950886 6134 solver.cpp:245] Train net output #16: total_confidence = 0.113242 | |
| I0401 20:03:12.950897 6134 sgd_solver.cpp:106] Iteration 100000, lr = 0.01 | |
| I0401 20:05:21.755398 6134 solver.cpp:229] Iteration 100500, loss = 3.20342 | |
| I0401 20:05:21.755547 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.306122 | |
| I0401 20:05:21.755568 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 20:05:21.755589 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.612245 | |
| I0401 20:05:21.755619 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.16509 (* 0.3 = 0.649528 loss) | |
| I0401 20:05:21.755650 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.66196 (* 0.3 = 0.198588 loss) | |
| I0401 20:05:21.755669 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.326531 | |
| I0401 20:05:21.755681 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0401 20:05:21.755694 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.673469 | |
| I0401 20:05:21.755708 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.03873 (* 0.3 = 0.61162 loss) | |
| I0401 20:05:21.755722 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.610981 (* 0.3 = 0.183294 loss) | |
| I0401 20:05:21.755735 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.836735 | |
| I0401 20:05:21.755748 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0401 20:05:21.755759 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.959184 | |
| I0401 20:05:21.755774 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.587665 (* 1 = 0.587665 loss) | |
| I0401 20:05:21.755789 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.175055 (* 1 = 0.175055 loss) | |
| I0401 20:05:21.755800 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 20:05:21.755813 6134 solver.cpp:245] Train net output #16: total_confidence = 0.241051 | |
| I0401 20:05:21.755825 6134 sgd_solver.cpp:106] Iteration 100500, lr = 0.01 | |
| I0401 20:07:30.516391 6134 solver.cpp:229] Iteration 101000, loss = 3.15056 | |
| I0401 20:07:30.516639 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.313726 | |
| I0401 20:07:30.516659 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 20:07:30.516671 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54902 | |
| I0401 20:07:30.516687 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.60599 (* 0.3 = 0.781796 loss) | |
| I0401 20:07:30.516702 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.785102 (* 0.3 = 0.235531 loss) | |
| I0401 20:07:30.516716 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.352941 | |
| I0401 20:07:30.516728 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 20:07:30.516741 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.686275 | |
| I0401 20:07:30.516754 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.04339 (* 0.3 = 0.613017 loss) | |
| I0401 20:07:30.516769 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.60341 (* 0.3 = 0.181023 loss) | |
| I0401 20:07:30.516782 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.705882 | |
| I0401 20:07:30.516793 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 20:07:30.516805 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.901961 | |
| I0401 20:07:30.516819 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.03586 (* 1 = 1.03586 loss) | |
| I0401 20:07:30.516834 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.304452 (* 1 = 0.304452 loss) | |
| I0401 20:07:30.516845 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 20:07:30.516858 6134 solver.cpp:245] Train net output #16: total_confidence = 0.155616 | |
| I0401 20:07:30.516870 6134 sgd_solver.cpp:106] Iteration 101000, lr = 0.01 | |
| I0401 20:09:39.458657 6134 solver.cpp:229] Iteration 101500, loss = 3.23611 | |
| I0401 20:09:39.458796 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26087 | |
| I0401 20:09:39.458820 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 20:09:39.458833 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.543478 | |
| I0401 20:09:39.458850 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.40327 (* 0.3 = 0.720981 loss) | |
| I0401 20:09:39.458865 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.705157 (* 0.3 = 0.211547 loss) | |
| I0401 20:09:39.458878 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.521739 | |
| I0401 20:09:39.458890 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0401 20:09:39.458901 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76087 | |
| I0401 20:09:39.458915 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.69295 (* 0.3 = 0.507885 loss) | |
| I0401 20:09:39.458930 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.482607 (* 0.3 = 0.144782 loss) | |
| I0401 20:09:39.458943 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.673913 | |
| I0401 20:09:39.458956 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 20:09:39.458967 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.934783 | |
| I0401 20:09:39.458981 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.948112 (* 1 = 0.948112 loss) | |
| I0401 20:09:39.458995 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.288369 (* 1 = 0.288369 loss) | |
| I0401 20:09:39.459007 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 20:09:39.459019 6134 solver.cpp:245] Train net output #16: total_confidence = 0.317184 | |
| I0401 20:09:39.459030 6134 sgd_solver.cpp:106] Iteration 101500, lr = 0.01 | |
| I0401 20:11:48.302608 6134 solver.cpp:229] Iteration 102000, loss = 3.21942 | |
| I0401 20:11:48.302713 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.422222 | |
| I0401 20:11:48.302742 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 20:11:48.302765 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0401 20:11:48.302795 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.03806 (* 0.3 = 0.611418 loss) | |
| I0401 20:11:48.302822 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.642956 (* 0.3 = 0.192887 loss) | |
| I0401 20:11:48.302845 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.333333 | |
| I0401 20:11:48.302867 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 20:11:48.302891 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0401 20:11:48.302916 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.88271 (* 0.3 = 0.564814 loss) | |
| I0401 20:11:48.302940 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.587018 (* 0.3 = 0.176105 loss) | |
| I0401 20:11:48.302961 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.555556 | |
| I0401 20:11:48.302983 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0401 20:11:48.303002 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.755556 | |
| I0401 20:11:48.303028 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.39764 (* 1 = 1.39764 loss) | |
| I0401 20:11:48.303061 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.443299 (* 1 = 0.443299 loss) | |
| I0401 20:11:48.303083 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 20:11:48.303104 6134 solver.cpp:245] Train net output #16: total_confidence = 0.231594 | |
| I0401 20:11:48.303125 6134 sgd_solver.cpp:106] Iteration 102000, lr = 0.01 | |
| I0401 20:13:57.050154 6134 solver.cpp:229] Iteration 102500, loss = 3.17863 | |
| I0401 20:13:57.050289 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.282609 | |
| I0401 20:13:57.050309 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 20:13:57.050323 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0401 20:13:57.050338 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.42524 (* 0.3 = 0.727572 loss) | |
| I0401 20:13:57.050354 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.700396 (* 0.3 = 0.210119 loss) | |
| I0401 20:13:57.050366 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.391304 | |
| I0401 20:13:57.050379 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0401 20:13:57.050390 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.673913 | |
| I0401 20:13:57.050405 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.95 (* 0.3 = 0.584999 loss) | |
| I0401 20:13:57.050418 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.563587 (* 0.3 = 0.169076 loss) | |
| I0401 20:13:57.050431 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.565217 | |
| I0401 20:13:57.050443 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0401 20:13:57.050456 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.73913 | |
| I0401 20:13:57.050469 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.49425 (* 1 = 1.49425 loss) | |
| I0401 20:13:57.050483 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.411637 (* 1 = 0.411637 loss) | |
| I0401 20:13:57.050496 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 20:13:57.050508 6134 solver.cpp:245] Train net output #16: total_confidence = 0.227529 | |
| I0401 20:13:57.050523 6134 sgd_solver.cpp:106] Iteration 102500, lr = 0.01 | |
| I0401 20:16:05.641904 6134 solver.cpp:229] Iteration 103000, loss = 3.20546 | |
| I0401 20:16:05.642001 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0401 20:16:05.642020 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 20:16:05.642036 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.533333 | |
| I0401 20:16:05.642052 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.39746 (* 0.3 = 0.71924 loss) | |
| I0401 20:16:05.642067 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.785214 (* 0.3 = 0.235564 loss) | |
| I0401 20:16:05.642081 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.311111 | |
| I0401 20:16:05.642092 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 20:16:05.642104 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0401 20:16:05.642118 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.12274 (* 0.3 = 0.636823 loss) | |
| I0401 20:16:05.642132 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.660691 (* 0.3 = 0.198207 loss) | |
| I0401 20:16:05.642145 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0401 20:16:05.642158 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 20:16:05.642169 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.888889 | |
| I0401 20:16:05.642184 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.13371 (* 1 = 1.13371 loss) | |
| I0401 20:16:05.642197 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.51384 (* 1 = 0.51384 loss) | |
| I0401 20:16:05.642210 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 20:16:05.642221 6134 solver.cpp:245] Train net output #16: total_confidence = 0.151087 | |
| I0401 20:16:05.642233 6134 sgd_solver.cpp:106] Iteration 103000, lr = 0.01 | |
| I0401 20:18:14.235375 6134 solver.cpp:229] Iteration 103500, loss = 3.18303 | |
| I0401 20:18:14.235697 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0401 20:18:14.235716 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 20:18:14.235729 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.541667 | |
| I0401 20:18:14.235745 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.36049 (* 0.3 = 0.708148 loss) | |
| I0401 20:18:14.235760 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.743305 (* 0.3 = 0.222991 loss) | |
| I0401 20:18:14.235772 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.416667 | |
| I0401 20:18:14.235785 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 20:18:14.235796 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.645833 | |
| I0401 20:18:14.235810 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.88649 (* 0.3 = 0.565947 loss) | |
| I0401 20:18:14.235824 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.611478 (* 0.3 = 0.183443 loss) | |
| I0401 20:18:14.235836 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.645833 | |
| I0401 20:18:14.235848 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 20:18:14.235860 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.791667 | |
| I0401 20:18:14.235874 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.35029 (* 1 = 1.35029 loss) | |
| I0401 20:18:14.235888 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.455845 (* 1 = 0.455845 loss) | |
| I0401 20:18:14.235900 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 20:18:14.235913 6134 solver.cpp:245] Train net output #16: total_confidence = 0.161204 | |
| I0401 20:18:14.235924 6134 sgd_solver.cpp:106] Iteration 103500, lr = 0.01 | |
| I0401 20:20:22.947991 6134 solver.cpp:229] Iteration 104000, loss = 3.16425 | |
| I0401 20:20:22.948129 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.244444 | |
| I0401 20:20:22.948150 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 20:20:22.948165 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.488889 | |
| I0401 20:20:22.948181 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.70294 (* 0.3 = 0.810882 loss) | |
| I0401 20:20:22.948195 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.754738 (* 0.3 = 0.226422 loss) | |
| I0401 20:20:22.948209 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.311111 | |
| I0401 20:20:22.948220 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 20:20:22.948233 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.511111 | |
| I0401 20:20:22.948247 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.41045 (* 0.3 = 0.723136 loss) | |
| I0401 20:20:22.948261 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.666784 (* 0.3 = 0.200035 loss) | |
| I0401 20:20:22.948274 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.488889 | |
| I0401 20:20:22.948287 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0401 20:20:22.948298 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.644444 | |
| I0401 20:20:22.948312 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.88195 (* 1 = 1.88195 loss) | |
| I0401 20:20:22.948326 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.517241 (* 1 = 0.517241 loss) | |
| I0401 20:20:22.948338 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 20:20:22.948350 6134 solver.cpp:245] Train net output #16: total_confidence = 0.172693 | |
| I0401 20:20:22.948362 6134 sgd_solver.cpp:106] Iteration 104000, lr = 0.01 | |
| I0401 20:22:31.697028 6134 solver.cpp:229] Iteration 104500, loss = 3.18663 | |
| I0401 20:22:31.697154 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0401 20:22:31.697175 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 20:22:31.697188 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.530612 | |
| I0401 20:22:31.697206 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.38002 (* 0.3 = 0.714005 loss) | |
| I0401 20:22:31.697221 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.737326 (* 0.3 = 0.221198 loss) | |
| I0401 20:22:31.697232 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.387755 | |
| I0401 20:22:31.697245 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 20:22:31.697258 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.795918 | |
| I0401 20:22:31.697271 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.89749 (* 0.3 = 0.569246 loss) | |
| I0401 20:22:31.697285 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.576328 (* 0.3 = 0.172898 loss) | |
| I0401 20:22:31.697298 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.714286 | |
| I0401 20:22:31.697309 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0401 20:22:31.697321 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.897959 | |
| I0401 20:22:31.697335 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.976941 (* 1 = 0.976941 loss) | |
| I0401 20:22:31.697350 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.306022 (* 1 = 0.306022 loss) | |
| I0401 20:22:31.697361 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 20:22:31.697373 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0990097 | |
| I0401 20:22:31.697386 6134 sgd_solver.cpp:106] Iteration 104500, lr = 0.01 | |
| I0401 20:24:40.310627 6134 solver.cpp:338] Iteration 105000, Testing net (#0) | |
| I0401 20:25:10.067342 6134 solver.cpp:393] Test loss: 2.58692 | |
| I0401 20:25:10.067397 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.445964 | |
| I0401 20:25:10.067414 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.851229 | |
| I0401 20:25:10.067427 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.735027 | |
| I0401 20:25:10.067445 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.88079 (* 0.3 = 0.564236 loss) | |
| I0401 20:25:10.067459 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.515861 (* 0.3 = 0.154758 loss) | |
| I0401 20:25:10.067472 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.572153 | |
| I0401 20:25:10.067484 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.882684 | |
| I0401 20:25:10.067497 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.835552 | |
| I0401 20:25:10.067509 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.42718 (* 0.3 = 0.428153 loss) | |
| I0401 20:25:10.067528 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.395267 (* 0.3 = 0.11858 loss) | |
| I0401 20:25:10.067539 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.73292 | |
| I0401 20:25:10.067553 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.926137 | |
| I0401 20:25:10.067564 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.875628 | |
| I0401 20:25:10.067577 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.03324 (* 1 = 1.03324 loss) | |
| I0401 20:25:10.067590 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.287956 (* 1 = 0.287956 loss) | |
| I0401 20:25:10.067602 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.321 | |
| I0401 20:25:10.067615 6134 solver.cpp:406] Test net output #16: total_confidence = 0.290055 | |
| I0401 20:25:10.217891 6134 solver.cpp:229] Iteration 105000, loss = 3.18891 | |
| I0401 20:25:10.217933 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.3125 | |
| I0401 20:25:10.217950 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 20:25:10.217963 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.520833 | |
| I0401 20:25:10.217978 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.37364 (* 0.3 = 0.712093 loss) | |
| I0401 20:25:10.217993 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.682644 (* 0.3 = 0.204793 loss) | |
| I0401 20:25:10.218005 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4375 | |
| I0401 20:25:10.218017 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 20:25:10.218029 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.770833 | |
| I0401 20:25:10.218045 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.85401 (* 0.3 = 0.556202 loss) | |
| I0401 20:25:10.218060 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.521482 (* 0.3 = 0.156445 loss) | |
| I0401 20:25:10.218072 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.708333 | |
| I0401 20:25:10.218085 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0401 20:25:10.218096 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.854167 | |
| I0401 20:25:10.218109 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.02862 (* 1 = 1.02862 loss) | |
| I0401 20:25:10.218123 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.292467 (* 1 = 0.292467 loss) | |
| I0401 20:25:10.218135 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 20:25:10.218147 6134 solver.cpp:245] Train net output #16: total_confidence = 0.286714 | |
| I0401 20:25:10.218159 6134 sgd_solver.cpp:106] Iteration 105000, lr = 0.01 | |
| I0401 20:27:18.889847 6134 solver.cpp:229] Iteration 105500, loss = 3.11883 | |
| I0401 20:27:18.890231 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0401 20:27:18.890254 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 20:27:18.890266 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5625 | |
| I0401 20:27:18.890282 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.25172 (* 0.3 = 0.675515 loss) | |
| I0401 20:27:18.890297 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.723 (* 0.3 = 0.2169 loss) | |
| I0401 20:27:18.890311 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.395833 | |
| I0401 20:27:18.890322 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0401 20:27:18.890334 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.666667 | |
| I0401 20:27:18.890348 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.05724 (* 0.3 = 0.617173 loss) | |
| I0401 20:27:18.890363 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.677055 (* 0.3 = 0.203116 loss) | |
| I0401 20:27:18.890375 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.5 | |
| I0401 20:27:18.890388 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0401 20:27:18.890400 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.791667 | |
| I0401 20:27:18.890414 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.43742 (* 1 = 1.43742 loss) | |
| I0401 20:27:18.890429 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.51086 (* 1 = 0.51086 loss) | |
| I0401 20:27:18.890441 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 20:27:18.890453 6134 solver.cpp:245] Train net output #16: total_confidence = 0.282611 | |
| I0401 20:27:18.890465 6134 sgd_solver.cpp:106] Iteration 105500, lr = 0.01 | |
| I0401 20:29:27.715417 6134 solver.cpp:229] Iteration 106000, loss = 3.13167 | |
| I0401 20:29:27.715541 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0401 20:29:27.715562 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 20:29:27.715575 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.619048 | |
| I0401 20:29:27.715591 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.18864 (* 0.3 = 0.656591 loss) | |
| I0401 20:29:27.715606 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.615251 (* 0.3 = 0.184575 loss) | |
| I0401 20:29:27.715620 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.428571 | |
| I0401 20:29:27.715631 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 20:29:27.715643 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.761905 | |
| I0401 20:29:27.715657 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.82594 (* 0.3 = 0.547781 loss) | |
| I0401 20:29:27.715672 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.510774 (* 0.3 = 0.153232 loss) | |
| I0401 20:29:27.715684 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.690476 | |
| I0401 20:29:27.715697 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0401 20:29:27.715708 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.880952 | |
| I0401 20:29:27.715723 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.884677 (* 1 = 0.884677 loss) | |
| I0401 20:29:27.715736 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.278475 (* 1 = 0.278475 loss) | |
| I0401 20:29:27.715749 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 20:29:27.715761 6134 solver.cpp:245] Train net output #16: total_confidence = 0.186882 | |
| I0401 20:29:27.715773 6134 sgd_solver.cpp:106] Iteration 106000, lr = 0.01 | |
| I0401 20:31:36.490885 6134 solver.cpp:229] Iteration 106500, loss = 3.25486 | |
| I0401 20:31:36.490993 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0401 20:31:36.491014 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 20:31:36.491026 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.68 | |
| I0401 20:31:36.491042 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.00207 (* 0.3 = 0.60062 loss) | |
| I0401 20:31:36.491058 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.609399 (* 0.3 = 0.18282 loss) | |
| I0401 20:31:36.491070 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 20:31:36.491082 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 20:31:36.491094 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.72 | |
| I0401 20:31:36.491109 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.68745 (* 0.3 = 0.506235 loss) | |
| I0401 20:31:36.491123 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.530981 (* 0.3 = 0.159294 loss) | |
| I0401 20:31:36.491137 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.64 | |
| I0401 20:31:36.491148 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 20:31:36.491160 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.86 | |
| I0401 20:31:36.491174 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.13941 (* 1 = 1.13941 loss) | |
| I0401 20:31:36.491189 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.337012 (* 1 = 0.337012 loss) | |
| I0401 20:31:36.491201 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 20:31:36.491214 6134 solver.cpp:245] Train net output #16: total_confidence = 0.21237 | |
| I0401 20:31:36.491225 6134 sgd_solver.cpp:106] Iteration 106500, lr = 0.01 | |
| I0401 20:33:45.152426 6134 solver.cpp:229] Iteration 107000, loss = 3.20772 | |
| I0401 20:33:45.152587 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.26 | |
| I0401 20:33:45.152608 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 20:33:45.152621 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54 | |
| I0401 20:33:45.152638 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.56301 (* 0.3 = 0.768903 loss) | |
| I0401 20:33:45.152653 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.784105 (* 0.3 = 0.235232 loss) | |
| I0401 20:33:45.152667 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.34 | |
| I0401 20:33:45.152678 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0401 20:33:45.152690 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.64 | |
| I0401 20:33:45.152705 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.12404 (* 0.3 = 0.637211 loss) | |
| I0401 20:33:45.152719 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.638616 (* 0.3 = 0.191585 loss) | |
| I0401 20:33:45.152732 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.7 | |
| I0401 20:33:45.152745 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 20:33:45.152755 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.82 | |
| I0401 20:33:45.152770 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.15874 (* 1 = 1.15874 loss) | |
| I0401 20:33:45.152783 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.34001 (* 1 = 0.34001 loss) | |
| I0401 20:33:45.152796 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 20:33:45.152808 6134 solver.cpp:245] Train net output #16: total_confidence = 0.25782 | |
| I0401 20:33:45.152820 6134 sgd_solver.cpp:106] Iteration 107000, lr = 0.01 | |
| I0401 20:35:53.846971 6134 solver.cpp:229] Iteration 107500, loss = 3.12294 | |
| I0401 20:35:53.847076 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.117647 | |
| I0401 20:35:53.847095 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.738636 | |
| I0401 20:35:53.847107 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.431373 | |
| I0401 20:35:53.847124 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.06462 (* 0.3 = 0.919387 loss) | |
| I0401 20:35:53.847139 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.943071 (* 0.3 = 0.282921 loss) | |
| I0401 20:35:53.847152 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.27451 | |
| I0401 20:35:53.847164 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.789773 | |
| I0401 20:35:53.847177 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.529412 | |
| I0401 20:35:53.847190 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.59189 (* 0.3 = 0.777568 loss) | |
| I0401 20:35:53.847204 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.798741 (* 0.3 = 0.239622 loss) | |
| I0401 20:35:53.847218 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.490196 | |
| I0401 20:35:53.847229 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0401 20:35:53.847241 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.627451 | |
| I0401 20:35:53.847255 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.01828 (* 1 = 2.01828 loss) | |
| I0401 20:35:53.847268 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.679109 (* 1 = 0.679109 loss) | |
| I0401 20:35:53.847281 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 20:35:53.847292 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0823212 | |
| I0401 20:35:53.847306 6134 sgd_solver.cpp:106] Iteration 107500, lr = 0.01 | |
| I0401 20:38:02.607471 6134 solver.cpp:229] Iteration 108000, loss = 3.14168 | |
| I0401 20:38:02.607777 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.156863 | |
| I0401 20:38:02.607798 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.75 | |
| I0401 20:38:02.607811 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.411765 | |
| I0401 20:38:02.607827 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.91557 (* 0.3 = 0.87467 loss) | |
| I0401 20:38:02.607842 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.889946 (* 0.3 = 0.266984 loss) | |
| I0401 20:38:02.607856 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.235294 | |
| I0401 20:38:02.607867 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 20:38:02.607879 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.568627 | |
| I0401 20:38:02.607892 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.62287 (* 0.3 = 0.786861 loss) | |
| I0401 20:38:02.607908 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.780722 (* 0.3 = 0.234217 loss) | |
| I0401 20:38:02.607920 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.392157 | |
| I0401 20:38:02.607933 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.818182 | |
| I0401 20:38:02.607944 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.666667 | |
| I0401 20:38:02.607959 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.05918 (* 1 = 2.05918 loss) | |
| I0401 20:38:02.607972 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.624329 (* 1 = 0.624329 loss) | |
| I0401 20:38:02.607985 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 20:38:02.607996 6134 solver.cpp:245] Train net output #16: total_confidence = 0.147684 | |
| I0401 20:38:02.608008 6134 sgd_solver.cpp:106] Iteration 108000, lr = 0.01 | |
| I0401 20:40:11.514422 6134 solver.cpp:229] Iteration 108500, loss = 3.15093 | |
| I0401 20:40:11.514534 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.304348 | |
| I0401 20:40:11.514554 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 20:40:11.514567 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.608696 | |
| I0401 20:40:11.514585 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.4011 (* 0.3 = 0.720329 loss) | |
| I0401 20:40:11.514600 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.702504 (* 0.3 = 0.210751 loss) | |
| I0401 20:40:11.514612 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 20:40:11.514624 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 20:40:11.514636 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.717391 | |
| I0401 20:40:11.514649 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.06731 (* 0.3 = 0.620193 loss) | |
| I0401 20:40:11.514664 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.674229 (* 0.3 = 0.202269 loss) | |
| I0401 20:40:11.514677 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.521739 | |
| I0401 20:40:11.514688 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.852273 | |
| I0401 20:40:11.514700 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.804348 | |
| I0401 20:40:11.514714 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.46116 (* 1 = 1.46116 loss) | |
| I0401 20:40:11.514727 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.481607 (* 1 = 0.481607 loss) | |
| I0401 20:40:11.514739 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 20:40:11.514751 6134 solver.cpp:245] Train net output #16: total_confidence = 0.150348 | |
| I0401 20:40:11.514763 6134 sgd_solver.cpp:106] Iteration 108500, lr = 0.01 | |
| I0401 20:42:20.158370 6134 solver.cpp:229] Iteration 109000, loss = 3.10781 | |
| I0401 20:42:20.158505 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.24 | |
| I0401 20:42:20.158529 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 20:42:20.158542 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.52 | |
| I0401 20:42:20.158560 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.65843 (* 0.3 = 0.79753 loss) | |
| I0401 20:42:20.158574 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.801586 (* 0.3 = 0.240476 loss) | |
| I0401 20:42:20.158586 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.38 | |
| I0401 20:42:20.158599 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 20:42:20.158612 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.7 | |
| I0401 20:42:20.158625 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.16385 (* 0.3 = 0.649154 loss) | |
| I0401 20:42:20.158639 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.694219 (* 0.3 = 0.208266 loss) | |
| I0401 20:42:20.158651 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.76 | |
| I0401 20:42:20.158663 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 20:42:20.158674 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.88 | |
| I0401 20:42:20.158689 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.08531 (* 1 = 1.08531 loss) | |
| I0401 20:42:20.158704 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.333656 (* 1 = 0.333656 loss) | |
| I0401 20:42:20.158715 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 20:42:20.158726 6134 solver.cpp:245] Train net output #16: total_confidence = 0.195671 | |
| I0401 20:42:20.158738 6134 sgd_solver.cpp:106] Iteration 109000, lr = 0.01 | |
| I0401 20:44:28.886112 6134 solver.cpp:229] Iteration 109500, loss = 3.08329 | |
| I0401 20:44:28.886229 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.327273 | |
| I0401 20:44:28.886258 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 20:44:28.886282 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.563636 | |
| I0401 20:44:28.886312 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.33893 (* 0.3 = 0.701678 loss) | |
| I0401 20:44:28.886338 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.758737 (* 0.3 = 0.227621 loss) | |
| I0401 20:44:28.886360 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.418182 | |
| I0401 20:44:28.886384 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 20:44:28.886406 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.763636 | |
| I0401 20:44:28.886431 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.88043 (* 0.3 = 0.564128 loss) | |
| I0401 20:44:28.886456 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.60427 (* 0.3 = 0.181281 loss) | |
| I0401 20:44:28.886477 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.654545 | |
| I0401 20:44:28.886502 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0401 20:44:28.886528 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.818182 | |
| I0401 20:44:28.886554 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.26468 (* 1 = 1.26468 loss) | |
| I0401 20:44:28.886579 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.410661 (* 1 = 0.410661 loss) | |
| I0401 20:44:28.886600 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 20:44:28.886620 6134 solver.cpp:245] Train net output #16: total_confidence = 0.196794 | |
| I0401 20:44:28.886641 6134 sgd_solver.cpp:106] Iteration 109500, lr = 0.01 | |
| I0401 20:46:37.804672 6134 solver.cpp:338] Iteration 110000, Testing net (#0) | |
| I0401 20:47:07.620965 6134 solver.cpp:393] Test loss: 2.74833 | |
| I0401 20:47:07.621009 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.403395 | |
| I0401 20:47:07.621026 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.831775 | |
| I0401 20:47:07.621038 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.68389 | |
| I0401 20:47:07.621068 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.02018 (* 0.3 = 0.606055 loss) | |
| I0401 20:47:07.621083 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.580927 (* 0.3 = 0.174278 loss) | |
| I0401 20:47:07.621095 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.585988 | |
| I0401 20:47:07.621107 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.879367 | |
| I0401 20:47:07.621119 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.821281 | |
| I0401 20:47:07.621134 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.44967 (* 0.3 = 0.4349 loss) | |
| I0401 20:47:07.621147 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.422477 (* 0.3 = 0.126743 loss) | |
| I0401 20:47:07.621160 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.713275 | |
| I0401 20:47:07.621171 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.926092 | |
| I0401 20:47:07.621183 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.869046 | |
| I0401 20:47:07.621196 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.11319 (* 1 = 1.11319 loss) | |
| I0401 20:47:07.621211 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.293159 (* 1 = 0.293159 loss) | |
| I0401 20:47:07.621222 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.309 | |
| I0401 20:47:07.621234 6134 solver.cpp:406] Test net output #16: total_confidence = 0.276958 | |
| I0401 20:47:07.772079 6134 solver.cpp:229] Iteration 110000, loss = 3.16454 | |
| I0401 20:47:07.772119 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.545455 | |
| I0401 20:47:07.772136 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0401 20:47:07.772150 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.75 | |
| I0401 20:47:07.772169 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.65756 (* 0.3 = 0.497268 loss) | |
| I0401 20:47:07.772186 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.534087 (* 0.3 = 0.160226 loss) | |
| I0401 20:47:07.772198 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.613636 | |
| I0401 20:47:07.772210 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0401 20:47:07.772222 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.772727 | |
| I0401 20:47:07.772236 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.4688 (* 0.3 = 0.440639 loss) | |
| I0401 20:47:07.772251 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.477818 (* 0.3 = 0.143345 loss) | |
| I0401 20:47:07.772264 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.818182 | |
| I0401 20:47:07.772275 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0401 20:47:07.772287 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.909091 | |
| I0401 20:47:07.772301 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.683135 (* 1 = 0.683135 loss) | |
| I0401 20:47:07.772315 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.25268 (* 1 = 0.25268 loss) | |
| I0401 20:47:07.772327 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 20:47:07.772338 6134 solver.cpp:245] Train net output #16: total_confidence = 0.286492 | |
| I0401 20:47:07.772351 6134 sgd_solver.cpp:106] Iteration 110000, lr = 0.01 | |
| I0401 20:49:16.602751 6134 solver.cpp:229] Iteration 110500, loss = 3.17201 | |
| I0401 20:49:16.602903 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0401 20:49:16.602934 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 20:49:16.602962 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.777778 | |
| I0401 20:49:16.602984 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.87451 (* 0.3 = 0.562353 loss) | |
| I0401 20:49:16.602999 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.541921 (* 0.3 = 0.162576 loss) | |
| I0401 20:49:16.603013 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.377778 | |
| I0401 20:49:16.603024 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0401 20:49:16.603036 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.755556 | |
| I0401 20:49:16.603050 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.76402 (* 0.3 = 0.529206 loss) | |
| I0401 20:49:16.603065 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.506193 (* 0.3 = 0.151858 loss) | |
| I0401 20:49:16.603076 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.733333 | |
| I0401 20:49:16.603088 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 20:49:16.603101 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.888889 | |
| I0401 20:49:16.603114 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.33698 (* 1 = 1.33698 loss) | |
| I0401 20:49:16.603128 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.437425 (* 1 = 0.437425 loss) | |
| I0401 20:49:16.603143 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 20:49:16.603163 6134 solver.cpp:245] Train net output #16: total_confidence = 0.278058 | |
| I0401 20:49:16.603176 6134 sgd_solver.cpp:106] Iteration 110500, lr = 0.01 | |
| I0401 20:51:25.384232 6134 solver.cpp:229] Iteration 111000, loss = 3.15266 | |
| I0401 20:51:25.384331 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.295455 | |
| I0401 20:51:25.384351 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 20:51:25.384363 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.522727 | |
| I0401 20:51:25.384379 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.59964 (* 0.3 = 0.779893 loss) | |
| I0401 20:51:25.384394 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.805549 (* 0.3 = 0.241665 loss) | |
| I0401 20:51:25.384407 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.454545 | |
| I0401 20:51:25.384419 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 20:51:25.384431 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.704545 | |
| I0401 20:51:25.384445 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.9665 (* 0.3 = 0.58995 loss) | |
| I0401 20:51:25.384459 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.642247 (* 0.3 = 0.192674 loss) | |
| I0401 20:51:25.384472 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.636364 | |
| I0401 20:51:25.384485 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 20:51:25.384495 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.840909 | |
| I0401 20:51:25.384510 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.12227 (* 1 = 1.12227 loss) | |
| I0401 20:51:25.384523 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.331644 (* 1 = 0.331644 loss) | |
| I0401 20:51:25.384536 6134 solver.cpp:245] Train net output #15: total_accuracy = 0 | |
| I0401 20:51:25.384547 6134 solver.cpp:245] Train net output #16: total_confidence = 0.156779 | |
| I0401 20:51:25.384559 6134 sgd_solver.cpp:106] Iteration 111000, lr = 0.01 | |
| I0401 20:53:34.049607 6134 solver.cpp:229] Iteration 111500, loss = 3.05809 | |
| I0401 20:53:34.049787 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0401 20:53:34.049809 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 20:53:34.049823 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.595238 | |
| I0401 20:53:34.049839 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.163 (* 0.3 = 0.648899 loss) | |
| I0401 20:53:34.049855 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.61917 (* 0.3 = 0.185751 loss) | |
| I0401 20:53:34.049868 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.404762 | |
| I0401 20:53:34.049881 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 20:53:34.049893 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.785714 | |
| I0401 20:53:34.049907 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.53042 (* 0.3 = 0.459125 loss) | |
| I0401 20:53:34.049922 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.424004 (* 0.3 = 0.127201 loss) | |
| I0401 20:53:34.049934 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.904762 | |
| I0401 20:53:34.049947 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.971591 | |
| I0401 20:53:34.049959 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.97619 | |
| I0401 20:53:34.049973 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.520416 (* 1 = 0.520416 loss) | |
| I0401 20:53:34.049988 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.139595 (* 1 = 0.139595 loss) | |
| I0401 20:53:34.050000 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 20:53:34.050014 6134 solver.cpp:245] Train net output #16: total_confidence = 0.271506 | |
| I0401 20:53:34.050025 6134 sgd_solver.cpp:106] Iteration 111500, lr = 0.01 | |
| I0401 20:55:42.800943 6134 solver.cpp:229] Iteration 112000, loss = 3.10554 | |
| I0401 20:55:42.801085 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.409091 | |
| I0401 20:55:42.801110 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0401 20:55:42.801122 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.659091 | |
| I0401 20:55:42.801139 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.11318 (* 0.3 = 0.633955 loss) | |
| I0401 20:55:42.801154 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.598827 (* 0.3 = 0.179648 loss) | |
| I0401 20:55:42.801167 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.590909 | |
| I0401 20:55:42.801179 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0401 20:55:42.801190 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.772727 | |
| I0401 20:55:42.801204 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.66819 (* 0.3 = 0.500458 loss) | |
| I0401 20:55:42.801218 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.476944 (* 0.3 = 0.143083 loss) | |
| I0401 20:55:42.801230 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0401 20:55:42.801242 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0401 20:55:42.801254 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.863636 | |
| I0401 20:55:42.801268 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.856965 (* 1 = 0.856965 loss) | |
| I0401 20:55:42.801282 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.238319 (* 1 = 0.238319 loss) | |
| I0401 20:55:42.801295 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 20:55:42.801306 6134 solver.cpp:245] Train net output #16: total_confidence = 0.314475 | |
| I0401 20:55:42.801319 6134 sgd_solver.cpp:106] Iteration 112000, lr = 0.01 | |
| I0401 20:57:51.608515 6134 solver.cpp:229] Iteration 112500, loss = 3.09701 | |
| I0401 20:57:51.608865 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.3 | |
| I0401 20:57:51.608885 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 20:57:51.608897 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0401 20:57:51.608914 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.39785 (* 0.3 = 0.719355 loss) | |
| I0401 20:57:51.608929 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.726942 (* 0.3 = 0.218083 loss) | |
| I0401 20:57:51.608942 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.34 | |
| I0401 20:57:51.608954 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0401 20:57:51.608966 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.7 | |
| I0401 20:57:51.608980 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.00956 (* 0.3 = 0.602867 loss) | |
| I0401 20:57:51.608995 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.601414 (* 0.3 = 0.180424 loss) | |
| I0401 20:57:51.609007 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.66 | |
| I0401 20:57:51.609019 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 20:57:51.609031 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.86 | |
| I0401 20:57:51.609062 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.28992 (* 1 = 1.28992 loss) | |
| I0401 20:57:51.609079 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.393715 (* 1 = 0.393715 loss) | |
| I0401 20:57:51.609091 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 20:57:51.609103 6134 solver.cpp:245] Train net output #16: total_confidence = 0.196358 | |
| I0401 20:57:51.609115 6134 sgd_solver.cpp:106] Iteration 112500, lr = 0.01 | |
| I0401 21:00:00.570731 6134 solver.cpp:229] Iteration 113000, loss = 3.07903 | |
| I0401 21:00:00.570842 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.319149 | |
| I0401 21:00:00.570859 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 21:00:00.570873 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.680851 | |
| I0401 21:00:00.570889 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.09445 (* 0.3 = 0.628334 loss) | |
| I0401 21:00:00.570904 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.584774 (* 0.3 = 0.175432 loss) | |
| I0401 21:00:00.570917 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.446809 | |
| I0401 21:00:00.570930 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 21:00:00.570941 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.765957 | |
| I0401 21:00:00.570955 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.72616 (* 0.3 = 0.517849 loss) | |
| I0401 21:00:00.570971 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.476698 (* 0.3 = 0.143009 loss) | |
| I0401 21:00:00.570982 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.765957 | |
| I0401 21:00:00.570994 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0401 21:00:00.571007 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.851064 | |
| I0401 21:00:00.571020 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.911548 (* 1 = 0.911548 loss) | |
| I0401 21:00:00.571033 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.266542 (* 1 = 0.266542 loss) | |
| I0401 21:00:00.571045 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 21:00:00.571058 6134 solver.cpp:245] Train net output #16: total_confidence = 0.305107 | |
| I0401 21:00:00.571069 6134 sgd_solver.cpp:106] Iteration 113000, lr = 0.01 | |
| I0401 21:02:09.611042 6134 solver.cpp:229] Iteration 113500, loss = 3.05009 | |
| I0401 21:02:09.611183 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.390244 | |
| I0401 21:02:09.611204 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0401 21:02:09.611225 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.609756 | |
| I0401 21:02:09.611243 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.27393 (* 0.3 = 0.682178 loss) | |
| I0401 21:02:09.611258 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.587908 (* 0.3 = 0.176373 loss) | |
| I0401 21:02:09.611269 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.512195 | |
| I0401 21:02:09.611290 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 21:02:09.611302 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.682927 | |
| I0401 21:02:09.611316 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.7364 (* 0.3 = 0.520921 loss) | |
| I0401 21:02:09.611330 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.502656 (* 0.3 = 0.150797 loss) | |
| I0401 21:02:09.611349 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.682927 | |
| I0401 21:02:09.611361 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0401 21:02:09.611373 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.804878 | |
| I0401 21:02:09.611387 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.04839 (* 1 = 1.04839 loss) | |
| I0401 21:02:09.611402 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.279643 (* 1 = 0.279643 loss) | |
| I0401 21:02:09.611414 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 21:02:09.611426 6134 solver.cpp:245] Train net output #16: total_confidence = 0.150653 | |
| I0401 21:02:09.611438 6134 sgd_solver.cpp:106] Iteration 113500, lr = 0.01 | |
| I0401 21:04:18.750056 6134 solver.cpp:229] Iteration 114000, loss = 3.11182 | |
| I0401 21:04:18.750179 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.326531 | |
| I0401 21:04:18.750207 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 21:04:18.750221 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.632653 | |
| I0401 21:04:18.750238 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.23263 (* 0.3 = 0.669788 loss) | |
| I0401 21:04:18.750253 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.650211 (* 0.3 = 0.195063 loss) | |
| I0401 21:04:18.750267 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.346939 | |
| I0401 21:04:18.750278 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 21:04:18.750291 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.755102 | |
| I0401 21:04:18.750305 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.93391 (* 0.3 = 0.580172 loss) | |
| I0401 21:04:18.750319 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.567262 (* 0.3 = 0.170179 loss) | |
| I0401 21:04:18.750331 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.755102 | |
| I0401 21:04:18.750344 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0401 21:04:18.750356 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.918367 | |
| I0401 21:04:18.750370 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.858558 (* 1 = 0.858558 loss) | |
| I0401 21:04:18.750385 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.254861 (* 1 = 0.254861 loss) | |
| I0401 21:04:18.750397 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 21:04:18.750409 6134 solver.cpp:245] Train net output #16: total_confidence = 0.251196 | |
| I0401 21:04:18.750422 6134 sgd_solver.cpp:106] Iteration 114000, lr = 0.01 | |
| I0401 21:06:27.730038 6134 solver.cpp:229] Iteration 114500, loss = 3.10463 | |
| I0401 21:06:27.730423 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.44186 | |
| I0401 21:06:27.730444 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 21:06:27.730458 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.627907 | |
| I0401 21:06:27.730473 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.37716 (* 0.3 = 0.713149 loss) | |
| I0401 21:06:27.730489 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.707148 (* 0.3 = 0.212144 loss) | |
| I0401 21:06:27.730501 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.511628 | |
| I0401 21:06:27.730514 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 21:06:27.730528 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.767442 | |
| I0401 21:06:27.730543 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.95458 (* 0.3 = 0.586375 loss) | |
| I0401 21:06:27.730557 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.574038 (* 0.3 = 0.172211 loss) | |
| I0401 21:06:27.730571 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.837209 | |
| I0401 21:06:27.730592 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0401 21:06:27.730603 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.930233 | |
| I0401 21:06:27.730618 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.618443 (* 1 = 0.618443 loss) | |
| I0401 21:06:27.730631 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.202066 (* 1 = 0.202066 loss) | |
| I0401 21:06:27.730650 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 21:06:27.730664 6134 solver.cpp:245] Train net output #16: total_confidence = 0.299399 | |
| I0401 21:06:27.730675 6134 sgd_solver.cpp:106] Iteration 114500, lr = 0.01 | |
| I0401 21:08:37.433696 6134 solver.cpp:338] Iteration 115000, Testing net (#0) | |
| I0401 21:09:07.348337 6134 solver.cpp:393] Test loss: 2.81477 | |
| I0401 21:09:07.348410 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.394146 | |
| I0401 21:09:07.348440 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.829592 | |
| I0401 21:09:07.348475 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.638331 | |
| I0401 21:09:07.348508 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 2.19721 (* 0.3 = 0.659163 loss) | |
| I0401 21:09:07.348556 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.621907 (* 0.3 = 0.186572 loss) | |
| I0401 21:09:07.348582 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.611769 | |
| I0401 21:09:07.348608 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.884048 | |
| I0401 21:09:07.348634 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.836601 | |
| I0401 21:09:07.348664 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.39204 (* 0.3 = 0.417612 loss) | |
| I0401 21:09:07.348698 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.409443 (* 0.3 = 0.122833 loss) | |
| I0401 21:09:07.348723 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.714787 | |
| I0401 21:09:07.348748 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.914865 | |
| I0401 21:09:07.348773 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.860527 | |
| I0401 21:09:07.348803 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.10204 (* 1 = 1.10204 loss) | |
| I0401 21:09:07.348831 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.326545 (* 1 = 0.326545 loss) | |
| I0401 21:09:07.348857 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.281 | |
| I0401 21:09:07.348882 6134 solver.cpp:406] Test net output #16: total_confidence = 0.218882 | |
| I0401 21:09:07.500085 6134 solver.cpp:229] Iteration 115000, loss = 3.07551 | |
| I0401 21:09:07.500205 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.372093 | |
| I0401 21:09:07.500226 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 21:09:07.500237 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.627907 | |
| I0401 21:09:07.500253 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.04294 (* 0.3 = 0.612883 loss) | |
| I0401 21:09:07.500268 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.606401 (* 0.3 = 0.18192 loss) | |
| I0401 21:09:07.500282 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.511628 | |
| I0401 21:09:07.500293 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 21:09:07.500305 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.860465 | |
| I0401 21:09:07.500319 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.56615 (* 0.3 = 0.469846 loss) | |
| I0401 21:09:07.500332 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.456576 (* 0.3 = 0.136973 loss) | |
| I0401 21:09:07.500344 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.813953 | |
| I0401 21:09:07.500356 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0401 21:09:07.500367 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.883721 | |
| I0401 21:09:07.500381 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.689037 (* 1 = 0.689037 loss) | |
| I0401 21:09:07.500396 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.195716 (* 1 = 0.195716 loss) | |
| I0401 21:09:07.500408 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 21:09:07.500421 6134 solver.cpp:245] Train net output #16: total_confidence = 0.386977 | |
| I0401 21:09:07.500432 6134 sgd_solver.cpp:106] Iteration 115000, lr = 0.01 | |
| I0401 21:11:16.642848 6134 solver.cpp:229] Iteration 115500, loss = 3.09862 | |
| I0401 21:11:16.642982 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.55 | |
| I0401 21:11:16.643002 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.869318 | |
| I0401 21:11:16.643016 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.75 | |
| I0401 21:11:16.643033 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.69997 (* 0.3 = 0.509992 loss) | |
| I0401 21:11:16.643048 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.511828 (* 0.3 = 0.153549 loss) | |
| I0401 21:11:16.643059 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.675 | |
| I0401 21:11:16.643072 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.909091 | |
| I0401 21:11:16.643085 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.825 | |
| I0401 21:11:16.643100 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.22351 (* 0.3 = 0.367053 loss) | |
| I0401 21:11:16.643113 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.37539 (* 0.3 = 0.112617 loss) | |
| I0401 21:11:16.643126 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.9 | |
| I0401 21:11:16.643137 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0401 21:11:16.643151 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.975 | |
| I0401 21:11:16.643164 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.330447 (* 1 = 0.330447 loss) | |
| I0401 21:11:16.643178 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.138885 (* 1 = 0.138885 loss) | |
| I0401 21:11:16.643190 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0401 21:11:16.643203 6134 solver.cpp:245] Train net output #16: total_confidence = 0.405467 | |
| I0401 21:11:16.643214 6134 sgd_solver.cpp:106] Iteration 115500, lr = 0.01 | |
| I0401 21:13:25.674757 6134 solver.cpp:229] Iteration 116000, loss = 3.06802 | |
| I0401 21:13:25.674906 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.416667 | |
| I0401 21:13:25.674926 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0401 21:13:25.674947 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.722222 | |
| I0401 21:13:25.674963 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.91941 (* 0.3 = 0.575822 loss) | |
| I0401 21:13:25.674978 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.590345 (* 0.3 = 0.177104 loss) | |
| I0401 21:13:25.674991 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 21:13:25.675004 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 21:13:25.675015 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.777778 | |
| I0401 21:13:25.675029 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.65033 (* 0.3 = 0.495099 loss) | |
| I0401 21:13:25.675050 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.566718 (* 0.3 = 0.170015 loss) | |
| I0401 21:13:25.675061 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0401 21:13:25.675076 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0401 21:13:25.675088 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.888889 | |
| I0401 21:13:25.675110 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.07086 (* 1 = 1.07086 loss) | |
| I0401 21:13:25.675124 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.396916 (* 1 = 0.396916 loss) | |
| I0401 21:13:25.675137 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 21:13:25.675149 6134 solver.cpp:245] Train net output #16: total_confidence = 0.263439 | |
| I0401 21:13:25.675161 6134 sgd_solver.cpp:106] Iteration 116000, lr = 0.01 | |
| I0401 21:15:35.049955 6134 solver.cpp:229] Iteration 116500, loss = 3.04055 | |
| I0401 21:15:35.050079 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.512821 | |
| I0401 21:15:35.050102 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.875 | |
| I0401 21:15:35.050117 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.769231 | |
| I0401 21:15:35.050133 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.53854 (* 0.3 = 0.461563 loss) | |
| I0401 21:15:35.050148 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.413181 (* 0.3 = 0.123954 loss) | |
| I0401 21:15:35.050161 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.538462 | |
| I0401 21:15:35.050173 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0401 21:15:35.050186 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.897436 | |
| I0401 21:15:35.050199 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.30515 (* 0.3 = 0.391545 loss) | |
| I0401 21:15:35.050215 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.384231 (* 0.3 = 0.115269 loss) | |
| I0401 21:15:35.050226 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.923077 | |
| I0401 21:15:35.050238 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.971591 | |
| I0401 21:15:35.050251 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0401 21:15:35.050264 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.295763 (* 1 = 0.295763 loss) | |
| I0401 21:15:35.050278 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.117372 (* 1 = 0.117372 loss) | |
| I0401 21:15:35.050290 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0401 21:15:35.050302 6134 solver.cpp:245] Train net output #16: total_confidence = 0.337304 | |
| I0401 21:15:35.050314 6134 sgd_solver.cpp:106] Iteration 116500, lr = 0.01 | |
| I0401 21:17:44.339299 6134 solver.cpp:229] Iteration 117000, loss = 2.99817 | |
| I0401 21:17:44.339704 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.32 | |
| I0401 21:17:44.339725 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 21:17:44.339740 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54 | |
| I0401 21:17:44.339756 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.57695 (* 0.3 = 0.773085 loss) | |
| I0401 21:17:44.339771 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.820458 (* 0.3 = 0.246137 loss) | |
| I0401 21:17:44.339783 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0401 21:17:44.339797 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 21:17:44.339808 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.64 | |
| I0401 21:17:44.339823 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.22796 (* 0.3 = 0.668389 loss) | |
| I0401 21:17:44.339838 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.69269 (* 0.3 = 0.207807 loss) | |
| I0401 21:17:44.339849 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.74 | |
| I0401 21:17:44.339861 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 21:17:44.339874 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.78 | |
| I0401 21:17:44.339889 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.82612 (* 1 = 1.82612 loss) | |
| I0401 21:17:44.339902 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.538617 (* 1 = 0.538617 loss) | |
| I0401 21:17:44.339915 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 21:17:44.339927 6134 solver.cpp:245] Train net output #16: total_confidence = 0.358872 | |
| I0401 21:17:44.339939 6134 sgd_solver.cpp:106] Iteration 117000, lr = 0.01 | |
| I0401 21:19:53.742383 6134 solver.cpp:229] Iteration 117500, loss = 3.00549 | |
| I0401 21:19:53.742493 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.382979 | |
| I0401 21:19:53.742514 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 21:19:53.742530 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.617021 | |
| I0401 21:19:53.742547 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.13448 (* 0.3 = 0.640344 loss) | |
| I0401 21:19:53.742561 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.653879 (* 0.3 = 0.196164 loss) | |
| I0401 21:19:53.742573 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.425532 | |
| I0401 21:19:53.742586 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 21:19:53.742597 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.723404 | |
| I0401 21:19:53.742611 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.74353 (* 0.3 = 0.52306 loss) | |
| I0401 21:19:53.742625 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.524017 (* 0.3 = 0.157205 loss) | |
| I0401 21:19:53.742638 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.808511 | |
| I0401 21:19:53.742650 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0401 21:19:53.742662 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.87234 | |
| I0401 21:19:53.742676 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.717668 (* 1 = 0.717668 loss) | |
| I0401 21:19:53.742691 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.197834 (* 1 = 0.197834 loss) | |
| I0401 21:19:53.742702 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0401 21:19:53.742714 6134 solver.cpp:245] Train net output #16: total_confidence = 0.337985 | |
| I0401 21:19:53.742727 6134 sgd_solver.cpp:106] Iteration 117500, lr = 0.01 | |
| I0401 21:22:03.066050 6134 solver.cpp:229] Iteration 118000, loss = 3.0734 | |
| I0401 21:22:03.066169 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.302326 | |
| I0401 21:22:03.066190 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 21:22:03.066202 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.651163 | |
| I0401 21:22:03.066218 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.21521 (* 0.3 = 0.664563 loss) | |
| I0401 21:22:03.066233 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.603712 (* 0.3 = 0.181114 loss) | |
| I0401 21:22:03.066246 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.511628 | |
| I0401 21:22:03.066257 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 21:22:03.066269 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.767442 | |
| I0401 21:22:03.066283 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.63865 (* 0.3 = 0.491594 loss) | |
| I0401 21:22:03.066298 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.466374 (* 0.3 = 0.139912 loss) | |
| I0401 21:22:03.066310 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.813953 | |
| I0401 21:22:03.066323 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0401 21:22:03.066334 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.883721 | |
| I0401 21:22:03.066349 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.799151 (* 1 = 0.799151 loss) | |
| I0401 21:22:03.066362 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.237368 (* 1 = 0.237368 loss) | |
| I0401 21:22:03.066375 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 21:22:03.066387 6134 solver.cpp:245] Train net output #16: total_confidence = 0.227966 | |
| I0401 21:22:03.066400 6134 sgd_solver.cpp:106] Iteration 118000, lr = 0.01 | |
| I0401 21:24:12.180469 6134 solver.cpp:229] Iteration 118500, loss = 3.08438 | |
| I0401 21:24:12.180603 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.206897 | |
| I0401 21:24:12.180625 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.727273 | |
| I0401 21:24:12.180639 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.362069 | |
| I0401 21:24:12.180655 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.96626 (* 0.3 = 0.889879 loss) | |
| I0401 21:24:12.180670 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.00356 (* 0.3 = 0.301067 loss) | |
| I0401 21:24:12.180683 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.275862 | |
| I0401 21:24:12.180696 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.761364 | |
| I0401 21:24:12.180708 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.517241 | |
| I0401 21:24:12.180721 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.92721 (* 0.3 = 0.878162 loss) | |
| I0401 21:24:12.180737 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.978717 (* 0.3 = 0.293615 loss) | |
| I0401 21:24:12.180748 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.482759 | |
| I0401 21:24:12.180762 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.829545 | |
| I0401 21:24:12.180773 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.62069 | |
| I0401 21:24:12.180788 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.39682 (* 1 = 2.39682 loss) | |
| I0401 21:24:12.180802 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.794365 (* 1 = 0.794365 loss) | |
| I0401 21:24:12.180814 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 21:24:12.180827 6134 solver.cpp:245] Train net output #16: total_confidence = 0.158034 | |
| I0401 21:24:12.180840 6134 sgd_solver.cpp:106] Iteration 118500, lr = 0.01 | |
| I0401 21:26:21.260782 6134 solver.cpp:229] Iteration 119000, loss = 3.02618 | |
| I0401 21:26:21.260920 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.487179 | |
| I0401 21:26:21.260941 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 21:26:21.260953 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0401 21:26:21.260969 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.15062 (* 0.3 = 0.645186 loss) | |
| I0401 21:26:21.260984 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.673331 (* 0.3 = 0.201999 loss) | |
| I0401 21:26:21.260998 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.512821 | |
| I0401 21:26:21.261009 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0401 21:26:21.261023 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.769231 | |
| I0401 21:26:21.261036 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.78918 (* 0.3 = 0.536753 loss) | |
| I0401 21:26:21.261067 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.533039 (* 0.3 = 0.159912 loss) | |
| I0401 21:26:21.261081 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.717949 | |
| I0401 21:26:21.261093 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 21:26:21.261106 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.897436 | |
| I0401 21:26:21.261119 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.00439 (* 1 = 1.00439 loss) | |
| I0401 21:26:21.261133 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.308866 (* 1 = 0.308866 loss) | |
| I0401 21:26:21.261147 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 21:26:21.261158 6134 solver.cpp:245] Train net output #16: total_confidence = 0.231927 | |
| I0401 21:26:21.261171 6134 sgd_solver.cpp:106] Iteration 119000, lr = 0.01 | |
| I0401 21:28:29.982239 6134 solver.cpp:229] Iteration 119500, loss = 2.96259 | |
| I0401 21:28:29.982497 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.375 | |
| I0401 21:28:29.982517 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 21:28:29.982529 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.520833 | |
| I0401 21:28:29.982545 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.49155 (* 0.3 = 0.747466 loss) | |
| I0401 21:28:29.982560 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.745781 (* 0.3 = 0.223734 loss) | |
| I0401 21:28:29.982574 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 21:28:29.982586 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 21:28:29.982599 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.708333 | |
| I0401 21:28:29.982612 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.9561 (* 0.3 = 0.586829 loss) | |
| I0401 21:28:29.982626 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.582508 (* 0.3 = 0.174753 loss) | |
| I0401 21:28:29.982638 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.625 | |
| I0401 21:28:29.982650 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 21:28:29.982662 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.833333 | |
| I0401 21:28:29.982676 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.47331 (* 1 = 1.47331 loss) | |
| I0401 21:28:29.982689 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.420362 (* 1 = 0.420362 loss) | |
| I0401 21:28:29.982702 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 21:28:29.982714 6134 solver.cpp:245] Train net output #16: total_confidence = 0.203283 | |
| I0401 21:28:29.982727 6134 sgd_solver.cpp:106] Iteration 119500, lr = 0.01 | |
| I0401 21:30:38.721177 6134 solver.cpp:338] Iteration 120000, Testing net (#0) | |
| I0401 21:31:08.614490 6134 solver.cpp:393] Test loss: 2.46792 | |
| I0401 21:31:08.614539 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.467879 | |
| I0401 21:31:08.614557 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.861276 | |
| I0401 21:31:08.614569 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.747206 | |
| I0401 21:31:08.614585 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.8082 (* 0.3 = 0.542461 loss) | |
| I0401 21:31:08.614600 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.478965 (* 0.3 = 0.14369 loss) | |
| I0401 21:31:08.614612 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.610903 | |
| I0401 21:31:08.614625 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.897367 | |
| I0401 21:31:08.614637 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.844399 | |
| I0401 21:31:08.614650 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.36246 (* 0.3 = 0.408739 loss) | |
| I0401 21:31:08.614665 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.362173 (* 0.3 = 0.108652 loss) | |
| I0401 21:31:08.614677 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.747257 | |
| I0401 21:31:08.614689 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.935774 | |
| I0401 21:31:08.614701 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.881838 | |
| I0401 21:31:08.614714 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.00661 (* 1 = 1.00661 loss) | |
| I0401 21:31:08.614728 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.257767 (* 1 = 0.257767 loss) | |
| I0401 21:31:08.614740 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.41 | |
| I0401 21:31:08.614751 6134 solver.cpp:406] Test net output #16: total_confidence = 0.360301 | |
| I0401 21:31:08.766552 6134 solver.cpp:229] Iteration 120000, loss = 3.06511 | |
| I0401 21:31:08.766652 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.340426 | |
| I0401 21:31:08.766672 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 21:31:08.766685 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.680851 | |
| I0401 21:31:08.766700 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.13186 (* 0.3 = 0.639559 loss) | |
| I0401 21:31:08.766716 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.60825 (* 0.3 = 0.182475 loss) | |
| I0401 21:31:08.766727 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.510638 | |
| I0401 21:31:08.766741 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 21:31:08.766752 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.744681 | |
| I0401 21:31:08.766767 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.63049 (* 0.3 = 0.489147 loss) | |
| I0401 21:31:08.766780 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.480539 (* 0.3 = 0.144162 loss) | |
| I0401 21:31:08.766791 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.829787 | |
| I0401 21:31:08.766803 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0401 21:31:08.766815 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.957447 | |
| I0401 21:31:08.766829 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.752787 (* 1 = 0.752787 loss) | |
| I0401 21:31:08.766844 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.211452 (* 1 = 0.211452 loss) | |
| I0401 21:31:08.766856 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 21:31:08.766868 6134 solver.cpp:245] Train net output #16: total_confidence = 0.384081 | |
| I0401 21:31:08.766880 6134 sgd_solver.cpp:106] Iteration 120000, lr = 0.01 | |
| I0401 21:33:17.790287 6134 solver.cpp:229] Iteration 120500, loss = 3.04799 | |
| I0401 21:33:17.790418 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.270833 | |
| I0401 21:33:17.790438 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 21:33:17.790452 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.458333 | |
| I0401 21:33:17.790467 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.68738 (* 0.3 = 0.806213 loss) | |
| I0401 21:33:17.790482 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.777644 (* 0.3 = 0.233293 loss) | |
| I0401 21:33:17.790494 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.375 | |
| I0401 21:33:17.790508 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0401 21:33:17.790521 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.645833 | |
| I0401 21:33:17.790536 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.35438 (* 0.3 = 0.706314 loss) | |
| I0401 21:33:17.790550 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.67296 (* 0.3 = 0.201888 loss) | |
| I0401 21:33:17.790563 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.708333 | |
| I0401 21:33:17.790575 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 21:33:17.790587 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.791667 | |
| I0401 21:33:17.790601 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.67818 (* 1 = 1.67818 loss) | |
| I0401 21:33:17.790616 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.492554 (* 1 = 0.492554 loss) | |
| I0401 21:33:17.790627 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 21:33:17.790639 6134 solver.cpp:245] Train net output #16: total_confidence = 0.210683 | |
| I0401 21:33:17.790652 6134 sgd_solver.cpp:106] Iteration 120500, lr = 0.01 | |
| I0401 21:35:26.687125 6134 solver.cpp:229] Iteration 121000, loss = 3.00929 | |
| I0401 21:35:26.687222 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.404762 | |
| I0401 21:35:26.687240 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 21:35:26.687253 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.547619 | |
| I0401 21:35:26.687270 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.41594 (* 0.3 = 0.724782 loss) | |
| I0401 21:35:26.687284 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.722074 (* 0.3 = 0.216622 loss) | |
| I0401 21:35:26.687297 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.452381 | |
| I0401 21:35:26.687312 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 21:35:26.687325 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.690476 | |
| I0401 21:35:26.687340 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.04233 (* 0.3 = 0.6127 loss) | |
| I0401 21:35:26.687355 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.58377 (* 0.3 = 0.175131 loss) | |
| I0401 21:35:26.687366 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.571429 | |
| I0401 21:35:26.687378 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 21:35:26.687391 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.690476 | |
| I0401 21:35:26.687404 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.56763 (* 1 = 1.56763 loss) | |
| I0401 21:35:26.687418 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.441366 (* 1 = 0.441366 loss) | |
| I0401 21:35:26.687430 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 21:35:26.687443 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0759415 | |
| I0401 21:35:26.687455 6134 sgd_solver.cpp:106] Iteration 121000, lr = 0.01 | |
| I0401 21:37:35.394323 6134 solver.cpp:229] Iteration 121500, loss = 3.02634 | |
| I0401 21:37:35.394645 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.292683 | |
| I0401 21:37:35.394666 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 21:37:35.394680 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.463415 | |
| I0401 21:37:35.394696 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.2495 (* 0.3 = 0.67485 loss) | |
| I0401 21:37:35.394711 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.673329 (* 0.3 = 0.201999 loss) | |
| I0401 21:37:35.394724 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.439024 | |
| I0401 21:37:35.394737 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 21:37:35.394749 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.682927 | |
| I0401 21:37:35.394763 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.99105 (* 0.3 = 0.597315 loss) | |
| I0401 21:37:35.394778 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.604402 (* 0.3 = 0.181321 loss) | |
| I0401 21:37:35.394789 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.536585 | |
| I0401 21:37:35.394801 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0401 21:37:35.394814 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.682927 | |
| I0401 21:37:35.394827 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.58517 (* 1 = 1.58517 loss) | |
| I0401 21:37:35.394841 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.474835 (* 1 = 0.474835 loss) | |
| I0401 21:37:35.394853 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 21:37:35.394865 6134 solver.cpp:245] Train net output #16: total_confidence = 0.255485 | |
| I0401 21:37:35.394876 6134 sgd_solver.cpp:106] Iteration 121500, lr = 0.01 | |
| I0401 21:39:44.038480 6134 solver.cpp:229] Iteration 122000, loss = 3.01625 | |
| I0401 21:39:44.038600 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.3125 | |
| I0401 21:39:44.038621 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.778409 | |
| I0401 21:39:44.038635 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.583333 | |
| I0401 21:39:44.038650 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.2486 (* 0.3 = 0.674581 loss) | |
| I0401 21:39:44.038664 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.729504 (* 0.3 = 0.218851 loss) | |
| I0401 21:39:44.038677 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.458333 | |
| I0401 21:39:44.038691 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 21:39:44.038702 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.625 | |
| I0401 21:39:44.038717 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.88599 (* 0.3 = 0.565798 loss) | |
| I0401 21:39:44.038730 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.587426 (* 0.3 = 0.176228 loss) | |
| I0401 21:39:44.038743 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0401 21:39:44.038755 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 21:39:44.038768 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.875 | |
| I0401 21:39:44.038781 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.029 (* 1 = 1.029 loss) | |
| I0401 21:39:44.038795 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.317776 (* 1 = 0.317776 loss) | |
| I0401 21:39:44.038807 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 21:39:44.038818 6134 solver.cpp:245] Train net output #16: total_confidence = 0.148153 | |
| I0401 21:39:44.038830 6134 sgd_solver.cpp:106] Iteration 122000, lr = 0.01 | |
| I0401 21:41:52.766170 6134 solver.cpp:229] Iteration 122500, loss = 3.07265 | |
| I0401 21:41:52.766324 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.3125 | |
| I0401 21:41:52.766345 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 21:41:52.766358 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.583333 | |
| I0401 21:41:52.766376 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.27269 (* 0.3 = 0.681808 loss) | |
| I0401 21:41:52.766389 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.722471 (* 0.3 = 0.216741 loss) | |
| I0401 21:41:52.766403 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.416667 | |
| I0401 21:41:52.766415 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0401 21:41:52.766427 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.729167 | |
| I0401 21:41:52.766441 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.76157 (* 0.3 = 0.528471 loss) | |
| I0401 21:41:52.766455 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.577845 (* 0.3 = 0.173354 loss) | |
| I0401 21:41:52.766468 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0401 21:41:52.766480 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0401 21:41:52.766491 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.8125 | |
| I0401 21:41:52.766505 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.33243 (* 1 = 1.33243 loss) | |
| I0401 21:41:52.766523 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.475626 (* 1 = 0.475626 loss) | |
| I0401 21:41:52.766535 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 21:41:52.766547 6134 solver.cpp:245] Train net output #16: total_confidence = 0.265638 | |
| I0401 21:41:52.766561 6134 sgd_solver.cpp:106] Iteration 122500, lr = 0.01 | |
| I0401 21:44:01.424077 6134 solver.cpp:229] Iteration 123000, loss = 3.03214 | |
| I0401 21:44:01.424186 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.347826 | |
| I0401 21:44:01.424206 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 21:44:01.424219 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.543478 | |
| I0401 21:44:01.424237 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.29497 (* 0.3 = 0.688492 loss) | |
| I0401 21:44:01.424252 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.69531 (* 0.3 = 0.208593 loss) | |
| I0401 21:44:01.424264 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 21:44:01.424276 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 21:44:01.424288 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.695652 | |
| I0401 21:44:01.424304 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.77653 (* 0.3 = 0.53296 loss) | |
| I0401 21:44:01.424319 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.532323 (* 0.3 = 0.159697 loss) | |
| I0401 21:44:01.424330 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.673913 | |
| I0401 21:44:01.424342 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 21:44:01.424355 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.76087 | |
| I0401 21:44:01.424368 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.17737 (* 1 = 1.17737 loss) | |
| I0401 21:44:01.424382 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.3504 (* 1 = 0.3504 loss) | |
| I0401 21:44:01.424396 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 21:44:01.424407 6134 solver.cpp:245] Train net output #16: total_confidence = 0.299731 | |
| I0401 21:44:01.424419 6134 sgd_solver.cpp:106] Iteration 123000, lr = 0.01 | |
| I0401 21:46:10.072324 6134 solver.cpp:229] Iteration 123500, loss = 2.99169 | |
| I0401 21:46:10.072460 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.416667 | |
| I0401 21:46:10.072481 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 21:46:10.072494 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.583333 | |
| I0401 21:46:10.072510 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.06729 (* 0.3 = 0.620187 loss) | |
| I0401 21:46:10.072526 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.613022 (* 0.3 = 0.183907 loss) | |
| I0401 21:46:10.072540 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 21:46:10.072552 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 21:46:10.072563 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0401 21:46:10.072577 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.65079 (* 0.3 = 0.495237 loss) | |
| I0401 21:46:10.072592 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.497878 (* 0.3 = 0.149363 loss) | |
| I0401 21:46:10.072603 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.791667 | |
| I0401 21:46:10.072616 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0401 21:46:10.072628 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.854167 | |
| I0401 21:46:10.072643 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.755394 (* 1 = 0.755394 loss) | |
| I0401 21:46:10.072655 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.214354 (* 1 = 0.214354 loss) | |
| I0401 21:46:10.072669 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0401 21:46:10.072679 6134 solver.cpp:245] Train net output #16: total_confidence = 0.395969 | |
| I0401 21:46:10.072691 6134 sgd_solver.cpp:106] Iteration 123500, lr = 0.01 | |
| I0401 21:48:18.757720 6134 solver.cpp:229] Iteration 124000, loss = 3.07176 | |
| I0401 21:48:18.758051 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.340426 | |
| I0401 21:48:18.758072 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 21:48:18.758085 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.744681 | |
| I0401 21:48:18.758101 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.9709 (* 0.3 = 0.591269 loss) | |
| I0401 21:48:18.758116 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.588817 (* 0.3 = 0.176645 loss) | |
| I0401 21:48:18.758128 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.489362 | |
| I0401 21:48:18.758141 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 21:48:18.758153 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.87234 | |
| I0401 21:48:18.758167 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.39174 (* 0.3 = 0.417521 loss) | |
| I0401 21:48:18.758182 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.404885 (* 0.3 = 0.121466 loss) | |
| I0401 21:48:18.758193 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.93617 | |
| I0401 21:48:18.758205 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.982955 | |
| I0401 21:48:18.758218 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.957447 | |
| I0401 21:48:18.758232 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.543056 (* 1 = 0.543056 loss) | |
| I0401 21:48:18.758246 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.15979 (* 1 = 0.15979 loss) | |
| I0401 21:48:18.758258 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0401 21:48:18.758270 6134 solver.cpp:245] Train net output #16: total_confidence = 0.236918 | |
| I0401 21:48:18.758282 6134 sgd_solver.cpp:106] Iteration 124000, lr = 0.01 | |
| I0401 21:50:27.483078 6134 solver.cpp:229] Iteration 124500, loss = 2.88758 | |
| I0401 21:50:27.483240 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.265306 | |
| I0401 21:50:27.483270 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 21:50:27.483294 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.591837 | |
| I0401 21:50:27.483319 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.43453 (* 0.3 = 0.730359 loss) | |
| I0401 21:50:27.483335 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.720591 (* 0.3 = 0.216177 loss) | |
| I0401 21:50:27.483351 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.44898 | |
| I0401 21:50:27.483364 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 21:50:27.483376 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.693878 | |
| I0401 21:50:27.483391 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.90128 (* 0.3 = 0.570385 loss) | |
| I0401 21:50:27.483404 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.562827 (* 0.3 = 0.168848 loss) | |
| I0401 21:50:27.483417 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.653061 | |
| I0401 21:50:27.483428 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 21:50:27.483440 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.795918 | |
| I0401 21:50:27.483455 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.17577 (* 1 = 1.17577 loss) | |
| I0401 21:50:27.483469 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.343231 (* 1 = 0.343231 loss) | |
| I0401 21:50:27.483481 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 21:50:27.483494 6134 solver.cpp:245] Train net output #16: total_confidence = 0.205269 | |
| I0401 21:50:27.483505 6134 sgd_solver.cpp:106] Iteration 124500, lr = 0.01 | |
| I0401 21:52:36.435792 6134 solver.cpp:338] Iteration 125000, Testing net (#0) | |
| I0401 21:53:06.286003 6134 solver.cpp:393] Test loss: 2.87532 | |
| I0401 21:53:06.286064 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.427454 | |
| I0401 21:53:06.286082 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.848911 | |
| I0401 21:53:06.286094 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.712302 | |
| I0401 21:53:06.286110 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.99076 (* 0.3 = 0.597229 loss) | |
| I0401 21:53:06.286125 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.539994 (* 0.3 = 0.161998 loss) | |
| I0401 21:53:06.286139 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.605635 | |
| I0401 21:53:06.286150 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.887776 | |
| I0401 21:53:06.286162 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.827654 | |
| I0401 21:53:06.286176 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.4482 (* 0.3 = 0.434459 loss) | |
| I0401 21:53:06.286190 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.415422 (* 0.3 = 0.124627 loss) | |
| I0401 21:53:06.286202 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.71342 | |
| I0401 21:53:06.286216 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.927047 | |
| I0401 21:53:06.286227 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.845268 | |
| I0401 21:53:06.286240 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.23418 (* 1 = 1.23418 loss) | |
| I0401 21:53:06.286254 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.32283 (* 1 = 0.32283 loss) | |
| I0401 21:53:06.286267 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.365 | |
| I0401 21:53:06.286278 6134 solver.cpp:406] Test net output #16: total_confidence = 0.329559 | |
| I0401 21:53:06.438040 6134 solver.cpp:229] Iteration 125000, loss = 3.01147 | |
| I0401 21:53:06.438189 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.422222 | |
| I0401 21:53:06.438228 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 21:53:06.438257 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.577778 | |
| I0401 21:53:06.438278 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.22906 (* 0.3 = 0.668719 loss) | |
| I0401 21:53:06.438302 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.697996 (* 0.3 = 0.209399 loss) | |
| I0401 21:53:06.438313 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.511111 | |
| I0401 21:53:06.438326 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 21:53:06.438338 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.777778 | |
| I0401 21:53:06.438359 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.82202 (* 0.3 = 0.546606 loss) | |
| I0401 21:53:06.438374 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.557892 (* 0.3 = 0.167368 loss) | |
| I0401 21:53:06.438386 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.755556 | |
| I0401 21:53:06.438398 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 21:53:06.438410 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.933333 | |
| I0401 21:53:06.438424 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.16355 (* 1 = 1.16355 loss) | |
| I0401 21:53:06.438438 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.33006 (* 1 = 0.33006 loss) | |
| I0401 21:53:06.438451 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 21:53:06.438462 6134 solver.cpp:245] Train net output #16: total_confidence = 0.29961 | |
| I0401 21:53:06.438474 6134 sgd_solver.cpp:106] Iteration 125000, lr = 0.01 | |
| I0401 21:55:15.379734 6134 solver.cpp:229] Iteration 125500, loss = 2.92376 | |
| I0401 21:55:15.379847 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0401 21:55:15.379868 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 21:55:15.379880 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.577778 | |
| I0401 21:55:15.379896 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.49343 (* 0.3 = 0.748029 loss) | |
| I0401 21:55:15.379911 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.689151 (* 0.3 = 0.206745 loss) | |
| I0401 21:55:15.379925 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.311111 | |
| I0401 21:55:15.379936 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 21:55:15.379948 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.577778 | |
| I0401 21:55:15.379962 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.49286 (* 0.3 = 0.747857 loss) | |
| I0401 21:55:15.379976 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.684915 (* 0.3 = 0.205475 loss) | |
| I0401 21:55:15.379989 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.533333 | |
| I0401 21:55:15.380002 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 21:55:15.380013 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.688889 | |
| I0401 21:55:15.380028 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.77827 (* 1 = 1.77827 loss) | |
| I0401 21:55:15.380040 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.493407 (* 1 = 0.493407 loss) | |
| I0401 21:55:15.380053 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 21:55:15.380065 6134 solver.cpp:245] Train net output #16: total_confidence = 0.316205 | |
| I0401 21:55:15.380077 6134 sgd_solver.cpp:106] Iteration 125500, lr = 0.01 | |
| I0401 21:57:24.380252 6134 solver.cpp:229] Iteration 126000, loss = 2.98182 | |
| I0401 21:57:24.380614 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.309524 | |
| I0401 21:57:24.380637 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 21:57:24.380650 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.714286 | |
| I0401 21:57:24.380666 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.79878 (* 0.3 = 0.539633 loss) | |
| I0401 21:57:24.380681 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.474093 (* 0.3 = 0.142228 loss) | |
| I0401 21:57:24.380694 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 21:57:24.380707 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0401 21:57:24.380723 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.833333 | |
| I0401 21:57:24.380753 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.41944 (* 0.3 = 0.425833 loss) | |
| I0401 21:57:24.380772 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.389385 (* 0.3 = 0.116815 loss) | |
| I0401 21:57:24.380785 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.642857 | |
| I0401 21:57:24.380806 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0401 21:57:24.380818 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.904762 | |
| I0401 21:57:24.380832 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.02744 (* 1 = 1.02744 loss) | |
| I0401 21:57:24.380847 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.291565 (* 1 = 0.291565 loss) | |
| I0401 21:57:24.380859 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 21:57:24.380879 6134 solver.cpp:245] Train net output #16: total_confidence = 0.216189 | |
| I0401 21:57:24.380892 6134 sgd_solver.cpp:106] Iteration 126000, lr = 0.01 | |
| I0401 21:59:33.295614 6134 solver.cpp:229] Iteration 126500, loss = 2.9302 | |
| I0401 21:59:33.295717 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0401 21:59:33.295737 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 21:59:33.295749 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6 | |
| I0401 21:59:33.295766 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.15714 (* 0.3 = 0.647142 loss) | |
| I0401 21:59:33.295780 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.646055 (* 0.3 = 0.193816 loss) | |
| I0401 21:59:33.295794 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.44 | |
| I0401 21:59:33.295805 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 21:59:33.295817 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.74 | |
| I0401 21:59:33.295831 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.7318 (* 0.3 = 0.519539 loss) | |
| I0401 21:59:33.295846 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.519978 (* 0.3 = 0.155993 loss) | |
| I0401 21:59:33.295858 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.7 | |
| I0401 21:59:33.295871 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 21:59:33.295882 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.92 | |
| I0401 21:59:33.295897 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.90286 (* 1 = 0.90286 loss) | |
| I0401 21:59:33.295910 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.263895 (* 1 = 0.263895 loss) | |
| I0401 21:59:33.295922 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 21:59:33.295934 6134 solver.cpp:245] Train net output #16: total_confidence = 0.326621 | |
| I0401 21:59:33.295946 6134 sgd_solver.cpp:106] Iteration 126500, lr = 0.01 | |
| I0401 22:01:42.270443 6134 solver.cpp:229] Iteration 127000, loss = 3.041 | |
| I0401 22:01:42.270577 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.5 | |
| I0401 22:01:42.270597 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0401 22:01:42.270611 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0401 22:01:42.270627 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.92485 (* 0.3 = 0.577454 loss) | |
| I0401 22:01:42.270642 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.568849 (* 0.3 = 0.170655 loss) | |
| I0401 22:01:42.270654 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.645833 | |
| I0401 22:01:42.270666 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0401 22:01:42.270678 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8125 | |
| I0401 22:01:42.270692 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.4123 (* 0.3 = 0.423689 loss) | |
| I0401 22:01:42.270706 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.453808 (* 0.3 = 0.136142 loss) | |
| I0401 22:01:42.270720 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.8125 | |
| I0401 22:01:42.270731 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0401 22:01:42.270743 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9375 | |
| I0401 22:01:42.270756 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.701566 (* 1 = 0.701566 loss) | |
| I0401 22:01:42.270771 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.209528 (* 1 = 0.209528 loss) | |
| I0401 22:01:42.270782 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 22:01:42.270794 6134 solver.cpp:245] Train net output #16: total_confidence = 0.353933 | |
| I0401 22:01:42.270807 6134 sgd_solver.cpp:106] Iteration 127000, lr = 0.01 | |
| I0401 22:03:50.969490 6134 solver.cpp:229] Iteration 127500, loss = 2.96203 | |
| I0401 22:03:50.969629 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.222222 | |
| I0401 22:03:50.969648 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 22:03:50.969661 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0401 22:03:50.969678 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.71318 (* 0.3 = 0.813955 loss) | |
| I0401 22:03:50.969693 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.853313 (* 0.3 = 0.255994 loss) | |
| I0401 22:03:50.969707 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.481481 | |
| I0401 22:03:50.969718 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 22:03:50.969730 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.722222 | |
| I0401 22:03:50.969744 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.24372 (* 0.3 = 0.673116 loss) | |
| I0401 22:03:50.969758 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.712573 (* 0.3 = 0.213772 loss) | |
| I0401 22:03:50.969771 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.62963 | |
| I0401 22:03:50.969784 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0401 22:03:50.969795 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.796296 | |
| I0401 22:03:50.969810 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.44815 (* 1 = 1.44815 loss) | |
| I0401 22:03:50.969825 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.494227 (* 1 = 0.494227 loss) | |
| I0401 22:03:50.969836 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 22:03:50.969848 6134 solver.cpp:245] Train net output #16: total_confidence = 0.335803 | |
| I0401 22:03:50.969861 6134 sgd_solver.cpp:106] Iteration 127500, lr = 0.01 | |
| I0401 22:05:59.843924 6134 solver.cpp:229] Iteration 128000, loss = 2.93912 | |
| I0401 22:05:59.844084 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.469388 | |
| I0401 22:05:59.844125 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0401 22:05:59.844148 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.795918 | |
| I0401 22:05:59.844177 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.77903 (* 0.3 = 0.533708 loss) | |
| I0401 22:05:59.844193 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.529492 (* 0.3 = 0.158848 loss) | |
| I0401 22:05:59.844215 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.44898 | |
| I0401 22:05:59.844228 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 22:05:59.844240 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.857143 | |
| I0401 22:05:59.844254 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.54712 (* 0.3 = 0.464136 loss) | |
| I0401 22:05:59.844274 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.441303 (* 0.3 = 0.132391 loss) | |
| I0401 22:05:59.844287 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.816327 | |
| I0401 22:05:59.844300 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0401 22:05:59.844311 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.938776 | |
| I0401 22:05:59.844326 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.483446 (* 1 = 0.483446 loss) | |
| I0401 22:05:59.844339 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.159461 (* 1 = 0.159461 loss) | |
| I0401 22:05:59.844352 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 22:05:59.844363 6134 solver.cpp:245] Train net output #16: total_confidence = 0.388001 | |
| I0401 22:05:59.844377 6134 sgd_solver.cpp:106] Iteration 128000, lr = 0.01 | |
| I0401 22:08:08.742363 6134 solver.cpp:229] Iteration 128500, loss = 2.91383 | |
| I0401 22:08:08.742667 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.574468 | |
| I0401 22:08:08.742688 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.869318 | |
| I0401 22:08:08.742702 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.851064 | |
| I0401 22:08:08.742718 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.36761 (* 0.3 = 0.410282 loss) | |
| I0401 22:08:08.742733 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.40311 (* 0.3 = 0.120933 loss) | |
| I0401 22:08:08.742746 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.659574 | |
| I0401 22:08:08.742758 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.903409 | |
| I0401 22:08:08.742771 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.893617 | |
| I0401 22:08:08.742785 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.20715 (* 0.3 = 0.362146 loss) | |
| I0401 22:08:08.742799 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.353282 (* 0.3 = 0.105985 loss) | |
| I0401 22:08:08.742811 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.851064 | |
| I0401 22:08:08.742823 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0401 22:08:08.742835 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.978723 | |
| I0401 22:08:08.742849 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.433975 (* 1 = 0.433975 loss) | |
| I0401 22:08:08.742863 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.157422 (* 1 = 0.157422 loss) | |
| I0401 22:08:08.742876 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 22:08:08.742887 6134 solver.cpp:245] Train net output #16: total_confidence = 0.364401 | |
| I0401 22:08:08.742899 6134 sgd_solver.cpp:106] Iteration 128500, lr = 0.01 | |
| I0401 22:10:17.633812 6134 solver.cpp:229] Iteration 129000, loss = 2.92474 | |
| I0401 22:10:17.633956 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.418605 | |
| I0401 22:10:17.633976 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 22:10:17.633996 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.627907 | |
| I0401 22:10:17.634012 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.98648 (* 0.3 = 0.595944 loss) | |
| I0401 22:10:17.634028 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.599941 (* 0.3 = 0.179982 loss) | |
| I0401 22:10:17.634042 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.44186 | |
| I0401 22:10:17.634053 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 22:10:17.634065 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.72093 | |
| I0401 22:10:17.634088 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.68642 (* 0.3 = 0.505926 loss) | |
| I0401 22:10:17.634101 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.51331 (* 0.3 = 0.153993 loss) | |
| I0401 22:10:17.634114 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.674419 | |
| I0401 22:10:17.634125 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 22:10:17.634137 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.883721 | |
| I0401 22:10:17.634158 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.03135 (* 1 = 1.03135 loss) | |
| I0401 22:10:17.634172 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.328161 (* 1 = 0.328161 loss) | |
| I0401 22:10:17.634184 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 22:10:17.634196 6134 solver.cpp:245] Train net output #16: total_confidence = 0.271084 | |
| I0401 22:10:17.634208 6134 sgd_solver.cpp:106] Iteration 129000, lr = 0.01 | |
| I0401 22:12:26.721822 6134 solver.cpp:229] Iteration 129500, loss = 2.97487 | |
| I0401 22:12:26.721920 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.25 | |
| I0401 22:12:26.721947 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 22:12:26.721961 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.625 | |
| I0401 22:12:26.721976 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.44684 (* 0.3 = 0.734051 loss) | |
| I0401 22:12:26.721992 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.656174 (* 0.3 = 0.196852 loss) | |
| I0401 22:12:26.722012 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.45 | |
| I0401 22:12:26.722024 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 22:12:26.722036 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0401 22:12:26.722051 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.6101 (* 0.3 = 0.48303 loss) | |
| I0401 22:12:26.722065 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.44425 (* 0.3 = 0.133275 loss) | |
| I0401 22:12:26.722077 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.675 | |
| I0401 22:12:26.722090 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 22:12:26.722105 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.825 | |
| I0401 22:12:26.722120 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.22261 (* 1 = 1.22261 loss) | |
| I0401 22:12:26.722134 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.327257 (* 1 = 0.327257 loss) | |
| I0401 22:12:26.722146 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 22:12:26.722159 6134 solver.cpp:245] Train net output #16: total_confidence = 0.071549 | |
| I0401 22:12:26.722172 6134 sgd_solver.cpp:106] Iteration 129500, lr = 0.01 | |
| I0401 22:14:35.718173 6134 solver.cpp:338] Iteration 130000, Testing net (#0) | |
| I0401 22:15:05.463304 6134 solver.cpp:393] Test loss: 2.46971 | |
| I0401 22:15:05.463353 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.467163 | |
| I0401 22:15:05.463371 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.857412 | |
| I0401 22:15:05.463382 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.749701 | |
| I0401 22:15:05.463398 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.80158 (* 0.3 = 0.540474 loss) | |
| I0401 22:15:05.463413 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.48685 (* 0.3 = 0.146055 loss) | |
| I0401 22:15:05.463425 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.611016 | |
| I0401 22:15:05.463438 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.897184 | |
| I0401 22:15:05.463449 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.844666 | |
| I0401 22:15:05.463464 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.35707 (* 0.3 = 0.40712 loss) | |
| I0401 22:15:05.463477 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.360636 (* 0.3 = 0.108191 loss) | |
| I0401 22:15:05.463490 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.752538 | |
| I0401 22:15:05.463501 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.934319 | |
| I0401 22:15:05.463512 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.882423 | |
| I0401 22:15:05.463531 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.99856 (* 1 = 0.99856 loss) | |
| I0401 22:15:05.463544 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.269315 (* 1 = 0.269315 loss) | |
| I0401 22:15:05.463557 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.372 | |
| I0401 22:15:05.463568 6134 solver.cpp:406] Test net output #16: total_confidence = 0.321816 | |
| I0401 22:15:05.615486 6134 solver.cpp:229] Iteration 130000, loss = 2.94054 | |
| I0401 22:15:05.615528 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.238095 | |
| I0401 22:15:05.615546 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 22:15:05.615558 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.642857 | |
| I0401 22:15:05.615574 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.07449 (* 0.3 = 0.622347 loss) | |
| I0401 22:15:05.615589 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.601412 (* 0.3 = 0.180423 loss) | |
| I0401 22:15:05.615602 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.404762 | |
| I0401 22:15:05.615614 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0401 22:15:05.615627 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.690476 | |
| I0401 22:15:05.615640 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.81395 (* 0.3 = 0.544185 loss) | |
| I0401 22:15:05.615654 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.539172 (* 0.3 = 0.161751 loss) | |
| I0401 22:15:05.615666 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.619048 | |
| I0401 22:15:05.615679 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0401 22:15:05.615690 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.880952 | |
| I0401 22:15:05.615705 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.07549 (* 1 = 1.07549 loss) | |
| I0401 22:15:05.615718 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.297945 (* 1 = 0.297945 loss) | |
| I0401 22:15:05.615736 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 22:15:05.615747 6134 solver.cpp:245] Train net output #16: total_confidence = 0.179404 | |
| I0401 22:15:05.615761 6134 sgd_solver.cpp:106] Iteration 130000, lr = 0.01 | |
| I0401 22:17:14.634042 6134 solver.cpp:229] Iteration 130500, loss = 2.90193 | |
| I0401 22:17:14.634490 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.282609 | |
| I0401 22:17:14.634511 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 22:17:14.634527 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.456522 | |
| I0401 22:17:14.634546 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.58042 (* 0.3 = 0.774127 loss) | |
| I0401 22:17:14.634559 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.736022 (* 0.3 = 0.220806 loss) | |
| I0401 22:17:14.634572 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.456522 | |
| I0401 22:17:14.634585 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 22:17:14.634598 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.586957 | |
| I0401 22:17:14.634613 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.05143 (* 0.3 = 0.615428 loss) | |
| I0401 22:17:14.634626 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.588315 (* 0.3 = 0.176495 loss) | |
| I0401 22:17:14.634639 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.608696 | |
| I0401 22:17:14.634651 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 22:17:14.634663 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.76087 | |
| I0401 22:17:14.634677 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.49257 (* 1 = 1.49257 loss) | |
| I0401 22:17:14.634692 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.422018 (* 1 = 0.422018 loss) | |
| I0401 22:17:14.634704 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 22:17:14.634716 6134 solver.cpp:245] Train net output #16: total_confidence = 0.196065 | |
| I0401 22:17:14.634728 6134 sgd_solver.cpp:106] Iteration 130500, lr = 0.01 | |
| I0401 22:19:23.616466 6134 solver.cpp:229] Iteration 131000, loss = 2.87173 | |
| I0401 22:19:23.616581 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.466667 | |
| I0401 22:19:23.616601 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0401 22:19:23.616614 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.644444 | |
| I0401 22:19:23.616638 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.17605 (* 0.3 = 0.652817 loss) | |
| I0401 22:19:23.616653 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.626032 (* 0.3 = 0.187809 loss) | |
| I0401 22:19:23.616667 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488889 | |
| I0401 22:19:23.616678 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 22:19:23.616699 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0401 22:19:23.616714 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.47588 (* 0.3 = 0.442764 loss) | |
| I0401 22:19:23.616729 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.417709 (* 0.3 = 0.125313 loss) | |
| I0401 22:19:23.616740 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.733333 | |
| I0401 22:19:23.616752 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0401 22:19:23.616765 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.866667 | |
| I0401 22:19:23.616778 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.04792 (* 1 = 1.04792 loss) | |
| I0401 22:19:23.616792 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.286518 (* 1 = 0.286518 loss) | |
| I0401 22:19:23.616806 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 22:19:23.616817 6134 solver.cpp:245] Train net output #16: total_confidence = 0.350021 | |
| I0401 22:19:23.616829 6134 sgd_solver.cpp:106] Iteration 131000, lr = 0.01 | |
| I0401 22:21:32.737262 6134 solver.cpp:229] Iteration 131500, loss = 2.90371 | |
| I0401 22:21:32.737387 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.209302 | |
| I0401 22:21:32.737406 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 22:21:32.737421 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.465116 | |
| I0401 22:21:32.737445 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.68061 (* 0.3 = 0.804184 loss) | |
| I0401 22:21:32.737462 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.739145 (* 0.3 = 0.221744 loss) | |
| I0401 22:21:32.737474 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.348837 | |
| I0401 22:21:32.737486 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0401 22:21:32.737498 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.674419 | |
| I0401 22:21:32.737516 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.01556 (* 0.3 = 0.604669 loss) | |
| I0401 22:21:32.737530 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.56865 (* 0.3 = 0.170595 loss) | |
| I0401 22:21:32.737542 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.581395 | |
| I0401 22:21:32.737555 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0401 22:21:32.737576 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.790698 | |
| I0401 22:21:32.737589 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.39499 (* 1 = 1.39499 loss) | |
| I0401 22:21:32.737604 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.390546 (* 1 = 0.390546 loss) | |
| I0401 22:21:32.737617 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 22:21:32.737628 6134 solver.cpp:245] Train net output #16: total_confidence = 0.276227 | |
| I0401 22:21:32.737640 6134 sgd_solver.cpp:106] Iteration 131500, lr = 0.01 | |
| I0401 22:23:41.965656 6134 solver.cpp:229] Iteration 132000, loss = 2.87519 | |
| I0401 22:23:41.965760 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.291667 | |
| I0401 22:23:41.965780 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 22:23:41.965791 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.645833 | |
| I0401 22:23:41.965807 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.18576 (* 0.3 = 0.655727 loss) | |
| I0401 22:23:41.965823 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.65905 (* 0.3 = 0.197715 loss) | |
| I0401 22:23:41.965836 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.291667 | |
| I0401 22:23:41.965848 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.784091 | |
| I0401 22:23:41.965860 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.583333 | |
| I0401 22:23:41.965874 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.24383 (* 0.3 = 0.673149 loss) | |
| I0401 22:23:41.965888 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.710943 (* 0.3 = 0.213283 loss) | |
| I0401 22:23:41.965900 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.625 | |
| I0401 22:23:41.965914 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0401 22:23:41.965924 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.75 | |
| I0401 22:23:41.965939 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.29871 (* 1 = 1.29871 loss) | |
| I0401 22:23:41.965953 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.365336 (* 1 = 0.365336 loss) | |
| I0401 22:23:41.965965 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 22:23:41.965977 6134 solver.cpp:245] Train net output #16: total_confidence = 0.240226 | |
| I0401 22:23:41.965989 6134 sgd_solver.cpp:106] Iteration 132000, lr = 0.01 | |
| I0401 22:25:51.152062 6134 solver.cpp:229] Iteration 132500, loss = 2.86724 | |
| I0401 22:25:51.152211 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.466667 | |
| I0401 22:25:51.152231 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0401 22:25:51.152252 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.8 | |
| I0401 22:25:51.152268 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.77735 (* 0.3 = 0.533206 loss) | |
| I0401 22:25:51.152283 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.514314 (* 0.3 = 0.154294 loss) | |
| I0401 22:25:51.152297 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.622222 | |
| I0401 22:25:51.152308 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0401 22:25:51.152328 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.933333 | |
| I0401 22:25:51.152343 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.15035 (* 0.3 = 0.345105 loss) | |
| I0401 22:25:51.152357 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.349946 (* 0.3 = 0.104984 loss) | |
| I0401 22:25:51.152369 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.844444 | |
| I0401 22:25:51.152390 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0401 22:25:51.152401 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.977778 | |
| I0401 22:25:51.152415 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.455901 (* 1 = 0.455901 loss) | |
| I0401 22:25:51.152429 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.136885 (* 1 = 0.136885 loss) | |
| I0401 22:25:51.152441 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 22:25:51.152453 6134 solver.cpp:245] Train net output #16: total_confidence = 0.201208 | |
| I0401 22:25:51.152465 6134 sgd_solver.cpp:106] Iteration 132500, lr = 0.01 | |
| I0401 22:28:00.395478 6134 solver.cpp:229] Iteration 133000, loss = 2.94169 | |
| I0401 22:28:00.395778 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.291667 | |
| I0401 22:28:00.395798 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 22:28:00.395812 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.541667 | |
| I0401 22:28:00.395828 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.92246 (* 0.3 = 0.876737 loss) | |
| I0401 22:28:00.395843 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.821809 (* 0.3 = 0.246543 loss) | |
| I0401 22:28:00.395856 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.416667 | |
| I0401 22:28:00.395869 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 22:28:00.395880 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.583333 | |
| I0401 22:28:00.395895 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.59991 (* 0.3 = 0.779974 loss) | |
| I0401 22:28:00.395910 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.730231 (* 0.3 = 0.219069 loss) | |
| I0401 22:28:00.395922 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.6875 | |
| I0401 22:28:00.395934 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0401 22:28:00.395946 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.791667 | |
| I0401 22:28:00.395961 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.00298 (* 1 = 2.00298 loss) | |
| I0401 22:28:00.395974 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.568301 (* 1 = 0.568301 loss) | |
| I0401 22:28:00.395987 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 22:28:00.395999 6134 solver.cpp:245] Train net output #16: total_confidence = 0.359794 | |
| I0401 22:28:00.396011 6134 sgd_solver.cpp:106] Iteration 133000, lr = 0.01 | |
| I0401 22:30:09.525424 6134 solver.cpp:229] Iteration 133500, loss = 2.92398 | |
| I0401 22:30:09.525574 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.266667 | |
| I0401 22:30:09.525604 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0401 22:30:09.525616 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0401 22:30:09.525631 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.01809 (* 0.3 = 0.605426 loss) | |
| I0401 22:30:09.525647 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.613436 (* 0.3 = 0.184031 loss) | |
| I0401 22:30:09.525660 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.577778 | |
| I0401 22:30:09.525681 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 22:30:09.525692 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.777778 | |
| I0401 22:30:09.525707 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.48218 (* 0.3 = 0.444653 loss) | |
| I0401 22:30:09.525720 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.458624 (* 0.3 = 0.137587 loss) | |
| I0401 22:30:09.525741 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.755556 | |
| I0401 22:30:09.525753 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0401 22:30:09.525765 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.866667 | |
| I0401 22:30:09.525779 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.889386 (* 1 = 0.889386 loss) | |
| I0401 22:30:09.525794 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.27019 (* 1 = 0.27019 loss) | |
| I0401 22:30:09.525806 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 22:30:09.525818 6134 solver.cpp:245] Train net output #16: total_confidence = 0.231932 | |
| I0401 22:30:09.525830 6134 sgd_solver.cpp:106] Iteration 133500, lr = 0.01 | |
| I0401 22:32:18.585121 6134 solver.cpp:229] Iteration 134000, loss = 2.87388 | |
| I0401 22:32:18.585233 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.227273 | |
| I0401 22:32:18.585253 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 22:32:18.585265 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.659091 | |
| I0401 22:32:18.585283 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.20166 (* 0.3 = 0.660499 loss) | |
| I0401 22:32:18.585296 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.634832 (* 0.3 = 0.19045 loss) | |
| I0401 22:32:18.585309 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.363636 | |
| I0401 22:32:18.585322 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 22:32:18.585335 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.772727 | |
| I0401 22:32:18.585348 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.67539 (* 0.3 = 0.502618 loss) | |
| I0401 22:32:18.585363 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.527066 (* 0.3 = 0.15812 loss) | |
| I0401 22:32:18.585376 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.659091 | |
| I0401 22:32:18.585388 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0401 22:32:18.585400 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.909091 | |
| I0401 22:32:18.585422 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.00757 (* 1 = 1.00757 loss) | |
| I0401 22:32:18.585435 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.375373 (* 1 = 0.375373 loss) | |
| I0401 22:32:18.585448 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 22:32:18.585460 6134 solver.cpp:245] Train net output #16: total_confidence = 0.22199 | |
| I0401 22:32:18.585480 6134 sgd_solver.cpp:106] Iteration 134000, lr = 0.01 | |
| I0401 22:34:27.687016 6134 solver.cpp:229] Iteration 134500, loss = 2.84337 | |
| I0401 22:34:27.687139 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.326531 | |
| I0401 22:34:27.687168 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0401 22:34:27.687192 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.612245 | |
| I0401 22:34:27.687209 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.20849 (* 0.3 = 0.662546 loss) | |
| I0401 22:34:27.687224 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.655565 (* 0.3 = 0.19667 loss) | |
| I0401 22:34:27.687237 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.469388 | |
| I0401 22:34:27.687250 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 22:34:27.687263 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.673469 | |
| I0401 22:34:27.687276 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.86069 (* 0.3 = 0.558207 loss) | |
| I0401 22:34:27.687291 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.557464 (* 0.3 = 0.167239 loss) | |
| I0401 22:34:27.687304 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.714286 | |
| I0401 22:34:27.687315 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0401 22:34:27.687327 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.857143 | |
| I0401 22:34:27.687342 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.03522 (* 1 = 1.03522 loss) | |
| I0401 22:34:27.687356 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.304922 (* 1 = 0.304922 loss) | |
| I0401 22:34:27.687369 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 22:34:27.687381 6134 solver.cpp:245] Train net output #16: total_confidence = 0.221719 | |
| I0401 22:34:27.687393 6134 sgd_solver.cpp:106] Iteration 134500, lr = 0.01 | |
| I0401 22:36:36.643805 6134 solver.cpp:338] Iteration 135000, Testing net (#0) | |
| I0401 22:37:06.470626 6134 solver.cpp:393] Test loss: 2.38915 | |
| I0401 22:37:06.470672 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.450222 | |
| I0401 22:37:06.470690 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.860003 | |
| I0401 22:37:06.470701 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.751273 | |
| I0401 22:37:06.470717 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.80765 (* 0.3 = 0.542295 loss) | |
| I0401 22:37:06.470731 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.47066 (* 0.3 = 0.141198 loss) | |
| I0401 22:37:06.470743 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.640893 | |
| I0401 22:37:06.470757 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.901639 | |
| I0401 22:37:06.470768 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.86025 | |
| I0401 22:37:06.470782 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.29257 (* 0.3 = 0.387771 loss) | |
| I0401 22:37:06.470795 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.357236 (* 0.3 = 0.107171 loss) | |
| I0401 22:37:06.470808 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.7605 | |
| I0401 22:37:06.470818 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.939137 | |
| I0401 22:37:06.470830 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.88615 | |
| I0401 22:37:06.470844 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.958686 (* 1 = 0.958686 loss) | |
| I0401 22:37:06.470857 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.252026 (* 1 = 0.252026 loss) | |
| I0401 22:37:06.470868 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.424 | |
| I0401 22:37:06.470880 6134 solver.cpp:406] Test net output #16: total_confidence = 0.408773 | |
| I0401 22:37:06.622740 6134 solver.cpp:229] Iteration 135000, loss = 2.82365 | |
| I0401 22:37:06.622788 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.44186 | |
| I0401 22:37:06.622805 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0401 22:37:06.622819 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.511628 | |
| I0401 22:37:06.622835 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.21766 (* 0.3 = 0.665297 loss) | |
| I0401 22:37:06.622850 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.623903 (* 0.3 = 0.187171 loss) | |
| I0401 22:37:06.622862 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.511628 | |
| I0401 22:37:06.622875 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0401 22:37:06.622886 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.744186 | |
| I0401 22:37:06.622900 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.73372 (* 0.3 = 0.520115 loss) | |
| I0401 22:37:06.622915 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.510135 (* 0.3 = 0.15304 loss) | |
| I0401 22:37:06.622927 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.697674 | |
| I0401 22:37:06.622939 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0401 22:37:06.622951 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.790698 | |
| I0401 22:37:06.622964 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.37813 (* 1 = 1.37813 loss) | |
| I0401 22:37:06.622978 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.357917 (* 1 = 0.357917 loss) | |
| I0401 22:37:06.622990 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 22:37:06.623003 6134 solver.cpp:245] Train net output #16: total_confidence = 0.32322 | |
| I0401 22:37:06.623014 6134 sgd_solver.cpp:106] Iteration 135000, lr = 0.01 | |
| I0401 22:39:15.658776 6134 solver.cpp:229] Iteration 135500, loss = 2.83606 | |
| I0401 22:39:15.658912 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.452381 | |
| I0401 22:39:15.658932 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0401 22:39:15.658946 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.619048 | |
| I0401 22:39:15.658969 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.04119 (* 0.3 = 0.612358 loss) | |
| I0401 22:39:15.658984 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.572231 (* 0.3 = 0.171669 loss) | |
| I0401 22:39:15.658998 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 22:39:15.659010 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0401 22:39:15.659023 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.833333 | |
| I0401 22:39:15.659036 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.4548 (* 0.3 = 0.436439 loss) | |
| I0401 22:39:15.659057 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.422773 (* 0.3 = 0.126832 loss) | |
| I0401 22:39:15.659070 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.880952 | |
| I0401 22:39:15.659081 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0401 22:39:15.659093 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.952381 | |
| I0401 22:39:15.659116 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.441938 (* 1 = 0.441938 loss) | |
| I0401 22:39:15.659129 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.146318 (* 1 = 0.146318 loss) | |
| I0401 22:39:15.659142 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 22:39:15.659154 6134 solver.cpp:245] Train net output #16: total_confidence = 0.304105 | |
| I0401 22:39:15.659167 6134 sgd_solver.cpp:106] Iteration 135500, lr = 0.01 | |
| I0401 22:41:24.846462 6134 solver.cpp:229] Iteration 136000, loss = 2.84893 | |
| I0401 22:41:24.846628 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.473684 | |
| I0401 22:41:24.846668 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.863636 | |
| I0401 22:41:24.846695 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.631579 | |
| I0401 22:41:24.846719 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.17303 (* 0.3 = 0.651908 loss) | |
| I0401 22:41:24.846745 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.550999 (* 0.3 = 0.1653 loss) | |
| I0401 22:41:24.846757 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.552632 | |
| I0401 22:41:24.846771 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0401 22:41:24.846782 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.789474 | |
| I0401 22:41:24.846804 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.73075 (* 0.3 = 0.519226 loss) | |
| I0401 22:41:24.846818 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.430891 (* 0.3 = 0.129267 loss) | |
| I0401 22:41:24.846832 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.605263 | |
| I0401 22:41:24.846843 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 22:41:24.846856 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.894737 | |
| I0401 22:41:24.846870 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.20571 (* 1 = 1.20571 loss) | |
| I0401 22:41:24.846884 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.302141 (* 1 = 0.302141 loss) | |
| I0401 22:41:24.846896 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 22:41:24.846909 6134 solver.cpp:245] Train net output #16: total_confidence = 0.188635 | |
| I0401 22:41:24.846921 6134 sgd_solver.cpp:106] Iteration 136000, lr = 0.01 | |
| I0401 22:43:34.329324 6134 solver.cpp:229] Iteration 136500, loss = 2.92602 | |
| I0401 22:43:34.329457 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.390244 | |
| I0401 22:43:34.329478 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0401 22:43:34.329493 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.829268 | |
| I0401 22:43:34.329509 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.56058 (* 0.3 = 0.468174 loss) | |
| I0401 22:43:34.329526 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.445221 (* 0.3 = 0.133566 loss) | |
| I0401 22:43:34.329540 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.707317 | |
| I0401 22:43:34.329552 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.903409 | |
| I0401 22:43:34.329565 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.878049 | |
| I0401 22:43:34.329578 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.906265 (* 0.3 = 0.271879 loss) | |
| I0401 22:43:34.329593 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.285285 (* 0.3 = 0.0855855 loss) | |
| I0401 22:43:34.329607 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.878049 | |
| I0401 22:43:34.329619 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0401 22:43:34.329632 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.951219 | |
| I0401 22:43:34.329645 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.348445 (* 1 = 0.348445 loss) | |
| I0401 22:43:34.329660 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.115312 (* 1 = 0.115312 loss) | |
| I0401 22:43:34.329674 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0401 22:43:34.329685 6134 solver.cpp:245] Train net output #16: total_confidence = 0.442387 | |
| I0401 22:43:34.329699 6134 sgd_solver.cpp:106] Iteration 136500, lr = 0.01 | |
| I0401 22:45:43.442328 6134 solver.cpp:229] Iteration 137000, loss = 2.88441 | |
| I0401 22:45:43.442456 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.488372 | |
| I0401 22:45:43.442476 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0401 22:45:43.442489 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.72093 | |
| I0401 22:45:43.442513 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.7862 (* 0.3 = 0.535861 loss) | |
| I0401 22:45:43.442528 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.506368 (* 0.3 = 0.15191 loss) | |
| I0401 22:45:43.442540 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.55814 | |
| I0401 22:45:43.442553 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0401 22:45:43.442566 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.906977 | |
| I0401 22:45:43.442580 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.37141 (* 0.3 = 0.411422 loss) | |
| I0401 22:45:43.442595 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.399851 (* 0.3 = 0.119955 loss) | |
| I0401 22:45:43.442607 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.906977 | |
| I0401 22:45:43.442627 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.971591 | |
| I0401 22:45:43.442639 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0401 22:45:43.442653 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.248194 (* 1 = 0.248194 loss) | |
| I0401 22:45:43.442668 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0745973 (* 1 = 0.0745973 loss) | |
| I0401 22:45:43.442689 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0401 22:45:43.442701 6134 solver.cpp:245] Train net output #16: total_confidence = 0.373763 | |
| I0401 22:45:43.442713 6134 sgd_solver.cpp:106] Iteration 137000, lr = 0.01 | |
| I0401 22:47:53.423318 6134 solver.cpp:229] Iteration 137500, loss = 2.8324 | |
| I0401 22:47:53.423575 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0401 22:47:53.423595 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 22:47:53.423609 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.56 | |
| I0401 22:47:53.423624 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.12947 (* 0.3 = 0.63884 loss) | |
| I0401 22:47:53.423640 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.63775 (* 0.3 = 0.191325 loss) | |
| I0401 22:47:53.423652 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.52 | |
| I0401 22:47:53.423666 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 22:47:53.423678 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.7 | |
| I0401 22:47:53.423692 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.69303 (* 0.3 = 0.50791 loss) | |
| I0401 22:47:53.423707 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.518181 (* 0.3 = 0.155454 loss) | |
| I0401 22:47:53.423718 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.78 | |
| I0401 22:47:53.423730 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 22:47:53.423743 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.92 | |
| I0401 22:47:53.423758 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.859546 (* 1 = 0.859546 loss) | |
| I0401 22:47:53.423771 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.290759 (* 1 = 0.290759 loss) | |
| I0401 22:47:53.423784 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 22:47:53.423796 6134 solver.cpp:245] Train net output #16: total_confidence = 0.219871 | |
| I0401 22:47:53.423809 6134 sgd_solver.cpp:106] Iteration 137500, lr = 0.01 | |
| I0401 22:50:02.621531 6134 solver.cpp:229] Iteration 138000, loss = 2.89598 | |
| I0401 22:50:02.621692 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.5 | |
| I0401 22:50:02.621722 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0401 22:50:02.621757 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.681818 | |
| I0401 22:50:02.621783 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.96605 (* 0.3 = 0.589816 loss) | |
| I0401 22:50:02.621800 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.612812 (* 0.3 = 0.183844 loss) | |
| I0401 22:50:02.621819 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.545455 | |
| I0401 22:50:02.621834 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 22:50:02.621845 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0401 22:50:02.621860 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.6399 (* 0.3 = 0.49197 loss) | |
| I0401 22:50:02.621883 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.499884 (* 0.3 = 0.149965 loss) | |
| I0401 22:50:02.621896 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.704545 | |
| I0401 22:50:02.621908 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 22:50:02.621920 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.863636 | |
| I0401 22:50:02.621934 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.93831 (* 1 = 0.93831 loss) | |
| I0401 22:50:02.621948 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.316685 (* 1 = 0.316685 loss) | |
| I0401 22:50:02.621960 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 22:50:02.621973 6134 solver.cpp:245] Train net output #16: total_confidence = 0.319394 | |
| I0401 22:50:02.621985 6134 sgd_solver.cpp:106] Iteration 138000, lr = 0.01 | |
| I0401 22:52:11.876436 6134 solver.cpp:229] Iteration 138500, loss = 2.87677 | |
| I0401 22:52:11.876556 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.414634 | |
| I0401 22:52:11.876577 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0401 22:52:11.876590 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.634146 | |
| I0401 22:52:11.876606 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.3023 (* 0.3 = 0.690689 loss) | |
| I0401 22:52:11.876622 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.61938 (* 0.3 = 0.185814 loss) | |
| I0401 22:52:11.876636 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.585366 | |
| I0401 22:52:11.876647 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0401 22:52:11.876659 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.780488 | |
| I0401 22:52:11.876673 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.7443 (* 0.3 = 0.52329 loss) | |
| I0401 22:52:11.876688 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.507001 (* 0.3 = 0.1521 loss) | |
| I0401 22:52:11.876700 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.804878 | |
| I0401 22:52:11.876713 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0401 22:52:11.876724 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.878049 | |
| I0401 22:52:11.876739 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.926091 (* 1 = 0.926091 loss) | |
| I0401 22:52:11.876754 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.236218 (* 1 = 0.236218 loss) | |
| I0401 22:52:11.876766 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 22:52:11.876778 6134 solver.cpp:245] Train net output #16: total_confidence = 0.306052 | |
| I0401 22:52:11.876791 6134 sgd_solver.cpp:106] Iteration 138500, lr = 0.01 | |
| I0401 22:54:21.075995 6134 solver.cpp:229] Iteration 139000, loss = 2.92673 | |
| I0401 22:54:21.076149 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.447368 | |
| I0401 22:54:21.076170 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.863636 | |
| I0401 22:54:21.076190 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.684211 | |
| I0401 22:54:21.076207 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.84192 (* 0.3 = 0.552577 loss) | |
| I0401 22:54:21.076222 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.507005 (* 0.3 = 0.152101 loss) | |
| I0401 22:54:21.076236 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.578947 | |
| I0401 22:54:21.076256 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0401 22:54:21.076267 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.815789 | |
| I0401 22:54:21.076280 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.34027 (* 0.3 = 0.40208 loss) | |
| I0401 22:54:21.076295 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.437226 (* 0.3 = 0.131168 loss) | |
| I0401 22:54:21.076314 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.868421 | |
| I0401 22:54:21.076326 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0401 22:54:21.076339 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.973684 | |
| I0401 22:54:21.076352 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.409513 (* 1 = 0.409513 loss) | |
| I0401 22:54:21.076366 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.121597 (* 1 = 0.121597 loss) | |
| I0401 22:54:21.076378 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 22:54:21.076390 6134 solver.cpp:245] Train net output #16: total_confidence = 0.289752 | |
| I0401 22:54:21.076402 6134 sgd_solver.cpp:106] Iteration 139000, lr = 0.01 | |
| I0401 22:56:30.044622 6134 solver.cpp:229] Iteration 139500, loss = 2.87404 | |
| I0401 22:56:30.044886 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.444444 | |
| I0401 22:56:30.044908 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0401 22:56:30.044920 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0401 22:56:30.044939 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.96516 (* 0.3 = 0.589548 loss) | |
| I0401 22:56:30.044975 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.559665 (* 0.3 = 0.167899 loss) | |
| I0401 22:56:30.044993 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.644444 | |
| I0401 22:56:30.045016 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0401 22:56:30.045027 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.822222 | |
| I0401 22:56:30.045055 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.2354 (* 0.3 = 0.370621 loss) | |
| I0401 22:56:30.045078 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.365959 (* 0.3 = 0.109788 loss) | |
| I0401 22:56:30.045090 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.822222 | |
| I0401 22:56:30.045102 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0401 22:56:30.045115 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0401 22:56:30.045130 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.432717 (* 1 = 0.432717 loss) | |
| I0401 22:56:30.045143 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.130789 (* 1 = 0.130789 loss) | |
| I0401 22:56:30.045156 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 22:56:30.045171 6134 solver.cpp:245] Train net output #16: total_confidence = 0.235513 | |
| I0401 22:56:30.045183 6134 sgd_solver.cpp:106] Iteration 139500, lr = 0.01 | |
| I0401 22:58:39.177711 6134 solver.cpp:338] Iteration 140000, Testing net (#0) | |
| I0401 22:59:09.082482 6134 solver.cpp:393] Test loss: 2.50581 | |
| I0401 22:59:09.082530 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.451338 | |
| I0401 22:59:09.082548 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.842093 | |
| I0401 22:59:09.082561 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.735262 | |
| I0401 22:59:09.082578 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.82402 (* 0.3 = 0.547207 loss) | |
| I0401 22:59:09.082593 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.533338 (* 0.3 = 0.160001 loss) | |
| I0401 22:59:09.082605 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.641451 | |
| I0401 22:59:09.082618 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.898229 | |
| I0401 22:59:09.082629 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.843793 | |
| I0401 22:59:09.082643 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.32488 (* 0.3 = 0.397464 loss) | |
| I0401 22:59:09.082658 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.378104 (* 0.3 = 0.113431 loss) | |
| I0401 22:59:09.082669 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.749237 | |
| I0401 22:59:09.082681 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.929091 | |
| I0401 22:59:09.082693 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.874111 | |
| I0401 22:59:09.082707 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 1.00193 (* 1 = 1.00193 loss) | |
| I0401 22:59:09.082721 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.285772 (* 1 = 0.285772 loss) | |
| I0401 22:59:09.082733 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.333 | |
| I0401 22:59:09.082746 6134 solver.cpp:406] Test net output #16: total_confidence = 0.313801 | |
| I0401 22:59:09.234396 6134 solver.cpp:229] Iteration 140000, loss = 2.87741 | |
| I0401 22:59:09.234488 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.327273 | |
| I0401 22:59:09.234513 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0401 22:59:09.234526 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.654545 | |
| I0401 22:59:09.234541 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.04697 (* 0.3 = 0.614091 loss) | |
| I0401 22:59:09.234556 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.695616 (* 0.3 = 0.208685 loss) | |
| I0401 22:59:09.234568 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.418182 | |
| I0401 22:59:09.234580 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0401 22:59:09.234592 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.690909 | |
| I0401 22:59:09.234606 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.77131 (* 0.3 = 0.531393 loss) | |
| I0401 22:59:09.234619 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.590841 (* 0.3 = 0.177252 loss) | |
| I0401 22:59:09.234635 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.672727 | |
| I0401 22:59:09.234647 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0401 22:59:09.234659 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.818182 | |
| I0401 22:59:09.234673 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.04543 (* 1 = 1.04543 loss) | |
| I0401 22:59:09.234688 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.365933 (* 1 = 0.365933 loss) | |
| I0401 22:59:09.234699 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 22:59:09.234711 6134 solver.cpp:245] Train net output #16: total_confidence = 0.217653 | |
| I0401 22:59:09.234724 6134 sgd_solver.cpp:106] Iteration 140000, lr = 0.01 | |
| I0401 23:01:18.345346 6134 solver.cpp:229] Iteration 140500, loss = 2.89481 | |
| I0401 23:01:18.345489 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.395349 | |
| I0401 23:01:18.345510 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 23:01:18.345527 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.674419 | |
| I0401 23:01:18.345544 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.10554 (* 0.3 = 0.631663 loss) | |
| I0401 23:01:18.345559 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.6298 (* 0.3 = 0.18894 loss) | |
| I0401 23:01:18.345572 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.534884 | |
| I0401 23:01:18.345585 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 23:01:18.345597 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.790698 | |
| I0401 23:01:18.345612 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.77278 (* 0.3 = 0.531834 loss) | |
| I0401 23:01:18.345625 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.551246 (* 0.3 = 0.165374 loss) | |
| I0401 23:01:18.345638 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.72093 | |
| I0401 23:01:18.345650 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 23:01:18.345661 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.883721 | |
| I0401 23:01:18.345676 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.03216 (* 1 = 1.03216 loss) | |
| I0401 23:01:18.345690 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.263117 (* 1 = 0.263117 loss) | |
| I0401 23:01:18.345703 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 23:01:18.345715 6134 solver.cpp:245] Train net output #16: total_confidence = 0.339467 | |
| I0401 23:01:18.345727 6134 sgd_solver.cpp:106] Iteration 140500, lr = 0.01 | |
| I0401 23:03:27.306005 6134 solver.cpp:229] Iteration 141000, loss = 2.84174 | |
| I0401 23:03:27.306110 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.361111 | |
| I0401 23:03:27.306129 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 23:03:27.306143 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.611111 | |
| I0401 23:03:27.306159 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.05446 (* 0.3 = 0.616338 loss) | |
| I0401 23:03:27.306174 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.690136 (* 0.3 = 0.207041 loss) | |
| I0401 23:03:27.306186 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.527778 | |
| I0401 23:03:27.306198 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0401 23:03:27.306210 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0401 23:03:27.306226 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.55242 (* 0.3 = 0.465727 loss) | |
| I0401 23:03:27.306241 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.489686 (* 0.3 = 0.146906 loss) | |
| I0401 23:03:27.306252 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.694444 | |
| I0401 23:03:27.306264 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0401 23:03:27.306277 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.777778 | |
| I0401 23:03:27.306290 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.08374 (* 1 = 1.08374 loss) | |
| I0401 23:03:27.306304 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.258721 (* 1 = 0.258721 loss) | |
| I0401 23:03:27.306316 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 23:03:27.306329 6134 solver.cpp:245] Train net output #16: total_confidence = 0.303887 | |
| I0401 23:03:27.306340 6134 sgd_solver.cpp:106] Iteration 141000, lr = 0.01 | |
| I0401 23:05:36.417520 6134 solver.cpp:229] Iteration 141500, loss = 2.83691 | |
| I0401 23:05:36.417665 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.392157 | |
| I0401 23:05:36.417697 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 23:05:36.417711 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54902 | |
| I0401 23:05:36.417727 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.25528 (* 0.3 = 0.676584 loss) | |
| I0401 23:05:36.417742 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.700296 (* 0.3 = 0.210089 loss) | |
| I0401 23:05:36.417754 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.470588 | |
| I0401 23:05:36.417768 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 23:05:36.417788 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.745098 | |
| I0401 23:05:36.417800 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.88755 (* 0.3 = 0.566264 loss) | |
| I0401 23:05:36.417815 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.603512 (* 0.3 = 0.181054 loss) | |
| I0401 23:05:36.417827 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0401 23:05:36.417847 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0401 23:05:36.417858 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.882353 | |
| I0401 23:05:36.417873 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.13192 (* 1 = 1.13192 loss) | |
| I0401 23:05:36.417887 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.3668 (* 1 = 0.3668 loss) | |
| I0401 23:05:36.417899 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 23:05:36.417912 6134 solver.cpp:245] Train net output #16: total_confidence = 0.155313 | |
| I0401 23:05:36.417923 6134 sgd_solver.cpp:106] Iteration 141500, lr = 0.01 | |
| I0401 23:07:45.431453 6134 solver.cpp:229] Iteration 142000, loss = 2.73288 | |
| I0401 23:07:45.431754 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.340426 | |
| I0401 23:07:45.431773 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0401 23:07:45.431787 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.510638 | |
| I0401 23:07:45.431803 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.33376 (* 0.3 = 0.700128 loss) | |
| I0401 23:07:45.431818 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.702919 (* 0.3 = 0.210876 loss) | |
| I0401 23:07:45.431831 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.489362 | |
| I0401 23:07:45.431843 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 23:07:45.431855 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.723404 | |
| I0401 23:07:45.431869 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.79781 (* 0.3 = 0.539343 loss) | |
| I0401 23:07:45.431885 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.557073 (* 0.3 = 0.167122 loss) | |
| I0401 23:07:45.431896 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.680851 | |
| I0401 23:07:45.431908 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0401 23:07:45.431921 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.787234 | |
| I0401 23:07:45.431934 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.23065 (* 1 = 1.23065 loss) | |
| I0401 23:07:45.431949 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.357192 (* 1 = 0.357192 loss) | |
| I0401 23:07:45.431962 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 23:07:45.431973 6134 solver.cpp:245] Train net output #16: total_confidence = 0.253087 | |
| I0401 23:07:45.431987 6134 sgd_solver.cpp:106] Iteration 142000, lr = 0.01 | |
| I0401 23:09:54.411468 6134 solver.cpp:229] Iteration 142500, loss = 2.80046 | |
| I0401 23:09:54.411623 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.365385 | |
| I0401 23:09:54.411650 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 23:09:54.411664 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.596154 | |
| I0401 23:09:54.411680 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.13391 (* 0.3 = 0.640172 loss) | |
| I0401 23:09:54.411695 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.645705 (* 0.3 = 0.193711 loss) | |
| I0401 23:09:54.411707 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.480769 | |
| I0401 23:09:54.411720 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 23:09:54.411732 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0401 23:09:54.411749 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.76004 (* 0.3 = 0.528011 loss) | |
| I0401 23:09:54.411763 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.541301 (* 0.3 = 0.16239 loss) | |
| I0401 23:09:54.411775 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.673077 | |
| I0401 23:09:54.411787 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0401 23:09:54.411808 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.826923 | |
| I0401 23:09:54.411823 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.10442 (* 1 = 1.10442 loss) | |
| I0401 23:09:54.411837 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.35224 (* 1 = 0.35224 loss) | |
| I0401 23:09:54.411849 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 23:09:54.411861 6134 solver.cpp:245] Train net output #16: total_confidence = 0.267956 | |
| I0401 23:09:54.411873 6134 sgd_solver.cpp:106] Iteration 142500, lr = 0.01 | |
| I0401 23:12:03.336230 6134 solver.cpp:229] Iteration 143000, loss = 2.85736 | |
| I0401 23:12:03.336386 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.348837 | |
| I0401 23:12:03.336408 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 23:12:03.336421 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.55814 | |
| I0401 23:12:03.336438 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.25544 (* 0.3 = 0.676632 loss) | |
| I0401 23:12:03.336453 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.651362 (* 0.3 = 0.195408 loss) | |
| I0401 23:12:03.336467 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488372 | |
| I0401 23:12:03.336479 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 23:12:03.336491 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.813953 | |
| I0401 23:12:03.336506 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.59884 (* 0.3 = 0.479653 loss) | |
| I0401 23:12:03.336524 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.525687 (* 0.3 = 0.157706 loss) | |
| I0401 23:12:03.336536 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.697674 | |
| I0401 23:12:03.336549 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 23:12:03.336561 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.860465 | |
| I0401 23:12:03.336575 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.947207 (* 1 = 0.947207 loss) | |
| I0401 23:12:03.336590 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.262341 (* 1 = 0.262341 loss) | |
| I0401 23:12:03.336602 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 23:12:03.336614 6134 solver.cpp:245] Train net output #16: total_confidence = 0.424609 | |
| I0401 23:12:03.336627 6134 sgd_solver.cpp:106] Iteration 143000, lr = 0.01 | |
| I0401 23:14:12.106567 6134 solver.cpp:229] Iteration 143500, loss = 2.7921 | |
| I0401 23:14:12.106703 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.479167 | |
| I0401 23:14:12.106722 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0401 23:14:12.106735 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0401 23:14:12.106751 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.12242 (* 0.3 = 0.636727 loss) | |
| I0401 23:14:12.106766 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.669855 (* 0.3 = 0.200956 loss) | |
| I0401 23:14:12.106778 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.541667 | |
| I0401 23:14:12.106791 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 23:14:12.106803 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.791667 | |
| I0401 23:14:12.106817 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.65537 (* 0.3 = 0.496611 loss) | |
| I0401 23:14:12.106832 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.526266 (* 0.3 = 0.15788 loss) | |
| I0401 23:14:12.106844 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.729167 | |
| I0401 23:14:12.106856 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0401 23:14:12.106868 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.895833 | |
| I0401 23:14:12.106883 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.842935 (* 1 = 0.842935 loss) | |
| I0401 23:14:12.106896 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.256486 (* 1 = 0.256486 loss) | |
| I0401 23:14:12.106909 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 23:14:12.106920 6134 solver.cpp:245] Train net output #16: total_confidence = 0.255138 | |
| I0401 23:14:12.106932 6134 sgd_solver.cpp:106] Iteration 143500, lr = 0.01 | |
| I0401 23:16:20.773715 6134 solver.cpp:229] Iteration 144000, loss = 2.82095 | |
| I0401 23:16:20.773844 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.395833 | |
| I0401 23:16:20.773864 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0401 23:16:20.773877 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0401 23:16:20.773895 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.07584 (* 0.3 = 0.622753 loss) | |
| I0401 23:16:20.773910 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.579638 (* 0.3 = 0.173891 loss) | |
| I0401 23:16:20.773921 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.645833 | |
| I0401 23:16:20.773934 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.903409 | |
| I0401 23:16:20.773947 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8125 | |
| I0401 23:16:20.773960 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.31893 (* 0.3 = 0.39568 loss) | |
| I0401 23:16:20.773975 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.373008 (* 0.3 = 0.111902 loss) | |
| I0401 23:16:20.773988 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.833333 | |
| I0401 23:16:20.773999 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0401 23:16:20.774011 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9375 | |
| I0401 23:16:20.774025 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.552025 (* 1 = 0.552025 loss) | |
| I0401 23:16:20.774039 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.155365 (* 1 = 0.155365 loss) | |
| I0401 23:16:20.774051 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 23:16:20.774065 6134 solver.cpp:245] Train net output #16: total_confidence = 0.326973 | |
| I0401 23:16:20.774077 6134 sgd_solver.cpp:106] Iteration 144000, lr = 0.01 | |
| I0401 23:18:29.466743 6134 solver.cpp:229] Iteration 144500, loss = 2.82515 | |
| I0401 23:18:29.467070 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.5 | |
| I0401 23:18:29.467092 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.863636 | |
| I0401 23:18:29.467105 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.659091 | |
| I0401 23:18:29.467123 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.72865 (* 0.3 = 0.518595 loss) | |
| I0401 23:18:29.467138 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.513828 (* 0.3 = 0.154149 loss) | |
| I0401 23:18:29.467150 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.545455 | |
| I0401 23:18:29.467164 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 23:18:29.467176 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.840909 | |
| I0401 23:18:29.467190 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.32778 (* 0.3 = 0.398335 loss) | |
| I0401 23:18:29.467205 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.39602 (* 0.3 = 0.118806 loss) | |
| I0401 23:18:29.467217 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.772727 | |
| I0401 23:18:29.467229 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0401 23:18:29.467242 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.909091 | |
| I0401 23:18:29.467255 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.01896 (* 1 = 1.01896 loss) | |
| I0401 23:18:29.467269 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.280863 (* 1 = 0.280863 loss) | |
| I0401 23:18:29.467283 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 23:18:29.467294 6134 solver.cpp:245] Train net output #16: total_confidence = 0.378135 | |
| I0401 23:18:29.467306 6134 sgd_solver.cpp:106] Iteration 144500, lr = 0.01 | |
| I0401 23:20:37.987339 6134 solver.cpp:338] Iteration 145000, Testing net (#0) | |
| I0401 23:21:07.728816 6134 solver.cpp:393] Test loss: 2.40444 | |
| I0401 23:21:07.728863 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.488808 | |
| I0401 23:21:07.728880 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.862094 | |
| I0401 23:21:07.728893 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.767153 | |
| I0401 23:21:07.728909 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.7142 (* 0.3 = 0.514259 loss) | |
| I0401 23:21:07.728924 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.468268 (* 0.3 = 0.14048 loss) | |
| I0401 23:21:07.728935 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.615201 | |
| I0401 23:21:07.728947 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.898594 | |
| I0401 23:21:07.728960 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.855929 | |
| I0401 23:21:07.728973 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.32451 (* 0.3 = 0.397353 loss) | |
| I0401 23:21:07.728988 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.35365 (* 0.3 = 0.106095 loss) | |
| I0401 23:21:07.729001 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.747602 | |
| I0401 23:21:07.729012 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.934137 | |
| I0401 23:21:07.729024 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.880776 | |
| I0401 23:21:07.729038 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.983366 (* 1 = 0.983366 loss) | |
| I0401 23:21:07.729066 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.262883 (* 1 = 0.262883 loss) | |
| I0401 23:21:07.729079 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.367 | |
| I0401 23:21:07.729091 6134 solver.cpp:406] Test net output #16: total_confidence = 0.318207 | |
| I0401 23:21:07.879640 6134 solver.cpp:229] Iteration 145000, loss = 2.84218 | |
| I0401 23:21:07.879679 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.608696 | |
| I0401 23:21:07.879696 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.892045 | |
| I0401 23:21:07.879709 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.717391 | |
| I0401 23:21:07.879724 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.66583 (* 0.3 = 0.499748 loss) | |
| I0401 23:21:07.879739 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.465786 (* 0.3 = 0.139736 loss) | |
| I0401 23:21:07.879750 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.717391 | |
| I0401 23:21:07.879762 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.920455 | |
| I0401 23:21:07.879775 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.978261 | |
| I0401 23:21:07.879788 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.982039 (* 0.3 = 0.294612 loss) | |
| I0401 23:21:07.879802 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.277326 (* 0.3 = 0.0831978 loss) | |
| I0401 23:21:07.879814 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.978261 | |
| I0401 23:21:07.879827 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.988636 | |
| I0401 23:21:07.879838 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0401 23:21:07.879851 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.175716 (* 1 = 0.175716 loss) | |
| I0401 23:21:07.879865 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0702877 (* 1 = 0.0702877 loss) | |
| I0401 23:21:07.879878 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.875 | |
| I0401 23:21:07.879889 6134 solver.cpp:245] Train net output #16: total_confidence = 0.539958 | |
| I0401 23:21:07.879901 6134 sgd_solver.cpp:106] Iteration 145000, lr = 0.01 | |
| I0401 23:23:16.490078 6134 solver.cpp:229] Iteration 145500, loss = 2.82648 | |
| I0401 23:23:16.490228 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.428571 | |
| I0401 23:23:16.490248 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 23:23:16.490262 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.714286 | |
| I0401 23:23:16.490278 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.78183 (* 0.3 = 0.534548 loss) | |
| I0401 23:23:16.490293 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.551941 (* 0.3 = 0.165582 loss) | |
| I0401 23:23:16.490305 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.469388 | |
| I0401 23:23:16.490317 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 23:23:16.490329 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.77551 | |
| I0401 23:23:16.490342 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.55057 (* 0.3 = 0.46517 loss) | |
| I0401 23:23:16.490356 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.46275 (* 0.3 = 0.138825 loss) | |
| I0401 23:23:16.490370 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.734694 | |
| I0401 23:23:16.490381 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0401 23:23:16.490392 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.836735 | |
| I0401 23:23:16.490406 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.936346 (* 1 = 0.936346 loss) | |
| I0401 23:23:16.490422 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.289769 (* 1 = 0.289769 loss) | |
| I0401 23:23:16.490433 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 23:23:16.490445 6134 solver.cpp:245] Train net output #16: total_confidence = 0.303071 | |
| I0401 23:23:16.490458 6134 sgd_solver.cpp:106] Iteration 145500, lr = 0.01 | |
| I0401 23:25:25.218041 6134 solver.cpp:229] Iteration 146000, loss = 2.76748 | |
| I0401 23:25:25.218204 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431373 | |
| I0401 23:25:25.218225 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 23:25:25.218238 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.745098 | |
| I0401 23:25:25.218255 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.80209 (* 0.3 = 0.540627 loss) | |
| I0401 23:25:25.218269 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.558115 (* 0.3 = 0.167434 loss) | |
| I0401 23:25:25.218282 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.607843 | |
| I0401 23:25:25.218294 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0401 23:25:25.218307 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.843137 | |
| I0401 23:25:25.218320 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.27051 (* 0.3 = 0.381153 loss) | |
| I0401 23:25:25.218334 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.375284 (* 0.3 = 0.112585 loss) | |
| I0401 23:25:25.218348 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.921569 | |
| I0401 23:25:25.218359 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.971591 | |
| I0401 23:25:25.218371 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.980392 | |
| I0401 23:25:25.218385 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.399456 (* 1 = 0.399456 loss) | |
| I0401 23:25:25.218400 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.126378 (* 1 = 0.126378 loss) | |
| I0401 23:25:25.218412 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0401 23:25:25.218425 6134 solver.cpp:245] Train net output #16: total_confidence = 0.31367 | |
| I0401 23:25:25.218436 6134 sgd_solver.cpp:106] Iteration 146000, lr = 0.01 | |
| I0401 23:27:33.874668 6134 solver.cpp:229] Iteration 146500, loss = 2.82705 | |
| I0401 23:27:33.874907 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.348837 | |
| I0401 23:27:33.874927 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 23:27:33.874940 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.627907 | |
| I0401 23:27:33.874956 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.24418 (* 0.3 = 0.673255 loss) | |
| I0401 23:27:33.874971 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.629645 (* 0.3 = 0.188893 loss) | |
| I0401 23:27:33.874984 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.488372 | |
| I0401 23:27:33.874996 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0401 23:27:33.875008 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.837209 | |
| I0401 23:27:33.875022 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.53762 (* 0.3 = 0.461286 loss) | |
| I0401 23:27:33.875036 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.43596 (* 0.3 = 0.130788 loss) | |
| I0401 23:27:33.875048 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.744186 | |
| I0401 23:27:33.875063 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 23:27:33.875075 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.860465 | |
| I0401 23:27:33.875090 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.971167 (* 1 = 0.971167 loss) | |
| I0401 23:27:33.875104 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.265258 (* 1 = 0.265258 loss) | |
| I0401 23:27:33.875116 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 23:27:33.875128 6134 solver.cpp:245] Train net output #16: total_confidence = 0.150196 | |
| I0401 23:27:33.875140 6134 sgd_solver.cpp:106] Iteration 146500, lr = 0.01 | |
| I0401 23:29:42.525481 6134 solver.cpp:229] Iteration 147000, loss = 2.83904 | |
| I0401 23:29:42.525647 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.283019 | |
| I0401 23:29:42.525670 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0401 23:29:42.525682 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.396226 | |
| I0401 23:29:42.525698 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.78791 (* 0.3 = 0.836374 loss) | |
| I0401 23:29:42.525714 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.906203 (* 0.3 = 0.271861 loss) | |
| I0401 23:29:42.525727 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.358491 | |
| I0401 23:29:42.525739 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.795455 | |
| I0401 23:29:42.525751 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.528302 | |
| I0401 23:29:42.525765 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.46108 (* 0.3 = 0.738325 loss) | |
| I0401 23:29:42.525779 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.767455 (* 0.3 = 0.230237 loss) | |
| I0401 23:29:42.525792 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.490566 | |
| I0401 23:29:42.525804 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.835227 | |
| I0401 23:29:42.525816 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.622642 | |
| I0401 23:29:42.525830 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.13676 (* 1 = 2.13676 loss) | |
| I0401 23:29:42.525845 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.671843 (* 1 = 0.671843 loss) | |
| I0401 23:29:42.525856 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0401 23:29:42.525869 6134 solver.cpp:245] Train net output #16: total_confidence = 0.0984711 | |
| I0401 23:29:42.525882 6134 sgd_solver.cpp:106] Iteration 147000, lr = 0.01 | |
| I0401 23:31:51.271040 6134 solver.cpp:229] Iteration 147500, loss = 2.81142 | |
| I0401 23:31:51.271149 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.354167 | |
| I0401 23:31:51.271170 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 23:31:51.271183 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.708333 | |
| I0401 23:31:51.271198 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.78569 (* 0.3 = 0.535706 loss) | |
| I0401 23:31:51.271214 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.520621 (* 0.3 = 0.156186 loss) | |
| I0401 23:31:51.271226 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.583333 | |
| I0401 23:31:51.271239 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0401 23:31:51.271250 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.875 | |
| I0401 23:31:51.271265 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.28278 (* 0.3 = 0.384835 loss) | |
| I0401 23:31:51.271280 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.366015 (* 0.3 = 0.109804 loss) | |
| I0401 23:31:51.271292 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.875 | |
| I0401 23:31:51.271304 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0401 23:31:51.271317 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9375 | |
| I0401 23:31:51.271329 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.504067 (* 1 = 0.504067 loss) | |
| I0401 23:31:51.271344 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.143748 (* 1 = 0.143748 loss) | |
| I0401 23:31:51.271356 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0401 23:31:51.271368 6134 solver.cpp:245] Train net output #16: total_confidence = 0.341807 | |
| I0401 23:31:51.271381 6134 sgd_solver.cpp:106] Iteration 147500, lr = 0.01 | |
| I0401 23:33:59.850605 6134 solver.cpp:229] Iteration 148000, loss = 2.77565 | |
| I0401 23:33:59.850745 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.254545 | |
| I0401 23:33:59.850777 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0401 23:33:59.850800 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.490909 | |
| I0401 23:33:59.850818 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.34429 (* 0.3 = 1.00329 loss) | |
| I0401 23:33:59.850833 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.10437 (* 0.3 = 0.331312 loss) | |
| I0401 23:33:59.850847 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.309091 | |
| I0401 23:33:59.850858 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.772727 | |
| I0401 23:33:59.850870 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.545455 | |
| I0401 23:33:59.850884 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.96383 (* 0.3 = 0.889149 loss) | |
| I0401 23:33:59.850898 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.968295 (* 0.3 = 0.290488 loss) | |
| I0401 23:33:59.850910 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.6 | |
| I0401 23:33:59.850922 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0401 23:33:59.850934 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.690909 | |
| I0401 23:33:59.850949 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.12284 (* 1 = 2.12284 loss) | |
| I0401 23:33:59.850962 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.677121 (* 1 = 0.677121 loss) | |
| I0401 23:33:59.850975 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 23:33:59.850986 6134 solver.cpp:245] Train net output #16: total_confidence = 0.187856 | |
| I0401 23:33:59.850998 6134 sgd_solver.cpp:106] Iteration 148000, lr = 0.01 | |
| I0401 23:36:08.606850 6134 solver.cpp:229] Iteration 148500, loss = 2.83109 | |
| I0401 23:36:08.606963 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.391304 | |
| I0401 23:36:08.606984 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 23:36:08.606997 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.673913 | |
| I0401 23:36:08.607013 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.79617 (* 0.3 = 0.538851 loss) | |
| I0401 23:36:08.607028 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.5089 (* 0.3 = 0.15267 loss) | |
| I0401 23:36:08.607040 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.543478 | |
| I0401 23:36:08.607053 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0401 23:36:08.607065 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.782609 | |
| I0401 23:36:08.607079 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.29862 (* 0.3 = 0.389585 loss) | |
| I0401 23:36:08.607094 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.379075 (* 0.3 = 0.113722 loss) | |
| I0401 23:36:08.607105 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.73913 | |
| I0401 23:36:08.607117 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0401 23:36:08.607131 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.956522 | |
| I0401 23:36:08.607143 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.747645 (* 1 = 0.747645 loss) | |
| I0401 23:36:08.607157 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.223841 (* 1 = 0.223841 loss) | |
| I0401 23:36:08.607169 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 23:36:08.607182 6134 solver.cpp:245] Train net output #16: total_confidence = 0.283068 | |
| I0401 23:36:08.607193 6134 sgd_solver.cpp:106] Iteration 148500, lr = 0.01 | |
| I0401 23:38:17.278573 6134 solver.cpp:229] Iteration 149000, loss = 2.75564 | |
| I0401 23:38:17.278887 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.382979 | |
| I0401 23:38:17.278908 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0401 23:38:17.278921 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.617021 | |
| I0401 23:38:17.278937 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.0687 (* 0.3 = 0.62061 loss) | |
| I0401 23:38:17.278951 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.60989 (* 0.3 = 0.182967 loss) | |
| I0401 23:38:17.278964 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.489362 | |
| I0401 23:38:17.278977 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 23:38:17.278990 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.702128 | |
| I0401 23:38:17.279002 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.77928 (* 0.3 = 0.533783 loss) | |
| I0401 23:38:17.279016 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.542357 (* 0.3 = 0.162707 loss) | |
| I0401 23:38:17.279028 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.680851 | |
| I0401 23:38:17.279042 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0401 23:38:17.279054 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.851064 | |
| I0401 23:38:17.279068 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.1221 (* 1 = 1.1221 loss) | |
| I0401 23:38:17.279083 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.346508 (* 1 = 0.346508 loss) | |
| I0401 23:38:17.279094 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 23:38:17.279106 6134 solver.cpp:245] Train net output #16: total_confidence = 0.148648 | |
| I0401 23:38:17.279119 6134 sgd_solver.cpp:106] Iteration 149000, lr = 0.01 | |
| I0401 23:40:25.931428 6134 solver.cpp:229] Iteration 149500, loss = 2.79422 | |
| I0401 23:40:25.931557 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.244444 | |
| I0401 23:40:25.931578 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0401 23:40:25.931591 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.422222 | |
| I0401 23:40:25.931607 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.73261 (* 0.3 = 0.819784 loss) | |
| I0401 23:40:25.931622 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.748035 (* 0.3 = 0.22441 loss) | |
| I0401 23:40:25.931634 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.533333 | |
| I0401 23:40:25.931648 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0401 23:40:25.931659 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.733333 | |
| I0401 23:40:25.931673 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.8296 (* 0.3 = 0.548879 loss) | |
| I0401 23:40:25.931687 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.487726 (* 0.3 = 0.146318 loss) | |
| I0401 23:40:25.931699 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.733333 | |
| I0401 23:40:25.931711 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 23:40:25.931722 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.911111 | |
| I0401 23:40:25.931737 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.872758 (* 1 = 0.872758 loss) | |
| I0401 23:40:25.931751 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.259262 (* 1 = 0.259262 loss) | |
| I0401 23:40:25.931762 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 23:40:25.931776 6134 solver.cpp:245] Train net output #16: total_confidence = 0.195861 | |
| I0401 23:40:25.931787 6134 sgd_solver.cpp:106] Iteration 149500, lr = 0.01 | |
| I0401 23:42:34.461205 6134 solver.cpp:456] Snapshotting to binary proto file /mnt/snapshots/mixed_lstm9_bn_iter_150000.caffemodel | |
| I0401 23:42:34.935806 6134 sgd_solver.cpp:273] Snapshotting solver state to binary proto file /mnt/snapshots/mixed_lstm9_bn_iter_150000.solverstate | |
| I0401 23:42:35.100368 6134 solver.cpp:338] Iteration 150000, Testing net (#0) | |
| I0401 23:43:04.853565 6134 solver.cpp:393] Test loss: 2.39619 | |
| I0401 23:43:04.853688 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.479191 | |
| I0401 23:43:04.853706 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.858002 | |
| I0401 23:43:04.853719 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.759394 | |
| I0401 23:43:04.853736 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.73463 (* 0.3 = 0.520389 loss) | |
| I0401 23:43:04.853751 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.481018 (* 0.3 = 0.144305 loss) | |
| I0401 23:43:04.853763 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.665072 | |
| I0401 23:43:04.853776 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.893275 | |
| I0401 23:43:04.853788 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.864958 | |
| I0401 23:43:04.853801 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.20784 (* 0.3 = 0.362353 loss) | |
| I0401 23:43:04.853817 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.376466 (* 0.3 = 0.11294 loss) | |
| I0401 23:43:04.853829 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.749671 | |
| I0401 23:43:04.853842 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.932864 | |
| I0401 23:43:04.853853 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.885665 | |
| I0401 23:43:04.853868 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.987377 (* 1 = 0.987377 loss) | |
| I0401 23:43:04.853880 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.268828 (* 1 = 0.268828 loss) | |
| I0401 23:43:04.853893 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.388 | |
| I0401 23:43:04.853904 6134 solver.cpp:406] Test net output #16: total_confidence = 0.373392 | |
| I0401 23:43:05.004815 6134 solver.cpp:229] Iteration 150000, loss = 2.75586 | |
| I0401 23:43:05.004856 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.44186 | |
| I0401 23:43:05.004874 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0401 23:43:05.004887 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.651163 | |
| I0401 23:43:05.004906 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.98139 (* 0.3 = 0.594418 loss) | |
| I0401 23:43:05.004920 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.547771 (* 0.3 = 0.164331 loss) | |
| I0401 23:43:05.004933 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.55814 | |
| I0401 23:43:05.004946 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0401 23:43:05.004957 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.767442 | |
| I0401 23:43:05.004971 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.48745 (* 0.3 = 0.446236 loss) | |
| I0401 23:43:05.004986 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.448884 (* 0.3 = 0.134665 loss) | |
| I0401 23:43:05.004997 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.790698 | |
| I0401 23:43:05.005009 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0401 23:43:05.005022 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.906977 | |
| I0401 23:43:05.005035 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.780462 (* 1 = 0.780462 loss) | |
| I0401 23:43:05.005074 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.222747 (* 1 = 0.222747 loss) | |
| I0401 23:43:05.005087 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 23:43:05.005107 6134 solver.cpp:245] Train net output #16: total_confidence = 0.335473 | |
| I0401 23:43:05.005120 6134 sgd_solver.cpp:106] Iteration 150000, lr = 0.01 | |
| I0401 23:45:14.031409 6134 solver.cpp:229] Iteration 150500, loss = 2.78698 | |
| I0401 23:45:14.031536 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.384615 | |
| I0401 23:45:14.031556 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 23:45:14.031569 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.564103 | |
| I0401 23:45:14.031585 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.51211 (* 0.3 = 0.753634 loss) | |
| I0401 23:45:14.031600 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.729157 (* 0.3 = 0.218747 loss) | |
| I0401 23:45:14.031613 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.461538 | |
| I0401 23:45:14.031625 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 23:45:14.031637 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.641026 | |
| I0401 23:45:14.031651 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.32058 (* 0.3 = 0.696175 loss) | |
| I0401 23:45:14.031666 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.602945 (* 0.3 = 0.180884 loss) | |
| I0401 23:45:14.031677 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.641026 | |
| I0401 23:45:14.031690 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0401 23:45:14.031702 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.794872 | |
| I0401 23:45:14.031716 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.23255 (* 1 = 2.23255 loss) | |
| I0401 23:45:14.031730 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.547196 (* 1 = 0.547196 loss) | |
| I0401 23:45:14.031743 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 23:45:14.031754 6134 solver.cpp:245] Train net output #16: total_confidence = 0.330419 | |
| I0401 23:45:14.031766 6134 sgd_solver.cpp:106] Iteration 150500, lr = 0.01 | |
| I0401 23:47:22.927872 6134 solver.cpp:229] Iteration 151000, loss = 2.78594 | |
| I0401 23:47:22.928154 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.386364 | |
| I0401 23:47:22.928174 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0401 23:47:22.928187 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.659091 | |
| I0401 23:47:22.928205 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.03704 (* 0.3 = 0.611113 loss) | |
| I0401 23:47:22.928220 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.581592 (* 0.3 = 0.174478 loss) | |
| I0401 23:47:22.928232 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.522727 | |
| I0401 23:47:22.928244 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0401 23:47:22.928256 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.909091 | |
| I0401 23:47:22.928270 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.44009 (* 0.3 = 0.432028 loss) | |
| I0401 23:47:22.928285 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.4246 (* 0.3 = 0.12738 loss) | |
| I0401 23:47:22.928306 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.886364 | |
| I0401 23:47:22.928325 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.971591 | |
| I0401 23:47:22.928339 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.977273 | |
| I0401 23:47:22.928354 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.434807 (* 1 = 0.434807 loss) | |
| I0401 23:47:22.928367 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.11643 (* 1 = 0.11643 loss) | |
| I0401 23:47:22.928380 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 23:47:22.928401 6134 solver.cpp:245] Train net output #16: total_confidence = 0.340153 | |
| I0401 23:47:22.928423 6134 sgd_solver.cpp:106] Iteration 151000, lr = 0.01 | |
| I0401 23:49:32.013300 6134 solver.cpp:229] Iteration 151500, loss = 2.76927 | |
| I0401 23:49:32.013509 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.352941 | |
| I0401 23:49:32.013540 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 23:49:32.013555 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.529412 | |
| I0401 23:49:32.013572 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.22883 (* 0.3 = 0.668648 loss) | |
| I0401 23:49:32.013587 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.663045 (* 0.3 = 0.198913 loss) | |
| I0401 23:49:32.013600 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.45098 | |
| I0401 23:49:32.013613 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0401 23:49:32.013627 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.705882 | |
| I0401 23:49:32.013640 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.83122 (* 0.3 = 0.549367 loss) | |
| I0401 23:49:32.013664 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.555914 (* 0.3 = 0.166774 loss) | |
| I0401 23:49:32.013676 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.666667 | |
| I0401 23:49:32.013689 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 23:49:32.013701 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.901961 | |
| I0401 23:49:32.013715 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.97007 (* 1 = 0.97007 loss) | |
| I0401 23:49:32.013738 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.295167 (* 1 = 0.295167 loss) | |
| I0401 23:49:32.013751 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 23:49:32.013763 6134 solver.cpp:245] Train net output #16: total_confidence = 0.231387 | |
| I0401 23:49:32.013777 6134 sgd_solver.cpp:106] Iteration 151500, lr = 0.01 | |
| I0401 23:51:41.074734 6134 solver.cpp:229] Iteration 152000, loss = 2.77101 | |
| I0401 23:51:41.074851 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.363636 | |
| I0401 23:51:41.074872 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0401 23:51:41.074885 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.636364 | |
| I0401 23:51:41.074900 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.09602 (* 0.3 = 0.628805 loss) | |
| I0401 23:51:41.074916 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.621232 (* 0.3 = 0.18637 loss) | |
| I0401 23:51:41.074929 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0401 23:51:41.074941 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 23:51:41.074954 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.840909 | |
| I0401 23:51:41.074967 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.68626 (* 0.3 = 0.505879 loss) | |
| I0401 23:51:41.074981 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.497148 (* 0.3 = 0.149144 loss) | |
| I0401 23:51:41.074993 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0401 23:51:41.075006 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0401 23:51:41.075017 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.886364 | |
| I0401 23:51:41.075031 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.936185 (* 1 = 0.936185 loss) | |
| I0401 23:51:41.075045 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.267438 (* 1 = 0.267438 loss) | |
| I0401 23:51:41.075057 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0401 23:51:41.075069 6134 solver.cpp:245] Train net output #16: total_confidence = 0.306197 | |
| I0401 23:51:41.075081 6134 sgd_solver.cpp:106] Iteration 152000, lr = 0.01 | |
| I0401 23:53:50.121984 6134 solver.cpp:229] Iteration 152500, loss = 2.77517 | |
| I0401 23:53:50.122119 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0401 23:53:50.122140 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0401 23:53:50.122160 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.636364 | |
| I0401 23:53:50.122176 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.10448 (* 0.3 = 0.631344 loss) | |
| I0401 23:53:50.122190 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.686609 (* 0.3 = 0.205983 loss) | |
| I0401 23:53:50.122203 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.490909 | |
| I0401 23:53:50.122222 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0401 23:53:50.122234 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.745455 | |
| I0401 23:53:50.122248 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.89672 (* 0.3 = 0.569017 loss) | |
| I0401 23:53:50.122262 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.617694 (* 0.3 = 0.185308 loss) | |
| I0401 23:53:50.122275 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.727273 | |
| I0401 23:53:50.122297 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0401 23:53:50.122308 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.836364 | |
| I0401 23:53:50.122323 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.05115 (* 1 = 1.05115 loss) | |
| I0401 23:53:50.122336 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.364064 (* 1 = 0.364064 loss) | |
| I0401 23:53:50.122349 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0401 23:53:50.122360 6134 solver.cpp:245] Train net output #16: total_confidence = 0.148337 | |
| I0401 23:53:50.122372 6134 sgd_solver.cpp:106] Iteration 152500, lr = 0.01 | |
| I0401 23:55:59.283843 6134 solver.cpp:229] Iteration 153000, loss = 2.745 | |
| I0401 23:55:59.283967 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.414634 | |
| I0401 23:55:59.283987 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0401 23:55:59.283999 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.731707 | |
| I0401 23:55:59.284015 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.83757 (* 0.3 = 0.55127 loss) | |
| I0401 23:55:59.284030 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.522915 (* 0.3 = 0.156875 loss) | |
| I0401 23:55:59.284044 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.512195 | |
| I0401 23:55:59.284055 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0401 23:55:59.284067 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.804878 | |
| I0401 23:55:59.284081 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.48716 (* 0.3 = 0.446147 loss) | |
| I0401 23:55:59.284096 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.450138 (* 0.3 = 0.135041 loss) | |
| I0401 23:55:59.284107 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.707317 | |
| I0401 23:55:59.284124 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0401 23:55:59.284137 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.853659 | |
| I0401 23:55:59.284150 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.906008 (* 1 = 0.906008 loss) | |
| I0401 23:55:59.284164 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.244095 (* 1 = 0.244095 loss) | |
| I0401 23:55:59.284184 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 23:55:59.284196 6134 solver.cpp:245] Train net output #16: total_confidence = 0.336511 | |
| I0401 23:55:59.284209 6134 sgd_solver.cpp:106] Iteration 153000, lr = 0.01 | |
| I0401 23:58:08.460147 6134 solver.cpp:229] Iteration 153500, loss = 2.77097 | |
| I0401 23:58:08.460539 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.552632 | |
| I0401 23:58:08.460559 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0401 23:58:08.460572 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.736842 | |
| I0401 23:58:08.460597 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.56747 (* 0.3 = 0.470242 loss) | |
| I0401 23:58:08.460613 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.489539 (* 0.3 = 0.146862 loss) | |
| I0401 23:58:08.460625 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.736842 | |
| I0401 23:58:08.460638 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0401 23:58:08.460650 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.894737 | |
| I0401 23:58:08.460664 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.982054 (* 0.3 = 0.294616 loss) | |
| I0401 23:58:08.460680 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.360584 (* 0.3 = 0.108175 loss) | |
| I0401 23:58:08.460691 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.868421 | |
| I0401 23:58:08.460703 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0401 23:58:08.460716 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.973684 | |
| I0401 23:58:08.460731 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.311908 (* 1 = 0.311908 loss) | |
| I0401 23:58:08.460744 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0894257 (* 1 = 0.0894257 loss) | |
| I0401 23:58:08.460757 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0401 23:58:08.460769 6134 solver.cpp:245] Train net output #16: total_confidence = 0.420631 | |
| I0401 23:58:08.460782 6134 sgd_solver.cpp:106] Iteration 153500, lr = 0.01 | |
| I0402 00:00:17.645659 6134 solver.cpp:229] Iteration 154000, loss = 2.77823 | |
| I0402 00:00:17.645774 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.458333 | |
| I0402 00:00:17.645794 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0402 00:00:17.645807 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.6875 | |
| I0402 00:00:17.645823 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.91001 (* 0.3 = 0.573002 loss) | |
| I0402 00:00:17.645845 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.598358 (* 0.3 = 0.179507 loss) | |
| I0402 00:00:17.645859 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.604167 | |
| I0402 00:00:17.645870 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 00:00:17.645882 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8125 | |
| I0402 00:00:17.645896 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.67889 (* 0.3 = 0.503667 loss) | |
| I0402 00:00:17.645920 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.501856 (* 0.3 = 0.150557 loss) | |
| I0402 00:00:17.645931 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.729167 | |
| I0402 00:00:17.645944 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0402 00:00:17.645956 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.895833 | |
| I0402 00:00:17.645970 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.814565 (* 1 = 0.814565 loss) | |
| I0402 00:00:17.645984 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.240524 (* 1 = 0.240524 loss) | |
| I0402 00:00:17.645997 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 00:00:17.646008 6134 solver.cpp:245] Train net output #16: total_confidence = 0.349816 | |
| I0402 00:00:17.646020 6134 sgd_solver.cpp:106] Iteration 154000, lr = 0.01 | |
| I0402 00:02:26.804237 6134 solver.cpp:229] Iteration 154500, loss = 2.76312 | |
| I0402 00:02:26.804388 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.270833 | |
| I0402 00:02:26.804409 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0402 00:02:26.804430 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.479167 | |
| I0402 00:02:26.804447 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.54877 (* 0.3 = 0.76463 loss) | |
| I0402 00:02:26.804462 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.753418 (* 0.3 = 0.226025 loss) | |
| I0402 00:02:26.804474 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.354167 | |
| I0402 00:02:26.804496 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.806818 | |
| I0402 00:02:26.804507 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.6875 | |
| I0402 00:02:26.804523 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.15345 (* 0.3 = 0.646036 loss) | |
| I0402 00:02:26.804538 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.671646 (* 0.3 = 0.201494 loss) | |
| I0402 00:02:26.804559 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.625 | |
| I0402 00:02:26.804571 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0402 00:02:26.804584 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.770833 | |
| I0402 00:02:26.804597 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.34074 (* 1 = 1.34074 loss) | |
| I0402 00:02:26.804611 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.408859 (* 1 = 0.408859 loss) | |
| I0402 00:02:26.804625 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 00:02:26.804643 6134 solver.cpp:245] Train net output #16: total_confidence = 0.140748 | |
| I0402 00:02:26.804656 6134 sgd_solver.cpp:106] Iteration 154500, lr = 0.01 | |
| I0402 00:04:35.857313 6134 solver.cpp:338] Iteration 155000, Testing net (#0) | |
| I0402 00:05:05.701642 6134 solver.cpp:393] Test loss: 2.32398 | |
| I0402 00:05:05.701699 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.510318 | |
| I0402 00:05:05.701716 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.869367 | |
| I0402 00:05:05.701730 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.772614 | |
| I0402 00:05:05.701746 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.67624 (* 0.3 = 0.502873 loss) | |
| I0402 00:05:05.701761 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.455017 (* 0.3 = 0.136505 loss) | |
| I0402 00:05:05.701773 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.645632 | |
| I0402 00:05:05.701786 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.907776 | |
| I0402 00:05:05.701798 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.861672 | |
| I0402 00:05:05.701812 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.25445 (* 0.3 = 0.376334 loss) | |
| I0402 00:05:05.701827 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.333876 (* 0.3 = 0.100163 loss) | |
| I0402 00:05:05.701838 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.771499 | |
| I0402 00:05:05.701849 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.938273 | |
| I0402 00:05:05.701861 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.884596 | |
| I0402 00:05:05.701874 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.948064 (* 1 = 0.948064 loss) | |
| I0402 00:05:05.701889 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.260042 (* 1 = 0.260042 loss) | |
| I0402 00:05:05.701900 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.411 | |
| I0402 00:05:05.701912 6134 solver.cpp:406] Test net output #16: total_confidence = 0.359422 | |
| I0402 00:05:05.853673 6134 solver.cpp:229] Iteration 155000, loss = 2.73481 | |
| I0402 00:05:05.853732 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.369565 | |
| I0402 00:05:05.853750 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0402 00:05:05.853763 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.673913 | |
| I0402 00:05:05.853780 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.9332 (* 0.3 = 0.57996 loss) | |
| I0402 00:05:05.853797 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.605147 (* 0.3 = 0.181544 loss) | |
| I0402 00:05:05.853811 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.521739 | |
| I0402 00:05:05.853823 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 00:05:05.853835 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76087 | |
| I0402 00:05:05.853849 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.56761 (* 0.3 = 0.470283 loss) | |
| I0402 00:05:05.853864 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.467943 (* 0.3 = 0.140383 loss) | |
| I0402 00:05:05.853876 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.782609 | |
| I0402 00:05:05.853889 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 00:05:05.853901 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.913043 | |
| I0402 00:05:05.853916 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.729948 (* 1 = 0.729948 loss) | |
| I0402 00:05:05.853930 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.225597 (* 1 = 0.225597 loss) | |
| I0402 00:05:05.853943 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 00:05:05.853955 6134 solver.cpp:245] Train net output #16: total_confidence = 0.339266 | |
| I0402 00:05:05.853967 6134 sgd_solver.cpp:106] Iteration 155000, lr = 0.01 | |
| I0402 00:07:14.725356 6134 solver.cpp:229] Iteration 155500, loss = 2.74292 | |
| I0402 00:07:14.725706 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.384615 | |
| I0402 00:07:14.725728 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0402 00:07:14.725742 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.615385 | |
| I0402 00:07:14.725759 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.06832 (* 0.3 = 0.620497 loss) | |
| I0402 00:07:14.725774 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.633492 (* 0.3 = 0.190048 loss) | |
| I0402 00:07:14.725785 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.442308 | |
| I0402 00:07:14.725798 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0402 00:07:14.725811 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0402 00:07:14.725826 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.7557 (* 0.3 = 0.526711 loss) | |
| I0402 00:07:14.725839 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.532358 (* 0.3 = 0.159707 loss) | |
| I0402 00:07:14.725852 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.903846 | |
| I0402 00:07:14.725864 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.971591 | |
| I0402 00:07:14.725877 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.980769 | |
| I0402 00:07:14.725891 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.336591 (* 1 = 0.336591 loss) | |
| I0402 00:07:14.725905 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.103802 (* 1 = 0.103802 loss) | |
| I0402 00:07:14.725917 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0402 00:07:14.725930 6134 solver.cpp:245] Train net output #16: total_confidence = 0.387976 | |
| I0402 00:07:14.725942 6134 sgd_solver.cpp:106] Iteration 155500, lr = 0.01 | |
| I0402 00:09:23.745364 6134 solver.cpp:229] Iteration 156000, loss = 2.73921 | |
| I0402 00:09:23.745522 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.380952 | |
| I0402 00:09:23.745553 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 00:09:23.745565 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.738095 | |
| I0402 00:09:23.745581 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.01489 (* 0.3 = 0.604468 loss) | |
| I0402 00:09:23.745597 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.599943 (* 0.3 = 0.179983 loss) | |
| I0402 00:09:23.745618 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0402 00:09:23.745631 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0402 00:09:23.745643 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.833333 | |
| I0402 00:09:23.745657 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.53334 (* 0.3 = 0.460001 loss) | |
| I0402 00:09:23.745678 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.43454 (* 0.3 = 0.130362 loss) | |
| I0402 00:09:23.745692 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.738095 | |
| I0402 00:09:23.745703 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 00:09:23.745715 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.904762 | |
| I0402 00:09:23.745730 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.0454 (* 1 = 1.0454 loss) | |
| I0402 00:09:23.745753 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.269795 (* 1 = 0.269795 loss) | |
| I0402 00:09:23.745764 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 00:09:23.745777 6134 solver.cpp:245] Train net output #16: total_confidence = 0.331052 | |
| I0402 00:09:23.745790 6134 sgd_solver.cpp:106] Iteration 156000, lr = 0.01 | |
| I0402 00:11:32.626065 6134 solver.cpp:229] Iteration 156500, loss = 2.73836 | |
| I0402 00:11:32.626189 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.390244 | |
| I0402 00:11:32.626209 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 00:11:32.626222 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.634146 | |
| I0402 00:11:32.626237 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.01779 (* 0.3 = 0.605336 loss) | |
| I0402 00:11:32.626261 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.58198 (* 0.3 = 0.174594 loss) | |
| I0402 00:11:32.626276 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.536585 | |
| I0402 00:11:32.626287 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0402 00:11:32.626299 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.731707 | |
| I0402 00:11:32.626313 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.49826 (* 0.3 = 0.449478 loss) | |
| I0402 00:11:32.626327 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.464068 (* 0.3 = 0.13922 loss) | |
| I0402 00:11:32.626340 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.780488 | |
| I0402 00:11:32.626353 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0402 00:11:32.626363 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.829268 | |
| I0402 00:11:32.626377 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.00415 (* 1 = 1.00415 loss) | |
| I0402 00:11:32.626391 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.298754 (* 1 = 0.298754 loss) | |
| I0402 00:11:32.626405 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 00:11:32.626416 6134 solver.cpp:245] Train net output #16: total_confidence = 0.256071 | |
| I0402 00:11:32.626428 6134 sgd_solver.cpp:106] Iteration 156500, lr = 0.01 | |
| I0402 00:13:41.596166 6134 solver.cpp:229] Iteration 157000, loss = 2.79315 | |
| I0402 00:13:41.596303 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.404255 | |
| I0402 00:13:41.596343 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0402 00:13:41.596370 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.680851 | |
| I0402 00:13:41.596395 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.09645 (* 0.3 = 0.628935 loss) | |
| I0402 00:13:41.596418 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.58342 (* 0.3 = 0.175026 loss) | |
| I0402 00:13:41.596431 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.489362 | |
| I0402 00:13:41.596443 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0402 00:13:41.596456 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.680851 | |
| I0402 00:13:41.596477 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.03459 (* 0.3 = 0.610377 loss) | |
| I0402 00:13:41.596492 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.555328 (* 0.3 = 0.166598 loss) | |
| I0402 00:13:41.596504 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.787234 | |
| I0402 00:13:41.596518 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0402 00:13:41.596531 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.851064 | |
| I0402 00:13:41.596562 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.6698 (* 1 = 1.6698 loss) | |
| I0402 00:13:41.596585 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.456847 (* 1 = 0.456847 loss) | |
| I0402 00:13:41.596596 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 00:13:41.596613 6134 solver.cpp:245] Train net output #16: total_confidence = 0.441317 | |
| I0402 00:13:41.596626 6134 sgd_solver.cpp:106] Iteration 157000, lr = 0.01 | |
| I0402 00:15:50.676112 6134 solver.cpp:229] Iteration 157500, loss = 2.65067 | |
| I0402 00:15:50.676228 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0402 00:15:50.676246 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0402 00:15:50.676259 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.5 | |
| I0402 00:15:50.676276 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.56799 (* 0.3 = 0.770398 loss) | |
| I0402 00:15:50.676291 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.720467 (* 0.3 = 0.21614 loss) | |
| I0402 00:15:50.676304 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.375 | |
| I0402 00:15:50.676316 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0402 00:15:50.676332 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.645833 | |
| I0402 00:15:50.676347 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.17371 (* 0.3 = 0.652113 loss) | |
| I0402 00:15:50.676362 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.618872 (* 0.3 = 0.185662 loss) | |
| I0402 00:15:50.676374 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.625 | |
| I0402 00:15:50.676386 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0402 00:15:50.676398 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.729167 | |
| I0402 00:15:50.676414 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.46403 (* 1 = 1.46403 loss) | |
| I0402 00:15:50.676427 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.408433 (* 1 = 0.408433 loss) | |
| I0402 00:15:50.676440 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 00:15:50.676452 6134 solver.cpp:245] Train net output #16: total_confidence = 0.218195 | |
| I0402 00:15:50.676470 6134 sgd_solver.cpp:106] Iteration 157500, lr = 0.01 | |
| I0402 00:17:59.723016 6134 solver.cpp:229] Iteration 158000, loss = 2.75658 | |
| I0402 00:17:59.723361 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0402 00:17:59.723383 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 00:17:59.723397 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.644444 | |
| I0402 00:17:59.723412 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.93929 (* 0.3 = 0.581787 loss) | |
| I0402 00:17:59.723428 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.564881 (* 0.3 = 0.169464 loss) | |
| I0402 00:17:59.723440 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.533333 | |
| I0402 00:17:59.723453 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0402 00:17:59.723464 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.822222 | |
| I0402 00:17:59.723479 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.42667 (* 0.3 = 0.428 loss) | |
| I0402 00:17:59.723494 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.448254 (* 0.3 = 0.134476 loss) | |
| I0402 00:17:59.723505 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.822222 | |
| I0402 00:17:59.723520 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0402 00:17:59.723533 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.933333 | |
| I0402 00:17:59.723547 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.730618 (* 1 = 0.730618 loss) | |
| I0402 00:17:59.723562 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.236218 (* 1 = 0.236218 loss) | |
| I0402 00:17:59.723573 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 00:17:59.723585 6134 solver.cpp:245] Train net output #16: total_confidence = 0.347608 | |
| I0402 00:17:59.723598 6134 sgd_solver.cpp:106] Iteration 158000, lr = 0.01 | |
| I0402 00:20:08.439046 6134 solver.cpp:229] Iteration 158500, loss = 2.69821 | |
| I0402 00:20:08.439146 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.395833 | |
| I0402 00:20:08.439165 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0402 00:20:08.439178 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.645833 | |
| I0402 00:20:08.439194 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.99444 (* 0.3 = 0.598333 loss) | |
| I0402 00:20:08.439209 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.603773 (* 0.3 = 0.181132 loss) | |
| I0402 00:20:08.439223 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.520833 | |
| I0402 00:20:08.439235 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0402 00:20:08.439247 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.791667 | |
| I0402 00:20:08.439261 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.63723 (* 0.3 = 0.49117 loss) | |
| I0402 00:20:08.439275 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.536154 (* 0.3 = 0.160846 loss) | |
| I0402 00:20:08.439288 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.770833 | |
| I0402 00:20:08.439301 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0402 00:20:08.439312 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.958333 | |
| I0402 00:20:08.439327 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.789183 (* 1 = 0.789183 loss) | |
| I0402 00:20:08.439340 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.280683 (* 1 = 0.280683 loss) | |
| I0402 00:20:08.439353 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 00:20:08.439365 6134 solver.cpp:245] Train net output #16: total_confidence = 0.344616 | |
| I0402 00:20:08.439378 6134 sgd_solver.cpp:106] Iteration 158500, lr = 0.01 | |
| I0402 00:22:17.146591 6134 solver.cpp:229] Iteration 159000, loss = 2.69187 | |
| I0402 00:22:17.146742 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.380952 | |
| I0402 00:22:17.146772 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 00:22:17.146793 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.619048 | |
| I0402 00:22:17.146821 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.08936 (* 0.3 = 0.626808 loss) | |
| I0402 00:22:17.146848 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.584524 (* 0.3 = 0.175357 loss) | |
| I0402 00:22:17.146872 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.571429 | |
| I0402 00:22:17.146896 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0402 00:22:17.146919 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.785714 | |
| I0402 00:22:17.146945 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.54683 (* 0.3 = 0.464049 loss) | |
| I0402 00:22:17.146970 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.447493 (* 0.3 = 0.134248 loss) | |
| I0402 00:22:17.146994 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.785714 | |
| I0402 00:22:17.147017 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0402 00:22:17.147043 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.928571 | |
| I0402 00:22:17.147070 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.730097 (* 1 = 0.730097 loss) | |
| I0402 00:22:17.147097 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.194703 (* 1 = 0.194703 loss) | |
| I0402 00:22:17.147120 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 00:22:17.147142 6134 solver.cpp:245] Train net output #16: total_confidence = 0.182505 | |
| I0402 00:22:17.147164 6134 sgd_solver.cpp:106] Iteration 159000, lr = 0.01 | |
| I0402 00:24:25.802490 6134 solver.cpp:229] Iteration 159500, loss = 2.69779 | |
| I0402 00:24:25.802603 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.414634 | |
| I0402 00:24:25.802635 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 00:24:25.802660 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.609756 | |
| I0402 00:24:25.802690 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.06962 (* 0.3 = 0.620886 loss) | |
| I0402 00:24:25.802721 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.58755 (* 0.3 = 0.176265 loss) | |
| I0402 00:24:25.802747 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.512195 | |
| I0402 00:24:25.802772 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0402 00:24:25.802794 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.829268 | |
| I0402 00:24:25.802821 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.4902 (* 0.3 = 0.447061 loss) | |
| I0402 00:24:25.802848 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.393208 (* 0.3 = 0.117962 loss) | |
| I0402 00:24:25.802872 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.829268 | |
| I0402 00:24:25.802894 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0402 00:24:25.802917 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.926829 | |
| I0402 00:24:25.802944 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.608442 (* 1 = 0.608442 loss) | |
| I0402 00:24:25.802970 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.16313 (* 1 = 0.16313 loss) | |
| I0402 00:24:25.802994 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 00:24:25.803015 6134 solver.cpp:245] Train net output #16: total_confidence = 0.349599 | |
| I0402 00:24:25.803037 6134 sgd_solver.cpp:106] Iteration 159500, lr = 0.01 | |
| I0402 00:26:34.597472 6134 solver.cpp:338] Iteration 160000, Testing net (#0) | |
| I0402 00:27:04.392379 6134 solver.cpp:393] Test loss: 2.3532 | |
| I0402 00:27:04.392431 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.500362 | |
| I0402 00:27:04.392448 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.86882 | |
| I0402 00:27:04.392462 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.768038 | |
| I0402 00:27:04.392477 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.71629 (* 0.3 = 0.514886 loss) | |
| I0402 00:27:04.392493 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.45672 (* 0.3 = 0.137016 loss) | |
| I0402 00:27:04.392505 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.648438 | |
| I0402 00:27:04.392519 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.906411 | |
| I0402 00:27:04.392532 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.855749 | |
| I0402 00:27:04.392546 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.27808 (* 0.3 = 0.383425 loss) | |
| I0402 00:27:04.392561 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.344675 (* 0.3 = 0.103403 loss) | |
| I0402 00:27:04.392573 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.758616 | |
| I0402 00:27:04.392586 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.936319 | |
| I0402 00:27:04.392598 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.885792 | |
| I0402 00:27:04.392612 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.957267 (* 1 = 0.957267 loss) | |
| I0402 00:27:04.392627 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.2572 (* 1 = 0.2572 loss) | |
| I0402 00:27:04.392638 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.409 | |
| I0402 00:27:04.392649 6134 solver.cpp:406] Test net output #16: total_confidence = 0.327894 | |
| I0402 00:27:04.543519 6134 solver.cpp:229] Iteration 160000, loss = 2.72644 | |
| I0402 00:27:04.543557 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.339623 | |
| I0402 00:27:04.543575 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0402 00:27:04.543587 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.641509 | |
| I0402 00:27:04.543603 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.15681 (* 0.3 = 0.647044 loss) | |
| I0402 00:27:04.543617 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.659387 (* 0.3 = 0.197816 loss) | |
| I0402 00:27:04.543630 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.433962 | |
| I0402 00:27:04.543642 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.829545 | |
| I0402 00:27:04.543654 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.735849 | |
| I0402 00:27:04.543668 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.93481 (* 0.3 = 0.580442 loss) | |
| I0402 00:27:04.543681 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.598594 (* 0.3 = 0.179578 loss) | |
| I0402 00:27:04.543694 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.716981 | |
| I0402 00:27:04.543706 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0402 00:27:04.543717 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.867925 | |
| I0402 00:27:04.543731 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.1664 (* 1 = 1.1664 loss) | |
| I0402 00:27:04.543745 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.370269 (* 1 = 0.370269 loss) | |
| I0402 00:27:04.543758 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 00:27:04.543771 6134 solver.cpp:245] Train net output #16: total_confidence = 0.142962 | |
| I0402 00:27:04.543782 6134 sgd_solver.cpp:106] Iteration 160000, lr = 0.01 | |
| I0402 00:29:13.483693 6134 solver.cpp:229] Iteration 160500, loss = 2.71124 | |
| I0402 00:29:13.484030 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.473684 | |
| I0402 00:29:13.484052 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0402 00:29:13.484066 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.815789 | |
| I0402 00:29:13.484081 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.73958 (* 0.3 = 0.521874 loss) | |
| I0402 00:29:13.484097 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.495733 (* 0.3 = 0.14872 loss) | |
| I0402 00:29:13.484110 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0402 00:29:13.484122 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0402 00:29:13.484134 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.868421 | |
| I0402 00:29:13.484148 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.35409 (* 0.3 = 0.406226 loss) | |
| I0402 00:29:13.484163 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.406208 (* 0.3 = 0.121862 loss) | |
| I0402 00:29:13.484182 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.763158 | |
| I0402 00:29:13.484206 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0402 00:29:13.484231 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.921053 | |
| I0402 00:29:13.484257 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.917228 (* 1 = 0.917228 loss) | |
| I0402 00:29:13.484272 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.254106 (* 1 = 0.254106 loss) | |
| I0402 00:29:13.484285 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 00:29:13.484297 6134 solver.cpp:245] Train net output #16: total_confidence = 0.169519 | |
| I0402 00:29:13.484309 6134 sgd_solver.cpp:106] Iteration 160500, lr = 0.01 | |
| I0402 00:31:22.253234 6134 solver.cpp:229] Iteration 161000, loss = 2.74253 | |
| I0402 00:31:22.253348 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.357143 | |
| I0402 00:31:22.253379 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0402 00:31:22.253406 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.738095 | |
| I0402 00:31:22.253434 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.96162 (* 0.3 = 0.588485 loss) | |
| I0402 00:31:22.253464 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.608065 (* 0.3 = 0.182419 loss) | |
| I0402 00:31:22.253491 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.595238 | |
| I0402 00:31:22.253515 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0402 00:31:22.253542 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.880952 | |
| I0402 00:31:22.253568 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.39921 (* 0.3 = 0.419763 loss) | |
| I0402 00:31:22.253597 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.477203 (* 0.3 = 0.143161 loss) | |
| I0402 00:31:22.253619 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.833333 | |
| I0402 00:31:22.253640 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0402 00:31:22.253662 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.97619 | |
| I0402 00:31:22.253690 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.613882 (* 1 = 0.613882 loss) | |
| I0402 00:31:22.253715 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.19148 (* 1 = 0.19148 loss) | |
| I0402 00:31:22.253739 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 00:31:22.253759 6134 solver.cpp:245] Train net output #16: total_confidence = 0.26895 | |
| I0402 00:31:22.253782 6134 sgd_solver.cpp:106] Iteration 161000, lr = 0.01 | |
| I0402 00:33:31.123132 6134 solver.cpp:229] Iteration 161500, loss = 2.70165 | |
| I0402 00:33:31.123255 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.270833 | |
| I0402 00:33:31.123276 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.784091 | |
| I0402 00:33:31.123289 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0402 00:33:31.123306 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.03668 (* 0.3 = 0.611005 loss) | |
| I0402 00:33:31.123322 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.610148 (* 0.3 = 0.183045 loss) | |
| I0402 00:33:31.123334 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5625 | |
| I0402 00:33:31.123347 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0402 00:33:31.123359 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.791667 | |
| I0402 00:33:31.123374 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.36009 (* 0.3 = 0.408026 loss) | |
| I0402 00:33:31.123389 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.396233 (* 0.3 = 0.11887 loss) | |
| I0402 00:33:31.123401 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.895833 | |
| I0402 00:33:31.123414 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.971591 | |
| I0402 00:33:31.123426 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.979167 | |
| I0402 00:33:31.123440 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.402912 (* 1 = 0.402912 loss) | |
| I0402 00:33:31.123455 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.114451 (* 1 = 0.114451 loss) | |
| I0402 00:33:31.123466 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 00:33:31.123478 6134 solver.cpp:245] Train net output #16: total_confidence = 0.332945 | |
| I0402 00:33:31.123491 6134 sgd_solver.cpp:106] Iteration 161500, lr = 0.01 | |
| I0402 00:35:40.253204 6134 solver.cpp:229] Iteration 162000, loss = 2.73316 | |
| I0402 00:35:40.253331 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.404255 | |
| I0402 00:35:40.253352 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 00:35:40.253365 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.723404 | |
| I0402 00:35:40.253381 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.96075 (* 0.3 = 0.588226 loss) | |
| I0402 00:35:40.253396 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.608462 (* 0.3 = 0.182539 loss) | |
| I0402 00:35:40.253409 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.553191 | |
| I0402 00:35:40.253422 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 00:35:40.253434 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.787234 | |
| I0402 00:35:40.253448 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.53658 (* 0.3 = 0.460975 loss) | |
| I0402 00:35:40.253463 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.465646 (* 0.3 = 0.139694 loss) | |
| I0402 00:35:40.253475 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.787234 | |
| I0402 00:35:40.253489 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0402 00:35:40.253499 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.893617 | |
| I0402 00:35:40.253515 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.783798 (* 1 = 0.783798 loss) | |
| I0402 00:35:40.253532 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.244069 (* 1 = 0.244069 loss) | |
| I0402 00:35:40.253545 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 00:35:40.253557 6134 solver.cpp:245] Train net output #16: total_confidence = 0.449321 | |
| I0402 00:35:40.253569 6134 sgd_solver.cpp:106] Iteration 162000, lr = 0.01 | |
| I0402 00:37:49.283774 6134 solver.cpp:229] Iteration 162500, loss = 2.72992 | |
| I0402 00:37:49.284157 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.439024 | |
| I0402 00:37:49.284188 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0402 00:37:49.284212 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.682927 | |
| I0402 00:37:49.284240 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.90449 (* 0.3 = 0.571346 loss) | |
| I0402 00:37:49.284267 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.557181 (* 0.3 = 0.167154 loss) | |
| I0402 00:37:49.284291 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.560976 | |
| I0402 00:37:49.284315 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 00:37:49.284342 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.829268 | |
| I0402 00:37:49.284368 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.43619 (* 0.3 = 0.430858 loss) | |
| I0402 00:37:49.284395 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.441513 (* 0.3 = 0.132454 loss) | |
| I0402 00:37:49.284417 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.780488 | |
| I0402 00:37:49.284440 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 00:37:49.284477 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.951219 | |
| I0402 00:37:49.284507 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.659978 (* 1 = 0.659978 loss) | |
| I0402 00:37:49.284538 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.198329 (* 1 = 0.198329 loss) | |
| I0402 00:37:49.284560 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 00:37:49.284582 6134 solver.cpp:245] Train net output #16: total_confidence = 0.24824 | |
| I0402 00:37:49.284605 6134 sgd_solver.cpp:106] Iteration 162500, lr = 0.01 | |
| I0402 00:39:58.299093 6134 solver.cpp:229] Iteration 163000, loss = 2.78234 | |
| I0402 00:39:58.299213 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.326923 | |
| I0402 00:39:58.299239 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.789773 | |
| I0402 00:39:58.299252 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.538462 | |
| I0402 00:39:58.299276 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 3.29931 (* 0.3 = 0.989794 loss) | |
| I0402 00:39:58.299291 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 1.0046 (* 0.3 = 0.301379 loss) | |
| I0402 00:39:58.299304 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.442308 | |
| I0402 00:39:58.299316 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.823864 | |
| I0402 00:39:58.299330 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.557692 | |
| I0402 00:39:58.299355 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.98301 (* 0.3 = 0.894903 loss) | |
| I0402 00:39:58.299371 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.911629 (* 0.3 = 0.273489 loss) | |
| I0402 00:39:58.299383 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.596154 | |
| I0402 00:39:58.299396 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0402 00:39:58.299407 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.730769 | |
| I0402 00:39:58.299432 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 2.39998 (* 1 = 2.39998 loss) | |
| I0402 00:39:58.299463 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.721565 (* 1 = 0.721565 loss) | |
| I0402 00:39:58.299477 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 00:39:58.299490 6134 solver.cpp:245] Train net output #16: total_confidence = 0.348948 | |
| I0402 00:39:58.299502 6134 sgd_solver.cpp:106] Iteration 163000, lr = 0.01 | |
| I0402 00:42:07.330874 6134 solver.cpp:229] Iteration 163500, loss = 2.75242 | |
| I0402 00:42:07.331069 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.386364 | |
| I0402 00:42:07.331090 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0402 00:42:07.331104 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.590909 | |
| I0402 00:42:07.331122 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.04289 (* 0.3 = 0.612867 loss) | |
| I0402 00:42:07.331138 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.57878 (* 0.3 = 0.173634 loss) | |
| I0402 00:42:07.331151 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.522727 | |
| I0402 00:42:07.331163 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0402 00:42:07.331176 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0402 00:42:07.331190 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.62389 (* 0.3 = 0.487167 loss) | |
| I0402 00:42:07.331204 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.467522 (* 0.3 = 0.140257 loss) | |
| I0402 00:42:07.331218 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.681818 | |
| I0402 00:42:07.331229 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0402 00:42:07.331241 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.795455 | |
| I0402 00:42:07.331256 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.34738 (* 1 = 1.34738 loss) | |
| I0402 00:42:07.331270 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.361529 (* 1 = 0.361529 loss) | |
| I0402 00:42:07.331284 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 00:42:07.331295 6134 solver.cpp:245] Train net output #16: total_confidence = 0.211874 | |
| I0402 00:42:07.331308 6134 sgd_solver.cpp:106] Iteration 163500, lr = 0.01 | |
| I0402 00:44:16.407882 6134 solver.cpp:229] Iteration 164000, loss = 2.71823 | |
| I0402 00:44:16.407994 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.512821 | |
| I0402 00:44:16.408025 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.880682 | |
| I0402 00:44:16.408051 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.692308 | |
| I0402 00:44:16.408079 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.58876 (* 0.3 = 0.476627 loss) | |
| I0402 00:44:16.408113 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.422772 (* 0.3 = 0.126832 loss) | |
| I0402 00:44:16.408139 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.564103 | |
| I0402 00:44:16.408164 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 00:44:16.408187 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.74359 | |
| I0402 00:44:16.408213 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.41001 (* 0.3 = 0.423003 loss) | |
| I0402 00:44:16.408241 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.383359 (* 0.3 = 0.115008 loss) | |
| I0402 00:44:16.408263 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.769231 | |
| I0402 00:44:16.408287 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 00:44:16.408308 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.871795 | |
| I0402 00:44:16.408334 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.656193 (* 1 = 0.656193 loss) | |
| I0402 00:44:16.408360 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.187656 (* 1 = 0.187656 loss) | |
| I0402 00:44:16.408382 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 00:44:16.408404 6134 solver.cpp:245] Train net output #16: total_confidence = 0.406609 | |
| I0402 00:44:16.408427 6134 sgd_solver.cpp:106] Iteration 164000, lr = 0.01 | |
| I0402 00:46:25.277341 6134 solver.cpp:229] Iteration 164500, loss = 2.71883 | |
| I0402 00:46:25.277470 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.285714 | |
| I0402 00:46:25.277490 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0402 00:46:25.277504 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.591837 | |
| I0402 00:46:25.277523 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.19844 (* 0.3 = 0.659531 loss) | |
| I0402 00:46:25.277539 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.660812 (* 0.3 = 0.198244 loss) | |
| I0402 00:46:25.277551 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.530612 | |
| I0402 00:46:25.277565 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0402 00:46:25.277576 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.795918 | |
| I0402 00:46:25.277590 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.48466 (* 0.3 = 0.445397 loss) | |
| I0402 00:46:25.277604 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.504047 (* 0.3 = 0.151214 loss) | |
| I0402 00:46:25.277616 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.755102 | |
| I0402 00:46:25.277628 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0402 00:46:25.277640 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.877551 | |
| I0402 00:46:25.277654 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.01002 (* 1 = 1.01002 loss) | |
| I0402 00:46:25.277669 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.361679 (* 1 = 0.361679 loss) | |
| I0402 00:46:25.277681 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 00:46:25.277693 6134 solver.cpp:245] Train net output #16: total_confidence = 0.207051 | |
| I0402 00:46:25.277705 6134 sgd_solver.cpp:106] Iteration 164500, lr = 0.01 | |
| I0402 00:48:33.813465 6134 solver.cpp:338] Iteration 165000, Testing net (#0) | |
| I0402 00:49:03.615823 6134 solver.cpp:393] Test loss: 2.14033 | |
| I0402 00:49:03.615885 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.517247 | |
| I0402 00:49:03.615902 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.868367 | |
| I0402 00:49:03.615916 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.790535 | |
| I0402 00:49:03.615932 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.59634 (* 0.3 = 0.478902 loss) | |
| I0402 00:49:03.615947 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.4488 (* 0.3 = 0.13464 loss) | |
| I0402 00:49:03.615960 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.683265 | |
| I0402 00:49:03.615972 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.907275 | |
| I0402 00:49:03.615984 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.879017 | |
| I0402 00:49:03.615998 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.11841 (* 0.3 = 0.335522 loss) | |
| I0402 00:49:03.616013 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.326522 (* 0.3 = 0.0979567 loss) | |
| I0402 00:49:03.616024 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.777614 | |
| I0402 00:49:03.616037 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.943546 | |
| I0402 00:49:03.616049 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.896927 | |
| I0402 00:49:03.616062 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.864108 (* 1 = 0.864108 loss) | |
| I0402 00:49:03.616076 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.229198 (* 1 = 0.229198 loss) | |
| I0402 00:49:03.616088 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.45 | |
| I0402 00:49:03.616101 6134 solver.cpp:406] Test net output #16: total_confidence = 0.393103 | |
| I0402 00:49:03.767796 6134 solver.cpp:229] Iteration 165000, loss = 2.69447 | |
| I0402 00:49:03.767866 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.479167 | |
| I0402 00:49:03.767885 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 00:49:03.767899 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.854167 | |
| I0402 00:49:03.767916 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.67002 (* 0.3 = 0.501005 loss) | |
| I0402 00:49:03.767931 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.491785 (* 0.3 = 0.147535 loss) | |
| I0402 00:49:03.767945 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.666667 | |
| I0402 00:49:03.767957 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0402 00:49:03.767969 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.895833 | |
| I0402 00:49:03.767984 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.14679 (* 0.3 = 0.344038 loss) | |
| I0402 00:49:03.767998 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.35397 (* 0.3 = 0.106191 loss) | |
| I0402 00:49:03.768012 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.770833 | |
| I0402 00:49:03.768024 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 00:49:03.768036 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.979167 | |
| I0402 00:49:03.768050 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.729422 (* 1 = 0.729422 loss) | |
| I0402 00:49:03.768069 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.206234 (* 1 = 0.206234 loss) | |
| I0402 00:49:03.768081 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 00:49:03.768095 6134 solver.cpp:245] Train net output #16: total_confidence = 0.38681 | |
| I0402 00:49:03.768107 6134 sgd_solver.cpp:106] Iteration 165000, lr = 0.01 | |
| I0402 00:51:12.507491 6134 solver.cpp:229] Iteration 165500, loss = 2.7357 | |
| I0402 00:51:12.507654 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.240741 | |
| I0402 00:51:12.507678 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.755682 | |
| I0402 00:51:12.507690 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.592593 | |
| I0402 00:51:12.507706 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.19659 (* 0.3 = 0.658978 loss) | |
| I0402 00:51:12.507721 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.720428 (* 0.3 = 0.216129 loss) | |
| I0402 00:51:12.507735 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.407407 | |
| I0402 00:51:12.507746 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.8125 | |
| I0402 00:51:12.507760 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.740741 | |
| I0402 00:51:12.507773 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.70566 (* 0.3 = 0.511699 loss) | |
| I0402 00:51:12.507787 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.553573 (* 0.3 = 0.166072 loss) | |
| I0402 00:51:12.507800 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.555556 | |
| I0402 00:51:12.507812 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0402 00:51:12.507824 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.833333 | |
| I0402 00:51:12.507838 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.19723 (* 1 = 1.19723 loss) | |
| I0402 00:51:12.507851 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.384597 (* 1 = 0.384597 loss) | |
| I0402 00:51:12.507864 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0402 00:51:12.507876 6134 solver.cpp:245] Train net output #16: total_confidence = 0.1991 | |
| I0402 00:51:12.507889 6134 sgd_solver.cpp:106] Iteration 165500, lr = 0.01 | |
| I0402 00:53:21.235625 6134 solver.cpp:229] Iteration 166000, loss = 2.65113 | |
| I0402 00:53:21.235755 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.295455 | |
| I0402 00:53:21.235775 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0402 00:53:21.235788 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.590909 | |
| I0402 00:53:21.235805 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.23346 (* 0.3 = 0.670037 loss) | |
| I0402 00:53:21.235818 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.604542 (* 0.3 = 0.181363 loss) | |
| I0402 00:53:21.235831 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.431818 | |
| I0402 00:53:21.235846 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0402 00:53:21.235857 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.659091 | |
| I0402 00:53:21.235872 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.09657 (* 0.3 = 0.62897 loss) | |
| I0402 00:53:21.235898 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.576207 (* 0.3 = 0.172862 loss) | |
| I0402 00:53:21.235924 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.613636 | |
| I0402 00:53:21.235950 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0402 00:53:21.235965 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.840909 | |
| I0402 00:53:21.235980 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.29697 (* 1 = 1.29697 loss) | |
| I0402 00:53:21.235996 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.35108 (* 1 = 0.35108 loss) | |
| I0402 00:53:21.236007 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0402 00:53:21.236019 6134 solver.cpp:245] Train net output #16: total_confidence = 0.109811 | |
| I0402 00:53:21.236032 6134 sgd_solver.cpp:106] Iteration 166000, lr = 0.01 | |
| I0402 00:55:29.891840 6134 solver.cpp:229] Iteration 166500, loss = 2.6918 | |
| I0402 00:55:29.891995 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.44186 | |
| I0402 00:55:29.892015 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 00:55:29.892030 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.790698 | |
| I0402 00:55:29.892046 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.71067 (* 0.3 = 0.513201 loss) | |
| I0402 00:55:29.892060 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.557459 (* 0.3 = 0.167238 loss) | |
| I0402 00:55:29.892073 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.581395 | |
| I0402 00:55:29.892086 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0402 00:55:29.892097 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.883721 | |
| I0402 00:55:29.892112 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.23412 (* 0.3 = 0.370235 loss) | |
| I0402 00:55:29.892125 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.439038 (* 0.3 = 0.131712 loss) | |
| I0402 00:55:29.892138 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.813953 | |
| I0402 00:55:29.892150 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0402 00:55:29.892163 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.976744 | |
| I0402 00:55:29.892176 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.805024 (* 1 = 0.805024 loss) | |
| I0402 00:55:29.892190 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.226461 (* 1 = 0.226461 loss) | |
| I0402 00:55:29.892204 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 00:55:29.892216 6134 solver.cpp:245] Train net output #16: total_confidence = 0.445229 | |
| I0402 00:55:29.892240 6134 sgd_solver.cpp:106] Iteration 166500, lr = 0.01 | |
| I0402 00:57:38.587137 6134 solver.cpp:229] Iteration 167000, loss = 2.652 | |
| I0402 00:57:38.587517 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.477273 | |
| I0402 00:57:38.587538 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 00:57:38.587551 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.704545 | |
| I0402 00:57:38.587568 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.77078 (* 0.3 = 0.531235 loss) | |
| I0402 00:57:38.587584 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.492426 (* 0.3 = 0.147728 loss) | |
| I0402 00:57:38.587597 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.659091 | |
| I0402 00:57:38.587610 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0402 00:57:38.587622 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.840909 | |
| I0402 00:57:38.587635 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.20106 (* 0.3 = 0.360317 loss) | |
| I0402 00:57:38.587651 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.372101 (* 0.3 = 0.11163 loss) | |
| I0402 00:57:38.587663 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.886364 | |
| I0402 00:57:38.587676 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.971591 | |
| I0402 00:57:38.587687 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.909091 | |
| I0402 00:57:38.587702 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.65163 (* 1 = 0.65163 loss) | |
| I0402 00:57:38.587715 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.186806 (* 1 = 0.186806 loss) | |
| I0402 00:57:38.587728 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 00:57:38.587739 6134 solver.cpp:245] Train net output #16: total_confidence = 0.377882 | |
| I0402 00:57:38.587750 6134 sgd_solver.cpp:106] Iteration 167000, lr = 0.01 | |
| I0402 00:59:47.314788 6134 solver.cpp:229] Iteration 167500, loss = 2.67652 | |
| I0402 00:59:47.314893 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.42 | |
| I0402 00:59:47.314913 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 00:59:47.314925 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.7 | |
| I0402 00:59:47.314941 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.91971 (* 0.3 = 0.575912 loss) | |
| I0402 00:59:47.314957 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.584016 (* 0.3 = 0.175205 loss) | |
| I0402 00:59:47.314970 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.54 | |
| I0402 00:59:47.314983 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 00:59:47.314996 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.72 | |
| I0402 00:59:47.315008 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.95415 (* 0.3 = 0.586245 loss) | |
| I0402 00:59:47.315022 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.569254 (* 0.3 = 0.170776 loss) | |
| I0402 00:59:47.315035 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.66 | |
| I0402 00:59:47.315047 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0402 00:59:47.315058 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.86 | |
| I0402 00:59:47.315073 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.39208 (* 1 = 1.39208 loss) | |
| I0402 00:59:47.315086 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.399634 (* 1 = 0.399634 loss) | |
| I0402 00:59:47.315099 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 00:59:47.315110 6134 solver.cpp:245] Train net output #16: total_confidence = 0.431425 | |
| I0402 00:59:47.315122 6134 sgd_solver.cpp:106] Iteration 167500, lr = 0.01 | |
| I0402 01:01:56.435559 6134 solver.cpp:229] Iteration 168000, loss = 2.67403 | |
| I0402 01:01:56.435698 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.357143 | |
| I0402 01:01:56.435734 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0402 01:01:56.435751 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0402 01:01:56.435767 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.89965 (* 0.3 = 0.569895 loss) | |
| I0402 01:01:56.435782 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.569979 (* 0.3 = 0.170994 loss) | |
| I0402 01:01:56.435796 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.619048 | |
| I0402 01:01:56.435808 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 01:01:56.435820 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.785714 | |
| I0402 01:01:56.435834 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.27127 (* 0.3 = 0.38138 loss) | |
| I0402 01:01:56.435848 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.408255 (* 0.3 = 0.122477 loss) | |
| I0402 01:01:56.435861 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.785714 | |
| I0402 01:01:56.435873 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0402 01:01:56.435885 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.928571 | |
| I0402 01:01:56.435899 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.694919 (* 1 = 0.694919 loss) | |
| I0402 01:01:56.435914 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.182857 (* 1 = 0.182857 loss) | |
| I0402 01:01:56.435925 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 01:01:56.435937 6134 solver.cpp:245] Train net output #16: total_confidence = 0.354334 | |
| I0402 01:01:56.435950 6134 sgd_solver.cpp:106] Iteration 168000, lr = 0.01 | |
| I0402 01:04:05.185392 6134 solver.cpp:229] Iteration 168500, loss = 2.73343 | |
| I0402 01:04:05.185503 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.609756 | |
| I0402 01:04:05.185523 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.880682 | |
| I0402 01:04:05.185535 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.804878 | |
| I0402 01:04:05.185551 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.25059 (* 0.3 = 0.375178 loss) | |
| I0402 01:04:05.185566 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.392778 (* 0.3 = 0.117833 loss) | |
| I0402 01:04:05.185580 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.731707 | |
| I0402 01:04:05.185592 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.920455 | |
| I0402 01:04:05.185605 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 1 | |
| I0402 01:04:05.185619 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.758736 (* 0.3 = 0.227621 loss) | |
| I0402 01:04:05.185633 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.242259 (* 0.3 = 0.0726776 loss) | |
| I0402 01:04:05.185645 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.926829 | |
| I0402 01:04:05.185658 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.982955 | |
| I0402 01:04:05.185670 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0402 01:04:05.185684 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.203627 (* 1 = 0.203627 loss) | |
| I0402 01:04:05.185698 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0570466 (* 1 = 0.0570466 loss) | |
| I0402 01:04:05.185711 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 01:04:05.185724 6134 solver.cpp:245] Train net output #16: total_confidence = 0.522122 | |
| I0402 01:04:05.185735 6134 sgd_solver.cpp:106] Iteration 168500, lr = 0.01 | |
| I0402 01:06:13.947065 6134 solver.cpp:229] Iteration 169000, loss = 2.6912 | |
| I0402 01:06:13.947196 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.5 | |
| I0402 01:06:13.947214 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 01:06:13.947228 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.8 | |
| I0402 01:06:13.947244 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.72477 (* 0.3 = 0.51743 loss) | |
| I0402 01:06:13.947259 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.512311 (* 0.3 = 0.153693 loss) | |
| I0402 01:06:13.947273 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.46 | |
| I0402 01:06:13.947285 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0402 01:06:13.947298 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.86 | |
| I0402 01:06:13.947311 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.37633 (* 0.3 = 0.4129 loss) | |
| I0402 01:06:13.947335 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.407174 (* 0.3 = 0.122152 loss) | |
| I0402 01:06:13.947361 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.78 | |
| I0402 01:06:13.947388 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 01:06:13.947408 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.94 | |
| I0402 01:06:13.947423 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.603055 (* 1 = 0.603055 loss) | |
| I0402 01:06:13.947438 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.179088 (* 1 = 0.179088 loss) | |
| I0402 01:06:13.947450 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 01:06:13.947463 6134 solver.cpp:245] Train net output #16: total_confidence = 0.256889 | |
| I0402 01:06:13.947474 6134 sgd_solver.cpp:106] Iteration 169000, lr = 0.01 | |
| I0402 01:08:22.864560 6134 solver.cpp:229] Iteration 169500, loss = 2.67771 | |
| I0402 01:08:22.864903 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.511111 | |
| I0402 01:08:22.864924 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 01:08:22.864938 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.733333 | |
| I0402 01:08:22.864954 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.6849 (* 0.3 = 0.505469 loss) | |
| I0402 01:08:22.864969 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.487633 (* 0.3 = 0.14629 loss) | |
| I0402 01:08:22.864984 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.577778 | |
| I0402 01:08:22.864996 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 01:08:22.865008 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.777778 | |
| I0402 01:08:22.865022 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.38093 (* 0.3 = 0.414278 loss) | |
| I0402 01:08:22.865036 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.446387 (* 0.3 = 0.133916 loss) | |
| I0402 01:08:22.865068 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.777778 | |
| I0402 01:08:22.865083 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0402 01:08:22.865095 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.933333 | |
| I0402 01:08:22.865118 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.751128 (* 1 = 0.751128 loss) | |
| I0402 01:08:22.865133 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.210804 (* 1 = 0.210804 loss) | |
| I0402 01:08:22.865144 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 01:08:22.865157 6134 solver.cpp:245] Train net output #16: total_confidence = 0.459429 | |
| I0402 01:08:22.865177 6134 sgd_solver.cpp:106] Iteration 169500, lr = 0.01 | |
| I0402 01:10:31.297937 6134 solver.cpp:338] Iteration 170000, Testing net (#0) | |
| I0402 01:11:01.050061 6134 solver.cpp:393] Test loss: 2.33927 | |
| I0402 01:11:01.050106 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.495465 | |
| I0402 01:11:01.050122 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.863548 | |
| I0402 01:11:01.050134 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.765233 | |
| I0402 01:11:01.050150 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.70155 (* 0.3 = 0.510466 loss) | |
| I0402 01:11:01.050165 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.465642 (* 0.3 = 0.139693 loss) | |
| I0402 01:11:01.050178 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.671708 | |
| I0402 01:11:01.050189 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.911547 | |
| I0402 01:11:01.050201 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.861488 | |
| I0402 01:11:01.050215 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.19372 (* 0.3 = 0.358115 loss) | |
| I0402 01:11:01.050228 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.323611 (* 0.3 = 0.0970834 loss) | |
| I0402 01:11:01.050241 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.756258 | |
| I0402 01:11:01.050253 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.935682 | |
| I0402 01:11:01.050264 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.876468 | |
| I0402 01:11:01.050277 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.972676 (* 1 = 0.972676 loss) | |
| I0402 01:11:01.050292 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.261235 (* 1 = 0.261235 loss) | |
| I0402 01:11:01.050303 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.408 | |
| I0402 01:11:01.050315 6134 solver.cpp:406] Test net output #16: total_confidence = 0.387686 | |
| I0402 01:11:01.201478 6134 solver.cpp:229] Iteration 170000, loss = 2.63396 | |
| I0402 01:11:01.201517 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.408163 | |
| I0402 01:11:01.201535 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0402 01:11:01.201548 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.693878 | |
| I0402 01:11:01.201563 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.90846 (* 0.3 = 0.572537 loss) | |
| I0402 01:11:01.201578 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.577989 (* 0.3 = 0.173397 loss) | |
| I0402 01:11:01.201591 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.591837 | |
| I0402 01:11:01.201604 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0402 01:11:01.201617 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.816327 | |
| I0402 01:11:01.201629 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.32515 (* 0.3 = 0.397546 loss) | |
| I0402 01:11:01.201643 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.382276 (* 0.3 = 0.114683 loss) | |
| I0402 01:11:01.201655 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.795918 | |
| I0402 01:11:01.201668 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 01:11:01.201679 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.938776 | |
| I0402 01:11:01.201694 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.574034 (* 1 = 0.574034 loss) | |
| I0402 01:11:01.201707 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.184977 (* 1 = 0.184977 loss) | |
| I0402 01:11:01.201719 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 01:11:01.201731 6134 solver.cpp:245] Train net output #16: total_confidence = 0.245015 | |
| I0402 01:11:01.201743 6134 sgd_solver.cpp:106] Iteration 170000, lr = 0.01 | |
| I0402 01:13:09.774139 6134 solver.cpp:229] Iteration 170500, loss = 2.66633 | |
| I0402 01:13:09.774281 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.410256 | |
| I0402 01:13:09.774302 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0402 01:13:09.774314 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.846154 | |
| I0402 01:13:09.774330 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.73075 (* 0.3 = 0.519225 loss) | |
| I0402 01:13:09.774345 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.481398 (* 0.3 = 0.144419 loss) | |
| I0402 01:13:09.774358 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.538462 | |
| I0402 01:13:09.774370 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 01:13:09.774384 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.794872 | |
| I0402 01:13:09.774397 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.31322 (* 0.3 = 0.393967 loss) | |
| I0402 01:13:09.774412 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.35273 (* 0.3 = 0.105819 loss) | |
| I0402 01:13:09.774425 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.820513 | |
| I0402 01:13:09.774438 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0402 01:13:09.774461 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.923077 | |
| I0402 01:13:09.774489 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.510509 (* 1 = 0.510509 loss) | |
| I0402 01:13:09.774521 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.153084 (* 1 = 0.153084 loss) | |
| I0402 01:13:09.774544 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 01:13:09.774559 6134 solver.cpp:245] Train net output #16: total_confidence = 0.353086 | |
| I0402 01:13:09.774570 6134 sgd_solver.cpp:106] Iteration 170500, lr = 0.01 | |
| I0402 01:15:18.487587 6134 solver.cpp:229] Iteration 171000, loss = 2.68242 | |
| I0402 01:15:18.487710 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.44186 | |
| I0402 01:15:18.487730 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0402 01:15:18.487743 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.72093 | |
| I0402 01:15:18.487759 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.59987 (* 0.3 = 0.479961 loss) | |
| I0402 01:15:18.487776 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.454242 (* 0.3 = 0.136273 loss) | |
| I0402 01:15:18.487788 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.581395 | |
| I0402 01:15:18.487800 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 01:15:18.487812 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.813953 | |
| I0402 01:15:18.487828 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.28674 (* 0.3 = 0.386023 loss) | |
| I0402 01:15:18.487843 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.382352 (* 0.3 = 0.114705 loss) | |
| I0402 01:15:18.487855 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.906977 | |
| I0402 01:15:18.487867 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.977273 | |
| I0402 01:15:18.487879 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.930233 | |
| I0402 01:15:18.487895 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.466982 (* 1 = 0.466982 loss) | |
| I0402 01:15:18.487918 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.126044 (* 1 = 0.126044 loss) | |
| I0402 01:15:18.487942 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 01:15:18.487958 6134 solver.cpp:245] Train net output #16: total_confidence = 0.355569 | |
| I0402 01:15:18.487970 6134 sgd_solver.cpp:106] Iteration 171000, lr = 0.01 | |
| I0402 01:17:27.137944 6134 solver.cpp:229] Iteration 171500, loss = 2.64813 | |
| I0402 01:17:27.138284 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.391304 | |
| I0402 01:17:27.138305 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 01:17:27.138319 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.76087 | |
| I0402 01:17:27.138335 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.93424 (* 0.3 = 0.580273 loss) | |
| I0402 01:17:27.138350 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.541187 (* 0.3 = 0.162356 loss) | |
| I0402 01:17:27.138362 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.565217 | |
| I0402 01:17:27.138375 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 01:17:27.138387 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76087 | |
| I0402 01:17:27.138401 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.5211 (* 0.3 = 0.45633 loss) | |
| I0402 01:17:27.138416 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.450932 (* 0.3 = 0.13528 loss) | |
| I0402 01:17:27.138427 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.630435 | |
| I0402 01:17:27.138439 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0402 01:17:27.138452 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.76087 | |
| I0402 01:17:27.138464 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.31045 (* 1 = 1.31045 loss) | |
| I0402 01:17:27.138478 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.390524 (* 1 = 0.390524 loss) | |
| I0402 01:17:27.138491 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 01:17:27.138504 6134 solver.cpp:245] Train net output #16: total_confidence = 0.363538 | |
| I0402 01:17:27.138515 6134 sgd_solver.cpp:106] Iteration 171500, lr = 0.01 | |
| I0402 01:19:35.834512 6134 solver.cpp:229] Iteration 172000, loss = 2.66084 | |
| I0402 01:19:35.834627 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.386364 | |
| I0402 01:19:35.834647 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0402 01:19:35.834661 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.681818 | |
| I0402 01:19:35.834677 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.17288 (* 0.3 = 0.651863 loss) | |
| I0402 01:19:35.834692 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.573352 (* 0.3 = 0.172006 loss) | |
| I0402 01:19:35.834705 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.545455 | |
| I0402 01:19:35.834718 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 01:19:35.834730 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0402 01:19:35.834744 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.8838 (* 0.3 = 0.565141 loss) | |
| I0402 01:19:35.834758 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.498962 (* 0.3 = 0.149689 loss) | |
| I0402 01:19:35.834771 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.840909 | |
| I0402 01:19:35.834784 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0402 01:19:35.834795 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.909091 | |
| I0402 01:19:35.834808 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.996922 (* 1 = 0.996922 loss) | |
| I0402 01:19:35.834822 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.266197 (* 1 = 0.266197 loss) | |
| I0402 01:19:35.834835 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 01:19:35.834846 6134 solver.cpp:245] Train net output #16: total_confidence = 0.23151 | |
| I0402 01:19:35.834858 6134 sgd_solver.cpp:106] Iteration 172000, lr = 0.01 | |
| I0402 01:21:44.557588 6134 solver.cpp:229] Iteration 172500, loss = 2.61568 | |
| I0402 01:21:44.557715 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.489796 | |
| I0402 01:21:44.557736 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0402 01:21:44.557749 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.714286 | |
| I0402 01:21:44.557765 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.80808 (* 0.3 = 0.542423 loss) | |
| I0402 01:21:44.557780 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.562917 (* 0.3 = 0.168875 loss) | |
| I0402 01:21:44.557792 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.673469 | |
| I0402 01:21:44.557806 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0402 01:21:44.557817 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.877551 | |
| I0402 01:21:44.557832 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.10216 (* 0.3 = 0.330648 loss) | |
| I0402 01:21:44.557845 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.342872 (* 0.3 = 0.102862 loss) | |
| I0402 01:21:44.557858 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.938776 | |
| I0402 01:21:44.557870 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.977273 | |
| I0402 01:21:44.557881 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0402 01:21:44.557896 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.19174 (* 1 = 0.19174 loss) | |
| I0402 01:21:44.557910 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0734843 (* 1 = 0.0734843 loss) | |
| I0402 01:21:44.557922 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0402 01:21:44.557934 6134 solver.cpp:245] Train net output #16: total_confidence = 0.508812 | |
| I0402 01:21:44.557946 6134 sgd_solver.cpp:106] Iteration 172500, lr = 0.01 | |
| I0402 01:23:53.248498 6134 solver.cpp:229] Iteration 173000, loss = 2.65029 | |
| I0402 01:23:53.248606 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.458333 | |
| I0402 01:23:53.248626 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 01:23:53.248639 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.583333 | |
| I0402 01:23:53.248656 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.04503 (* 0.3 = 0.613508 loss) | |
| I0402 01:23:53.248670 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.622748 (* 0.3 = 0.186825 loss) | |
| I0402 01:23:53.248683 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.458333 | |
| I0402 01:23:53.248697 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0402 01:23:53.248708 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.770833 | |
| I0402 01:23:53.248723 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.70118 (* 0.3 = 0.510355 loss) | |
| I0402 01:23:53.248736 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.488249 (* 0.3 = 0.146475 loss) | |
| I0402 01:23:53.248749 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.770833 | |
| I0402 01:23:53.248760 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0402 01:23:53.248772 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9375 | |
| I0402 01:23:53.248787 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.901333 (* 1 = 0.901333 loss) | |
| I0402 01:23:53.248801 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.28303 (* 1 = 0.28303 loss) | |
| I0402 01:23:53.248813 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 01:23:53.248826 6134 solver.cpp:245] Train net output #16: total_confidence = 0.24481 | |
| I0402 01:23:53.248838 6134 sgd_solver.cpp:106] Iteration 173000, lr = 0.01 | |
| I0402 01:26:01.941668 6134 solver.cpp:229] Iteration 173500, loss = 2.63202 | |
| I0402 01:26:01.941807 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.439024 | |
| I0402 01:26:01.941826 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 01:26:01.941839 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.658537 | |
| I0402 01:26:01.941855 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.07344 (* 0.3 = 0.622032 loss) | |
| I0402 01:26:01.941870 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.554584 (* 0.3 = 0.166375 loss) | |
| I0402 01:26:01.941884 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.536585 | |
| I0402 01:26:01.941895 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0402 01:26:01.941908 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.756098 | |
| I0402 01:26:01.941922 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.63381 (* 0.3 = 0.490144 loss) | |
| I0402 01:26:01.941936 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.474302 (* 0.3 = 0.142291 loss) | |
| I0402 01:26:01.941948 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.634146 | |
| I0402 01:26:01.941961 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0402 01:26:01.941972 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.829268 | |
| I0402 01:26:01.941987 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.06787 (* 1 = 1.06787 loss) | |
| I0402 01:26:01.942000 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.278262 (* 1 = 0.278262 loss) | |
| I0402 01:26:01.942013 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 01:26:01.942025 6134 solver.cpp:245] Train net output #16: total_confidence = 0.214706 | |
| I0402 01:26:01.942037 6134 sgd_solver.cpp:106] Iteration 173500, lr = 0.01 | |
| I0402 01:28:10.723630 6134 solver.cpp:229] Iteration 174000, loss = 2.6685 | |
| I0402 01:28:10.723997 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0402 01:28:10.724019 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0402 01:28:10.724031 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.64 | |
| I0402 01:28:10.724047 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.00592 (* 0.3 = 0.601775 loss) | |
| I0402 01:28:10.724062 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.627005 (* 0.3 = 0.188101 loss) | |
| I0402 01:28:10.724076 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.54 | |
| I0402 01:28:10.724087 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0402 01:28:10.724099 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.9 | |
| I0402 01:28:10.724114 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.3027 (* 0.3 = 0.390811 loss) | |
| I0402 01:28:10.724128 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.409884 (* 0.3 = 0.122965 loss) | |
| I0402 01:28:10.724141 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.76 | |
| I0402 01:28:10.724153 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0402 01:28:10.724165 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.92 | |
| I0402 01:28:10.724179 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.792803 (* 1 = 0.792803 loss) | |
| I0402 01:28:10.724192 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.248389 (* 1 = 0.248389 loss) | |
| I0402 01:28:10.724205 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 01:28:10.724216 6134 solver.cpp:245] Train net output #16: total_confidence = 0.367279 | |
| I0402 01:28:10.724228 6134 sgd_solver.cpp:106] Iteration 174000, lr = 0.01 | |
| I0402 01:30:19.327945 6134 solver.cpp:229] Iteration 174500, loss = 2.69312 | |
| I0402 01:30:19.328066 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.55102 | |
| I0402 01:30:19.328085 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.869318 | |
| I0402 01:30:19.328099 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.734694 | |
| I0402 01:30:19.328114 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.57184 (* 0.3 = 0.471553 loss) | |
| I0402 01:30:19.328130 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.483671 (* 0.3 = 0.145101 loss) | |
| I0402 01:30:19.328142 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.612245 | |
| I0402 01:30:19.328155 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 01:30:19.328166 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.918367 | |
| I0402 01:30:19.328181 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.19949 (* 0.3 = 0.359846 loss) | |
| I0402 01:30:19.328194 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.367443 (* 0.3 = 0.110233 loss) | |
| I0402 01:30:19.328207 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.836735 | |
| I0402 01:30:19.328218 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0402 01:30:19.328230 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.959184 | |
| I0402 01:30:19.328244 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.587251 (* 1 = 0.587251 loss) | |
| I0402 01:30:19.328258 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.170106 (* 1 = 0.170106 loss) | |
| I0402 01:30:19.328270 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 01:30:19.328282 6134 solver.cpp:245] Train net output #16: total_confidence = 0.456811 | |
| I0402 01:30:19.328294 6134 sgd_solver.cpp:106] Iteration 174500, lr = 0.01 | |
| I0402 01:32:27.870632 6134 solver.cpp:338] Iteration 175000, Testing net (#0) | |
| I0402 01:32:57.627775 6134 solver.cpp:393] Test loss: 2.23417 | |
| I0402 01:32:57.627821 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.500389 | |
| I0402 01:32:57.627838 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.864503 | |
| I0402 01:32:57.627851 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.778926 | |
| I0402 01:32:57.627868 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.68566 (* 0.3 = 0.505699 loss) | |
| I0402 01:32:57.627882 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.463202 (* 0.3 = 0.138961 loss) | |
| I0402 01:32:57.627895 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.695283 | |
| I0402 01:32:57.627907 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.911957 | |
| I0402 01:32:57.627919 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.874038 | |
| I0402 01:32:57.627933 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.10492 (* 0.3 = 0.331476 loss) | |
| I0402 01:32:57.627948 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.316394 (* 0.3 = 0.0949183 loss) | |
| I0402 01:32:57.627960 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.775739 | |
| I0402 01:32:57.627972 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.940091 | |
| I0402 01:32:57.627985 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.886399 | |
| I0402 01:32:57.627997 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.916199 (* 1 = 0.916199 loss) | |
| I0402 01:32:57.628012 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.246915 (* 1 = 0.246915 loss) | |
| I0402 01:32:57.628024 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.429 | |
| I0402 01:32:57.628036 6134 solver.cpp:406] Test net output #16: total_confidence = 0.406118 | |
| I0402 01:32:57.779409 6134 solver.cpp:229] Iteration 175000, loss = 2.63505 | |
| I0402 01:32:57.779453 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.435897 | |
| I0402 01:32:57.779471 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0402 01:32:57.779484 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.692308 | |
| I0402 01:32:57.779500 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.90701 (* 0.3 = 0.572102 loss) | |
| I0402 01:32:57.779520 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.59884 (* 0.3 = 0.179652 loss) | |
| I0402 01:32:57.779532 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.641026 | |
| I0402 01:32:57.779544 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 01:32:57.779556 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.794872 | |
| I0402 01:32:57.779570 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.40231 (* 0.3 = 0.420693 loss) | |
| I0402 01:32:57.779585 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.428874 (* 0.3 = 0.128662 loss) | |
| I0402 01:32:57.779597 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.897436 | |
| I0402 01:32:57.779609 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0402 01:32:57.779621 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.923077 | |
| I0402 01:32:57.779635 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.635092 (* 1 = 0.635092 loss) | |
| I0402 01:32:57.779649 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.177076 (* 1 = 0.177076 loss) | |
| I0402 01:32:57.779661 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 01:32:57.779675 6134 solver.cpp:245] Train net output #16: total_confidence = 0.356422 | |
| I0402 01:32:57.779686 6134 sgd_solver.cpp:106] Iteration 175000, lr = 0.01 | |
| I0402 01:35:06.563580 6134 solver.cpp:229] Iteration 175500, loss = 2.5722 | |
| I0402 01:35:06.563787 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.44186 | |
| I0402 01:35:06.563809 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 01:35:06.563822 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.627907 | |
| I0402 01:35:06.563840 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.06084 (* 0.3 = 0.618251 loss) | |
| I0402 01:35:06.563856 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.544697 (* 0.3 = 0.163409 loss) | |
| I0402 01:35:06.563869 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.534884 | |
| I0402 01:35:06.563881 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 01:35:06.563894 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.697674 | |
| I0402 01:35:06.563907 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.75201 (* 0.3 = 0.525603 loss) | |
| I0402 01:35:06.563922 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.457938 (* 0.3 = 0.137381 loss) | |
| I0402 01:35:06.563935 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.72093 | |
| I0402 01:35:06.563957 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0402 01:35:06.563982 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.837209 | |
| I0402 01:35:06.563999 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.15855 (* 1 = 1.15855 loss) | |
| I0402 01:35:06.564014 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.290421 (* 1 = 0.290421 loss) | |
| I0402 01:35:06.564026 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 01:35:06.564038 6134 solver.cpp:245] Train net output #16: total_confidence = 0.395635 | |
| I0402 01:35:06.564051 6134 sgd_solver.cpp:106] Iteration 175500, lr = 0.01 | |
| I0402 01:37:15.404570 6134 solver.cpp:229] Iteration 176000, loss = 2.67871 | |
| I0402 01:37:15.404901 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.477273 | |
| I0402 01:37:15.404922 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 01:37:15.404934 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.636364 | |
| I0402 01:37:15.404952 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.75647 (* 0.3 = 0.52694 loss) | |
| I0402 01:37:15.404966 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.503518 (* 0.3 = 0.151056 loss) | |
| I0402 01:37:15.404979 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.636364 | |
| I0402 01:37:15.404991 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0402 01:37:15.405004 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.840909 | |
| I0402 01:37:15.405017 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.20387 (* 0.3 = 0.361161 loss) | |
| I0402 01:37:15.405031 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.35857 (* 0.3 = 0.107571 loss) | |
| I0402 01:37:15.405058 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.818182 | |
| I0402 01:37:15.405074 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0402 01:37:15.405086 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.909091 | |
| I0402 01:37:15.405100 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.524005 (* 1 = 0.524005 loss) | |
| I0402 01:37:15.405115 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.157898 (* 1 = 0.157898 loss) | |
| I0402 01:37:15.405128 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 01:37:15.405139 6134 solver.cpp:245] Train net output #16: total_confidence = 0.318845 | |
| I0402 01:37:15.405151 6134 sgd_solver.cpp:106] Iteration 176000, lr = 0.01 | |
| I0402 01:39:23.989951 6134 solver.cpp:229] Iteration 176500, loss = 2.60251 | |
| I0402 01:39:23.990061 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431818 | |
| I0402 01:39:23.990080 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0402 01:39:23.990093 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.636364 | |
| I0402 01:39:23.990110 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.96974 (* 0.3 = 0.590922 loss) | |
| I0402 01:39:23.990125 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.636917 (* 0.3 = 0.191075 loss) | |
| I0402 01:39:23.990139 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0402 01:39:23.990150 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0402 01:39:23.990162 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.727273 | |
| I0402 01:39:23.990176 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.5999 (* 0.3 = 0.479969 loss) | |
| I0402 01:39:23.990191 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.540641 (* 0.3 = 0.162192 loss) | |
| I0402 01:39:23.990203 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.681818 | |
| I0402 01:39:23.990216 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.909091 | |
| I0402 01:39:23.990228 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.75 | |
| I0402 01:39:23.990242 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.15296 (* 1 = 1.15296 loss) | |
| I0402 01:39:23.990255 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.331926 (* 1 = 0.331926 loss) | |
| I0402 01:39:23.990267 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 01:39:23.990279 6134 solver.cpp:245] Train net output #16: total_confidence = 0.375515 | |
| I0402 01:39:23.990291 6134 sgd_solver.cpp:106] Iteration 176500, lr = 0.01 | |
| I0402 01:41:32.713969 6134 solver.cpp:229] Iteration 177000, loss = 2.64763 | |
| I0402 01:41:32.714102 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.298246 | |
| I0402 01:41:32.714123 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.772727 | |
| I0402 01:41:32.714136 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54386 | |
| I0402 01:41:32.714153 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.40601 (* 0.3 = 0.721804 loss) | |
| I0402 01:41:32.714167 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.793753 (* 0.3 = 0.238126 loss) | |
| I0402 01:41:32.714180 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.315789 | |
| I0402 01:41:32.714192 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.778409 | |
| I0402 01:41:32.714205 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.561404 | |
| I0402 01:41:32.714218 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.70011 (* 0.3 = 0.810033 loss) | |
| I0402 01:41:32.714233 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.877082 (* 0.3 = 0.263125 loss) | |
| I0402 01:41:32.714246 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.578947 | |
| I0402 01:41:32.714258 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.863636 | |
| I0402 01:41:32.714270 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.736842 | |
| I0402 01:41:32.714284 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.5262 (* 1 = 1.5262 loss) | |
| I0402 01:41:32.714298 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.506094 (* 1 = 0.506094 loss) | |
| I0402 01:41:32.714310 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 01:41:32.714323 6134 solver.cpp:245] Train net output #16: total_confidence = 0.194682 | |
| I0402 01:41:32.714334 6134 sgd_solver.cpp:106] Iteration 177000, lr = 0.01 | |
| I0402 01:43:41.294534 6134 solver.cpp:229] Iteration 177500, loss = 2.53935 | |
| I0402 01:43:41.294656 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.45098 | |
| I0402 01:43:41.294685 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0402 01:43:41.294709 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0402 01:43:41.294740 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.01959 (* 0.3 = 0.605878 loss) | |
| I0402 01:43:41.294766 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.658582 (* 0.3 = 0.197575 loss) | |
| I0402 01:43:41.294788 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.431373 | |
| I0402 01:43:41.294813 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.801136 | |
| I0402 01:43:41.294836 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.686275 | |
| I0402 01:43:41.294862 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.90285 (* 0.3 = 0.570854 loss) | |
| I0402 01:43:41.294888 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.629397 (* 0.3 = 0.188819 loss) | |
| I0402 01:43:41.294909 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.588235 | |
| I0402 01:43:41.294932 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.875 | |
| I0402 01:43:41.294955 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.745098 | |
| I0402 01:43:41.294981 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.26468 (* 1 = 1.26468 loss) | |
| I0402 01:43:41.295006 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.392614 (* 1 = 0.392614 loss) | |
| I0402 01:43:41.295027 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 01:43:41.295048 6134 solver.cpp:245] Train net output #16: total_confidence = 0.356158 | |
| I0402 01:43:41.295069 6134 sgd_solver.cpp:106] Iteration 177500, lr = 0.01 | |
| I0402 01:45:50.085403 6134 solver.cpp:229] Iteration 178000, loss = 2.59907 | |
| I0402 01:45:50.085553 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.333333 | |
| I0402 01:45:50.085578 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0402 01:45:50.085593 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.644444 | |
| I0402 01:45:50.085609 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.06404 (* 0.3 = 0.619212 loss) | |
| I0402 01:45:50.085625 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.597329 (* 0.3 = 0.179199 loss) | |
| I0402 01:45:50.085638 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.4 | |
| I0402 01:45:50.085650 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0402 01:45:50.085662 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.777778 | |
| I0402 01:45:50.085676 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.54163 (* 0.3 = 0.46249 loss) | |
| I0402 01:45:50.085690 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.468385 (* 0.3 = 0.140515 loss) | |
| I0402 01:45:50.085702 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.844444 | |
| I0402 01:45:50.085714 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 01:45:50.085726 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.955556 | |
| I0402 01:45:50.085741 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.594494 (* 1 = 0.594494 loss) | |
| I0402 01:45:50.085754 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.199818 (* 1 = 0.199818 loss) | |
| I0402 01:45:50.085767 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 01:45:50.085778 6134 solver.cpp:245] Train net output #16: total_confidence = 0.159295 | |
| I0402 01:45:50.085791 6134 sgd_solver.cpp:106] Iteration 178000, lr = 0.01 | |
| I0402 01:47:58.812901 6134 solver.cpp:229] Iteration 178500, loss = 2.58668 | |
| I0402 01:47:58.813230 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.386364 | |
| I0402 01:47:58.813249 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 01:47:58.813261 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.659091 | |
| I0402 01:47:58.813277 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.10511 (* 0.3 = 0.631533 loss) | |
| I0402 01:47:58.813293 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.571875 (* 0.3 = 0.171563 loss) | |
| I0402 01:47:58.813307 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.545455 | |
| I0402 01:47:58.813318 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 01:47:58.813330 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.727273 | |
| I0402 01:47:58.813344 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.65638 (* 0.3 = 0.496913 loss) | |
| I0402 01:47:58.813359 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.479764 (* 0.3 = 0.143929 loss) | |
| I0402 01:47:58.813370 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.659091 | |
| I0402 01:47:58.813382 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0402 01:47:58.813395 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.795455 | |
| I0402 01:47:58.813408 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.19346 (* 1 = 1.19346 loss) | |
| I0402 01:47:58.813422 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.311541 (* 1 = 0.311541 loss) | |
| I0402 01:47:58.813434 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 01:47:58.813447 6134 solver.cpp:245] Train net output #16: total_confidence = 0.27072 | |
| I0402 01:47:58.813459 6134 sgd_solver.cpp:106] Iteration 178500, lr = 0.01 | |
| I0402 01:50:07.642112 6134 solver.cpp:229] Iteration 179000, loss = 2.55088 | |
| I0402 01:50:07.642271 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.404255 | |
| I0402 01:50:07.642302 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 01:50:07.642314 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.617021 | |
| I0402 01:50:07.642330 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.3504 (* 0.3 = 0.705119 loss) | |
| I0402 01:50:07.642354 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.660249 (* 0.3 = 0.198075 loss) | |
| I0402 01:50:07.642367 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.468085 | |
| I0402 01:50:07.642379 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0402 01:50:07.642391 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.680851 | |
| I0402 01:50:07.642405 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.84605 (* 0.3 = 0.553815 loss) | |
| I0402 01:50:07.642426 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.559483 (* 0.3 = 0.167845 loss) | |
| I0402 01:50:07.642439 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.531915 | |
| I0402 01:50:07.642452 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.869318 | |
| I0402 01:50:07.642463 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.787234 | |
| I0402 01:50:07.642477 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.42939 (* 1 = 1.42939 loss) | |
| I0402 01:50:07.642490 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.412995 (* 1 = 0.412995 loss) | |
| I0402 01:50:07.642503 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 01:50:07.642518 6134 solver.cpp:245] Train net output #16: total_confidence = 0.306105 | |
| I0402 01:50:07.642531 6134 sgd_solver.cpp:106] Iteration 179000, lr = 0.01 | |
| I0402 01:52:16.588376 6134 solver.cpp:229] Iteration 179500, loss = 2.63031 | |
| I0402 01:52:16.588479 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.416667 | |
| I0402 01:52:16.588497 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0402 01:52:16.588510 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.770833 | |
| I0402 01:52:16.588527 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.54177 (* 0.3 = 0.462532 loss) | |
| I0402 01:52:16.588542 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.441461 (* 0.3 = 0.132438 loss) | |
| I0402 01:52:16.588554 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.604167 | |
| I0402 01:52:16.588567 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 01:52:16.588579 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.958333 | |
| I0402 01:52:16.588593 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.02056 (* 0.3 = 0.306167 loss) | |
| I0402 01:52:16.588608 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.318633 (* 0.3 = 0.0955899 loss) | |
| I0402 01:52:16.588620 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.854167 | |
| I0402 01:52:16.588632 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0402 01:52:16.588645 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.979167 | |
| I0402 01:52:16.588660 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.404942 (* 1 = 0.404942 loss) | |
| I0402 01:52:16.588673 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.14464 (* 1 = 0.14464 loss) | |
| I0402 01:52:16.588685 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 01:52:16.588697 6134 solver.cpp:245] Train net output #16: total_confidence = 0.352587 | |
| I0402 01:52:16.588709 6134 sgd_solver.cpp:106] Iteration 179500, lr = 0.01 | |
| I0402 01:54:25.125720 6134 solver.cpp:338] Iteration 180000, Testing net (#0) | |
| I0402 01:54:54.896297 6134 solver.cpp:393] Test loss: 2.23829 | |
| I0402 01:54:54.896347 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.523174 | |
| I0402 01:54:54.896363 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.874413 | |
| I0402 01:54:54.896375 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.796818 | |
| I0402 01:54:54.896391 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.61872 (* 0.3 = 0.485616 loss) | |
| I0402 01:54:54.896406 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.433549 (* 0.3 = 0.130065 loss) | |
| I0402 01:54:54.896420 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.697814 | |
| I0402 01:54:54.896431 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.920183 | |
| I0402 01:54:54.896443 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.879918 | |
| I0402 01:54:54.896457 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.1293 (* 0.3 = 0.33879 loss) | |
| I0402 01:54:54.896471 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.306052 (* 0.3 = 0.0918155 loss) | |
| I0402 01:54:54.896483 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.77584 | |
| I0402 01:54:54.896497 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.940818 | |
| I0402 01:54:54.896507 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.892329 | |
| I0402 01:54:54.896524 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.93923 (* 1 = 0.93923 loss) | |
| I0402 01:54:54.896538 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.252776 (* 1 = 0.252776 loss) | |
| I0402 01:54:54.896550 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.442 | |
| I0402 01:54:54.896563 6134 solver.cpp:406] Test net output #16: total_confidence = 0.393443 | |
| I0402 01:54:55.047623 6134 solver.cpp:229] Iteration 180000, loss = 2.61947 | |
| I0402 01:54:55.047665 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.533333 | |
| I0402 01:54:55.047683 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.875 | |
| I0402 01:54:55.047696 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.8 | |
| I0402 01:54:55.047713 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.54213 (* 0.3 = 0.46264 loss) | |
| I0402 01:54:55.047726 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.419607 (* 0.3 = 0.125882 loss) | |
| I0402 01:54:55.047739 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.622222 | |
| I0402 01:54:55.047751 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.903409 | |
| I0402 01:54:55.047763 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.866667 | |
| I0402 01:54:55.047777 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.40304 (* 0.3 = 0.420913 loss) | |
| I0402 01:54:55.047792 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.377345 (* 0.3 = 0.113204 loss) | |
| I0402 01:54:55.047804 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.844444 | |
| I0402 01:54:55.047816 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0402 01:54:55.047828 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.933333 | |
| I0402 01:54:55.047842 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.653125 (* 1 = 0.653125 loss) | |
| I0402 01:54:55.047857 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.181416 (* 1 = 0.181416 loss) | |
| I0402 01:54:55.047868 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 01:54:55.047880 6134 solver.cpp:245] Train net output #16: total_confidence = 0.517053 | |
| I0402 01:54:55.047893 6134 sgd_solver.cpp:106] Iteration 180000, lr = 0.01 | |
| I0402 01:57:03.741307 6134 solver.cpp:229] Iteration 180500, loss = 2.60978 | |
| I0402 01:57:03.741750 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.478261 | |
| I0402 01:57:03.741771 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 01:57:03.741786 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.804348 | |
| I0402 01:57:03.741802 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.57353 (* 0.3 = 0.472059 loss) | |
| I0402 01:57:03.741818 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.506671 (* 0.3 = 0.152001 loss) | |
| I0402 01:57:03.741832 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.673913 | |
| I0402 01:57:03.741843 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0402 01:57:03.741857 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.847826 | |
| I0402 01:57:03.741870 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.14392 (* 0.3 = 0.343177 loss) | |
| I0402 01:57:03.741884 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.353633 (* 0.3 = 0.10609 loss) | |
| I0402 01:57:03.741897 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.73913 | |
| I0402 01:57:03.741910 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0402 01:57:03.741922 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.891304 | |
| I0402 01:57:03.741936 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.78782 (* 1 = 0.78782 loss) | |
| I0402 01:57:03.741950 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.213183 (* 1 = 0.213183 loss) | |
| I0402 01:57:03.741963 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 01:57:03.741976 6134 solver.cpp:245] Train net output #16: total_confidence = 0.351851 | |
| I0402 01:57:03.741988 6134 sgd_solver.cpp:106] Iteration 180500, lr = 0.01 | |
| I0402 01:59:12.529578 6134 solver.cpp:229] Iteration 181000, loss = 2.60925 | |
| I0402 01:59:12.529690 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431818 | |
| I0402 01:59:12.529711 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 01:59:12.529723 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.681818 | |
| I0402 01:59:12.529739 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.74141 (* 0.3 = 0.522423 loss) | |
| I0402 01:59:12.529754 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.468585 (* 0.3 = 0.140576 loss) | |
| I0402 01:59:12.529767 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.454545 | |
| I0402 01:59:12.529778 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0402 01:59:12.529791 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.75 | |
| I0402 01:59:12.529805 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.4618 (* 0.3 = 0.438541 loss) | |
| I0402 01:59:12.529819 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.413547 (* 0.3 = 0.124064 loss) | |
| I0402 01:59:12.529832 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0402 01:59:12.529844 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 01:59:12.529855 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.840909 | |
| I0402 01:59:12.529870 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.814359 (* 1 = 0.814359 loss) | |
| I0402 01:59:12.529883 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.209962 (* 1 = 0.209962 loss) | |
| I0402 01:59:12.529896 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 01:59:12.529908 6134 solver.cpp:245] Train net output #16: total_confidence = 0.343017 | |
| I0402 01:59:12.529919 6134 sgd_solver.cpp:106] Iteration 181000, lr = 0.01 | |
| I0402 02:01:21.397717 6134 solver.cpp:229] Iteration 181500, loss = 2.6238 | |
| I0402 02:01:21.397858 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.461538 | |
| I0402 02:01:21.397881 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0402 02:01:21.397893 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.794872 | |
| I0402 02:01:21.397909 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.63988 (* 0.3 = 0.491965 loss) | |
| I0402 02:01:21.397924 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.464716 (* 0.3 = 0.139415 loss) | |
| I0402 02:01:21.397936 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.74359 | |
| I0402 02:01:21.397949 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.909091 | |
| I0402 02:01:21.397961 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.897436 | |
| I0402 02:01:21.397974 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.06046 (* 0.3 = 0.318139 loss) | |
| I0402 02:01:21.397989 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.352677 (* 0.3 = 0.105803 loss) | |
| I0402 02:01:21.398002 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.923077 | |
| I0402 02:01:21.398015 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.977273 | |
| I0402 02:01:21.398026 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0402 02:01:21.398041 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.297505 (* 1 = 0.297505 loss) | |
| I0402 02:01:21.398054 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0842105 (* 1 = 0.0842105 loss) | |
| I0402 02:01:21.398066 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 02:01:21.398078 6134 solver.cpp:245] Train net output #16: total_confidence = 0.507073 | |
| I0402 02:01:21.398090 6134 sgd_solver.cpp:106] Iteration 181500, lr = 0.01 | |
| I0402 02:03:30.313899 6134 solver.cpp:229] Iteration 182000, loss = 2.59862 | |
| I0402 02:03:30.313994 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.413043 | |
| I0402 02:03:30.314014 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 02:03:30.314028 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.630435 | |
| I0402 02:03:30.314044 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.06334 (* 0.3 = 0.619001 loss) | |
| I0402 02:03:30.314059 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.601879 (* 0.3 = 0.180564 loss) | |
| I0402 02:03:30.314074 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.586957 | |
| I0402 02:03:30.314087 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 02:03:30.314100 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.826087 | |
| I0402 02:03:30.314113 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.46803 (* 0.3 = 0.440409 loss) | |
| I0402 02:03:30.314128 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.426515 (* 0.3 = 0.127954 loss) | |
| I0402 02:03:30.314141 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.869565 | |
| I0402 02:03:30.314152 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0402 02:03:30.314164 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.891304 | |
| I0402 02:03:30.314178 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.849067 (* 1 = 0.849067 loss) | |
| I0402 02:03:30.314193 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.23175 (* 1 = 0.23175 loss) | |
| I0402 02:03:30.314204 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 02:03:30.314216 6134 solver.cpp:245] Train net output #16: total_confidence = 0.358069 | |
| I0402 02:03:30.314229 6134 sgd_solver.cpp:106] Iteration 182000, lr = 0.01 | |
| I0402 02:05:39.308456 6134 solver.cpp:229] Iteration 182500, loss = 2.63084 | |
| I0402 02:05:39.308650 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431818 | |
| I0402 02:05:39.308672 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0402 02:05:39.308686 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.659091 | |
| I0402 02:05:39.308702 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.89027 (* 0.3 = 0.56708 loss) | |
| I0402 02:05:39.308717 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.562816 (* 0.3 = 0.168845 loss) | |
| I0402 02:05:39.308730 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.613636 | |
| I0402 02:05:39.308743 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0402 02:05:39.308755 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.795455 | |
| I0402 02:05:39.308769 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.32192 (* 0.3 = 0.396575 loss) | |
| I0402 02:05:39.308784 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.368307 (* 0.3 = 0.110492 loss) | |
| I0402 02:05:39.308797 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.772727 | |
| I0402 02:05:39.308809 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0402 02:05:39.308821 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.977273 | |
| I0402 02:05:39.308835 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.636027 (* 1 = 0.636027 loss) | |
| I0402 02:05:39.308850 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.202725 (* 1 = 0.202725 loss) | |
| I0402 02:05:39.308861 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 02:05:39.308874 6134 solver.cpp:245] Train net output #16: total_confidence = 0.347608 | |
| I0402 02:05:39.308887 6134 sgd_solver.cpp:106] Iteration 182500, lr = 0.01 | |
| I0402 02:07:48.189590 6134 solver.cpp:229] Iteration 183000, loss = 2.66984 | |
| I0402 02:07:48.189923 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.488372 | |
| I0402 02:07:48.189941 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0402 02:07:48.189954 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.767442 | |
| I0402 02:07:48.189971 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.65479 (* 0.3 = 0.496437 loss) | |
| I0402 02:07:48.189985 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.481842 (* 0.3 = 0.144553 loss) | |
| I0402 02:07:48.189998 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.813953 | |
| I0402 02:07:48.190011 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.926136 | |
| I0402 02:07:48.190023 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.953488 | |
| I0402 02:07:48.190037 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.749616 (* 0.3 = 0.224885 loss) | |
| I0402 02:07:48.190052 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.258203 (* 0.3 = 0.0774608 loss) | |
| I0402 02:07:48.190067 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.976744 | |
| I0402 02:07:48.190079 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.994318 | |
| I0402 02:07:48.190091 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0402 02:07:48.190105 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.15831 (* 1 = 0.15831 loss) | |
| I0402 02:07:48.190119 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0420598 (* 1 = 0.0420598 loss) | |
| I0402 02:07:48.190132 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.875 | |
| I0402 02:07:48.190145 6134 solver.cpp:245] Train net output #16: total_confidence = 0.544107 | |
| I0402 02:07:48.190156 6134 sgd_solver.cpp:106] Iteration 183000, lr = 0.01 | |
| I0402 02:09:57.062791 6134 solver.cpp:229] Iteration 183500, loss = 2.60526 | |
| I0402 02:09:57.062933 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.42 | |
| I0402 02:09:57.062953 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 02:09:57.062966 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.62 | |
| I0402 02:09:57.062983 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.87425 (* 0.3 = 0.562276 loss) | |
| I0402 02:09:57.062999 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.566225 (* 0.3 = 0.169867 loss) | |
| I0402 02:09:57.063010 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.52 | |
| I0402 02:09:57.063024 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0402 02:09:57.063035 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.82 | |
| I0402 02:09:57.063048 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.64248 (* 0.3 = 0.492743 loss) | |
| I0402 02:09:57.063063 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.490811 (* 0.3 = 0.147243 loss) | |
| I0402 02:09:57.063076 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.74 | |
| I0402 02:09:57.063088 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0402 02:09:57.063100 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.88 | |
| I0402 02:09:57.063114 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.916602 (* 1 = 0.916602 loss) | |
| I0402 02:09:57.063128 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.26959 (* 1 = 0.26959 loss) | |
| I0402 02:09:57.063140 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 02:09:57.063153 6134 solver.cpp:245] Train net output #16: total_confidence = 0.29982 | |
| I0402 02:09:57.063164 6134 sgd_solver.cpp:106] Iteration 183500, lr = 0.01 | |
| I0402 02:12:05.964361 6134 solver.cpp:229] Iteration 184000, loss = 2.61697 | |
| I0402 02:12:05.964475 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.387755 | |
| I0402 02:12:05.964495 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0402 02:12:05.964509 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.755102 | |
| I0402 02:12:05.964526 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.80176 (* 0.3 = 0.540528 loss) | |
| I0402 02:12:05.964541 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.550611 (* 0.3 = 0.165183 loss) | |
| I0402 02:12:05.964555 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.55102 | |
| I0402 02:12:05.964567 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0402 02:12:05.964579 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.755102 | |
| I0402 02:12:05.964592 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.47839 (* 0.3 = 0.443517 loss) | |
| I0402 02:12:05.964607 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.47083 (* 0.3 = 0.141249 loss) | |
| I0402 02:12:05.964619 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.734694 | |
| I0402 02:12:05.964632 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0402 02:12:05.964643 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.877551 | |
| I0402 02:12:05.964658 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.3406 (* 1 = 1.3406 loss) | |
| I0402 02:12:05.964671 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.410167 (* 1 = 0.410167 loss) | |
| I0402 02:12:05.964684 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 02:12:05.964695 6134 solver.cpp:245] Train net output #16: total_confidence = 0.22648 | |
| I0402 02:12:05.964707 6134 sgd_solver.cpp:106] Iteration 184000, lr = 0.01 | |
| I0402 02:14:14.932867 6134 solver.cpp:229] Iteration 184500, loss = 2.56436 | |
| I0402 02:14:14.932996 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.489796 | |
| I0402 02:14:14.933017 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 02:14:14.933030 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.816327 | |
| I0402 02:14:14.933045 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.39784 (* 0.3 = 0.419352 loss) | |
| I0402 02:14:14.933060 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.41354 (* 0.3 = 0.124062 loss) | |
| I0402 02:14:14.933073 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.755102 | |
| I0402 02:14:14.933087 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.931818 | |
| I0402 02:14:14.933099 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.959184 | |
| I0402 02:14:14.933114 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.938615 (* 0.3 = 0.281585 loss) | |
| I0402 02:14:14.933142 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.274458 (* 0.3 = 0.0823373 loss) | |
| I0402 02:14:14.933156 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.938776 | |
| I0402 02:14:14.933168 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.977273 | |
| I0402 02:14:14.933181 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.979592 | |
| I0402 02:14:14.933194 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.201359 (* 1 = 0.201359 loss) | |
| I0402 02:14:14.933209 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0694399 (* 1 = 0.0694399 loss) | |
| I0402 02:14:14.933221 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 02:14:14.933233 6134 solver.cpp:245] Train net output #16: total_confidence = 0.509486 | |
| I0402 02:14:14.933245 6134 sgd_solver.cpp:106] Iteration 184500, lr = 0.01 | |
| I0402 02:16:23.454079 6134 solver.cpp:338] Iteration 185000, Testing net (#0) | |
| I0402 02:16:53.321988 6134 solver.cpp:393] Test loss: 2.25526 | |
| I0402 02:16:53.322032 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.52536 | |
| I0402 02:16:53.322048 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.874275 | |
| I0402 02:16:53.322062 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.788177 | |
| I0402 02:16:53.322077 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.65927 (* 0.3 = 0.497781 loss) | |
| I0402 02:16:53.322093 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.441612 (* 0.3 = 0.132484 loss) | |
| I0402 02:16:53.322104 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.658484 | |
| I0402 02:16:53.322118 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.914183 | |
| I0402 02:16:53.322129 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.86378 | |
| I0402 02:16:53.322144 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.23793 (* 0.3 = 0.371379 loss) | |
| I0402 02:16:53.322157 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.314326 (* 0.3 = 0.0942978 loss) | |
| I0402 02:16:53.322170 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.783371 | |
| I0402 02:16:53.322181 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.942682 | |
| I0402 02:16:53.322193 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.888609 | |
| I0402 02:16:53.322206 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.919635 (* 1 = 0.919635 loss) | |
| I0402 02:16:53.322221 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.239684 (* 1 = 0.239684 loss) | |
| I0402 02:16:53.322232 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.431 | |
| I0402 02:16:53.322244 6134 solver.cpp:406] Test net output #16: total_confidence = 0.408926 | |
| I0402 02:16:53.472124 6134 solver.cpp:229] Iteration 185000, loss = 2.64308 | |
| I0402 02:16:53.472496 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.595745 | |
| I0402 02:16:53.472519 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 02:16:53.472533 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.87234 | |
| I0402 02:16:53.472548 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.33846 (* 0.3 = 0.401539 loss) | |
| I0402 02:16:53.472563 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.461656 (* 0.3 = 0.138497 loss) | |
| I0402 02:16:53.472575 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.765957 | |
| I0402 02:16:53.472589 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.903409 | |
| I0402 02:16:53.472599 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.957447 | |
| I0402 02:16:53.472614 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 0.78138 (* 0.3 = 0.234414 loss) | |
| I0402 02:16:53.472627 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.27812 (* 0.3 = 0.083436 loss) | |
| I0402 02:16:53.472640 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.957447 | |
| I0402 02:16:53.472652 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.977273 | |
| I0402 02:16:53.472663 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0402 02:16:53.472676 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.244932 (* 1 = 0.244932 loss) | |
| I0402 02:16:53.472690 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0946138 (* 1 = 0.0946138 loss) | |
| I0402 02:16:53.472702 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 02:16:53.472713 6134 solver.cpp:245] Train net output #16: total_confidence = 0.434287 | |
| I0402 02:16:53.472725 6134 sgd_solver.cpp:106] Iteration 185000, lr = 0.01 | |
| I0402 02:19:02.156186 6134 solver.cpp:229] Iteration 185500, loss = 2.59329 | |
| I0402 02:19:02.156322 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.372549 | |
| I0402 02:19:02.156342 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.806818 | |
| I0402 02:19:02.156355 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.588235 | |
| I0402 02:19:02.156371 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.26422 (* 0.3 = 0.679266 loss) | |
| I0402 02:19:02.156388 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.7163 (* 0.3 = 0.21489 loss) | |
| I0402 02:19:02.156400 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.45098 | |
| I0402 02:19:02.156414 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0402 02:19:02.156425 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.72549 | |
| I0402 02:19:02.156440 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.7135 (* 0.3 = 0.514049 loss) | |
| I0402 02:19:02.156455 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.517793 (* 0.3 = 0.155338 loss) | |
| I0402 02:19:02.156467 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.647059 | |
| I0402 02:19:02.156479 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.892045 | |
| I0402 02:19:02.156491 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.784314 | |
| I0402 02:19:02.156505 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.05788 (* 1 = 1.05788 loss) | |
| I0402 02:19:02.156522 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.332993 (* 1 = 0.332993 loss) | |
| I0402 02:19:02.156535 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 02:19:02.156548 6134 solver.cpp:245] Train net output #16: total_confidence = 0.370239 | |
| I0402 02:19:02.156560 6134 sgd_solver.cpp:106] Iteration 185500, lr = 0.01 | |
| I0402 02:21:10.661631 6134 solver.cpp:229] Iteration 186000, loss = 2.58071 | |
| I0402 02:21:10.661788 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431818 | |
| I0402 02:21:10.661809 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 02:21:10.661823 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.659091 | |
| I0402 02:21:10.661839 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.83737 (* 0.3 = 0.551212 loss) | |
| I0402 02:21:10.661854 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.559457 (* 0.3 = 0.167837 loss) | |
| I0402 02:21:10.661867 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.522727 | |
| I0402 02:21:10.661880 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0402 02:21:10.661891 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.772727 | |
| I0402 02:21:10.661906 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.58713 (* 0.3 = 0.476138 loss) | |
| I0402 02:21:10.661919 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.515448 (* 0.3 = 0.154634 loss) | |
| I0402 02:21:10.661932 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.681818 | |
| I0402 02:21:10.661944 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0402 02:21:10.661957 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.75 | |
| I0402 02:21:10.661972 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.5948 (* 1 = 1.5948 loss) | |
| I0402 02:21:10.661985 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.436508 (* 1 = 0.436508 loss) | |
| I0402 02:21:10.661998 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 02:21:10.662010 6134 solver.cpp:245] Train net output #16: total_confidence = 0.413604 | |
| I0402 02:21:10.662022 6134 sgd_solver.cpp:106] Iteration 186000, lr = 0.01 | |
| I0402 02:23:19.244269 6134 solver.cpp:229] Iteration 186500, loss = 2.55349 | |
| I0402 02:23:19.244387 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.380952 | |
| I0402 02:23:19.244407 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0402 02:23:19.244421 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.761905 | |
| I0402 02:23:19.244437 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.66588 (* 0.3 = 0.499763 loss) | |
| I0402 02:23:19.244452 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.526542 (* 0.3 = 0.157962 loss) | |
| I0402 02:23:19.244465 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.642857 | |
| I0402 02:23:19.244477 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0402 02:23:19.244490 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.857143 | |
| I0402 02:23:19.244504 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.23403 (* 0.3 = 0.370209 loss) | |
| I0402 02:23:19.244521 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.430102 (* 0.3 = 0.129031 loss) | |
| I0402 02:23:19.244534 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.809524 | |
| I0402 02:23:19.244545 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0402 02:23:19.244557 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.952381 | |
| I0402 02:23:19.244571 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.575805 (* 1 = 0.575805 loss) | |
| I0402 02:23:19.244585 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.186599 (* 1 = 0.186599 loss) | |
| I0402 02:23:19.244598 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 02:23:19.244611 6134 solver.cpp:245] Train net output #16: total_confidence = 0.275197 | |
| I0402 02:23:19.244623 6134 sgd_solver.cpp:106] Iteration 186500, lr = 0.01 | |
| I0402 02:25:28.022889 6134 solver.cpp:229] Iteration 187000, loss = 2.60199 | |
| I0402 02:25:28.023017 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.454545 | |
| I0402 02:25:28.023048 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0402 02:25:28.023072 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.840909 | |
| I0402 02:25:28.023090 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.46989 (* 0.3 = 0.440966 loss) | |
| I0402 02:25:28.023105 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.448743 (* 0.3 = 0.134623 loss) | |
| I0402 02:25:28.023118 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.545455 | |
| I0402 02:25:28.023131 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0402 02:25:28.023144 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.931818 | |
| I0402 02:25:28.023157 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.15583 (* 0.3 = 0.346749 loss) | |
| I0402 02:25:28.023172 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.369188 (* 0.3 = 0.110756 loss) | |
| I0402 02:25:28.023185 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.886364 | |
| I0402 02:25:28.023196 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0402 02:25:28.023208 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.977273 | |
| I0402 02:25:28.023222 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.428467 (* 1 = 0.428467 loss) | |
| I0402 02:25:28.023236 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.123117 (* 1 = 0.123117 loss) | |
| I0402 02:25:28.023248 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 02:25:28.023262 6134 solver.cpp:245] Train net output #16: total_confidence = 0.366129 | |
| I0402 02:25:28.023273 6134 sgd_solver.cpp:106] Iteration 187000, lr = 0.01 | |
| I0402 02:27:36.627903 6134 solver.cpp:229] Iteration 187500, loss = 2.50041 | |
| I0402 02:27:36.628231 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.466667 | |
| I0402 02:27:36.628249 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0402 02:27:36.628262 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.8 | |
| I0402 02:27:36.628278 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.7356 (* 0.3 = 0.52068 loss) | |
| I0402 02:27:36.628293 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.504511 (* 0.3 = 0.151353 loss) | |
| I0402 02:27:36.628306 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.6 | |
| I0402 02:27:36.628319 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0402 02:27:36.628330 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.866667 | |
| I0402 02:27:36.628345 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.3669 (* 0.3 = 0.410071 loss) | |
| I0402 02:27:36.628360 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.396422 (* 0.3 = 0.118927 loss) | |
| I0402 02:27:36.628371 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.844444 | |
| I0402 02:27:36.628383 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.960227 | |
| I0402 02:27:36.628396 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.888889 | |
| I0402 02:27:36.628409 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.691505 (* 1 = 0.691505 loss) | |
| I0402 02:27:36.628423 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.185228 (* 1 = 0.185228 loss) | |
| I0402 02:27:36.628435 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 02:27:36.628448 6134 solver.cpp:245] Train net output #16: total_confidence = 0.482235 | |
| I0402 02:27:36.628459 6134 sgd_solver.cpp:106] Iteration 187500, lr = 0.01 | |
| I0402 02:29:45.251060 6134 solver.cpp:229] Iteration 188000, loss = 2.54921 | |
| I0402 02:29:45.251222 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.425532 | |
| I0402 02:29:45.251243 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 02:29:45.251256 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.638298 | |
| I0402 02:29:45.251273 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.5669 (* 0.3 = 0.770071 loss) | |
| I0402 02:29:45.251288 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.803239 (* 0.3 = 0.240972 loss) | |
| I0402 02:29:45.251301 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.510638 | |
| I0402 02:29:45.251313 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.846591 | |
| I0402 02:29:45.251325 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.680851 | |
| I0402 02:29:45.251339 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 2.00756 (* 0.3 = 0.602269 loss) | |
| I0402 02:29:45.251353 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.667793 (* 0.3 = 0.200338 loss) | |
| I0402 02:29:45.251365 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.723404 | |
| I0402 02:29:45.251379 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0402 02:29:45.251389 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.765957 | |
| I0402 02:29:45.251405 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.57008 (* 1 = 1.57008 loss) | |
| I0402 02:29:45.251418 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.554635 (* 1 = 0.554635 loss) | |
| I0402 02:29:45.251431 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 02:29:45.251443 6134 solver.cpp:245] Train net output #16: total_confidence = 0.363831 | |
| I0402 02:29:45.251454 6134 sgd_solver.cpp:106] Iteration 188000, lr = 0.01 | |
| I0402 02:31:54.012687 6134 solver.cpp:229] Iteration 188500, loss = 2.52728 | |
| I0402 02:31:54.012805 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.355556 | |
| I0402 02:31:54.012830 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0402 02:31:54.012842 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.666667 | |
| I0402 02:31:54.012859 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.11389 (* 0.3 = 0.634167 loss) | |
| I0402 02:31:54.012874 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.653943 (* 0.3 = 0.196183 loss) | |
| I0402 02:31:54.012887 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.622222 | |
| I0402 02:31:54.012898 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0402 02:31:54.012910 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.755556 | |
| I0402 02:31:54.012924 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.42391 (* 0.3 = 0.427173 loss) | |
| I0402 02:31:54.012939 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.421048 (* 0.3 = 0.126315 loss) | |
| I0402 02:31:54.012951 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.777778 | |
| I0402 02:31:54.012964 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 02:31:54.012975 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.888889 | |
| I0402 02:31:54.012989 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.853155 (* 1 = 0.853155 loss) | |
| I0402 02:31:54.013003 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.249357 (* 1 = 0.249357 loss) | |
| I0402 02:31:54.013015 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 02:31:54.013027 6134 solver.cpp:245] Train net output #16: total_confidence = 0.442347 | |
| I0402 02:31:54.013039 6134 sgd_solver.cpp:106] Iteration 188500, lr = 0.01 | |
| I0402 02:34:02.739753 6134 solver.cpp:229] Iteration 189000, loss = 2.61694 | |
| I0402 02:34:02.739881 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.38 | |
| I0402 02:34:02.739902 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.818182 | |
| I0402 02:34:02.739915 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.72 | |
| I0402 02:34:02.739931 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.92821 (* 0.3 = 0.578463 loss) | |
| I0402 02:34:02.739946 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.583344 (* 0.3 = 0.175003 loss) | |
| I0402 02:34:02.739959 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0402 02:34:02.739971 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0402 02:34:02.739984 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.76 | |
| I0402 02:34:02.739997 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.55156 (* 0.3 = 0.465469 loss) | |
| I0402 02:34:02.740013 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.496165 (* 0.3 = 0.148849 loss) | |
| I0402 02:34:02.740025 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.72 | |
| I0402 02:34:02.740038 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0402 02:34:02.740049 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.88 | |
| I0402 02:34:02.740063 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.887985 (* 1 = 0.887985 loss) | |
| I0402 02:34:02.740077 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.306735 (* 1 = 0.306735 loss) | |
| I0402 02:34:02.740090 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 02:34:02.740103 6134 solver.cpp:245] Train net output #16: total_confidence = 0.226298 | |
| I0402 02:34:02.740114 6134 sgd_solver.cpp:106] Iteration 189000, lr = 0.01 | |
| I0402 02:36:11.404500 6134 solver.cpp:229] Iteration 189500, loss = 2.54933 | |
| I0402 02:36:11.404618 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.377778 | |
| I0402 02:36:11.404639 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 02:36:11.404651 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.622222 | |
| I0402 02:36:11.404666 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.90802 (* 0.3 = 0.572407 loss) | |
| I0402 02:36:11.404681 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.557916 (* 0.3 = 0.167375 loss) | |
| I0402 02:36:11.404695 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.577778 | |
| I0402 02:36:11.404706 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 02:36:11.404719 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.866667 | |
| I0402 02:36:11.404732 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.26519 (* 0.3 = 0.379556 loss) | |
| I0402 02:36:11.404747 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.352762 (* 0.3 = 0.105829 loss) | |
| I0402 02:36:11.404760 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.822222 | |
| I0402 02:36:11.404772 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0402 02:36:11.404784 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.955556 | |
| I0402 02:36:11.404798 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.504882 (* 1 = 0.504882 loss) | |
| I0402 02:36:11.404811 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.143264 (* 1 = 0.143264 loss) | |
| I0402 02:36:11.404824 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 02:36:11.404836 6134 solver.cpp:245] Train net output #16: total_confidence = 0.367848 | |
| I0402 02:36:11.404849 6134 sgd_solver.cpp:106] Iteration 189500, lr = 0.01 | |
| I0402 02:38:20.125980 6134 solver.cpp:338] Iteration 190000, Testing net (#0) | |
| I0402 02:38:49.869405 6134 solver.cpp:393] Test loss: 2.26244 | |
| I0402 02:38:49.869449 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.513228 | |
| I0402 02:38:49.869467 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.870957 | |
| I0402 02:38:49.869479 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.784534 | |
| I0402 02:38:49.869495 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.6877 (* 0.3 = 0.506311 loss) | |
| I0402 02:38:49.869510 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.456084 (* 0.3 = 0.136825 loss) | |
| I0402 02:38:49.869526 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.679799 | |
| I0402 02:38:49.869539 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.917229 | |
| I0402 02:38:49.869550 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.865481 | |
| I0402 02:38:49.869563 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.19656 (* 0.3 = 0.358968 loss) | |
| I0402 02:38:49.869578 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.314087 (* 0.3 = 0.0942262 loss) | |
| I0402 02:38:49.869590 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.770363 | |
| I0402 02:38:49.869602 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.940455 | |
| I0402 02:38:49.869613 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.88923 | |
| I0402 02:38:49.869627 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.923241 (* 1 = 0.923241 loss) | |
| I0402 02:38:49.869642 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.242866 (* 1 = 0.242866 loss) | |
| I0402 02:38:49.869652 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.444 | |
| I0402 02:38:49.869664 6134 solver.cpp:406] Test net output #16: total_confidence = 0.412445 | |
| I0402 02:38:50.020536 6134 solver.cpp:229] Iteration 190000, loss = 2.60416 | |
| I0402 02:38:50.020586 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.454545 | |
| I0402 02:38:50.020602 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 02:38:50.020615 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.590909 | |
| I0402 02:38:50.020632 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.08606 (* 0.3 = 0.625817 loss) | |
| I0402 02:38:50.020647 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.581542 (* 0.3 = 0.174463 loss) | |
| I0402 02:38:50.020659 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.613636 | |
| I0402 02:38:50.020671 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0402 02:38:50.020684 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.772727 | |
| I0402 02:38:50.020697 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.58663 (* 0.3 = 0.47599 loss) | |
| I0402 02:38:50.020711 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.430009 (* 0.3 = 0.129003 loss) | |
| I0402 02:38:50.020723 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.772727 | |
| I0402 02:38:50.020735 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0402 02:38:50.020747 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.818182 | |
| I0402 02:38:50.020761 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.87964 (* 1 = 0.87964 loss) | |
| I0402 02:38:50.020776 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.236985 (* 1 = 0.236985 loss) | |
| I0402 02:38:50.020787 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 02:38:50.020799 6134 solver.cpp:245] Train net output #16: total_confidence = 0.486673 | |
| I0402 02:38:50.020812 6134 sgd_solver.cpp:106] Iteration 190000, lr = 0.01 | |
| I0402 02:40:58.680814 6134 solver.cpp:229] Iteration 190500, loss = 2.54403 | |
| I0402 02:40:58.680943 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.405405 | |
| I0402 02:40:58.680963 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0402 02:40:58.680975 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.621622 | |
| I0402 02:40:58.680992 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.28437 (* 0.3 = 0.685311 loss) | |
| I0402 02:40:58.681007 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.603057 (* 0.3 = 0.180917 loss) | |
| I0402 02:40:58.681020 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.378378 | |
| I0402 02:40:58.681032 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0402 02:40:58.681044 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.648649 | |
| I0402 02:40:58.681058 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.93007 (* 0.3 = 0.579022 loss) | |
| I0402 02:40:58.681090 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.513597 (* 0.3 = 0.154079 loss) | |
| I0402 02:40:58.681104 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.567568 | |
| I0402 02:40:58.681116 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.897727 | |
| I0402 02:40:58.681128 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.72973 | |
| I0402 02:40:58.681143 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.4602 (* 1 = 1.4602 loss) | |
| I0402 02:40:58.681157 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.370564 (* 1 = 0.370564 loss) | |
| I0402 02:40:58.681169 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 02:40:58.681181 6134 solver.cpp:245] Train net output #16: total_confidence = 0.277891 | |
| I0402 02:40:58.681193 6134 sgd_solver.cpp:106] Iteration 190500, lr = 0.01 | |
| I0402 02:43:07.294528 6134 solver.cpp:229] Iteration 191000, loss = 2.63624 | |
| I0402 02:43:07.294636 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.425532 | |
| I0402 02:43:07.294656 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0402 02:43:07.294669 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.617021 | |
| I0402 02:43:07.294687 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.92177 (* 0.3 = 0.576531 loss) | |
| I0402 02:43:07.294702 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.569456 (* 0.3 = 0.170837 loss) | |
| I0402 02:43:07.294714 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.510638 | |
| I0402 02:43:07.294726 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0402 02:43:07.294739 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.829787 | |
| I0402 02:43:07.294754 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.54427 (* 0.3 = 0.463281 loss) | |
| I0402 02:43:07.294767 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.454247 (* 0.3 = 0.136274 loss) | |
| I0402 02:43:07.294780 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.723404 | |
| I0402 02:43:07.294791 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0402 02:43:07.294802 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.808511 | |
| I0402 02:43:07.294817 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.0222 (* 1 = 1.0222 loss) | |
| I0402 02:43:07.294831 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.288435 (* 1 = 0.288435 loss) | |
| I0402 02:43:07.294843 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 02:43:07.294855 6134 solver.cpp:245] Train net output #16: total_confidence = 0.27923 | |
| I0402 02:43:07.294867 6134 sgd_solver.cpp:106] Iteration 191000, lr = 0.01 | |
| I0402 02:45:16.254225 6134 solver.cpp:229] Iteration 191500, loss = 2.53745 | |
| I0402 02:45:16.254348 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.478261 | |
| I0402 02:45:16.254367 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0402 02:45:16.254380 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.847826 | |
| I0402 02:45:16.254396 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.57466 (* 0.3 = 0.472397 loss) | |
| I0402 02:45:16.254411 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.434561 (* 0.3 = 0.130368 loss) | |
| I0402 02:45:16.254425 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.717391 | |
| I0402 02:45:16.254437 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.920455 | |
| I0402 02:45:16.254448 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.913043 | |
| I0402 02:45:16.254462 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.03473 (* 0.3 = 0.31042 loss) | |
| I0402 02:45:16.254477 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.301724 (* 0.3 = 0.0905172 loss) | |
| I0402 02:45:16.254490 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.913043 | |
| I0402 02:45:16.254503 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.977273 | |
| I0402 02:45:16.254514 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.956522 | |
| I0402 02:45:16.254528 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.427243 (* 1 = 0.427243 loss) | |
| I0402 02:45:16.254542 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.117487 (* 1 = 0.117487 loss) | |
| I0402 02:45:16.254554 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 02:45:16.254566 6134 solver.cpp:245] Train net output #16: total_confidence = 0.414072 | |
| I0402 02:45:16.254577 6134 sgd_solver.cpp:106] Iteration 191500, lr = 0.01 | |
| I0402 02:47:24.933358 6134 solver.cpp:229] Iteration 192000, loss = 2.5141 | |
| I0402 02:47:24.933714 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.439024 | |
| I0402 02:47:24.933744 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 02:47:24.933768 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.658537 | |
| I0402 02:47:24.933799 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.09013 (* 0.3 = 0.627039 loss) | |
| I0402 02:47:24.933825 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.588139 (* 0.3 = 0.176442 loss) | |
| I0402 02:47:24.933847 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.560976 | |
| I0402 02:47:24.933871 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 02:47:24.933892 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.804878 | |
| I0402 02:47:24.933917 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.42827 (* 0.3 = 0.42848 loss) | |
| I0402 02:47:24.933943 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.404409 (* 0.3 = 0.121323 loss) | |
| I0402 02:47:24.933964 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.731707 | |
| I0402 02:47:24.933997 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.920455 | |
| I0402 02:47:24.934031 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.853659 | |
| I0402 02:47:24.934058 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.940584 (* 1 = 0.940584 loss) | |
| I0402 02:47:24.934087 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.268154 (* 1 = 0.268154 loss) | |
| I0402 02:47:24.934109 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 02:47:24.934130 6134 solver.cpp:245] Train net output #16: total_confidence = 0.265713 | |
| I0402 02:47:24.934151 6134 sgd_solver.cpp:106] Iteration 192000, lr = 0.01 | |
| I0402 02:49:33.614536 6134 solver.cpp:229] Iteration 192500, loss = 2.50409 | |
| I0402 02:49:33.614717 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.510204 | |
| I0402 02:49:33.614738 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.857955 | |
| I0402 02:49:33.614751 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.632653 | |
| I0402 02:49:33.614768 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.04432 (* 0.3 = 0.613295 loss) | |
| I0402 02:49:33.614784 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.600523 (* 0.3 = 0.180157 loss) | |
| I0402 02:49:33.614795 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.571429 | |
| I0402 02:49:33.614809 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 02:49:33.614820 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.836735 | |
| I0402 02:49:33.614833 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.43378 (* 0.3 = 0.430134 loss) | |
| I0402 02:49:33.614848 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.427602 (* 0.3 = 0.128281 loss) | |
| I0402 02:49:33.614861 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.836735 | |
| I0402 02:49:33.614872 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0402 02:49:33.614884 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.897959 | |
| I0402 02:49:33.614898 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.62765 (* 1 = 0.62765 loss) | |
| I0402 02:49:33.614912 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.182117 (* 1 = 0.182117 loss) | |
| I0402 02:49:33.614924 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 02:49:33.614936 6134 solver.cpp:245] Train net output #16: total_confidence = 0.359157 | |
| I0402 02:49:33.614948 6134 sgd_solver.cpp:106] Iteration 192500, lr = 0.01 | |
| I0402 02:51:42.066972 6134 solver.cpp:229] Iteration 193000, loss = 2.49777 | |
| I0402 02:51:42.067087 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.425532 | |
| I0402 02:51:42.067117 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0402 02:51:42.067140 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.744681 | |
| I0402 02:51:42.067170 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.77516 (* 0.3 = 0.532549 loss) | |
| I0402 02:51:42.067198 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.511704 (* 0.3 = 0.153511 loss) | |
| I0402 02:51:42.067220 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.680851 | |
| I0402 02:51:42.067244 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.903409 | |
| I0402 02:51:42.067266 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.914894 | |
| I0402 02:51:42.067291 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.08664 (* 0.3 = 0.325991 loss) | |
| I0402 02:51:42.067318 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.319452 (* 0.3 = 0.0958355 loss) | |
| I0402 02:51:42.067339 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.93617 | |
| I0402 02:51:42.067361 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.982955 | |
| I0402 02:51:42.067381 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 1 | |
| I0402 02:51:42.067409 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.261076 (* 1 = 0.261076 loss) | |
| I0402 02:51:42.067436 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0742366 (* 1 = 0.0742366 loss) | |
| I0402 02:51:42.067458 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0402 02:51:42.067481 6134 solver.cpp:245] Train net output #16: total_confidence = 0.38126 | |
| I0402 02:51:42.067502 6134 sgd_solver.cpp:106] Iteration 193000, lr = 0.01 | |
| I0402 02:53:50.758524 6134 solver.cpp:229] Iteration 193500, loss = 2.52772 | |
| I0402 02:53:50.758652 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.340426 | |
| I0402 02:53:50.758682 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.795455 | |
| I0402 02:53:50.758705 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.680851 | |
| I0402 02:53:50.758736 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.20515 (* 0.3 = 0.661545 loss) | |
| I0402 02:53:50.758764 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.670271 (* 0.3 = 0.201081 loss) | |
| I0402 02:53:50.758786 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.489362 | |
| I0402 02:53:50.758810 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0402 02:53:50.758832 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.787234 | |
| I0402 02:53:50.758857 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.68216 (* 0.3 = 0.504648 loss) | |
| I0402 02:53:50.758882 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.503708 (* 0.3 = 0.151113 loss) | |
| I0402 02:53:50.758904 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.765957 | |
| I0402 02:53:50.758924 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0402 02:53:50.758945 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.893617 | |
| I0402 02:53:50.758972 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.980722 (* 1 = 0.980722 loss) | |
| I0402 02:53:50.759001 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.288701 (* 1 = 0.288701 loss) | |
| I0402 02:53:50.759021 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 02:53:50.759043 6134 solver.cpp:245] Train net output #16: total_confidence = 0.363504 | |
| I0402 02:53:50.759063 6134 sgd_solver.cpp:106] Iteration 193500, lr = 0.01 | |
| I0402 02:55:59.379551 6134 solver.cpp:229] Iteration 194000, loss = 2.49342 | |
| I0402 02:55:59.379683 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.525 | |
| I0402 02:55:59.379703 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.869318 | |
| I0402 02:55:59.379716 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.775 | |
| I0402 02:55:59.379731 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.45659 (* 0.3 = 0.436976 loss) | |
| I0402 02:55:59.379747 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.425143 (* 0.3 = 0.127543 loss) | |
| I0402 02:55:59.379760 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.6 | |
| I0402 02:55:59.379772 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0402 02:55:59.379784 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.9 | |
| I0402 02:55:59.379798 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.10399 (* 0.3 = 0.331198 loss) | |
| I0402 02:55:59.379813 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.306423 (* 0.3 = 0.0919268 loss) | |
| I0402 02:55:59.379827 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.95 | |
| I0402 02:55:59.379838 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.988636 | |
| I0402 02:55:59.379849 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.95 | |
| I0402 02:55:59.379863 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.262864 (* 1 = 0.262864 loss) | |
| I0402 02:55:59.379878 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.0752398 (* 1 = 0.0752398 loss) | |
| I0402 02:55:59.379890 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.875 | |
| I0402 02:55:59.379901 6134 solver.cpp:245] Train net output #16: total_confidence = 0.531969 | |
| I0402 02:55:59.379914 6134 sgd_solver.cpp:106] Iteration 194000, lr = 0.01 | |
| I0402 02:58:08.004564 6134 solver.cpp:229] Iteration 194500, loss = 2.50357 | |
| I0402 02:58:08.004935 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.5 | |
| I0402 02:58:08.004964 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 02:58:08.004988 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.694444 | |
| I0402 02:58:08.005018 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.89805 (* 0.3 = 0.569415 loss) | |
| I0402 02:58:08.005065 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.614486 (* 0.3 = 0.184346 loss) | |
| I0402 02:58:08.005094 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.472222 | |
| I0402 02:58:08.005118 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0402 02:58:08.005141 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.777778 | |
| I0402 02:58:08.005167 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.6776 (* 0.3 = 0.503281 loss) | |
| I0402 02:58:08.005193 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.532132 (* 0.3 = 0.15964 loss) | |
| I0402 02:58:08.005218 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.861111 | |
| I0402 02:58:08.005239 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0402 02:58:08.005259 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.972222 | |
| I0402 02:58:08.005285 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.644364 (* 1 = 0.644364 loss) | |
| I0402 02:58:08.005309 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.157445 (* 1 = 0.157445 loss) | |
| I0402 02:58:08.005331 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 02:58:08.005352 6134 solver.cpp:245] Train net output #16: total_confidence = 0.435561 | |
| I0402 02:58:08.005373 6134 sgd_solver.cpp:106] Iteration 194500, lr = 0.01 | |
| I0402 03:00:16.600250 6134 solver.cpp:338] Iteration 195000, Testing net (#0) | |
| I0402 03:00:46.364195 6134 solver.cpp:393] Test loss: 2.21434 | |
| I0402 03:00:46.364240 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.547498 | |
| I0402 03:00:46.364258 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.874048 | |
| I0402 03:00:46.364270 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.795258 | |
| I0402 03:00:46.364285 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.59432 (* 0.3 = 0.478295 loss) | |
| I0402 03:00:46.364300 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.444611 (* 0.3 = 0.133383 loss) | |
| I0402 03:00:46.364312 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.695619 | |
| I0402 03:00:46.364325 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.91432 | |
| I0402 03:00:46.364336 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.870314 | |
| I0402 03:00:46.364349 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.13668 (* 0.3 = 0.341004 loss) | |
| I0402 03:00:46.364363 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.320404 (* 0.3 = 0.0961213 loss) | |
| I0402 03:00:46.364375 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.777778 | |
| I0402 03:00:46.364387 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.938137 | |
| I0402 03:00:46.364399 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.88199 | |
| I0402 03:00:46.364413 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.908771 (* 1 = 0.908771 loss) | |
| I0402 03:00:46.364426 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.256766 (* 1 = 0.256766 loss) | |
| I0402 03:00:46.364439 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.427 | |
| I0402 03:00:46.364449 6134 solver.cpp:406] Test net output #16: total_confidence = 0.372638 | |
| I0402 03:00:46.515811 6134 solver.cpp:229] Iteration 195000, loss = 2.61633 | |
| I0402 03:00:46.515851 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.42 | |
| I0402 03:00:46.515868 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0402 03:00:46.515880 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.54 | |
| I0402 03:00:46.515895 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.07939 (* 0.3 = 0.623818 loss) | |
| I0402 03:00:46.515909 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.698789 (* 0.3 = 0.209637 loss) | |
| I0402 03:00:46.515929 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.48 | |
| I0402 03:00:46.515954 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.818182 | |
| I0402 03:00:46.515969 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.68 | |
| I0402 03:00:46.515983 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.88376 (* 0.3 = 0.565128 loss) | |
| I0402 03:00:46.515997 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.654275 (* 0.3 = 0.196283 loss) | |
| I0402 03:00:46.516010 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.64 | |
| I0402 03:00:46.516021 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.880682 | |
| I0402 03:00:46.516032 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.84 | |
| I0402 03:00:46.516049 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.10041 (* 1 = 1.10041 loss) | |
| I0402 03:00:46.516063 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.359171 (* 1 = 0.359171 loss) | |
| I0402 03:00:46.516077 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 03:00:46.516088 6134 solver.cpp:245] Train net output #16: total_confidence = 0.202482 | |
| I0402 03:00:46.516100 6134 sgd_solver.cpp:106] Iteration 195000, lr = 0.01 | |
| I0402 03:02:55.098968 6134 solver.cpp:229] Iteration 195500, loss = 2.536 | |
| I0402 03:02:55.099125 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.486486 | |
| I0402 03:02:55.099146 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.852273 | |
| I0402 03:02:55.099159 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.675676 | |
| I0402 03:02:55.099175 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.87027 (* 0.3 = 0.561081 loss) | |
| I0402 03:02:55.099190 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.556108 (* 0.3 = 0.166832 loss) | |
| I0402 03:02:55.099202 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.621622 | |
| I0402 03:02:55.099215 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.886364 | |
| I0402 03:02:55.099227 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.837838 | |
| I0402 03:02:55.099241 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.24419 (* 0.3 = 0.373257 loss) | |
| I0402 03:02:55.099256 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.359819 (* 0.3 = 0.107946 loss) | |
| I0402 03:02:55.099268 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.810811 | |
| I0402 03:02:55.099280 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.943182 | |
| I0402 03:02:55.099292 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.945946 | |
| I0402 03:02:55.099305 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.522753 (* 1 = 0.522753 loss) | |
| I0402 03:02:55.099319 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.174824 (* 1 = 0.174824 loss) | |
| I0402 03:02:55.099333 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 03:02:55.099344 6134 solver.cpp:245] Train net output #16: total_confidence = 0.400479 | |
| I0402 03:02:55.099356 6134 sgd_solver.cpp:106] Iteration 195500, lr = 0.01 | |
| I0402 03:05:03.879400 6134 solver.cpp:229] Iteration 196000, loss = 2.54486 | |
| I0402 03:05:03.879549 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.613636 | |
| I0402 03:05:03.879570 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.869318 | |
| I0402 03:05:03.879583 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.727273 | |
| I0402 03:05:03.879598 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.6647 (* 0.3 = 0.49941 loss) | |
| I0402 03:05:03.879613 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.514321 (* 0.3 = 0.154296 loss) | |
| I0402 03:05:03.879626 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.613636 | |
| I0402 03:05:03.879638 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.892045 | |
| I0402 03:05:03.879650 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.863636 | |
| I0402 03:05:03.879664 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.26587 (* 0.3 = 0.379762 loss) | |
| I0402 03:05:03.879679 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.364552 (* 0.3 = 0.109366 loss) | |
| I0402 03:05:03.879691 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.818182 | |
| I0402 03:05:03.879703 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0402 03:05:03.879715 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.863636 | |
| I0402 03:05:03.879729 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.02997 (* 1 = 1.02997 loss) | |
| I0402 03:05:03.879744 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.265341 (* 1 = 0.265341 loss) | |
| I0402 03:05:03.879755 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 03:05:03.879767 6134 solver.cpp:245] Train net output #16: total_confidence = 0.521304 | |
| I0402 03:05:03.879779 6134 sgd_solver.cpp:106] Iteration 196000, lr = 0.01 | |
| I0402 03:07:12.477133 6134 solver.cpp:229] Iteration 196500, loss = 2.58673 | |
| I0402 03:07:12.477461 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.520833 | |
| I0402 03:07:12.477480 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0402 03:07:12.477494 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.770833 | |
| I0402 03:07:12.477510 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.74339 (* 0.3 = 0.523016 loss) | |
| I0402 03:07:12.477525 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.567596 (* 0.3 = 0.170279 loss) | |
| I0402 03:07:12.477538 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.604167 | |
| I0402 03:07:12.477550 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 03:07:12.477563 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.875 | |
| I0402 03:07:12.477578 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.13198 (* 0.3 = 0.339595 loss) | |
| I0402 03:07:12.477592 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.340537 (* 0.3 = 0.102161 loss) | |
| I0402 03:07:12.477605 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.895833 | |
| I0402 03:07:12.477617 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.965909 | |
| I0402 03:07:12.477629 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.979167 | |
| I0402 03:07:12.477643 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.461942 (* 1 = 0.461942 loss) | |
| I0402 03:07:12.477658 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.152893 (* 1 = 0.152893 loss) | |
| I0402 03:07:12.477674 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 03:07:12.477697 6134 solver.cpp:245] Train net output #16: total_confidence = 0.468722 | |
| I0402 03:07:12.477725 6134 sgd_solver.cpp:106] Iteration 196500, lr = 0.01 | |
| I0402 03:09:21.426103 6134 solver.cpp:229] Iteration 197000, loss = 2.58499 | |
| I0402 03:09:21.426232 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.479167 | |
| I0402 03:09:21.426251 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.846591 | |
| I0402 03:09:21.426265 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.729167 | |
| I0402 03:09:21.426280 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.79966 (* 0.3 = 0.539897 loss) | |
| I0402 03:09:21.426295 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.552162 (* 0.3 = 0.165648 loss) | |
| I0402 03:09:21.426307 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5625 | |
| I0402 03:09:21.426321 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 03:09:21.426331 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.791667 | |
| I0402 03:09:21.426347 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.3849 (* 0.3 = 0.415469 loss) | |
| I0402 03:09:21.426360 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.430488 (* 0.3 = 0.129147 loss) | |
| I0402 03:09:21.426373 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.791667 | |
| I0402 03:09:21.426385 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0402 03:09:21.426396 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.958333 | |
| I0402 03:09:21.426410 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.733774 (* 1 = 0.733774 loss) | |
| I0402 03:09:21.426424 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.239431 (* 1 = 0.239431 loss) | |
| I0402 03:09:21.426436 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0402 03:09:21.426450 6134 solver.cpp:245] Train net output #16: total_confidence = 0.354256 | |
| I0402 03:09:21.426461 6134 sgd_solver.cpp:106] Iteration 197000, lr = 0.01 | |
| I0402 03:11:30.228564 6134 solver.cpp:229] Iteration 197500, loss = 2.5767 | |
| I0402 03:11:30.228694 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.264151 | |
| I0402 03:11:30.228715 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.767045 | |
| I0402 03:11:30.228729 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.660377 | |
| I0402 03:11:30.228744 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.08257 (* 0.3 = 0.62477 loss) | |
| I0402 03:11:30.228760 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.660695 (* 0.3 = 0.198208 loss) | |
| I0402 03:11:30.228775 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.471698 | |
| I0402 03:11:30.228787 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.840909 | |
| I0402 03:11:30.228799 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.811321 | |
| I0402 03:11:30.228813 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.82381 (* 0.3 = 0.547143 loss) | |
| I0402 03:11:30.228827 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.557035 (* 0.3 = 0.167111 loss) | |
| I0402 03:11:30.228840 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.716981 | |
| I0402 03:11:30.228852 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.914773 | |
| I0402 03:11:30.228864 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.981132 | |
| I0402 03:11:30.228878 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.828247 (* 1 = 0.828247 loss) | |
| I0402 03:11:30.228893 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.252868 (* 1 = 0.252868 loss) | |
| I0402 03:11:30.228905 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 03:11:30.228917 6134 solver.cpp:245] Train net output #16: total_confidence = 0.318088 | |
| I0402 03:11:30.228930 6134 sgd_solver.cpp:106] Iteration 197500, lr = 0.01 | |
| I0402 03:13:38.857164 6134 solver.cpp:229] Iteration 198000, loss = 2.5127 | |
| I0402 03:13:38.857297 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.571429 | |
| I0402 03:13:38.857328 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.875 | |
| I0402 03:13:38.857352 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.755102 | |
| I0402 03:13:38.857383 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.71974 (* 0.3 = 0.515923 loss) | |
| I0402 03:13:38.857410 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.499995 (* 0.3 = 0.149999 loss) | |
| I0402 03:13:38.857432 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.591837 | |
| I0402 03:13:38.857456 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.875 | |
| I0402 03:13:38.857480 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.857143 | |
| I0402 03:13:38.857506 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.27939 (* 0.3 = 0.383817 loss) | |
| I0402 03:13:38.857535 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.377871 (* 0.3 = 0.113361 loss) | |
| I0402 03:13:38.857558 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.857143 | |
| I0402 03:13:38.857579 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0402 03:13:38.857602 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.979592 | |
| I0402 03:13:38.857630 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.42635 (* 1 = 0.42635 loss) | |
| I0402 03:13:38.857656 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.127894 (* 1 = 0.127894 loss) | |
| I0402 03:13:38.857677 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 03:13:38.857698 6134 solver.cpp:245] Train net output #16: total_confidence = 0.436467 | |
| I0402 03:13:38.857719 6134 sgd_solver.cpp:106] Iteration 198000, lr = 0.01 | |
| I0402 03:15:47.514930 6134 solver.cpp:229] Iteration 198500, loss = 2.51121 | |
| I0402 03:15:47.515043 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.42 | |
| I0402 03:15:47.515072 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 03:15:47.515097 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.68 | |
| I0402 03:15:47.515127 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.79699 (* 0.3 = 0.539096 loss) | |
| I0402 03:15:47.515154 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.560798 (* 0.3 = 0.16824 loss) | |
| I0402 03:15:47.515177 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.66 | |
| I0402 03:15:47.515200 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.897727 | |
| I0402 03:15:47.515223 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.82 | |
| I0402 03:15:47.515249 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.20534 (* 0.3 = 0.361601 loss) | |
| I0402 03:15:47.515275 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.360897 (* 0.3 = 0.108269 loss) | |
| I0402 03:15:47.515295 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.88 | |
| I0402 03:15:47.515317 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0402 03:15:47.515341 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.98 | |
| I0402 03:15:47.515365 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.413055 (* 1 = 0.413055 loss) | |
| I0402 03:15:47.515391 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.166285 (* 1 = 0.166285 loss) | |
| I0402 03:15:47.515413 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 03:15:47.515434 6134 solver.cpp:245] Train net output #16: total_confidence = 0.297636 | |
| I0402 03:15:47.515455 6134 sgd_solver.cpp:106] Iteration 198500, lr = 0.01 | |
| I0402 03:17:56.275885 6134 solver.cpp:229] Iteration 199000, loss = 2.49523 | |
| I0402 03:17:56.276258 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.527778 | |
| I0402 03:17:56.276278 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.869318 | |
| I0402 03:17:56.276291 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.722222 | |
| I0402 03:17:56.276307 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.52195 (* 0.3 = 0.456584 loss) | |
| I0402 03:17:56.276322 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.463241 (* 0.3 = 0.138972 loss) | |
| I0402 03:17:56.276335 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.583333 | |
| I0402 03:17:56.276348 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 03:17:56.276360 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.833333 | |
| I0402 03:17:56.276373 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.16978 (* 0.3 = 0.350935 loss) | |
| I0402 03:17:56.276388 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.421333 (* 0.3 = 0.1264 loss) | |
| I0402 03:17:56.276401 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.833333 | |
| I0402 03:17:56.276412 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0402 03:17:56.276424 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.888889 | |
| I0402 03:17:56.276437 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.574556 (* 1 = 0.574556 loss) | |
| I0402 03:17:56.276451 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.187843 (* 1 = 0.187843 loss) | |
| I0402 03:17:56.276463 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 03:17:56.276475 6134 solver.cpp:245] Train net output #16: total_confidence = 0.38023 | |
| I0402 03:17:56.276487 6134 sgd_solver.cpp:106] Iteration 199000, lr = 0.01 | |
| I0402 03:20:04.967550 6134 solver.cpp:229] Iteration 199500, loss = 2.55946 | |
| I0402 03:20:04.967667 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.409091 | |
| I0402 03:20:04.967689 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.829545 | |
| I0402 03:20:04.967701 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.704545 | |
| I0402 03:20:04.967717 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.07558 (* 0.3 = 0.622673 loss) | |
| I0402 03:20:04.967732 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.582607 (* 0.3 = 0.174782 loss) | |
| I0402 03:20:04.967746 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.545455 | |
| I0402 03:20:04.967757 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0402 03:20:04.967769 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.772727 | |
| I0402 03:20:04.967783 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.53426 (* 0.3 = 0.460277 loss) | |
| I0402 03:20:04.967797 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.461792 (* 0.3 = 0.138538 loss) | |
| I0402 03:20:04.967810 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.75 | |
| I0402 03:20:04.967823 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0402 03:20:04.967834 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.954545 | |
| I0402 03:20:04.967849 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.837031 (* 1 = 0.837031 loss) | |
| I0402 03:20:04.967864 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.234632 (* 1 = 0.234632 loss) | |
| I0402 03:20:04.967875 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.25 | |
| I0402 03:20:04.967887 6134 solver.cpp:245] Train net output #16: total_confidence = 0.213076 | |
| I0402 03:20:04.967900 6134 sgd_solver.cpp:106] Iteration 199500, lr = 0.01 | |
| I0402 03:22:13.488392 6134 solver.cpp:456] Snapshotting to binary proto file /mnt/snapshots/mixed_lstm9_bn_iter_200000.caffemodel | |
| I0402 03:22:13.996386 6134 sgd_solver.cpp:273] Snapshotting solver state to binary proto file /mnt/snapshots/mixed_lstm9_bn_iter_200000.solverstate | |
| I0402 03:22:14.161216 6134 solver.cpp:338] Iteration 200000, Testing net (#0) | |
| I0402 03:22:43.942984 6134 solver.cpp:393] Test loss: 2.16217 | |
| I0402 03:22:43.943099 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.517091 | |
| I0402 03:22:43.943119 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.873412 | |
| I0402 03:22:43.943132 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.798996 | |
| I0402 03:22:43.943148 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.67018 (* 0.3 = 0.501054 loss) | |
| I0402 03:22:43.943163 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.439184 (* 0.3 = 0.131755 loss) | |
| I0402 03:22:43.943176 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.71731 | |
| I0402 03:22:43.943188 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.91841 | |
| I0402 03:22:43.943200 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.884759 | |
| I0402 03:22:43.943213 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.09016 (* 0.3 = 0.327049 loss) | |
| I0402 03:22:43.943228 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.305799 (* 0.3 = 0.0917397 loss) | |
| I0402 03:22:43.943239 6134 solver.cpp:406] Test net output #10: loss3/accuracy = 0.789189 | |
| I0402 03:22:43.943253 6134 solver.cpp:406] Test net output #11: loss3/accuracy_incl_empty = 0.947092 | |
| I0402 03:22:43.943264 6134 solver.cpp:406] Test net output #12: loss3/accuracy_top3 = 0.89831 | |
| I0402 03:22:43.943277 6134 solver.cpp:406] Test net output #13: loss3/cross_entropy_loss = 0.884634 (* 1 = 0.884634 loss) | |
| I0402 03:22:43.943291 6134 solver.cpp:406] Test net output #14: loss3/cross_entropy_loss_incl_empty = 0.22594 (* 1 = 0.22594 loss) | |
| I0402 03:22:43.943303 6134 solver.cpp:406] Test net output #15: total_accuracy = 0.48 | |
| I0402 03:22:43.943315 6134 solver.cpp:406] Test net output #16: total_confidence = 0.462319 | |
| I0402 03:22:44.093876 6134 solver.cpp:229] Iteration 200000, loss = 2.47136 | |
| I0402 03:22:44.093919 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.4 | |
| I0402 03:22:44.093936 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 03:22:44.093950 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.72 | |
| I0402 03:22:44.093966 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.02691 (* 0.3 = 0.608072 loss) | |
| I0402 03:22:44.093981 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.604296 (* 0.3 = 0.181289 loss) | |
| I0402 03:22:44.093993 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.52 | |
| I0402 03:22:44.094008 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0402 03:22:44.094022 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.8 | |
| I0402 03:22:44.094035 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.75081 (* 0.3 = 0.525242 loss) | |
| I0402 03:22:44.094049 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.507632 (* 0.3 = 0.15229 loss) | |
| I0402 03:22:44.094061 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.78 | |
| I0402 03:22:44.094074 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 03:22:44.094085 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.9 | |
| I0402 03:22:44.094099 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.947642 (* 1 = 0.947642 loss) | |
| I0402 03:22:44.094113 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.274615 (* 1 = 0.274615 loss) | |
| I0402 03:22:44.094125 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 03:22:44.094137 6134 solver.cpp:245] Train net output #16: total_confidence = 0.290559 | |
| I0402 03:22:44.094149 6134 sgd_solver.cpp:106] Iteration 200000, lr = 0.01 | |
| I0402 03:24:53.116948 6134 solver.cpp:229] Iteration 200500, loss = 2.55607 | |
| I0402 03:24:53.117094 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.346939 | |
| I0402 03:24:53.117115 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.8125 | |
| I0402 03:24:53.117127 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.571429 | |
| I0402 03:24:53.117144 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.14299 (* 0.3 = 0.642897 loss) | |
| I0402 03:24:53.117159 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.629219 (* 0.3 = 0.188766 loss) | |
| I0402 03:24:53.117172 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.510204 | |
| I0402 03:24:53.117184 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.852273 | |
| I0402 03:24:53.117197 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.77551 | |
| I0402 03:24:53.117210 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.48743 (* 0.3 = 0.44623 loss) | |
| I0402 03:24:53.117224 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.477481 (* 0.3 = 0.143244 loss) | |
| I0402 03:24:53.117236 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.714286 | |
| I0402 03:24:53.117249 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.903409 | |
| I0402 03:24:53.117260 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.816327 | |
| I0402 03:24:53.117274 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.18177 (* 1 = 1.18177 loss) | |
| I0402 03:24:53.117288 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.410686 (* 1 = 0.410686 loss) | |
| I0402 03:24:53.117300 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0402 03:24:53.117312 6134 solver.cpp:245] Train net output #16: total_confidence = 0.387233 | |
| I0402 03:24:53.117324 6134 sgd_solver.cpp:106] Iteration 200500, lr = 0.01 | |
| I0402 03:27:01.896141 6134 solver.cpp:229] Iteration 201000, loss = 2.55826 | |
| I0402 03:27:01.896494 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.431818 | |
| I0402 03:27:01.896515 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 03:27:01.896528 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.704545 | |
| I0402 03:27:01.896544 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.92826 (* 0.3 = 0.578478 loss) | |
| I0402 03:27:01.896560 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.5632 (* 0.3 = 0.16896 loss) | |
| I0402 03:27:01.896574 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.5 | |
| I0402 03:27:01.896585 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0402 03:27:01.896597 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.818182 | |
| I0402 03:27:01.896611 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.64273 (* 0.3 = 0.492819 loss) | |
| I0402 03:27:01.896625 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.483973 (* 0.3 = 0.145192 loss) | |
| I0402 03:27:01.896638 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.590909 | |
| I0402 03:27:01.896651 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.886364 | |
| I0402 03:27:01.896662 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.863636 | |
| I0402 03:27:01.896677 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 1.30246 (* 1 = 1.30246 loss) | |
| I0402 03:27:01.896692 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.348827 (* 1 = 0.348827 loss) | |
| I0402 03:27:01.896703 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.125 | |
| I0402 03:27:01.896715 6134 solver.cpp:245] Train net output #16: total_confidence = 0.178756 | |
| I0402 03:27:01.896728 6134 sgd_solver.cpp:106] Iteration 201000, lr = 0.01 | |
| I0402 03:29:10.523797 6134 solver.cpp:229] Iteration 201500, loss = 2.50932 | |
| I0402 03:29:10.523936 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.411765 | |
| I0402 03:29:10.523955 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 03:29:10.523968 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.588235 | |
| I0402 03:29:10.523985 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.01919 (* 0.3 = 0.605758 loss) | |
| I0402 03:29:10.524000 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.636453 (* 0.3 = 0.190936 loss) | |
| I0402 03:29:10.524013 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.509804 | |
| I0402 03:29:10.524025 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.835227 | |
| I0402 03:29:10.524037 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.745098 | |
| I0402 03:29:10.524051 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.71504 (* 0.3 = 0.514511 loss) | |
| I0402 03:29:10.524065 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.554398 (* 0.3 = 0.166319 loss) | |
| I0402 03:29:10.524077 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.843137 | |
| I0402 03:29:10.524090 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.948864 | |
| I0402 03:29:10.524101 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.921569 | |
| I0402 03:29:10.524116 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.649084 (* 1 = 0.649084 loss) | |
| I0402 03:29:10.524130 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.214108 (* 1 = 0.214108 loss) | |
| I0402 03:29:10.524142 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 03:29:10.524154 6134 solver.cpp:245] Train net output #16: total_confidence = 0.234556 | |
| I0402 03:29:10.524168 6134 sgd_solver.cpp:106] Iteration 201500, lr = 0.01 | |
| I0402 03:31:19.161463 6134 solver.cpp:229] Iteration 202000, loss = 2.56188 | |
| I0402 03:31:19.161573 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.34 | |
| I0402 03:31:19.161592 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.801136 | |
| I0402 03:31:19.161605 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.64 | |
| I0402 03:31:19.161623 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.12496 (* 0.3 = 0.637489 loss) | |
| I0402 03:31:19.161638 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.652134 (* 0.3 = 0.19564 loss) | |
| I0402 03:31:19.161650 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.56 | |
| I0402 03:31:19.161664 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0402 03:31:19.161675 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.78 | |
| I0402 03:31:19.161689 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.51941 (* 0.3 = 0.455822 loss) | |
| I0402 03:31:19.161705 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.47001 (* 0.3 = 0.141003 loss) | |
| I0402 03:31:19.161716 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.86 | |
| I0402 03:31:19.161730 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0402 03:31:19.161741 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.92 | |
| I0402 03:31:19.161756 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.663185 (* 1 = 0.663185 loss) | |
| I0402 03:31:19.161769 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.200125 (* 1 = 0.200125 loss) | |
| I0402 03:31:19.161782 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.375 | |
| I0402 03:31:19.161793 6134 solver.cpp:245] Train net output #16: total_confidence = 0.254977 | |
| I0402 03:31:19.161805 6134 sgd_solver.cpp:106] Iteration 202000, lr = 0.01 | |
| I0402 03:33:27.959579 6134 solver.cpp:229] Iteration 202500, loss = 2.51791 | |
| I0402 03:33:27.959703 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.395349 | |
| I0402 03:33:27.959723 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 03:33:27.959738 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.744186 | |
| I0402 03:33:27.959753 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.01303 (* 0.3 = 0.603908 loss) | |
| I0402 03:33:27.959769 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.559704 (* 0.3 = 0.167911 loss) | |
| I0402 03:33:27.959781 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.534884 | |
| I0402 03:33:27.959794 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 03:33:27.959805 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.744186 | |
| I0402 03:33:27.959820 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.47678 (* 0.3 = 0.443035 loss) | |
| I0402 03:33:27.959835 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.417148 (* 0.3 = 0.125144 loss) | |
| I0402 03:33:27.959846 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.837209 | |
| I0402 03:33:27.959858 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.954545 | |
| I0402 03:33:27.959870 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.953488 | |
| I0402 03:33:27.959884 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.550316 (* 1 = 0.550316 loss) | |
| I0402 03:33:27.959898 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.181522 (* 1 = 0.181522 loss) | |
| I0402 03:33:27.959910 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 03:33:27.959923 6134 solver.cpp:245] Train net output #16: total_confidence = 0.237636 | |
| I0402 03:33:27.959934 6134 sgd_solver.cpp:106] Iteration 202500, lr = 0.01 | |
| I0402 03:35:36.583065 6134 solver.cpp:229] Iteration 203000, loss = 2.56245 | |
| I0402 03:35:36.583205 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.488889 | |
| I0402 03:35:36.583236 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.835227 | |
| I0402 03:35:36.583258 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.688889 | |
| I0402 03:35:36.583287 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.13503 (* 0.3 = 0.640508 loss) | |
| I0402 03:35:36.583317 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.7526 (* 0.3 = 0.22578 loss) | |
| I0402 03:35:36.583338 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.577778 | |
| I0402 03:35:36.583359 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.857955 | |
| I0402 03:35:36.583380 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.866667 | |
| I0402 03:35:36.583405 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.60854 (* 0.3 = 0.482563 loss) | |
| I0402 03:35:36.583432 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.607431 (* 0.3 = 0.182229 loss) | |
| I0402 03:35:36.583456 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.822222 | |
| I0402 03:35:36.583477 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.926136 | |
| I0402 03:35:36.583498 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.933333 | |
| I0402 03:35:36.583526 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.901389 (* 1 = 0.901389 loss) | |
| I0402 03:35:36.583554 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.396582 (* 1 = 0.396582 loss) | |
| I0402 03:35:36.583575 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.5 | |
| I0402 03:35:36.583595 6134 solver.cpp:245] Train net output #16: total_confidence = 0.250669 | |
| I0402 03:35:36.583617 6134 sgd_solver.cpp:106] Iteration 203000, lr = 0.01 | |
| I0402 03:37:45.342015 6134 solver.cpp:229] Iteration 203500, loss = 2.48302 | |
| I0402 03:37:45.342363 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.454545 | |
| I0402 03:37:45.342384 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0402 03:37:45.342397 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.704545 | |
| I0402 03:37:45.342413 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.92115 (* 0.3 = 0.576345 loss) | |
| I0402 03:37:45.342428 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.576635 (* 0.3 = 0.172991 loss) | |
| I0402 03:37:45.342440 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.590909 | |
| I0402 03:37:45.342453 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.880682 | |
| I0402 03:37:45.342465 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.840909 | |
| I0402 03:37:45.342478 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.4259 (* 0.3 = 0.42777 loss) | |
| I0402 03:37:45.342494 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.432336 (* 0.3 = 0.129701 loss) | |
| I0402 03:37:45.342505 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.909091 | |
| I0402 03:37:45.342520 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.977273 | |
| I0402 03:37:45.342533 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.977273 | |
| I0402 03:37:45.342547 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.400654 (* 1 = 0.400654 loss) | |
| I0402 03:37:45.342561 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.106647 (* 1 = 0.106647 loss) | |
| I0402 03:37:45.342573 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.75 | |
| I0402 03:37:45.342586 6134 solver.cpp:245] Train net output #16: total_confidence = 0.403202 | |
| I0402 03:37:45.342597 6134 sgd_solver.cpp:106] Iteration 203500, lr = 0.01 | |
| I0402 03:39:54.126682 6134 solver.cpp:229] Iteration 204000, loss = 2.49187 | |
| I0402 03:39:54.126793 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.409091 | |
| I0402 03:39:54.126813 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.823864 | |
| I0402 03:39:54.126826 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.75 | |
| I0402 03:39:54.126842 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 1.6022 (* 0.3 = 0.480661 loss) | |
| I0402 03:39:54.126857 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.488088 (* 0.3 = 0.146426 loss) | |
| I0402 03:39:54.126869 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.522727 | |
| I0402 03:39:54.126881 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.869318 | |
| I0402 03:39:54.126893 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.795455 | |
| I0402 03:39:54.126907 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.34766 (* 0.3 = 0.404298 loss) | |
| I0402 03:39:54.126921 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.392043 (* 0.3 = 0.117613 loss) | |
| I0402 03:39:54.126934 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.795455 | |
| I0402 03:39:54.126945 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.9375 | |
| I0402 03:39:54.126957 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.863636 | |
| I0402 03:39:54.126971 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.787345 (* 1 = 0.787345 loss) | |
| I0402 03:39:54.126984 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.219588 (* 1 = 0.219588 loss) | |
| I0402 03:39:54.126996 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 03:39:54.127008 6134 solver.cpp:245] Train net output #16: total_confidence = 0.476849 | |
| I0402 03:39:54.127020 6134 sgd_solver.cpp:106] Iteration 204000, lr = 0.01 | |
| I0402 03:42:02.704972 6134 solver.cpp:229] Iteration 204500, loss = 2.51691 | |
| I0402 03:42:02.705121 6134 solver.cpp:245] Train net output #0: loss1/accuracy = 0.456522 | |
| I0402 03:42:02.705142 6134 solver.cpp:245] Train net output #1: loss1/accuracy_incl_empty = 0.840909 | |
| I0402 03:42:02.705154 6134 solver.cpp:245] Train net output #2: loss1/accuracy_top3 = 0.73913 | |
| I0402 03:42:02.705171 6134 solver.cpp:245] Train net output #3: loss1/cross_entropy_loss = 2.02285 (* 0.3 = 0.606856 loss) | |
| I0402 03:42:02.705186 6134 solver.cpp:245] Train net output #4: loss1/cross_entropy_loss_incl_empty = 0.601571 (* 0.3 = 0.180471 loss) | |
| I0402 03:42:02.705199 6134 solver.cpp:245] Train net output #5: loss2/accuracy = 0.543478 | |
| I0402 03:42:02.705211 6134 solver.cpp:245] Train net output #6: loss2/accuracy_incl_empty = 0.863636 | |
| I0402 03:42:02.705224 6134 solver.cpp:245] Train net output #7: loss2/accuracy_top3 = 0.826087 | |
| I0402 03:42:02.705238 6134 solver.cpp:245] Train net output #8: loss2/cross_entropy_loss = 1.58409 (* 0.3 = 0.475228 loss) | |
| I0402 03:42:02.705252 6134 solver.cpp:245] Train net output #9: loss2/cross_entropy_loss_incl_empty = 0.488011 (* 0.3 = 0.146403 loss) | |
| I0402 03:42:02.705265 6134 solver.cpp:245] Train net output #10: loss3/accuracy = 0.782609 | |
| I0402 03:42:02.705277 6134 solver.cpp:245] Train net output #11: loss3/accuracy_incl_empty = 0.931818 | |
| I0402 03:42:02.705289 6134 solver.cpp:245] Train net output #12: loss3/accuracy_top3 = 0.934783 | |
| I0402 03:42:02.705303 6134 solver.cpp:245] Train net output #13: loss3/cross_entropy_loss = 0.83162 (* 1 = 0.83162 loss) | |
| I0402 03:42:02.705318 6134 solver.cpp:245] Train net output #14: loss3/cross_entropy_loss_incl_empty = 0.242162 (* 1 = 0.242162 loss) | |
| I0402 03:42:02.705330 6134 solver.cpp:245] Train net output #15: total_accuracy = 0.625 | |
| I0402 03:42:02.705343 6134 solver.cpp:245] Train net output #16: total_confidence = 0.48424 | |
| I0402 03:42:02.705354 6134 sgd_solver.cpp:106] Iteration 204500, lr = 0.01 | |
| I0402 03:44:11.268452 6134 solver.cpp:338] Iteration 205000, Testing net (#0) | |
| I0402 03:44:41.022145 6134 solver.cpp:393] Test loss: 2.33548 | |
| I0402 03:44:41.022191 6134 solver.cpp:406] Test net output #0: loss1/accuracy = 0.528769 | |
| I0402 03:44:41.022207 6134 solver.cpp:406] Test net output #1: loss1/accuracy_incl_empty = 0.871958 | |
| I0402 03:44:41.022220 6134 solver.cpp:406] Test net output #2: loss1/accuracy_top3 = 0.782335 | |
| I0402 03:44:41.022235 6134 solver.cpp:406] Test net output #3: loss1/cross_entropy_loss = 1.66212 (* 0.3 = 0.498636 loss) | |
| I0402 03:44:41.022250 6134 solver.cpp:406] Test net output #4: loss1/cross_entropy_loss_incl_empty = 0.459256 (* 0.3 = 0.137777 loss) | |
| I0402 03:44:41.022263 6134 solver.cpp:406] Test net output #5: loss2/accuracy = 0.679723 | |
| I0402 03:44:41.022275 6134 solver.cpp:406] Test net output #6: loss2/accuracy_incl_empty = 0.917728 | |
| I0402 03:44:41.022287 6134 solver.cpp:406] Test net output #7: loss2/accuracy_top3 = 0.857787 | |
| I0402 03:44:41.022300 6134 solver.cpp:406] Test net output #8: loss2/cross_entropy_loss = 1.2162 (* 0.3 = 0.364859 loss) | |
| I0402 03:44:41.022315 6134 solver.cpp:406] Test net output #9: loss2/cross_entropy_loss_incl_empty = 0.322315 (* 0.3 = 0.0966946 loss) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment